Complexity

Complexity and Project Management: Challenges, Opportunities, and Future Research

Lead Guest Editor: José Ramón San Cristóbal Mateo Guest Editors: Emma Diaz Ruiz de Navamuel, Luis Carral Couce, José Ángel Fraguela Formoso, and Gregorio Iglesias Complexity and Project Management: Challenges, Opportunities, and Future Research Complexity

Complexity and Project Management: Challenges, Opportunities, and Future Research

Lead Guest Editor: José Ramón San Cristóbal Mateo Guest Editors: Emma Diaz Ruiz de Navamuel, Luis Carral Couce, José Ángel Fraguela Formoso, and Gregorio Iglesias Copyright © 2019 . All rights reserved.

This is a special issue published in “Complexity.” All articles are articles distributed under the Creative Commons Attribu- tion License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Editorial Board

José A. Acosta, Spain Carlos Gershenson, Mexico Irene Otero-Muras, Spain Carlos F. Aguilar-Ibáñez, Mexico Peter Giesl, UK Yongping Pan, Singapore Mojtaba Ahmadieh Khanesar, UK Sergio Gómez, Spain Daniela Paolotti, Italy Tarek Ahmed-Ali, France Lingzhong Guo, UK Cornelio Posadas-Castillo, Mexico Alex Alexandridis, Greece Xianggui Guo, China Mahardhika Pratama, Singapore Basil M. Al-Hadithi, Spain Sigurdur F. Hafstein, Iceland Luis M. Rocha, USA Juan A. Almendral, Spain Chittaranjan Hens, Israel Miguel Romance, Spain Diego R. Amancio, Brazil Giacomo Innocenti, Italy Avimanyu Sahoo, USA David Arroyo, Spain Sarangapani Jagannathan, USA Matilde Santos, Spain Mohamed Boutayeb, France Mahdi Jalili, Australia Josep Sardanyés Cayuela, Spain Átila Bueno, Brazil Jeffrey H. Johnson, UK Ramaswamy Savitha, Singapore Arturo Buscarino, Italy M. Hassan Khooban, Denmark Hiroki Sayama, USA Guido Caldarelli, Italy Abbas Khosravi, Australia Michele Scarpiniti, Italy Eric Campos-Canton, Mexico Toshikazu Kuniya, Japan Enzo Pasquale Scilingo, Italy Mohammed Chadli, France Vincent Labatut, France Dan Selişteanu, Romania Émile J. L. Chappin, Netherlands Lucas Lacasa, UK Dehua Shen, China Diyi Chen, China Guang Li, UK Dimitrios Stamovlasis, Greece Yu-Wang Chen, UK Qingdu Li, Germany Samuel Stanton, USA Giulio Cimini, Italy Chongyang Liu, China Roberto Tonelli, Italy Danilo Comminiello, Italy Xiaoping Liu, Canada Shahadat Uddin, Australia Sara Dadras, USA Xinzhi Liu, Canada Gaetano Valenza, Italy Sergey Dashkovskiy, Germany Rosa M. Lopez Gutierrez, Mexico Dimitri Volchenkov, USA Manlio De Domenico, Italy Vittorio Loreto, Italy Christos Volos, Greece Pietro De Lellis, Italy Noureddine Manamanni, France Zidong Wang, UK Albert Diaz-Guilera, Spain Didier Maquin, France Yan-Ling Wei, Singapore Thach Ngoc Dinh, France Eulalia Martínez, Spain Honglei Xu, Australia Jordi Duch, Spain Marcelo Messias, Brazil Yong Xu, China Marcio Eisencraft, Brazil Ana Meštrović, Croatia Xinggang Yan, UK Joshua Epstein, USA Ludovico Minati, Japan Baris Yuce, UK Mondher Farza, France Ch. P. Monterola, Philippines Massimiliano Zanin, Spain Thierry Floquet, France Marcin Mrugalski, Poland Hassan Zargarzadeh, USA Mattia Frasca, Italy Roberto Natella, Italy Rongqing Zhang, USA José Manuel Galán, Spain Sing Kiong Nguang, New Zealand Xianming Zhang, Australia Lucia Valentina Gambuzza, Italy Nam-Phong Nguyen, USA Xiaopeng Zhao, USA Bernhard C. Geiger, Austria B. M. Ombuki-Berman, Canada Quanmin Zhu, UK Contents

Complexity and Project Management: Challenges, Opportunities, and Future Research José R. San Cristóbal ,EmmaDiaz,LuisCarral ,JoséA.Fraguela,andGregorioIglesias Editorial (2 pages), Article ID 6979721, Volume 2019 (2019)

Complexity and Project Management: A General Overview José R. San Cristóbal , Luis Carral ,EmmaDiaz,JoséA.Fraguela,andGregorioIglesias Review Article (10 pages), Article ID 4891286, Volume 2018 (2019)

Comparing Project Complexity across Different Industry Sectors Marian Bosch-Rekveldt , Hans Bakker, and Marcel Hertogh Research Article (15 pages), Article ID 3246508, Volume 2018 (2019)

Exploring Project Complexity through Project Failure Factors: Analysis of Cluster Patterns Using Self-Organizing Maps Vicente Rodríguez Montequín , Joaquín Villanueva Balsera, Sonia María Cousillas Fernández, and Francisco Ortega Fernández Research Article (17 pages), Article ID 9496731, Volume 2018 (2019)

Measuring the Project Management Complexity: The Case of Information Projects Rocio Poveda-Bautista ,Jose-AntonioDiego-Mas , and Diego Leon-Medina Research Article (19 pages), Article ID 6058480, Volume 2018 (2019)

Strategies for Managing the Structural and Dynamic Consequences of Project Complexity Serghei Floricel , Sorin Piperca, and Richard Tee Research Article (17 pages), Article ID 3190251, Volume 2018 (2019) Hindawi Complexity Volume 2019, Article ID 6979721, 2 pages https://doi.org/10.1155/2019/6979721

Editorial Complexity and Project Management: Challenges, Opportunities, and Future Research

José R. San Cristóbal ,1 Emma Diaz,2 Luis Carral ,3 José A. Fraguela,3 and Gregorio Iglesias4

1 Project Management Research Group, Universidad de Cantabria, Santander 39004, Spain 2Escuela Tecnica´ Superior de Nautica,´ Universidad de Cantabria, Santander 39004, Spain 3Department of Naval and Industrial Engineering, GEM, Universidade da Coruna,˜ Ferrol 15403, Spain 4University of Plymouth, Plymouth, UK

Correspondence should be addressed to JoseR.SanCrist´ obal;´ [email protected]

Received 23 December 2018; Accepted 24 December 2018; Published 6 January 2019

Copyright ©  JoseR.SanCrist´ obal´ et al. is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Owing to its societal and economic relevance, Project Man- complexity. A large amount of required resources, a turbulent agement has become an important and relevant discipline environment, working on the edge of technology, and a large and a key concept in modern private and public organiza- number and diversity of actors working and communicating tions. Modern Project Management appeared during World with each other are all factors that affect project outcome. War II and, since then, it has grown up and spread around is complex environment influences project planning, coor- the world to become what it is today, that is, a set of practices, dination, and control; it can also affect the selection of an principles, theories, and methodologies. appropriate project organization structure and hinder the Project Management is the application of knowledge, clear identification of project goals. skills, tools, and techniques to project activities in order When problems fundamentally dynamic are treated stat- to meet project objectives. It is more applied and inter- ically, delays and cost overruns are common. Experience disciplinary than other management disciplines. Nowadays, suggests that the interrelationships between the project’s Project Management is a well-recognized discipline practiced components are more complex than it is suggested by tra- by almost all organizations which has accumulated exten- ditional approaches. ese, traditional approaches, using a sive knowledge and wide-industry based experience. Today, static approach, provide project managers with unrealistic project managers have gained recognition and employment estimations that ignore the nonlinear relationships of a opportunities beyond construction, aerospace, and defense, projectand,thus,areinadequatetothechallengeoftoday’s in pharmaceuticals, information systems, and manufactur- dynamic and complex projects. ing. Complex projects demand an exceptional level of man- However, paradoxically, many projects do not meet cus- agement and the application of the traditional tools and tomer expectations, and cost and schedule overruns are quite techniques developed for ordinary projects have been found common. Why then is so much effort spent today on projects to be inappropriate for complex projects. Complex Project to achieve only moderate levels of success? What is missing Management is a specialist profession that requires a specific in Project Management? Project Management is a complex set of competencies and a deep understanding of the project activity and a risky organizational adventure, different than and its environment. If project managers want to execute any functional activity or ongoing operation. a project successfully, in a context of increased complexity, Projects have become more and more complex because it is not only necessary that they attend the demands of of the increasing factors that are considered source of an increasingly complex environment or they develop the  Complexity right strategies to address the new challenges, a willingness to change leadership style will also be required. Project managers must be able to make decisions in the dynamic and unstable environments that are continuously changing and evolving in a random fashion and are hard to predict. To achieve this objective, more integrated approaches for managing projects and new methods of planning, scheduling, executing, and controlling projects must be investigated. e aim of this special issue is to publish selective ongoing research contributions that contribute and stimulate the debate in the topic.

Conflicts of Interest Regarding this special issue, the lead guest editor and the others guest editors do not have any possible conflicts of interest or private agreements with companies. JoseR.SanCrist´ obal´ Emma Diaz Luis Carral JoseA.Fraguela´ Gregorio Iglesias Hindawi Complexity Volume 2018, Article ID 4891286, 10 pages https://doi.org/10.1155/2018/4891286

Review Article Complexity and Project Management: A General Overview

1 2 3 2 José R. San Cristóbal , Luis Carral , Emma Diaz, José A. Fraguela, 4 and Gregorio Iglesias

1Project Management Research Group, Universidad de Cantabria, Santander 39004, Spain 2Department of Naval and Industrial Engineering, GEM, Universidade da Coruña, Ferrol 15403, Spain 3Escuela Técnica Superior de Náutica, Universidad de Cantabria, Santander 39004, Spain 4University of Plymouth, Plymouth, UK

Correspondence should be addressed to José R. San Cristóbal; [email protected]

Received 25 April 2018; Accepted 25 July 2018; Published 10 October 2018

Academic Editor: Roberto Natella

Copyright © 2018 José R. San Cristóbal et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

As projects have become more and more complex, there has been an increasing concern about the concept of project complexity. An understanding of project complexity and how it might be managed is of significant importance for project managers because of the differences associated with decision-making and goal attainment that are related to complexity. Complexity influences project planning and control; it can hinder the clear identification of goals and objectives, it can affect the selection of an appropriate project organization form, or it can even affect project outcomes. Identifying the different concepts associated to project complexity, its main factors and characteristics, the different types of project complexity, and the main project complexity models, can be of great support in assisting the global project management community. In this paper, we give a general overview of how complexity has been investigated by the project management community and propose several ideas to address this topic in the future.

1. Introduction arrangement; and (v) it can affect different project outcomes (time, cost, quality, safety, etc.). The origins of complexity theory applied to project man- An understanding of project complexity and how it agement can be traced back to the works by Morris [1, 2], might be managed is of significant importance for project Bennet and Fine [3], Bubshait and Selen [4], Bennet and managers because of the differences associated with Cropper [5], Gidado [6], Wozniak [7], and Baccarini [8]. decision-making and goal attainment that appear to be All these works highlight the importance of complexity in related to complexity [8, 9]. As projects have become more project contexts in general and in particular its effects on and more complex, there has been an increasing concern project goals and objectives, project organization form and about the concept of project complexity and the application arrangement, and in the experience requirements for the of traditional tools and techniques developed for simple management personnel. projects has been found to be inappropriate for complex The importance of complexity to the project manage- projects [1, 8]. According to Parsons-Hann and Liu [10], ment process is widely acknowledged for several reasons it is evident that complexity contributes to project failure [1–8]: (i) it influences project planning, coordination, and in organizations; what is not clear is to what degree this control; (ii) it hinders the clear identification of goals and statement holds true. Identifying and characterizing different objectives of major projects; (iii) it can affect the selection aspects of project complexity in order to understand more of an appropriate project organization form and experience efficiently the stakes of project management complexity requirements of management personnel; (iv) it can be used can be of great support in assisting the global project as criteria in the selection of a suitable project management management community. 2 Complexity

Complexity can have both a negative and a positive [35] associates linearity with complicated projects and influence on projects. The negative influence, in terms of nonlinearity with complex projects, which implies that difficulty to be understood and controlled, is because of the nonlinearity makes the relationship between inputs and emergence of new properties that none of the elements of outputs unpredictable. Remington et al. [9] defines a com- the system owns. The positive influence is due to the appari- plex project as one that demonstrates a number of charac- tion of phenomena that could not be predicted due to the sole teristics to a degree or level of severity that makes it knowing, even complete, of the behaviour and interactions extremely difficult to predict project outcomes, to control of the elements of the system. In order to properly man- or manage the project. Girmscheid and Brockmann [34] age complexity, project managers must know how to seize define project complexity as a set of problems that consists the opportunities emerging from complexity and to know of many parts with a multitude of possible interrelations, how to avoid or at least diminish the negative effects of most of them being of high consequence in the decision- complexity [11]. making process that brings about the final result. In this paper, we give a general overview of how complex- ity, which is the main purpose of this special issue, has been 3. Project Complexity Factors addressed to date in the project management literature. We and Characteristics begin by discussing the different definitions of complexity in project contexts. Next, a summary of the project complex- Experience suggests that the interrelationships between the ity factors and characteristics is presented. Then, the different project’s components are more complex than is suggested types of project complexity and the main project complexity by the traditional work breakdown structure of project models are presented. Finally, the current and the future network. Identifying the sources and factors that contribute management approaches to address this topic in the future or increase project complexity is paramount for project are proposed. managers. Gidado [36] determines four different sources of complexity: employed resources, environment, level of 2. Definitions of Project Complexity scientific and technological knowledge required, and number of different parts in the workflow. Thus, a large amount of In project contexts, there is a lack of consensus on what com- required resources, a turbulent environment, working on plexity really is [12–20]. There does not even seem to be a the edge of technology, and innumerable possible interac- single definition of project complexity that can capture the tions are certainly identifiable factors for complex projects. whole concept [11, 20–24]. Within the Luhmannian system Since there has been a lack of consensus and difficulty in theory, complexity is the sum of the following components defining complexity, some authors have focused on identify- [25]: differentiation of functions between project partici- ing the factors that contribute or increase project complexity. pants, dependencies between systems and subsystems, and Remington et al. [9] suggest to differentiate between dimen- the consequential impact of a decision field. Project complex- sions, characteristics, or sources of complexity, and severity ity can also be interpreted and operationalized in terms of factors, those factors that increase or decrease the severity differentiation (number of elements in a project) and inter- of complexity. Vidal and Marle [11] consider the following dependencies and connectivity (degree of interrelatedness factors as necessary but nonsufficient conditions for pro- between these elements), which are managed by integra- ject complexity: size, variety, interdependences and interre- tion, that is, by coordination, communication, and control lations within the project system, and context dependence. [1, 8, 26–29]. Custovic [30] defines complexity as that Remington et al. [9] group a number of factors that seem property of a system which makes it difficult to formulate to contribute to the perception of project complexity under its overall behaviour in a given language, even when given the following headings: goals, stakeholders, interfaces and reasonable complete information about its atomic compo- interdependencies, technology, management process, work nents and their interrelations. In a similar context, Vidal practices, and time. Table 1 shows the main factors that are and Marle [11] define project complexity as that property considered in the literature as drivers of project complexity. of a project which makes it difficult to understand, foresee, and keep under control its overall behaviour. Tatikonda 3.1. Size. Size has traditionally been considered the primary and Rosenthal [31] view complexity as consisting of interde- cause of complexity in organizations [37–40]. However, to pendencies among the product and process and consider size an indication of complexity, the organizational novelty and difficulty of goals. Pich et al. [32] define complex- structure of a system should be over a minimum critical size ity as information inadequacy when too many variables and their elements need to be interrelated [41]. Substantial interact. Ward and Chapman [33] view the number of relationships have been found in both cross-sectorial and influencing factors and their interdependencies as constitu- longitudinal studies in many different samples of organiza- ents of complexity. tions between size and various components of complexity Some authors associate complex or complicated projects such as personal specialization, division of labor, and with the number of elements and with the concept of linear- structural differentiation [38]. A large number of studies have ity. Girmscheid and Brockmann [34] argue that any differ- found that size is related to structural differentiation, but the ence between a complicated project and a complex project relationship between size and complexity is less clear [37, 40, has to do with the number of elements as opposed to the 42] . According to a study performed by Beyer and Trice [38] relationships between the elements (complex). Richardson on several departments of the US governments, size is a more Complexity 3

Table 1: Main factors affecting project complexity.

Factor To consider it an indication of complexity, the organizational structure of the project should be over Size a minimum critical size and their elements need to be interrelated. An event in an interconnected structure can cause totally unknown effects on another entity inside Interdependence and interrelations the structure. Goals and objectives They must be adequately and properly defined both at a strategic and at an operational level. The number of project participants and how the information flows between them are a key factor Stakeholders affecting project complexity. Relationships between project participants, suppliers, overlapping of activities, methods, and Management practices techniques are factors that affect project complexity. Adding project organizational structure by dividing labor, the way for personnel selection, and Division of labor the level of pressure on this personnel to achieve project objectives are factors that increase project complexity. Task scope or the variety of tasks that need to be accomplished is the most critical dimension of Technology technology. It explains why there is a need for a variety of technologies and a given level specialization in each of them. It breaks down functional and departmental barriers by integrating team members with different Concurrent engineering discipline backgrounds often known as cross-functional teams. boots complexity by the erosion of boundaries, higher mobility, heterarchy, and Globalization and context dependence higher dynamics. It can be an essential feature of complexity. Diversity A higher number of elements and a higher variety across elements increase complexity. Ambiguity It expresses uncertainty of meaning in which multiple interpretations are plausible. Flux is affected by external and internal influences. It also implies constant change and adaptation to Flux changing conditions. important predictor of complexity while in a similar study 3.5. Management Practices. Organizational and interactive from state employment agencies, Blau and Schoenherr [37] management is one of the riskiest parts of a project. found that division of labor is a more important predictor Contractor relationships and ethics, supplier monopolies, of complexity. overlapping of processes and activities, methodologies, and techniques based on either hard or soft approaches that can 3.2. Interdependence and Interrelations. It creates a link or affect the degree of definition of project goals and objectives influence of different types between entities in such a are all factors that can influence project complexity. way that an event in an interconnected structure can cause totally unknown effects on another entity inside the struc- 3.6. Division of Labor. Dividing labor into distinct tasks and ture [43]. The number of systems and subsystems that coordinating these tasks define the structure of an organiza- integrate the project, the different methodological and tion [44]. Adding project organizational structure by philosophical assumptions across these systems, the cross- dividing labor into smaller and more specialized tasks, the organizational and schedule interdependencies between way for personnel selection, and the level of pressure on the activities, the upgrading and retrofitting works, and the sheer personnel to achieve project objectives are all factors that size and entanglement in the project are all key factors can increase project complexity. influencing complexity. 3.7. Technology. Broadly speaking, technology can be defined 3.3. Goals and Objectives. Goals and objectives must be as the transformation process which converts inputs into adequately and properly defined, both at a strategic and at outputs using materials, means, techniques, knowledge, and an operational level. In addition, all project participants skills [8, 26]. The most critical dimension of technology is including owners, managers, contractors, and consultants the variety of tasks that need to be accomplished, what is must be clear about these goals and objectives. sometimes called task scope and is proposed as a determinant of horizontal differentiation [42]. It explains why there is a 3.4. Stakeholders. The number of project participants and need for a variety of technologies and a given level special- how the information flows between them are a key factor ization in each of them. Baccarini [8] proposes to define affecting project complexity. If the project is politically technological complexity in terms of differentiation and sensitive and of high visibility, project complexity can con- interdependencies. Technological complexity by differenti- siderably be increased. Managing conflicting agendas of ation refers to the variety and diversity of some aspects various stakeholder management strategies and processes, of a task such as number and diversity of inputs/outputs, which is linked to structural complexity, can also amplify number and diversity of tasks to undertake, and number the complexity of a project. of specialities and contractors, involved in the project. 4 Complexity

Technological complexity by interdependency encompasses seems to be in line with Baccarini’s [8] opinion on organiza- interdependencies between tasks, within a network of tasks, tional complexity which, according to him, is influenced by between teams, between different technologies, and between differentiation and operative interdependencies. inputs (technological interdependence can be one of three According to Vidal and Marle [11], there are historically types, pooled, sequential, and reciprocal, with reciprocal two main approaches of complexity. The one, usually known interdependency the prevalent type in construction projects). as the field of descriptive complexity, considers complexity as an intrinsic property of a system, a vision which invited 3.8. Concurrent Engineering. The ever increasing pressure researchers to try to quantify or measure complexity. The to execute projects more rapidly has led many companies other one, usually known as the field of perceived complexity, to deploy project organizations comprised of distributed considers complexity as subjective, since the complexity of a and often outsourced teams and in many cases to execute system is improperly understood through the perception of concurrently many activities [45]. Concurrent Engineering an observer. For all practical purposes, a project manager breaks down functional and departmental barriers by inte- deals with perceived complexity as he cannot understand grating team members with different discipline back- and deal with the whole reality and complexity of the project. grounds often known as cross-functional teams [46]. This According to this perceived complexity, project managers process requires changes in the organizational structure make the corresponding decisions and take the correspond- and a more vigorous communication, coordination, and ing actions to influence the project evolution and reach the collaboration [47]. desired project state [11, 49]. How complexity is perceived and interpreted by project 3.9. Globalization and Context Dependence. Globalization managers may result in different types of project complexity. boots complexity by the erosion of boundaries, higher mobil- Baccarini [8] considers technological and organizational ity, heterarchy, and higher dynamics [46]. The context and complexities as the core components of project complexity. environment under which the project is undertaken can be According to [25, 34], four different types of project com- an essential feature of complexity. In fact, the methods and plexity, overall, task, social, and cultural, help to best under- practices applicable to a project may not be directly transfer- stand and prevent projects from failure. Task complexity able to other projects with different institutional, language, fi refers to the density of the units, causal links, and conse- and cultural con gurations. quences within a temporal and spatial frame. Social complex- ity describes the number of members communicating and 3.10. Diversity. Diversity is defined as the plurality of working with each other and the differentiation of their tasks, elements. It encompasses two components, the number of while cultural complexity encompasses the number of differ- elements (multiplicity) and their dissimilarity (variety). A ent historical experiences and sense-making processes that higher number of elements and a higher variety across confront each other in a project. Cultural complexity com- elements increase complexity. presses the history, experience, and sense-making processes ff ff 3.11. Ambiguity. Ambiguity can be defined as too much of di erent groups that joint the e ort in a project. Overall information with less and less clarity on how to interpret and task complexity can be managed by a functional orga- and apply findings [43]. Ambiguity expresses uncertainty of nization with decentralized decision-making and social meaning in which multiple interpretations are plausible complexity by trust and commitment, whereas cultural which leads to the existence of multiple, often conflicting complexity by sense-making processes. situations, goals, and processes [46]. Pollack and Remington and Pollack [50, 51] emphasize that a clear distinction on the type of complexity helps in 3.12. Flux. Flux implies constant change and adaptation to selecting the appropriate model to manage a project. Based changing conditions making temporary solutions regarding on the source of complexity, the authors suggest four types interdependence, diversity, and ambiguity outdated from of project complexity: structural, technical, directional, and one day to another [48]. Flux is affected by external and temporal complexity. Structural complexity stems from internal influences. External influences can either be political large-scale projects which are typically broken down into or -related changes, while internal influences come small tasks and separate contracts. Projects in the engineer- from changes in strategy, in individual behaviour, etc. ing, construction, IT, and defence sectors where the complex- ity stems from the difficulty in managing and keeping track of 4. Types of Project Complexity the high number of interconnected tasks and activities are likely to have this type of complexity [51]. Technical com- Bosch-Rekveldt et al. [16] conducted an online survey using plexity is found in architectural, industrial design, and R&D the TOE framework (technical, organizational, and environ- projects which have design characteristics or technical mental) and came to determine the position of the aspects that are unknown or untried and where complexity respondents about the nature of the complexity of the arises because of the uncertainty regarding the outcome for organization in engineering projects. They concluded that many independent design solutions [51]. Baccarini [8] cate- project managers were more concerned with organizational gorizes technological complexity in terms of differentiation complexity than technical or environmental complexities. and interdependence, which is further categorized into three Vidal and Marle [11] argued that approximately 70% of the types given in an ascending order of complexity: (i) pooled, complexity factors of the project are organizational. This in which each element gives a discrete contribution to the Complexity 5 project; (ii) sequential, where one element’s output becomes another’s input; and (iii) reciprocal, where each element’s Type 2 projects Type 4 projects output becomes inputs for other elements [51, 52]. Direc- No tional complexity is often found in change projects where Product development R & D and the direction of the project is not understood and when it is organizational change clear that something must be done to improve a problematic Methods situation [51]. Temporal complexity results in projects where well-defined due to unexpected legislative changes of rapid changes in Type 1 projects Type 3 projects technology, there is a high level of uncertainty regarding Yes Engineering & Application soware future constraints that could destabilize the project. Opera- construction development tive complexity, i.e., the degree to which organizations of the project are independent when defining their operations Yes No to achieve given goals, and cognitive complexity which Goals well-defined identifies the degree to which self-reflection, sense-making processes, the emergence of an identity, or even an organi- Figure 1: Goals and methods matrix [53]. zational culture is possible, are also different types of complexity identified in the literature [36]. used to solve this type of situations; and (iv) far from agree- 5. Project Complexity Models ment far from certainty: this is the zone of anarchy with a high level of uncertainty and where traditional management Trying to find the most appropriate model for managing a techniques will not work. project can be a difficult task. If the model is too simple, it is not enough close to reality. On the contrary, if it is too 5.3. William’s Model. Williams and Hillson [55] extend complex, it can be useless to project managers. Next, some Baccarini’s model by one additional dimension. In addition of the most relevant complexity models in the project to the two components of complexity suggested by Baccarini, management literature will be revised. i.e., the number of elements and the interdependency of these elements, the authors introduce uncertainty and attributes 5.1. Goals and Methods Matrix. Based on how well-defined the increasing complexity in projects to two compounding are the goals and methods of achieving these goals in a pro- causes, the relationship between product complexity and ject, Turner and Cochrane [53] developed the goals and project complexity and the length of projects. The resulting methods matrix shown in Figure 1 where four types of pro- model is shown in Figure 3 where, as can be seen, project jects can be found: (i) type 1 projects are projects in which complexity is characterized by two dimensions, structural goals and methods are well-defined and understood. In this complexity and uncertainty, each of one having two subdi- case, the role of the project manager is that of a conductor; mensions, number and interdependency of elements, and (ii) type 2 projects are projects with well-defined goals but uncertainty in goals and methods, respectively. poorly defined methods. In this case, the role of the project manager is that of a coach; (iii) type 3 projects are projects 5.4. Kahane’s Approach. Kahane’s [56] approach to complex- planned in life-cycle stages with poorly defined goals but ity is deeply rooted in a social environment. He introduces well-defined methods; and (iv) type 4 projects are projects the U-process as a methodology for addressing complex with no defined goals and no defined methods. Typically, challenges and distinguishes complexity in three ways: engineering and construction projects fall within the cate- (i) dynamic complexity: the cause and effect of complexity gory of type 1 projects. Product development projects belong are far apart and it is hard to grasp from first-hand experi- to type 2, while application software development and R&D ence; (ii) generative complexity: a situation where the solu- and organizational change projects belong to type 3 and type tion cannot be calculated in advance based only on what 4 projects, respectively. has worked in the past; and (iii) social complexity: the people involved who have different perspectives and interests must 5.2. Stacey’s Agreement and Certainty Matrix. Stacey [54] participate in creating and implementing the solution. When analysed complexity on two dimensions, the degree of cer- using the U-process developed by Kahane [56], project tainty and the level of agreement and, based on these two managers undertake three activities: (i) sensing the current dimensions, developed the matrix shown in Figure 2 with reality of the project; (ii) reflecting about what is going on the following zones: (i) close to agreement, close to certainty: and what they have to do; and (iii) realizing and acting in this zone, we can find simple projects where traditional quickly to bring forth a new reality. project management techniques work well and the goal is to identify the right process to maximize efficiency and 5.5. Cynefin Decision-Making Framework. Snowden and effectives; (ii) far from agreement, close to certainty: in this Boone [57] developed the Cynefin framework, which allows case, coalitions, compromise, and negotiation are used to executives to see new things from new viewpoints, assimilate solve this type of situations; (iii) close to agreement, far from complex concepts, and address real-world problems and certainty: in this case, traditional project management tech- opportunities. The framework sorts it into five domains, niques may not work and leadership approaches must be simple, complicated, complex, chaotic, and disorder, each 6 Complexity

(ii) Scope 2: System. A complex collection of inter- Anarchy active units jointly performing a wide range of functions Far from agreement Complex (iii) Scope 3: Array. A large collection of systems func- Complicated tioning together to achieve a common purpose (c) The pace dimension

Close to Simple Complicated

agreement (i) Regular Projects. Projects that, although con- fined to a limited time-frame, still can achieve Close to certainty Far from certainty their objectives

Figure 2: Agreement and certainty matrix [54]. (ii) Fast-Competitive Projects. Projects conceived to create strategic positions, address market oppor- tunities, etc. In this type of projects, since time to of one requires different actions based on cause and effect. market is directly associated with competitive- The simple and complicated domains are characterized by ness, missing the deadline might not be fatal cause and effect relationships, and right answers can be deter- but it could hurt competitive positions mined based on facts. The complex and chaotic domains do (iii) Critical-blitz projects are the most urgent and ff not have a clear cause and e ect relationship, and decisions most time-critical projects in which meeting must be made based on incomplete data. The last domain, schedule is critical to success and project delay disorder, is applied when it is unclear which of the four is means project failure. dominant and is tackled by breaking it down into smaller components and then assigning them to the other four domains. Table 2 shows the characteristics of each context, 6. Current and Future Approaches to the leader’s job, the danger signals, and the response to these Manage Complexity danger signals [57]. Understanding how project managers deal with the different types of complexity and how they reply to these different fi 5.6. The UCP Model. The UCP model classi es projects types can help to prevent projects from failure. Stacey [54], according to uncertainty, complexity, and pace. Further- Kahane [56], and Snowden and Boone [57] focus on how more, uncertainty has been broken down into four levels complexity, particularly messy or ill-structured problems, of technological uncertainty (low-, medium-, high-, and might influence leadership style and decision-making in super high-technology projects). Complexity into three periods of organizational change. Clift and Vandenbosch levels of system scope is based on a hierarchy of systems [61], in a survey conducted with project manager leaders of and subsystems (assembly, system, and array) and pace new product development teams, found that long-cycle into three levels (regular, fast-competitive, and critical-blitz complex projects were run by autocratic leaders, adhered to – projects) [58 60]: a well-defined standardized, serial processing approach. In contrast, short-cycle complex projects were run by project (a) The technological dimension managers who used a more participative management style with many external sources of information. (i) Low-Technology Projects. Projects based on Project complexity has been addressed by researchers existing and well-established technologies from different perspectives and approaches. Early methods (ii) Medium-Technology Projects. Projects based from the general management literature include Declerck mainly on existing technologies but incorporat- and Eymery’s [62] method for analysing ill-structured ing a single new technology or feature problems and Turner and Cochrane’s goals and methods matrix [53]. Part of the literature has focused on uncertainty (iii) High-Technology Projects. Projects that inte- [32, 63]. Williams [64] views the number of elements and grate a collection of new but existing technologies their interrelationships as constituents of structural uncer- (iv) Super High-Technology Projects. Projects based tainty which is proposed as an element of complexity. on non-yet existing technologies in which, Shenhar [65] regards complexity and uncertainty as orthog- although the project goal is clear, no technology onal to each other. Atkinson et al. [66] considers complexity is known to achieve the final product as an element of uncertainty while Geraldi et al. and Müller et al. [17, 67] support uncertainty as an element of complex- (b) The system scope dimension (complexity) ity. Perminova et al. [68] equate complexity to systematic uncertainty. Pich et al. [32] associate categories of uncer- (i) Scope 1: Assembly. A collection of components tainty with variations, foreseen uncertainty, unforeseen in a single unit, performing a well-defined uncertainty, and chaos. Sommer and Loch [12] treat com- limited function plexity and unforeseeable uncertainty as separate constructs. Complexity 7

Size, number of elements Interactions in Structural complex ways, complexity total is more than sum of parts Interdependence of elements Project complexity Uncertainty in goals Structural Uncertainty complexity compounded by uncertainty Uncertainty in methods

Figure 3: William’s model [55].

Table 2: Context’s characteristics, leader’s job, danger signals, and response to danger signals.

Context’s characteristics Leader’s job Danger signals Response to danger signals Repeating patterns Sense, categorize, respond, Complacency and comfort Create communication channels Clear cause-and-effect and delegate Make simple problems Do not assume things are simple Simple relationships Use best practices complex Recognize both the value and Known knowns Communicate in clear and Entrained thinking limitations of best practice Fact-based management direct ways Expert diagnosis required Encourage external and internal Experts overconfident in Cause-and-effect Sense, analyse, and respond stakeholders to challenge their own solutions relationships not Create panels of experts expert opinions Complicated Analysis paralysis immediately apparent Listen to conflicting Use experiments and games Viewpoints of nonexperts Known unknowns objectives to force people to think outside excluded Fact-based management the familiar Probe, sense, and respond Flux and unpredictability Create environments and No right answers Temptation to fall back experiments that allow Allow time for reflection Unknown unknowns into habitual, command- patterns to emerge Use approaches that encourage Complex Many competing ideas and-control model Increase levels of interaction interaction so patterns A need for creative and Desire for accelerated and communication can emerge innovative approaches resolution of problems Use methods that can Pattern-based leadership generate ideas High turbulence Act, sense, and respond No clear cause-and-effect Applying a command- Set up mechanisms to take Look for what works instead relationships and-control approach advantage of the opportunities of seeking right answers Unknowables longer than needed afforded by a chaotic Chaotic Take immediate action to High tension Missed opportunity for environment reestablish order Many decisions to make and innovation Work to shift the context to Provide clear and direct no time to think Chaos unabated chaotic to complex communication Pattern-based leadership

Williams [69] defines two additional types of uncertainty, the organization, and the parties concerned. Laufer et al. aleatoric uncertainty relating to the reliability of calculations [71] explore the evolution of management styles associated and existence uncertainty stemming from lack of knowledge with the organizational complicacy of simple and complex and leading to project complexity. projects. Tatikonda and Rosenthal [31] and Pundir et al. Other approaches used to deal with complexity in project [72] relate technological novelty to technological maturity management contexts include systems theory to help under- of the organization; immaturity leads to task uncertainty. stand how different aspects affect the project as a system The increasingly fast-paced systems of today’s business [8, 51, 55]. Payne [70] takes a perspective which combines and social environment, characterized by discontinuity and difficulty and systems thinking, associating complexity change, force organizations to make decisions and take the with the multiple interfaces between individual projects, corresponding actions based on multiple unknown variables. 8 Complexity

According to Pundir et al. [72], since projects exhibit the [4] K. A. Bubshait and W. J. Selen, Project Characteristics that characteristics of complex systems, the method to manage Influence the Implementation of Project Management Tech- them cannot be predicted in advance, it will emerge from niques: a Survey, Project Management Institute, 1992. the interactions between the project elements and the [5] P. Bennett and S. Cropper, “Uncertainty and conflict: combin- environment. Richardson [35] explores the implications of ing conflict analysis and strategic choice,” Journal of Behav- complexity from the management of organizations and how ioral Decision Making, vol. 3, no. 1, pp. 29–45, 1990. “thinking complexity” may affect the way in which project [6] K. Gidado, Numerical Index of Complexity in Building managers do their jobs. According to the author, if there Construction to Its Effect on Production Time, University of are limits to what we can know about our organization, there Brighton, UK, 1993. are limits to what we can achieve in a predetermined and [7] T. M. Wozniak, Significance vs. Capability:“Fit for Use” Project planned way. H. Singh and A. Singh [73] argue that it is at Controls, AACE International Transactions, 1993. the edge of chaos, where linear systems begin to fail and [8] D. Baccarini, “The concept of project complexity—a review,” nonlinear systems begin to dominate and where project International Journal of Project Management, vol. 14, no. 4, – managers must begin to pay greater attention to the non- pp. 201 204, 1996. linear and subtle influences in their planning and manage- [9] K. Remington, R. Zolin, and R. Turner, “A model of project ment styles. complexity: distinguishing dimensions of complexity from severity,” in Proceedings of the 9th International Research Network of Project Management Conference, Berlin, IRNOP, 7. Conclusions 2009. [10] H. Parsons-Hann and K. Liu, “Measuring requirements When problems fundamentally dynamic are treated stati- complexity to increase the probability of project success,” in cally, delays and cost overruns are common. Traditional Proceedings of the Seventh International Conference on Enter- project management tools and techniques, based on the prise Information Systems - Volume 3: ICEIS, pp. 434–438, assumptions that a set of tasks can be discrete, with well- Miami, USA, 2005. fi de ned information about time, cost, and resources, and [11] L.-A. Vidal and F. Marle, “Understanding project complexity: with extensive preplanning and control, are often found implications on project management,” Kybernetes, vol. 37, inadequate. These traditional approaches that utilize a no. 8, pp. 1094–1110, 2008. static approach provide project managers with unrealistic [12] S. C. Sommer and C. H. Loch, “Selectionism and learning in estimations ignoring multiple feedback processes and non- projects with complexity and unforeseeable uncertainty,” linear relationships of the project. The interrelationships Management Science, vol. 50, no. 10, pp. 1334–1347, 2004. between the components of a project are more complex [13] H. Maylor, R. Vidgen, and S. Carver, “Managerial complexity that is suggested by traditional techniques, which makes in project-based operations: a grounded model and its implica- them inadequate to the challenges of today’s dynamic pro- tions for practice,” Project Management Journal, vol. 39, ject environment. 1_supplement, pp. S15–S26, 2008. The new complex and dynamic environments require [14] T. Cooke-Davies, S. Cicmil, L. Crawford, and K. Richardson, project managers to rethink the traditional definition of a “We’re not in Kansas anymore, Toto: mapping the strange project and the ways to manage it. Project managers must landscape of complexity theory, and its relationship to project ” be able to make decisions in these dynamic yet unstable sys- management, Project Management Journal, vol. 38, no. 2, – tems that are continuously changing and evolving in a ran- pp. 50 61, 2007. “ dom fashion and are hard to predict, very different from [15] D. McLain, Quantifying project characteristics related to uncertainty,” Project Management Journal, vol. 40, no. 4, the linear, predictable systems traditionally studied. To – achieve this objective, more integrated approaches for man- pp. 60 73, 2009. [16] M. Bosch-Rekveldt, Y. Jongkind, H. Mooi, H. Bakker, and aging projects in complex environments and new methods “ of planning, scheduling, executing, and controlling projects A. Verbraeck, Grasping project complexity in large engi- neering projects: the TOE (technical, organizational and must be investigated. environmental) framework,” International Journal of Project Management, vol. 29, no. 6, pp. 728–739, 2011. Conflicts of Interest [17] J. Geraldi, H. Maylor, and T. Williams, “Now, let’s make it really complex (complicated) a systematic review of the The authors declare that they have no conflicts of interest. complexities of projects,” International Journal of Operations & Production Management, vol. 31, no. 9, pp. 966–990, 2011. References [18] L.-A. Vidal, F. Marle, and J.-C. Bocquet, “Measuring project complexity using the analytic hierarchy process,” International [1] P. W. G. Morris, The Management of Projects, Thomas Journal of Project Management, vol. 29, no. 6, pp. 718– Telford, London, 1997. 727, 2011. [2] P. W. G. Morris, “Science, objective knowledge and the theory [19] T. Brady and A. Davies, “Managing structural and dynamic of project management,” Civil Engineering, vol. 150, no. 2, complexity: a tale of two projects,” Project Management pp. 82–90, 2002. Journal, vol. 45, no. 4, pp. 21–38, 2014. [3] J. Bennett and B. Fine, “Measurement of complexity in con- [20] M. Padalkar and S. Gopinath, “Are complexity and struction projects,” in Department of Construction Manage- uncertainty distinct concepts in project management? A ment, University of Reading, 1980. taxonomical examination from literature,” International Complexity 9

Journal of Project Management, vol. 34, no. 4, pp. 688–700, [39] P. De Meyer, F. Maes, and A. Volckaert, “Emissions from 2016. international shipping in the Belgian part of the North Sea ” [21] R. V. Ramasesh and T. R. Browning, “A conceptual framework and the Belgian seaports, Atmospheric Environment, vol. 42, – for tackling knowable unknown unknowns in project manage- no. 1, pp. 196 206, 2008. ment,” Journal of Operations Management, vol. 32, no. 4, [40] A. R. Meyer and L. J. Stockmeyer, “The equivalence problem pp. 190–204, 2014. for regular expressions with squaring requires exponential ” [22] S. M. Qureshi and C. Kang, “Analysing the organizational space, in 13th Annual Symposium on Switching and Autom- – factors of project complexity using structural equation model- ata Theory (swat 1972), pp. 125 129, USA, October 1972. ling,” International Journal of Project Management, vol. 33, [41] L. M. Corbett, J. Brockelsby, and C. Campbell-Hunt, Tackling no. 1, pp. 165–176, 2015. Industrial Complexity, Tackling Industrial Complexity, Insti- [23] F. C. Saunders, A. W. Gale, and A. H. Sherry, “Conceptualising tute for , Cambridge, 2002. uncertainty in safety-critical projects: a practitioner perspec- [42] R. Dewar and J. Hage, “Size, technology, complexity, and tive,” International Journal of Project Management, vol. 33, structural differentiation: toward a theoretical synthesis,” no. 2, pp. 467–478, 2015. Administrative Science Quarterly, vol. 23, no. 1, pp. 111– [24] S. Sinha, A. I. Thomson, and B. Kumar, “A complexity index 136, 1978. for the design process,” WDK Publications, vol. 1, pp. 157– [43] U. Steger, W. Amann, and M. L. Maznevski, Managing Com- 163, 2001. plexity in Global Organizations, John & Sons, 2007. [25] C. Brockmann and G. Girmscheid, Complexity of Megaproj- [44] H. Mintzberg, The Structuring of Organisations: a Synthesis of ects, in: CIB World Building Congress: Construction for the Research, University of Illinois, Champaign, IL, USA, 1979. Development: 14–17 May 2007, Cape Town International [45] R. E. Levitt, J. Thomsen, T. R. Christiansen, J. C. Kunz, Y. Jin, Convention Centre, South Africa, CIB, 2007. and C. Nass, “Simulating project work processes and organiza- tions: toward a micro-contingency theory of organizational [26] H. Mintzberg, Mintzberg on Management: Inside our Strange ” – World of Organizations, Simon and Schuster, 1989. design, Management Science, vol. 45, no. 11, pp. 1479 1495, 1999. [27] A. D. Hall, A Methodology for Systems Engineering, 1962. [46] E. C. Conforto, E. Rebentisch, and D. Amaral, “Learning the “ [28] J. S. Russell, E. J. Jaselskis, and S. P. Lawrence, Continuous art of business improvisation,” MIT Sloan Management ” assessment of project performance, Journal of Construction Review, vol. 57, p. 8, 2016. Engineering and Management., vol. 123, no. 1, pp. 64–71, 1997. [47] J. Priest and J. Sanchez, Product Development and Design for “ ff [29] P. R. Lawrence and J. W. Lorsch, Di erentiation and Manufacturing: a Collaborative Approach to Producibility ” integration in complex organizations, Administrative Science and Reliability, CRC Press, 2012. Quarterly, vol. 12, no. 1, p. 1, 1967. [48] D. Woodward and H. M. College, Understanding Complexity: [30] E. Custovic, “Engineering management: old story, new a Critique and Synthesis, Henley Management College, 1992. demands,” IEEE Engineering Management Review, vol. 43, “ – [49] A. Jaafari, Project management in the age of complexity and no. 2, pp. 21 23, 2015. change,” in Project Management Journal, vol. 34, no. 4, [31] M. V. Tatikonda and S. R. Rosenthal, “Technology novelty, pp. 47–57, 2003. project complexity, and product development project execu- [50] J. Pollack, “The changing paradigms of project management,” tion success: a deeper look at task uncertainty in product inno- International Journal of Project Management, vol. 25, no. 3, ” vation, IEEE Transactions on Engineering Management, pp. 266–274, 2007. vol. 47, no. 1, pp. 74–87, 2000. [51] K. Remington and J. Pollack, Tools for Complex Projects, “ [32] M. T. Pich, C. H. Loch, and A. D. Meyer, On uncertainty, Routledge, 2016. ambiguity, and complexity in project management,” Manage- – [52] J. D. Thompson, Organizations in Action: Social Science Bases ment Science, vol. 48, no. 8, pp. 1008 1023, 2002. of Administration, McGraw-Hill, New York City, NY, USA, [33] S. Ward and C. Chapman, “Transforming project risk 1967. ” management into project uncertainty management, Interna- [53] J. R. Turner and R. A. Cochrane, “Goals-and-methods matrix: – tional Journal of Project Management, vol. 21, no. 2, pp. 97 coping with projects with ill defined goals and/or methods of 105, 2003. achieving them,” International Journal of Project Manage- [34] G. Girmscheid and C. Brockmann, “The inherent complexity ment, vol. 11, no. 2, pp. 93–102, 1993. ” of large scale engineering projects, Project Perspectives, [54] R. D. Stacey, Complexity and Creativity in Organizations, – vol. 29, pp. 22 26, 2008. Berrett-Koehler Publishers, 1996. [35] K. A. Richardson, “Managing complex organizations: com- [55] T. Williams and D. Hillson, “Editorial–PMI Europe 2001,” plexity thinking and the science and art of management,” International Journal of Project Management, vol. 20, no. 3, Emergence: Complexity and Organization, vol. 10, p. 13, 2008. pp. 183-184, 2002. [36] K. I. Gidado, “Project complexity: the focal point of construc- [56] A. Kahane, Solving Tough Problems: an Open Way of Talking, tion production planning,” Construction Management and Listening, and Creating New Realities, Berrett-Koehler Economics, vol. 14, no. 3, pp. 213–225, 1996. Publishers, 2004. [37] P. M. Blau and R. A. Schoenherr, The Structure of Organiza- [57] D. J. Snowden and M. E. Boone, “A leader’s framework tions, Basic Books (AZ), 1971. for decision making,” Harvard Business Review, vol. 85, [38] J. M. Beyer and H. M. Trice, “A reexamination of the relations p. 68, 2007. between size and various components of organizational [58] A. J. Shenhar, Integrating Product Development and Project complexity,” Administrative Science Quarterly, vol. 24, no. 1, Management, in: 28th Annual Symposium of PMI, Project pp. 48–64, 1979. Management Institute, Chicago, IL, 1997. 10 Complexity

[59] A. J. Shenhar, “Strategic project management: the new frame- work,” in PICMET '99: Portland International Conference on Management of Engineering and Technology. Proceedings Vol-1: Book of Summaries (IEEE Cat. No.99CH36310), pp. 382–386, Portland, OR, USA, July 1999. [60] A. J. Shenhar and D. Dvir, “Toward a typological theory of project management,” Research Policy, vol. 25, no. 4, pp. 607–632, 1996. [61] T. B. Clift and M. B. Vandenbosch, “Project complexity and efforts to reduce product development cycle time,” Journal of Business Research, vol. 45, no. 2, pp. 187–198, 1999. [62] P. Declerck and R. P. et Eymery, Le management et l’ana- lyse des projets, Editions Hommes et Techniques, Paris, France, 1976. [63] T. Little, “Context-adaptive agility: managing complexity and uncertainty,” IEEE Software, vol. 22, no. 3, pp. 28–35, 2005. [64] T. M. Williams, “The need for new paradigms for complex projects,” International Journal of Project Management, vol. 17, no. 5, pp. 269–273, 1999. [65] A. J. Shenhar, “One size does not fit all projects: exploring classical contingency domains,” Management Science, vol. 47, no. 3, pp. 394–414, 2001. [66] R. Atkinson, L. Crawford, and S. Ward, “Fundamental uncer- tainties in projects and the scope of project management,” International Journal of Project Management, vol. 24, no. 8, pp. 687–698, 2006. [67] R. Müller, J. G. Geraldi, and J. R. Turner, “Linking complexity and leadership competences of project managers,” in Proceed- ings of IRNOP VIII (International Research Network for Organizing by Projects) Conference, Brighton, UK, CD-ROM. Universal Publishers, 2007. [68] O. Perminova, M. Gustafsson, and K. Wikström, “Defining uncertainty in projects–a new perspective,” International Jour- nal of Project Management, vol. 26, no. 1, pp. 73–79, 2008. [69] T. Williams, “Assessing and moving on from the dominant project management discourse in the light of project over- runs,” IEEE Transactions on Engineering Management, vol. 52, no. 4, pp. 497–508, 2005. [70] J. H. Payne, “Management of multiple simultaneous projects: a state-of-the-art review,” International Journal of Project Management, vol. 13, no. 3, pp. 163–168, 1995. [71] A. Laufer, G. R. Denker, and A. J. Shenhar, “Simultaneous management: the key to excellence in capital projects,” Inter- national Journal of Project Management, vol. 14, no. 4, pp. 189–199, 1996. [72] A. K. Pundir, L. Ganapathy, and N. Sambandam, “Towards a complexity framework for managing projects,” Emergence: Complexity and Organization, vol. 9, p. 17, 2007. [73] H. Singh and A. Singh, “Principles of complexity and in project execution: a new approach to management,” Cost Engineering, vol. 44, p. 23, 2002. Hindawi Complexity Volume 2018, Article ID 3246508, 15 pages https://doi.org/10.1155/2018/3246508

Research Article Comparing Project Complexity across Different Industry Sectors

Marian Bosch-Rekveldt , Hans Bakker, and Marcel Hertogh

Faculty of Civil Engineering and Geosciences, Delft University of Technology, Stevinweg 1, 2628 CN Delft, Netherlands

Correspondence should be addressed to Marian Bosch-Rekveldt; [email protected]

Received 26 October 2017; Accepted 9 May 2018; Published 24 June 2018

Academic Editor: Emma Diaz Ruiz de Navamuel

Copyright © 2018 Marian Bosch-Rekveldt et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Increasing complexity of projects is mentioned as one of the reasons for project failure—still. This paper presents a comparative research to investigate how project complexity was perceived by project practitioners in different industry sectors. Five sectors were included: process industry, construction industry, ICT, high-tech product development, and food processing industry. In total, more than 140 projects were included in the research, hence providing a broad view on Dutch project practice. From the complexity assessments, it is concluded that only one complexity element was present in the top complexity elements of projects across the five sectors: the high project schedule drive. The variety of external stakeholders’ perspectives, a lack of resources and skills availability, and interference with existing site were found in the top lists of three sectors. It was concluded that a framework to grasp project complexity could support the management of complex projects by creating awareness for the (expected) complexities. Further research could be focused on the subjective character of complexity as well as on the application of cross-sector learning, since this research does show similarities between large technical projects in different sectors.

1. Introduction complexities in one sector would compare to project com- plexities in other sectors. Are similar problems being faced? There has been a lot of attention for assessing project com- Whereas literature did report on theoretical insights and plexity in literature in the previous years [1–4]. Several stud- debates in the field of complexity [11, 12], insight in real pro- ies show the potential for, and opportunities of, project ject practice was lacking. Therefore, comparative research complexity [5, 6] in an attempt to exploit certain complexi- was performed to investigate how project complexity was ties and/or explicitly choose to increase complexity, but more perceived by project practitioners in different industry sec- often the potential negative consequences of project com- tors. Also, it was investigated how a framework to grasp pro- plexity are emphasized [7]. Since complexity is potentially ject complexity could support the management of complex hindering project performance, better managing project projects. The following research questions were formulated: complexity is considered an important research topic, still. ff In the tradition of project management, where the dom- (1) How do practitioners in di erent industry sectors inant paradigm is shifting from a one-size-fits-all approach perceive project complexity in large technical in the 1950s towards a more contingent approach, more projects? and more it is realized that projects are unique and should (2) How could a framework to grasp project complexity be treated as such, explicitly taking into account contextual be used to improve project performance? influences [8–10]. The project management approach should be chosen to best accommodate specific project circum- For this study, the earlier published TOE framework to stances and context. grasp project complexity, with TOE referring to technical, Despite such a supposed “fit-for-purpose” approach, it organizational, and external, was selected [1]. This frame- is felt that there could be similarities in projects in different work, based on extensive literature study as well as empirical sectors and these very different projects possibly could data, provided a broad base and enabled a rich view on the learn from each other. This raised the question how project potential aspects causing complexity in very different projects. 2 Complexity

Empirical research was performed in different sectors: 2.1. Further Developments of the TOE Framework. The process industry, construction industry, ICT industry, framework as published earlier [1] was slightly modified as high-tech product development industry, and food process- a result of subsequent research [7]. To modify the frame- ing industry. These sectors have in common that engineer- work, a mixed-methods approach was used, combining qual- ing tasks constitute a main part of all projects, albeit itative and quantitative methods [16, 17]. As a result, some different types of engineering, like software engineering, elements of the framework were adapted or reformulated, industrial engineering, mechanical engineering, and civil some switched category, and the majority of the elements engineering. These sectors also have in common that project were simply confirmed. In the version of the framework as performance is disappointing [13, 14]. Given the relation presented in Figure 1, the T-elements represent the potential between project complexity and project performance [7], complexity causes in the project related to the project scope exploring similarities and differences in the complexities or content of the project. The O-elements represent the faced in projects in these sectors might provide opportuni- potential complexity causes in the project related to the pro- ties for cross-sectoral learning. ject internal organization. The E-elements represent all the Using both case studies and broader surveys, data was potential external complexity causes in the project, related gathered from several respondents in these five industries. to external issues or external organizational complexities. In the different researches, slightly different approaches were The intended use of the TOE framework consists of pro- used because of the different specific scopes, but the com- viding the project team a means to create a complexity foot- mon factor in all researches was the use of the TOE frame- print of the project at hand. By sharing the (expectedly) work. More details on the data gathering are provided in different views in the project team and with other stake- the next section. holders involved, discussion is facilitated and awareness is The relevance of the research is considered from a scien- created for the expected complexities in the project. It is rec- tific point of view and a social point of view. To start with the ognized that this should not be a one-off exercise as the pro- latter: the research is aimed at contributing to the improve- ject complexities are expected to evolve during the different ment of business practice by a better understanding of pro- phases of the project. ject complexity in different industries, thereby improving project performance. Given the high failure rate of projects, 2.2. Data Gathering and Methods per Substudy. This section social benefits seem evident. From the scientific point of view, describes the data gathering in the different industry sectors this research embraces some of the project management including an elaboration on the specific methods used per “schools of thought” as distinguished in literature [15]: partic- substudy. Table 1 shows when the data was gathered and ularly the factor school and the contingency school, hence summarizes the number of respondents in each of the stud- illustrating a pluralistic approach in project management ies. Given the fact that in four studies, the data was gathered research. Also, it contributes to improving the understanding in 2011/2012, one could argue that the value of this study is of the notion of complexity in projects. limited. The data for the fifth sector was gathered more This paper is structured as follows. In Section 2, the recently. How time could have influenced the overall picture applied methods are discussed, starting with presenting will be discussed after providing the results. the further developed TOE framework to grasp project Large differences are observed in Table 1 regarding the complexity which is used as the main source supporting number of projects/respondents involved in the different the data gathering. Next, the set-up of the data gathering studies. The nature of the research—evaluating complexity in the five industry sectors is described. Section 3 presents in different industries using an existing framework—domi- the results of the five studies with regard to the complexity nantly asks for a quantitative approach [18]. We are inter- assessments. This includes a cross-sector comparison of the ested in general information on relative large numbers of results, highlighting similarities and differences between the projects, as opposed to detailed in-depth information on different sectors. In Section 4, the potential value of the small numbers of projects, and therefore, a survey is a suit- TOE framework for practice is discussed. This leads to able research tool [19, 20]. Indeed a web-based survey was the implications for managing complex projects and learn- applied in studies A and B. However, in the researches car- ing across projects as discussed in Section 5. This paper is ried out in IT, high-tech industry, and food processing indus- concluded with the conclusions and recommendations in try, only part of the research was related to measuring project Section 6. complexity, and those researches were organized around in- depth case studies, hence explaining the relative limited amount of data from sectors C, D, and E. Details per 2. Methods research, in terms of methods used and data gathered, are provided below. This paper has the character of a meta-study; the results of five separate researches, which have in common the appli- 2.2.1. Sector A: Process Industry. The study described in this cation of the TOE framework to grasp project complexity, section draws heavily upon Chapter 9 of a dissertation [7]. are compared thoroughly. Bringing together these results In the survey, respondents were asked to indicate their com- is expected to contribute to the understanding of practi- pany’s role (owner/contractor/other) and their experience tioners’ perspectives on elements causing complexity in level (as a project manager and working for their company). their projects. Next, the respondents had to score each of the 47 elements of Complexity 3

Technical complexity Organizational complexity (17 elements) (17 elements) High number of project goals High project schedule drive Nonalignment of project goals Lack of resource & skills availability Unclarity of project goals Lack of exprience with parties involved Uncertainties in scope Lack of HSSE awareness Strict quality requirements Interfaces between diferent disciplines Project duration Number of fnancial sources Size in CAPEX Number of contracts Number of locations Type of contract Newness of technology (worldwide) Number of diferent nationalities Lack of experience with technology Number of diferent languages High number of tasks Presence of JV partner High variety of tasks Involvement of diferent time zones Dependencies between tasks Size of project team Uncertainty in methods Incompatibility between diferent pm method/tools Involvement of diferent technical disciplines Lack of trust in project team Conficting norms and standards Lack of trust in contractor Technical risks Organizational risks

External risks Number of external stakeholders Variety of external stakeholders’ perspectives Dependencies on external stakeholders Political infuence Lack of company internal support Required local content Interference with exiting site Remoteness of location Lack of experience in the country Company internal strategic pressure Instability of project environment Level of competition External complexity (13 elements)

Figure 1: TOE model as used in this study [7].

Table 1: Summary of data gathered.

Study ID Industry Data gathered in Number of projects Number of respondents A Process 2011 64 64 B Construction and infra 2012 35 164 C IT 2011 8 8 D High-tech product development 2011 16 16 E Food 2015 21 25 the TOE framework on their potential contribution to the project complexity. Also, questions were posted regarding complexity of a project (not–little–some–substantial–very treating project complexity, but these are outside the scope much). They were not explicitly asked for one specific pro- of the current paper. Finally, the respondents were asked ject; they could answer the question for any project in mind. for their opinion about the potential use of the TOE com- Subsequently, the elements that were scored “substantial” plexity framework by means of open questions: or “very much” by the respondent were listed on the screen and the respondent had to select those three elements that (1) How would you apply the TOE project complexity in their opinion contribute most to a project’s complexity. framework in your daily practice? Next, the elements that were scored “none” or “little” by the respondent were listed on the screen, and the respondent (2) What would be the added value of using the TOE had to select which three of these would contribute least to framework in your projects, if any? 4 Complexity

For the latter part of the survey, open questions were used Table 2: Overview of responses in study A. as they do not restrict the respondent in their answers. The mixture of open questions and closed questions in one survey Group Number of requests Respondents Response rate perfectly fits a mixed-methods approach [21]. Total 111 64 58% In the development and design of the survey, several Contractor 35 24 69% measures were taken to ensure the internal validity. Before Owner 76 40 53% the survey was published on the internet, several experts were asked to test concept versions of the survey. Based on their feedback, questions were reformulated and terminology was or more responses were obtained providing 164 completed clarified. The external validity of this study was positively surveys in total (response rate of 36% overall). influenced by the fact that the survey was distributed Again, the web-based application NetQ was used (pro- amongst four totally different companies, all actively gram name in the meantime changed to Collector). Data involved in the NAP network, the competence network of was stored in SPSS compatible format. All data was gathered the Dutch process industry. between July 1st 2012 and September 21st 2012. Four companies, key players of the NAP network, were selected for participation: two owner companies and two 2.2.3. Sector C: ICT. A MSc study was undertaken in 2011 to contractor companies. All four approached companies were investigate which (combination of) factors determine the willing to participate, and the department heads distributed complexity of IT projects and how to manage these complex- the link to the web-based survey amongst their project man- ities [22]. Case studies were done in which eight IT projects agers. In total, 111 survey requests were sent and the survey in the financial services were analyzed. Cases were selected was started by 68 respondents. Of these respondents, 64 to represent a broad portfolio of projects in the IT sector indeed completed and submitted the completed results, (infrastructure, application, middleware, and other) and with hence obtaining a high overall response rate of 58%. For different performance scores. The case analysis was based on the contractor group (smaller in size), the response rate was semistructured interviews, held with the project managers of a little higher than for the owner group. An overview of the the IT service provider, and detailed project documentation. response rate, overall as well as per group (owner/contrac- In the interviews, project managers were asked about their tor), is given in Table 2. projects: their challenges in projects, their view on the com- While completing the survey, the progress was saved on plexity of projects, and how project complexity was actually the participant’s computer. Measures were taken to prevent managed. To identify their view on what factors contributed double submissions from one participant. Apart from their to the complexity of their projects, the TOE framework as typical company role and their work experience, no specific provided in Figure 1 was used. information about the respondents was included in the data In the interviews, respondents were asked to indicate to analysis. The survey was developed and executed in the what extent the different elements of the TOE framework web-based application NetQ. The majority of the respon- contributed to the complexity of the project (not (1)–little– dents needed 30 minutes to complete the survey. Data was some–substantial–very much (5)). Data was stored in written stored in SPSS compatible format. All data was gathered interview transcripts and Excel files for storage of the TOE between February 25th 2011 and March 21st 2011. scores. Data was gathered in the Summer of 2011. 2.2.2. Sector B: Construction and Infra. The study described in this section was performed in collaboration with KING 2.2.4. Sector D: High-Tech Product Development. Another (a network consisting of project management teams of MSc study was undertaken in 2011 to investigate the benefits large-scale infrastructural projects) and the RijksProjectAca- of applying the TOE framework in a company developing demie (an academy for project managers of public construc- high-tech products [23, 24]. Case studies were performed in tion projects). which 16 high-tech projects were investigated. The amount The survey contained similar questions as the survey of 16 cases was considered to provide a good balance between described in Section 2.2.1, without the part on application obtaining a broad overview of the projects and in-depth data of the TOE framework and dealing with project complexi- gathering. Cases were selected from different business lines of ties. Both the survey and the TOE framework had to be the company involved. Selection criteria also included the translated to Dutch because of the dominant use of this lan- project nature (product development or process develop- guage in the Dutch construction sector. Again, an internet ment oriented) and the product design characteristics (new/ survey was used. old technology). Project managers of the 16 projects were After consultation with several experts in the field of con- interviewed, with the assumption that the project manager struction projects, some elements were added to the TOE has the most extensive knowledge about the project. framework of Figure 1 while translating it to Dutch in order In the interviews, project managers were asked general to increase its applicability to construction industry (see questions about the project, about the project’s complexity, Table 3). As will be shown in Section 3, only 2 of them actu- and about its management. Respondents were asked to iden- ally proved relevant for describing project complexity. tify and scale complexities from the TOE framework in rela- In total, 454 project practitioners from 35 projects were tion to their projects (not applicable–very much applicable). invited to participate in the research. For all projects, one Data was gathered in the Summer of 2011. Complexity 5

Table 3: Comparison TOE elements.

A: Process industry B: Construction High number of project goals Aantal projectdoelstellingen Nonalignment of project goals Incongruentie van projectdoelstellingen Unclarity of project goals Onduidelijkheid over projectdoelstellingen Uncertainties in scope Onzekerheid over de scope Strict quality requirements Niveau van kwaliteitseisen Project duration Projectduur Size in CAPEX Investeringskosten Number of locations Aantal locaties T Newness of technology (worldwide) Gebruik nieuwe technologie Lack of experience with technology Ervaring met toegepaste technieken High number of tasks Aantal deelprojecten High variety of tasks Diversiteit van deelprojecten Dependencies between tasks Afhankelijkheid tussen deelprojecten Uncertainty in methods Onzekerheid over technische methoden Involvement of different technical disciplines Diversiteit van technische disciplines Conflicting norms and standards - Technical risks Technische risico’s High project schedule drive Druk op de tijdsplanning Lack of resource and skills availability Beschikbaarheid van capaciteit en vaardigheden Beschikbaarheid van middelen Discontinuïteit in bemensing Lack of experience with parties involved Ervaring met projectpartijen Lack of HSSE awareness VGM-bewustzijn Interfaces between different disciplines Interfaces tussen verschillende disciplines Number of financial sources Aantal financieringsbronnen Number of contracts Aantal uitvoeringscontracten en interfaces daartussen Type of contract Contractvorm Kwaliteit van het hoofdcontract O Number of different nationalities Aantal verschillende nationaliteiten Number of different languages Aantal verschillende talen Presence of JV partner Samenwerking tussen aannemers Aantal opdrachtgevers Involvement of different time zones Werktijden Bereikbaarheid en bouwlogistiek Size of project team Aantal projectmedewerkers Incompatibility different PM methods/tools Aansluiting tussen gebruikte PM tools & technieken Lack of trust in project team Vertrouwen tussen projectteam en opdrachtgever Lack of trust in contractor Vertrouwen tussen projectteam en aannemer(s) Cultuurverschillen Organizational risks Organisatorische risico’s External risks Externe risico’s Number of external stakeholders Aantal externe stakeholders Variety of external stakeholders’ perspectives Diversiteit in belangen van externe stakeholders Dependencies on external stakeholders Afhankelijkheid van externe stakeholders E Political influence Politieke invloed Lack of company internal support Management support vanuit de eigen organisatie Required local content - Interference with existing site Interfaces met andere projecten 6 Complexity

Table 3: Continued.

A: Process industry B: Construction Remoteness of location Aard van de omgeving Lack of experience in the country - Company internal strategic pressure Invloed van stakeholders van binnen de organisatie Instability of project environment Discontinuïteit bemensing stakeholders Economische omstandigheden Level of competition Marktomstandigheden BLVC bewustzijn Ervaring van omgevingspartijen met grote projecten Media invloed Sociale impact Conflicterende wet- en regelgeving Planologisch / juridische procedures

Less than 5 years

Between 5 and 10 years

Between 10 and 15 years

Between 15 and 20 years

Between 20 and 25 years

More than 25 years

024681012141618 Number of respondents

Figure 2: Project management experience of respondents (N =64).

2.2.5. Sector E: Food Processing Industry. In the course of 3.1. Sector A: Process Industry. Respondents were asked for 2015, a study was performed in the food processing industry their project management experience. As can be seen in in one specific company. In total, 21 projects were selected to Figure 2, the vast majority of the survey respondents did have represent all four business sectors for the project manage- considerable project management experience (only 10 out of ment community in company E. In total, 25 project man- 64 had less than 5 years of project management experience), agers participated in the research. thereby increasing the value of this study. Semistructured interviews were held including questions The respondents indicated to what extent the elements of about the application of project management processes and the TOE framework (potentially) contributed to the project’s the importance of project success criteria. As part of the complexity (table with results added in Figure 3). Amongst interview session, the interviewees completed a written the highest-scoring elements were the elements related to TOE complexity assessment for their project (scoring the ele- project goals and scope (unclarity of goals, nonalignment of ments on a 1 (not contributing) to a 5 (most contributing). goals, and uncertainties in scope), boundary conditions for Data was stored on answering sheets. Complexity scores the project (lack of resource and skills scarcity), and softer fac- were analyzed using Excel. tors (a lack of trust in the project team, a lack of trust in the contractor). The vast majority of the elements were scored 3. Results: Complexity Assessments between “some” and “substantial,” indicating the perceived relevance of these elements in their contribution to project Although the various researches provide a rich set of empir- complexity. Subsequently, respondents indicated their top- ical data, this paper will dominantly focus on the results 3s of most contributing elements (see Table 4). related to the complexity assessments. For each sector, the Comparing the results in Figure 3 and Table 4, all following findings are discussed: background of the respon- highest-scoring elements of Figure 3 appear in the top-3, dents, highest scoring complexity elements in T-, O-, and except for the element lack of trust in the contractor. E-categories, and overall impression of complexity scores. Table 4 shows that the top-3 of most often mentioned Complexity 7

Cumulative score per element (N = 64) (Not) (Little) (Some) (Substantial) (Very much) 64 128 192 256 320 Number of project goals Nonalignment of project goals Unclarity of project goals Uncertainties in scope Strict quality requirements Project duration Size in CAPEX Number of locatios Newness of technology Lack of experience with technology Number of tasks Variety of tasks Dependencies between tasks Uncertainty in methods Involvement of different technical disciplines Conflicting norms and standards Technical risks

High project schedule drive Lack of resource and skill availability Lack of experience with parties involved Lack of HSSE awareness Interfaces between different disciplines Number of financial sources Number of contracts Number of different nationalities Number of different languages Presence of a JV partner Involvement of different time zones Size of project team Incompatibility between different PM... Lack of trust in project team Lack of trust in contractor Organizational risks

Number of external stakeholder Variety of stakeholder perspectives Dependencies on external stakeholders Political influence Lack company internal support Required local content (forced cooperation with... Interference with existing site Weather conditions Remoteness of location Lack of experience in the country Company internal strategic pressure Instability of project environment Level of competition Risks from environment

Figure 3: Results sector A ([7], Figure 9.3).

T-elements includes elements related to project goals and literature highlights, improved project portfolio management project scope and that these elements were mentioned by might be needed to optimally distribute the available more than half of the respondents. There seems considerable resources [26]. The availability of resources and skills is out- agreement amongst the respondents about the importance of side the responsibility of the project manager [27]. From the these elements, which also confirms research in the construc- O-complexity elements, also a lack of trust in the project team tion industry [25]. The most often mentioned O-element is a was mentioned often (by 50% of the respondents), indicating lack of resource and skills availability, mentioned by about the importance of obtaining trust in a project team, which 70% of the respondents. This seems a trivial element contrib- also is stressed in literature [28]. For the E-elements, the ele- uting to complexity of a project: if resources are lacking, real- ment variety of stakeholders’ perspectives was mentioned izing project objectives becomes troublesome. Also, it might most often, by almost 60% of the respondents. This is the only highlight a serious problem that occurs in current project E-element for which such high agreement was found under practice which has to deal with constrained resources. As the respondents; other elements scored lower than 50%. 8 Complexity

Table 4: Top-3 of most contributing complexity elements—sector A. accessibility are mentioned, and in the E-category the ele- ments interference with existing site/projects and remoteness Most contributing to project Percentage of respondents of location are mentioned. In the O-category, a high sched- N =64 complexity ( ) ule drive is mentioned by more than a third of the respon- Technical dents, which is in combination with a lack of resources Nonalignment of project goals 58% and skills availability rather problematic. In the E-cate- Uncertainties in scope 56% gory, the influence of external stakeholders seems to cause Unclarity of project goals 55% complexity, given the high scores of variety of external stakeholders’ perspectives, political influence, and number Organizational of external stakeholders. Lack of resource and skills availability 70% Lack of trust in project team 50% 3.3. Sector C: ICT. The respondents’ experience in project man- High schedule drive 38% agement in research sector C is given in Figure 5. Although External the number of respondents is limited (N =8), they have con- fi Variety of stakeholders’ perspectives 58% siderable experience in the eld of project management. The respondents were asked to score the complexity ele- Lack of company internal support 44% ments, but in contrast to the investigations in sectors A and Interference with existing site 28% B, they were not asked to identify their top-3s. Therefore, Lack of experience in the country 28% Table 6 only shows the highest-scoring elements in view of the respondents. The ICT respondents score the T-elements generally low: 3.2. Sector B: Construction and Infra. Generally, the respon- apparently, they do not expect particular complexities from dents in research sector B had less experience in the role the technical area, apart from dependencies between tasks. they were playing in the project management field. However, They do experience complexity as a result of the O- the total number of respondents was considerably higher elements high project schedule drive and interfaces between (see Figure 4). different disciplines. They, however, expect most complexities The respondents indicated to what extent the TOE ele- from E-elements, particularly related to stakeholder involve- ments (potentially) contributed to their project’s complexity. ment, both internal and external to the company. Amongst the highest-scoring elements are no (!) T-elements, one O-element (schedule drive), and several E-elements: 3.4. Sector D: High-Tech Product Development. The work stakeholder-related (number external stakeholders, variety experience of the respondents in sector D ranged between 7 in stakeholder’s perspectives) or physical environment related and 28 years, with an average working experience of 18 years. (remoteness of environment). Apparently, technology is not On top of this relatively long work experience, the majority of posing the major challenges in the projects under con- the respondents was PMP certified (11 respondents) and two sideration. The general average scores tended to “some” con- were busy attaining a certification. tribution to project complexity (and not the higher Similar to the research in sector C, the respondents were substantial or very much). asked to score the complexity elements on a scale from 1 to 5 From the elements the respondents rated “substantial” or (not contributing to very much contributing). Results are “very much,” they selected their top-3 elements (see Table 5). given in Table 7. Note that a considerable number of the respondents did not The respondents score the T-elements relatively high in score project complexity high enough to answer this question their contribution to project complexity. Amongst the (20 for the T-elements, 18 for the O-elements, and 18 for the highest-scoring elements, there is only one E-element (level E-elements). of competition) and one O-element (high project schedule Amongst the top-3 elements as presented in Table 5, drive) whereas there are five T-elements. These high- there is only one element that was specifically added to the scoring T-elements seem to reflect the high-tech products TOE framework to better capture construction complexities that are being created in their major projects (size in CAPEX), (building and accessibility). Hence, the “original” requiring the involvement of different technical disciplines, TOE framework (Figure 1) seems to cover the most contrib- organized with a high number and variety of tasks and uting complexity elements reasonably well. In other words, involving technical risks. The O-element and the E-element aspects most contributing to project complexity seem to have are probably related as well: because of high competition in a rather generic character. the business under investigation, the projects are primarily We could summarize the results in Table 5 in complexi- (and strongly) schedule-driven: if you are not the first, you ties related to interfaces, complexities related to planning will lose the market. and resourcing, and complexities related to content and stakeholders. Particularly, interfaces appear often in the 3.5. Sector E: Food Processing Industry. The work experience top-3 elements: in the T-category, the elements dependencies of the respondents in the food processing industry ranged between tasks and involvement of different technical disci- from 5 to 34 years with an average of 21 years. The project plines are mentioned, in the O-category the elements inter- management experience of the participants was between 2 faces between different disciplines and building logistics and and 30 years averaging at 13 years (see Figure 6). Complexity 9

No experience

Less than 5 years

Between 5 and 10 years

Between 10 and 15 years

Between 15 and 20 years

Between 20 and 25 years

More than 25 years

0 5 10 15 20 25 30 35 40 45 50 55 Number of respondents

Figure 4: Experience of respondents (N = 164).

fi Table — sectors/studies: interference with existing site. The nal two 5: Top-3 of most contributing complexity elements sector B. elements out of the top seven are closely related company Percentage of internal strategic pressure and internal stakeholders. Most contributing to project complexity respondents 3.6. Comparing the Complexities in the Five Industries. Technical (N = 144) Table 9 presents a summarizing overview of the results of Dependencies between tasks 38% the preceding sections. All high-scoring elements are dis- Uncertainties in scope 28% played in the rows of this table; an “X” means that this com- Project duration 26% plexity element belonged to the highest-scoring elements in Involvement of different technical that specific sector. Given the different character of the data- 23% disciplines sets (in terms of number and character of the data points Organizational (N = 146) gathered), it was not possible to enhance this comparison High project schedule drive 36% with more elaborated statistical analysis. However, this com- parison does provide insight in high-scoring elements across Interfaces between different disciplines 28% the five sectors. Lack of resource and skills availability 23% For the highest-scoring T-elements as listed in Table 9, Building logistics and accessibility 22% only four of the eleven complexity elements were concluded External (N = 146) from two industry sectors. These are uncertainties in scope, Remoteness of location 50% dependencies between tasks, involvement of different technical Variety of external stakeholders’ disciplines, and project duration. Complexities in the high- 36% perspectives tech industry seem to be driven more by content-related ele- Political influence 23% ments than complexities in the other industries. The ICT sec- tor seems not to bother about technical complexities, Interference with existing site/projects 23% according to the current study. In the process industry, the Number of external stakeholders 21% complexity elements related to the project goals are empha- sized: unclarity of goals and unalignment of goals are amongst Similar to the research in the sectors C and D, the respon- the highest-scoring T-elements. The element project duration dents were asked to score each of the complexity elements on seems a complexity element specifically related to the con- a scale from 1 to 5 (not contributing to very much contribut- struction industry where projects with a 20-year or 30-year ing). The results for the seven (7) highest-ranking elements contract period become more and more common and a com- are presented in Table 8. plexity element related to the food processing industry. The The respondents scored the T-elements relatively mod- strict quality requirements that were found as a high- erate compared to the O- and the E-elements where the scoring complexity element in the food processing industry highest scores were found (all top 5 scores). Amongst the are clearly related to their specific context (food safety). top seven elements, there are only two T elements, strict Based on this data, the T-complexities seem rather indus- quality requirements and project duration. The first of these try-specific. is clearly sector-specific and related to food safety: the strict More alignment between industries is found for the O- food quality requirements that this industry has to deal with. elements, contributing to project complexity. From the six The top two scoring elements are from the organizational highest-scoring O-elements, three are shared by two or more category and are similar to those found in other industries: sectors. In all five researches, the complexity element high lack of resources and skills availability and high project sched- project schedule drive is amongst the highest-scoring com- ule drive. The third highest-ranking element is from the plexity elements, hence stressing the importance of realisti- external category and is also recognizable from the earlier cally estimating project schedules. The element lack of 10 Complexity

Less than 5 years

Between 5 and 10 years

Between 10 and 15 years

Between 15 and 20 years

Between 20 and 25 years

More than 25 years

012345 Number of respondents

Figure 5: Project management experience of respondents (N =8).

Table 6: Highest-scoring complexity elements—sector C.

Elements contributing to project complexity SUM Case 1 Case 2 Case 3 Case 4 Case 5 Case 6 Case 7 Case 8 Technical Dependencies between tasks 28 5 4 443134 Organizational High project schedule drive 28 5 4 252244 Interfaces between different disciplines 25 2 4 331444 External Variety of external stakeholders’ perspectives 24 4 4 144241 Dependencies on external stakeholders 23 5 4 152141 Political influence 21 4 4 155200 Company internal strategic pressure 25 5 4 154321

Table 7: Highest scoring complexity elements—sector D. related to interfaces is easily understood for construction and ICT: projects in both sectors have close connections Elements contributing to Average Standard N to other sectors and/or rely on these interfaces. To give project complexity score deviation some examples: the construction of a new station is next Technical to the construction itself closely related to the related infra- Involvement of different structure and future users (exploitation of the station), and 4.06 0.68 16 technical disciplines the development of a computer program in ICT itself is Technical risks 4.00 1.15 16 more an enabler for obtaining a wider goal, often overarch- Size in CAPEX 3.56 1.26 16 ing disciplines. Number of tasks 3.56 1.21 16 Almost similar to the T-elements (eleven), ten E- elements made it into the list of highest-scoring complexity Variety of tasks 3.63 0.96 16 elements in Table 9, four of which are shared by at least Organizational two industries. The element variety of external stakeholders’ High project schedule 4.44 0.63 16 perspectives is amongst the highest-scoring complexity ele- drive ments in the process, construction, and ICT industries. Only External in the high-tech, new product development industry, which Level of competition 4.07 0.59 15 is dominantly internally focused for the development of new products, this complexity element is not scored high. That a high number of parties involved is causing complexi- resource and skills availability is concluded from three ties is also visible from some other elements in Table 9: num- researches: process, construction, and food processing. ber of external stakeholders involved (construction industry) Scarcity in resources is a known problem in projects, and it and dependencies on external stakeholders (ICT industry). A is surprising that it does not appear in all five sectors under specific party causing complexities seem the politicians: the consideration. The element interfaces between different disci- element political influence appears amongst the highest- plines is shared by construction and ICT. The complexity scoring elements in both construction and ICT. The projects Complexity 11

Less than 5 years

Between 5 and 10 years

Between 10 and 15 years

Between 15 and 20 years

Between 20 and 25 years

More than 25 years

012345678 Number of respondents

Figure 6: Project management experience of respondents (N =25).

fi Table 8: Highest-scoring complexity elements—sector E. (i) A few elements appear in three or more of the ve industries: high project schedule drive, lack of resource Elements contributing to Average Standard and skills availability, variety of external stakeholders’ N complexity score deviation perspectives, and interference with existing site. These Technical elements demonstrate their broad applicability and Project duration 3.40 1.15 25 importance in determining project complexity. Strict quality requirements 3.48 0.87 25 (ii) Other elements appear in one sector and seem to Organizational have a more sector-specific character: for example, High project schedule drive 4.00 0.83 25 remoteness of location for the construction industry, technical risk for high-end product development, Lack of resources and skills 4.04 0.81 25 availability unclarity and unalignment of project goals for the process industry, and strict quality requirements for Internal stakeholders 3.58 1.13 25 the food processing industry. External Interference with existing site 3.88 0.95 25 4. Results: The Value of Using a Company internal strategic 3.60 1.03 25 pressure Complexity Framework A necessary condition to create support for implementation of a new complexity framework is its value for the people under investigation in these sectors both did have a (strong) who work with it. Once potential users see the value of a link with the public parties, in contrast to the high-tech prod- new tool, it is more likely they will explore it and actually uct development projects or the projects in the process indus- start using it [29, 30]. In the sector A research (process indus- try. The importance of the element interference with the try), respondents were specifically asked for their view on the existing site was found in both the construction industry, added value of the TOE framework. Results, clustered into the process industry, and the food processing industry. In overarching themes, are provided in Table 10. the process industry, this element relates, for example, to Only four of the respondents did not see any added the difference between Greenfield or Brownfield projects. A value. Some others did not know or did not understand Greenfield project implies that all is to be built from scratch, the question, but the majority of the respondents did see a clear whereas Brownfield projects have a direct link to existing added value, for example, in the areas of structuring, decision- facilities with ongoing operations (and related complexities). making, and stakeholder alignment and management. Similarly, in construction there is the difference between con- The TOE framework could help in achieving better structing a brand new road in a rural area or reconstructing a alignment in the team and better communication with train station where the train service cannot be interrupted. stakeholders. Also, the structured approach of the TOE Complexity caused by strategic issues like a lack of company framework adds value to the project, in view of the internal support or company internal strategic pressure was respondents. Based on the outcome of a TOE complexity seen in the process industry and the ICT sector, respectively. assessment, activities could be selected to manage those Only the most innovative sector (high-tech product develop- complexities that gained the highest priority. One of the ment) experienced the level of competition as an element con- actions could be to staff the project accordingly, similar tributing to project complexity. to earlier publications on matching the project manager’s Summarizing the findings as presented in Table 9: competences to the particular project complexity [31, 32]. 12 Complexity

Table 9: Elements of the TOE framework most contributing to complexity.

Sector A Sector B Sector C Sector D Sector E Those elements most contributing to Process Construction ICT High-tech Food complexity industry industry industry industry industry T-elements Uncertainties in scope X X Dependencies between tasks X X Involvement of different technical disciplines X X Unclarity of project goals X Nonalignment of project goals X Project duration X X Technical risks X Size in CAPEX X Number of tasks X Variety of tasks X Strict quality requirements X O-elements High project schedule drive X X X X X Lack of resource and skills availability X X X Interfaces between different disciplines X X Lack of trust in project team X Building logistics and accessibility X Internal stakeholders X E-elements Variety of external stakeholders’ perspectives X X X Interference with existing site/projects X X X Political influence X X Number of external stakeholders X Remoteness of location X Lack of company internal support X Lack of experience in the country X Dependencies on external stakeholders X Company internal strategic pressure X X Level of competition X

Table 10: Added value of the TOE complexity framework, adapted 5. Discussion from [7]. In this section, the findings and implications of this research Answers related to: Sector A responses (N =64) are discussed. Also, the limitations of the current research Better alignment 8 are described. Structured approach 12 Communication with stakeholders 4 5.1. Differences and Similarities: Opportunities for Learning Support decision-making 8 Across Sectors? Projects are unique, and at the same time Identify priorities 7 they are not unique in all their aspects. This research shows Integrate 3 there are similarities in complexities across projects in dif- Awareness 3 ferent sectors ((food) process, ICT, high-tech products, No added value 4 and construction industry) in terms of high-scoring com- plexity elements, but also differences are found. Both imply Not applicable 3 opportunities for cross-sector learning and show the appli- ’ Don t know 12 cability of the TOE framework in other industries than Question not well understood 4 the process industry for which the framework initially was developed [7]. Complexity 13

Cross-sector learning does not imply that we are in favor elements—overall—were included in each of the five of a one-size-fits-all approach: rather, we propose a fit-for- researches. Future research could be more strict in type and purpose management approach in which we carefully select amount of data to be gathered in different sectors in order the appropriate management tools, techniques, and processes to allow for more enhanced statistical analysis of the compar- based on the specific characteristics of the project. The com- ative data. plexity of the project could be the characteristic to base the The research has a qualitative character; we did not approach upon. Obviously, it is important then to under- attempt to objectively classify projects in terms of complex- stand the rationale of complexity in projects. ity. Although further research could investigate how an Referring back to project management literature, we dis- objective measure of project complexity would look like, we tinguish two main streams. The first stream considers com- do not think this is a relevant way to go given the inherent plexity as a subjective phenomenon; the second stream subjective character of complexity. considers complexity as a descriptive property of a system The data for the current analysis was mainly gathered [12]. The TOE complexity framework adopts the first stream some years ago, which could be considered as a limitation. by emphasizing the subjective nature of complexity and its Although the current analysis cannot guarantee that no other dynamic character. This dynamism is difficult to grasp objec- project complexity elements would be more prominent now- tively, although the complexity theory does describe the adays, still the research indicates the parallels as discussed behavior of complex systems over time [33]. More important earlier, opening up the potential for cross-sectorial learning. is the question how we can learn in and from projects that are Also, based on additional longitudinal research (2012–2015, considered complex and how to manage project complexity. unpublished), we have indications that the development of In the infrastructure sector, complex projects seem to ask complexity “ingredients” is rather stable since then, with an for a different management approach, characterized by flexi- exception for the growing importance of political influence. bility rather than the more traditional predict and control Another limitation of this research is found in the fact [34, 35]. A more flexible project management approach that only Dutch projects, companies, and organizations are includes the facilitation of collaboration, explorative learn- involved, thereby specifically focusing on Dutch project prac- ing, and adaptation and was shown to positively influence tice. It would be interesting to broaden the research to inter- project performance of complex projects [36]. Instead of national contexts, although it is the question to what extent focusing on reducing project complexity, “playing with com- country culture dominates project management culture [38]. plexity” could be considered [6]. This general idea of focusing on embracing complexity 6. Conclusions and Recommendations rather than reducing complexity could be applied in cross- sectoral contexts. And maybe this “embracement” is a lesson This research investigated perceptions of complexity in large to be drawn, given the parallels of the complexity elements technical projects amongst five industry sectors in the Neth- that were present in at least three of the sectors considered erlands. It is concluded that some complexity elements in this study (high schedule drive, lack of resources and skills appear relevant in three or more of the industries investi- availability, variety of stakeholders’ perspectives, and interfer- gated: high project schedule drive, lack of resources and skills ence with existing site). For the element high schedule drive, availability, variety of external stakeholders’ perspectives, the idea of complexity reduction seems attractive. High time and interference with existing site. These elements demon- pressure is a universal phenomenon in nowadays’ society, strate their broad applicability and importance in determin- and rather than trying to comply to unrealistic deadlines, ing project complexity. Other complexity elements appear better preparation of the project should be allowed for, hence in one sector and seem to have a more sector-specific charac- avoiding this high schedule drive. Similarly, it is advisable ter: for example, remoteness of location for the construction (and difficult) to avoid a lack of resources and skills avail- industry, technical risk for high-end product development, ability by timely taking appropriate actions (like realistic unclarity and unalignment of project goals for the process resource planning and training of staff). For the element industry, and strict quality requirements for the food process- variety of stakeholders’ perspectives, active involvement of ing industry. Both similarities and differences offer opportu- stakeholders is suggested since broader views might bring nities for (further) cross-sectoral learning, which could be the surprising insights, hence adding value to the project and focus of subsequent research. improving project performance [37]. Also, the element It is concluded that applying a framework to grasp pro- interference with existing site could benefit from an embrac- ject complexity could contribute to creating awareness for ing complexity approach by active involvement of stake- the actual (expected) complexities in the project. This aware- holders, assuming that the interference as such is part of ness could help in improving communication between the project anyway. relevant stakeholders in order to achieve better alignment. Also, the TOE framework provides a structured approach 5.2. Limitations of the Research. This (meta-) research to identifying the complexities in today’s projects, without brings together five separate researches all focused on an claiming an objective view on complexity. evaluation of the TOE complexity framework. Ideally, the Further research is foreseen in the area of measuring the set-up of the five researches would perfectly replicate each subjectivity of complexity and in the area of cross-sectoral other; however, this was not the case. Still, the comparison is learning. Why can the process industry implement success- considered meaningful since the highest-scoring complexity fully all types of ICT systems in their plants while tunnel 14 Complexity technical installations in large infrastructure projects still are [12] F. Marle and L.-A. Vidal, Managing Complex, High Risk challenging, to say the least? Projects, Springer, London, UK, 2016. [13] K. E. Emam and A. G. Koru, “A replicated survey of IT ” Disclosure software project failures, IEEE Software, vol. 25, no. 5, pp. 84–90, 2008. A previous version of this paper was orally presented at the [14] M. G. Mc Kenna, H. Wilczynski, and D. van der Schee, ÖTMC Conference 2017 (Porec, Croatia). “Capital project execution in the oil and gas industry,” 2006, https://www.boozallen.com/media/file/Capital_Project_ Execution.pdf. Conflicts of Interest [15] J. Söderlund, “Pluralism in project management: navigating ” fl the crossroads of specialization and fragmentation, Inter- The authors declare that they have no con icts of interest. national Journal of Management Reviews, vol. 13, no. 2, pp. 153–176, 2011. Acknowledgments [16] M. G. C. Bosch-Rekveldt, “Applying mixed methods for researching project management in engineering projects,” in The authors acknowledge the participants of the separate Designs, Methods and Practices for Research of Project Manage- researches on which this meta-analysis is based. The authors ment, pp. 315–325, Gower Publishing, 2015. thank the conference attendees for their valuable feedback. [17] A. Tashakkori and C. Teddlie, Mixed Methodology: Combining Qualitative and Quantitative Approaches, vol. 46, Sage Publi- References cations, London, UK, 1998. [18] W. L. Wallace, The Logic of Science in Sociology, Aldine De [1] M. Bosch-Rekveldt, Y. Jongkind, H. Mooi, H. Bakker, and Gruyter, 1971. A. Verbraeck, “Grasping project complexity in large engi- [19] D. B. Baarda and M. P. M. de Goede, Basisboek Methoden en neering projects: the TOE (technical, organizational and Technieken, Wolters-Noordhoff, Groningen, Netherlands, environmental) framework,” International Journal of Project 2006. – Management, vol. 29, no. 6, pp. 728 739, 2011. [20] P. Verschuren and H. Doorewaard, Designing a Research [2] J. G. Geraldi, “What complexity assessments can tell us about Project, Lemma, Utrecht, Netherlands, 1999. ” projects: dialogue between conception and perception, [21] N. Blaikie, Designing Social Research, Polity Press, Cambridge, Technology Analysis and , vol. 21, no. 5, UK, 2nd edition, 2009. pp. 665–678, 2009. [22] J. Helvoort, Complexity in IT projects, [M.S. Thesis], Delft [3] H. Maylor, R. Vidgen, and S. Carver, “Managerial complexity University of Technology, Delft, Netherlands, 2011. in project based operations: a grounded model and its implications for practice,” Project Management Journal, [23] S. Nauta, Investigating Project Complexity in High-Tech vol. 39, no. S1, pp. S15–S26, 2008. Projects, [M.S. Thesis], Delft University of Technology, [4] L.-A. Vidal and F. Marle, “Understanding project complexity: Delft, Netherlands, 2011. implications on project management,” Kybernetes, vol. 37, [24] S. Nauta, M. Bosch-Rekveldt, and H. G. Mooi, “Project com- no. 8, pp. 1094–1110, 2008. plexity in the semiconductor industry: a case study approach,” [5] M. J. C. M. Hertogh, S. K. Baker, P. L. Staal, and E. Westerveld, in Paper presented at PMI® Research and Education Confer- Managing Large Infrastructure Projects: Netlipse, AT Osborne ence, Limerick, Munster, Ireland, Project Management Insti- BV, 2008. tute, Newtown Square, PA, USA, 2012. [6] M. Hertogh and E. Westerveld, Playing with Complexity, [25] S. M. H. M. Al-Tmeemy, H. Abdul-Rahman, and Z. Harun, “ ” [Ph.D. Thesis], Erasmus Universiteit Rotterdam, Rotterdam, Future criteria for success of building projects in Malaysia, Netherlands, 2010. International Journal of Project Management, vol. 29, no. 3, – [7] M. G. C. Bosch-Rekveldt, Managing Project Complexity. A pp. 337 348, 2011. Study into Adapting Early Project Phases to Improve Project [26] Z. Laslo, “Project portfolio management: an integrated method Performance in Large Engineering Projects, [Ph.D. Thesis], for resource planning and scheduling to minimize planning/ Delft University of Technology, Delft, Netherlands, 2011. scheduling-dependent expenses,” International Journal of Pro- – [8] M. Engwall, “No project is an island: linking projects to history ject Management, vol. 28, no. 6, pp. 609 618, 2010. and context,” Research Policy, vol. 32, no. 5, pp. 789–808, 2003. [27] A. Murray, Managing Successful Projects with PRINCE2 (Office [9] D. Howell, C. Windahl, and R. Seidel, “A project contingency of Government Commerce), The Stationery Office, Norwich, framework based on uncertainty and its consequences,” Inter- UK, 5th edition, 2009. national Journal of Project Management, vol. 28, no. 3, [28] J. K. Pinto, D. P. Slevin, and B. English, “Trust in projects: an pp. 256–264, 2010. empirical assessment of owner/contractor relationships,” [10] A. J. Shenhar, “One size does not fit all projects: exploring clas- International Journal of Project Management, vol. 27, no. 6, sical contingency domains,” Management Science, vol. 47, pp. 638–648, 2009. no. 3, pp. 394–414, 2001. [29] R. O. Briggs, A. J. Davis, J. D. Murphy, L. Steinhauser, and [11] T. Cooke-Davies, S. Cicmil, L. Crawford, and K. Richardson, T. F. Carlisle, “Transferring a collaborative work practice to “We’re not in Kansas anymore, Toto: mapping the strange practitioners: a field study of the value frequency model for landscape of complexity theory, and its relationship to project change-of-practice,” in Groupware: Design, Implementation, management,” Project Management Journal, vol. 38, no. 2, and Use, vol. 4715 of Lecture Notes in Computer Science, pp. 50–61, 2007. Springer, Berlin, Heidelberg, 2007. Complexity 15

[30] R. O. Briggs, J. D. Murphy, T. F. Carlisle, and A. J. Davis, “Pre- dicting change: a study of the value frequency model for change of practice,” in 2009 42nd Hawaii International Con- ference on System Sciences, pp. 1–10, Big Island, HI, USA, 2009. [31] M. G. C. Bosch-Rekveldt, H. G. Mooi, A. Verbraeck, E. Sjoer, B. Wolsing, and C. Gulden, “Mapping project manager’s com- petences to project complexity,” in IPMA 23rd WorldCongress, Research Track Human Side of Projects in Modern Business, pp. 85–96, Helsinki, Finland, 2009. [32] J. Thomas and T. Mengel, “Preparing project managers to deal with complexity – advanced project management education,” International Journal of Project Management, vol. 26, no. 3, pp. 304–315, 2008. [33] B. Xia and A. P. C. Chan, “Measuring complexity for building projects: a Delphi study,” Engineering, Construction and Archi- tectural Management, vol. 19, no. 1, pp. 7–24, 2012. [34] M. Giezen, “Keeping it simple? A case study into the advan- tages and disadvantages of reducing complexity in mega pro- ject planning,” International Journal of Project Management, vol. 30, no. 7, pp. 781–790, 2012. [35] J. Koppenjan, W. Veeneman, H. Van der Voort, E. Ten Heuvelhof, and M. Leijten, “Competing management approaches in large engineering projects: the Dutch Rand- stadRail project,” International Journal of Project Manage- ment, vol. 29, no. 6, pp. 740–750, 2011. [36] P. E. Eriksson, J. Larsson, and O. Pesämaa, “Managing com- plex projects in the infrastructure sector — a structural equa- tion model for flexibility-focused project management,” International Journal of Project Management, vol. 35, no. 8, pp. 1512–1523, 2017. [37] F. Di Maddaloni and K. Davis, “The influence of local commu- nity stakeholders in megaprojects: rethinking their inclusive- ness to improve project performance,” International Journal of Project Management, vol. 35, no. 8, pp. 1537–1556, 2017. [38] L. Koops, M. Bosch-Rekveldt, H. Bakker, and M. Hertogh, “Exploring the influence of external actors on the cooperation in public–private project organizations for constructing infra- structure,” International Journal of Project Management, vol. 35, no. 4, pp. 618–632, 2017. Hindawi Complexity Volume 2018, Article ID 9496731, 17 pages https://doi.org/10.1155/2018/9496731

Research Article Exploring Project Complexity through Project Failure Factors: Analysis of Cluster Patterns Using Self-Organizing Maps

Vicente Rodríguez Montequín , Joaquín Villanueva Balsera, Sonia María Cousillas Fernández, and Francisco Ortega Fernández Project Engineering Area, University of Oviedo, C/Independencia 13, 33004 Oviedo, Spain

Correspondence should be addressed to Vicente Rodr´ıguez Montequ´ın; [email protected]

Received 27 October 2017; Accepted 2 April 2018; Published 22 May 2018

Academic Editor: Luis Carral Couce

Copyright © 2018 Vicente Rodr´ıguez Montequ´ın et al. Tis is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

In the feld of project management, complexity is closely related to project outcomes and hence project success and failure factors. Subjectivity is inherent to these concepts, which are also infuenced by sectorial, cultural, and geographical diferences. While theoretical frameworks to identify organizational complexity factors do exist, a thorough and multidimensional account of organizational complexity must take into account the behavior and interrelatedness of these factors. Our study is focused on analyzing the combinations of failure factors by means of self-organizing maps (SOM) and clustering techniques, thus getting diferent patterns about the project managers perception on infuencing project failure causes and hence project complexity. Te analysis is based on a survey conducted among project manager practitioners from all over the world to gather information on the degree of infuence of diferent factors on the projects failure causes. Te study is cross-sectorial. Behavioral patterns were found, concluding that in the sampled population there are fve clearly diferentiated groups (clusters) and at least three clear patterns of answers. Te prevalent order of infuence is project factors, organization related factors, project manager and team members factors,andexternalfactors.

1. Introduction logistics/market conditions, geopolitical and social issues, and permitting and approvals [1]. Several studies focus on As projects have become more and more complex, there is project complexity and the factors that infuence its efect an increasing concern about the concept of project com- on project success. In general, there is not a clear diference plexity and its infuence upon the project management pro- between complexity and success factors when considering cess. Projects have certain critical characteristics that deter- the literature of project management. For instance, Gidado mine the appropriate actions to manage them successfully. defned project complexity and identifed the factors that Project complexity (organizational, technological, informa- infuence its efect on project success in relation to estimated tional, etc.) is one such project dimension. Te project dimen- production time and cost, based on literature search and sion of complexity is widespread within project management structured interviewing of practitioners [2]. Kermanshachi literature. In the feld of projects, complexity is closed related et al. [3] consider that when complexity is poorly under- to the causal factors to get the project outcomes, which in stood and managed, project failure becomes the norm. Tey the project management feld is usually referred to as “project focused on strategies to manage complexity in order to success/failure factors”. Better understanding of project suc- increase the likelihood of project success. Vidal and Marle [4] cess/failure factors is a key point for creating a strategy to defne project complexity as a property of a project, which manage complexity. makes it difcult to understand, foresee, and control the Most attributes of complexity are known to be constantly project’s overall behavior. Remington et al. [5] believe that a changing variables such as project type, project size, project complex project demonstrates a number of characteristics to location, project team experience, interfaces within a project, a degree, or level of severity, that makes it difcult to predict 2 Complexity project outcomes or manage projects. One of the project on infuencing project failure causes and hence project complexity defnitions that best fts the aim of this study was complexity. the one given by Kermanshachi et al.: “Project complexity is Te work here presented is part of a more global study the degree of interrelatedness between project attributes and analyzing success factors and failure causes in projects. A interfaces, and their consequential impact on predictability questionnaire has been aimed specifcally to project man- and functionality” [3]. Tis defnition can guide the iden- agement practitioners to gather information on the degree tifcation of project complexity indicators and management of infuence of diferent factors on the failure or success in strategies which reduce the undesired outcomes ofen related a project. Te questionnaire inquires about their perception to project complexity. Under this context, we can expect on the most infuential factors to be considered to reach a relationship among project failure causes (derived from success, as well as the common failure causes they have project attributes and interfaces) and project complexity. most frequently encountered. A selection of critical success Success in projects is more complex than just meeting factors and failure causes were selected as a basis for the cost, deadlines, and specifcations. In fact, satisfac- questionnaire, compiling previous research work results [16, tion with the fnal result has a lot to do with the perception of 17] with the most frequent causes refected in the literature. success or failure in a project. In the end, what really matters Te questionnaire is generic, not intended for any specifc is whether the parties associated with and afected by aproject sector or geographical area. Although it is not focused are satisfed or not [6]. Meeting deadlines and costs does on any particular project or feld, it gathers this type of not really matter if the fnal project outcomes do not meet information to be able to correlate it. Te survey was dis- expectations. tributed anonymously to recipients through LinkedIn, an However, our work is not focused on the concepts of Internet professional network. Such study determines the success or failure but on the study of the aspects that lead most frequent failure causes and the most important success to the failure in projects and their combinations, considering factors in real world projects. In the initial stage, a statistical that the complexity of a project is determined by these factors. analysis of the sample data was conducted with the aim of Tere are plenty of factors whose application is signifcant answering the question of whether the valuations depend for the success or failure of a project. In the literature, these on the geographical areas of the respondents or on the are called critical success factors,andmanystudieshavebeen types of projects that have been carried out [16]. Te study devotedtodefne,clarify,andanalyzesuchfactors.Success shows that there is no absolute criterion and that subjectivity factors are subjected to the perceptions of the ones involved is the inherent characteristic of those valuations. So, as a in the project development, depending not only on the stake- complementary study, clustering techniques are applied in holder but also on cultural or geographical diferences, which order to fnd patterns in the set of received answers. Tis are refected in the context of the organization [7]. Tere are paper presents the obtained results. Our study is a cross- also a lot of sectorial infuences. For example, Huysegoms et sectorial study, which is a remarkable improvement over the al. have identifed the causes of failure in sofware product majority of studies that consider only one sector, with the lines context [8]. Other examples can be found in [9–12]. construction one being of the most studied. Obviously, projects fail due to many diferent reasons, Te remainder of this paper is organized as follows. if we understand “failure” as the systematic and widespread Section 2 reviews the literature related to the environment noncompliance of the criteria which defne a successful of this work. Section 3 describes the research methodology. project [13]. Nevertheless, due to the inner subjectivity of the Sections 4 and 5 present and discuss the results and fnally concept, each person working in the same project has a Section 6 exposes the conclusions. personal opinion about the determining causes of its failure. Tese opinions can also vary depending on the type and 2. Literature Review sector of projects, so that distinctive patterns of causes are associated with the failure of specifc kinds of projects. Te Te literature is explored according to the three main points most usual is that a combination of several factors with of this work: background of complexity in the feld of diferent levels of infuence in diferent stages of the life cycle project management, project success/failure causes and their of a project result in its success or failure. connection with project complexity, and applications of self- Most of the studies in the literature focus on determining organizing maps (SOM) in the feld of project management. lists of factors and their categorization, ranking them accord- ing to their infuence level. Nevertheless, projects behavior 2.1. Project Complexity Teory. Complexity has been recog- derives from “systematic interrelated sets of factors” rather nizedasoneofthemostrelevanttopicsinprojectmanage- than single causal factors and the fact that true causes of ment research. According to Baccarini [18], one of the frst project outcomes are difcult to identify [14, 15]. Te interac- reviews about complexity in the project management feld, tions between the diferent factors seem to be as important as project complexity can be defned in terms of diferentiation each factor separately. However, there seems to be no formal and interdependency and it is managed by integration. In way to account for these interrelations, which may be the the defnition, diferentiation refers to the number of varied reason why this point is weakly treated in the literature. Our components of the project (tasks, specialists, subsystems, study is focused on analyzing the combination of these factors and parts), and interdependency refers to the degree of by means of clustering techniques, thus getting diferent interlinkages among these components. He established the patterns about the project manager practitioners perception dichotomy considering that project complexity is composed Complexity 3 of technological complexity and organizational complexity. technical complexity per se that made the management of Complexity can take various forms, namely social, technolog- the projects complex. Rather, the primary determinants of ical, environmental, and organizational. Worth mentioning the complexity of the project management process stemmed here is the work of Bosch-Rekveldt et al. [19] who proposed from changes in the markets, regulatory context, and knowl- the TOE framework, consisting of ffy factors in three edge requirements facing the project. families: technical, organizational, and environmental. Te Tere are connections between the project complexity authors have also concluded that organizational complexity indicators identifed in the literature and the project failure worried project managers more than technical or environ- causes. For example, Kermanshachi et al. [3] identifed 37 mental complexities. Vidal and Marle [4] argued that approx- indicators. Among them, several examples usually found imately 70% of project complexity factors are organizational. in the literature as project failure causes are included (i.e., We follow these assertions by focusing our study mainly on impact of external agencies on the project execution plan, the organizational factors, knowing that this covers a major impact of required approvals from external stakeholders, area of complexity in projects. level of difculty in obtaining permits, level of project design Scholars have focused on the identifcation of complexity changes derived by Request for Information (RFI), etc.). In attributes more than any other topic in the feld of project fact, there are several examples of authors that are referenced complexity. Studies in this area have evolved signifcantly over frequentlyinthefeldofbothprojectcomplexityandproject the past twenty years. Cicmil et al. [20] identifed complexity success factors. Shenhar et al. [25], Dvir and Lechler [26], as a factor that helps determine planning and control prac- Cooke-Davies [27], and Pinto and Prescott [28] are some tices and hinders the identifcation of goals and objectives, or examples of those. a factor that infuences time, cost, and quality of a project. Gidado [2] defned project complexity and identifed the 2.2. Project Success/Failure Factors. One of the most relevant factors that infuence its efect on project success. Also, the feldsofstudyinprojectmanagementissuccessfactorsand study proposes an approach that measures the complexity of failure causes in projects. Tey were frst identifed at the PMI the production process in construction. Annual Seminar and Symposium in 1986 [29] and became Project complexity has been analyzed in the book edited one of the most discussed themes within specialized literature by Cooke-Davies [21] from three diferent perspectives: peo- [30]. Each stakeholder working on a project has his/her own ple who manage programs and projects (practitioners), line opinion on what is determinant for failure and it is much managers in organizations to which programs and projects more complex than adhering to the traditional criteria of make a substantial contribution (managers), and members of time, cost, and quality. If a project can be considered failed the academic research community who have an interest in when it has not delivered what was expected afer its com- how complexity shapes and infuences the practice of pro- pletion, causes leading to this unsatisfactory result are also gram and project management (researchers). Te book con- subjected to the diferent points of view of the stakeholders stitutes a valuable resource to put together what is currently involved. known and understood about the topic, to help practitioners Te diferent lists of success factors and failure causes and their managers improve future practice, and to guide intheliteratureshownoconsensus.Temostusualwould research into answering those questions that will best help to be a combination of several factors, with diferent levels of improve understanding of the topic. infuence in diferent stages of the project’s life cycle, resulting Although most authors emphasized the infuence of inter- in its success or failure. Te concepts of success factors and dependencies and interactions of various elements on project failure causes are closely related, but a failure cause is not complexity [22], few works analyze those dependencies. For necessarily the negation of, or the opposite to, a success example, TOE model [19] does not allow for an understand- factor. Tere is not always such correspondence among them. ing of how various elements contribute to overall complexity. Tis study is considering failure causes rather than successful Nevertheless, other authors regarded project complexity as factors because we are taking the assumption that the more having nonlinear, highly dynamic, and emerging features. the failure causes concur in a project, the higher the complex- Vidal et al. [23], for example, proposed the defnition of ity is. project complexity as “the property of a project which makes Te literature identifying success/failure factors is very it difcult to understand, foresee, and keep under control extensive. Some examples, especially regarding failure causes, its overall behavior, even when given reasonably complete are[31–35],allofthemrelatedtotheconstructionsector. information about the project system”. Lu et al. [24] propose Another sector with many references is the IT (information that project complexity can be defned as “consisting of many technology) projects. Some examples are [36–39]. Tere are varied interrelated parts, and has dynamic and emerging several frameworks classifying the factors. Belassi and Tukel features”. Lessard et al. built the House of Project Complexity [40] suggested a scheme that classifes the critical factors (HoPC), a combined structural and process-based theoretical in four diferent dimensions (Table 1) and describes the framework for understanding contributors to complexity. impacts of these factors on project performance. Shenhar et Te connection between failure and complexity in project al. [25] have identifed also four distinct dimensions: project management has also been established in the literature. Ivory efciency, impact on the customer, direct and business suc- andAlderman[22]studiedthefailureincomplexsystems cess, and preparing for the future. Tey stated that the exact in order to shed some critical light on the management of content of each dimension and its relative importance complex projects. Te authors conclude that it was not the may change with time and are contingent on the specifc 4 Complexity

Table1:BelassiandTukeltaxonomy. Table 2: Failure causes considered in the study.

Groups of factors Cause Code Related to project Irruption of competitors C1 Related to the project managers and the team members Continuous or dramatic changes to the initial requirements C2 Factors related to the organization Customer’s requirements inaccurate, incomplete or not C3 Factors related to the external environment defned Disagreement or confict of interest among departments C4 Inaccurate cost estimations C5 stakeholder. Lim and Mohamed [41] consider two view- points of project success: macro and micro. Regarding the Inaccurate time estimations C6 micro, they have identifed technical, commercial, fnance, Defcient management of suppliers and C7 risk, environmental, and human related factors. Lack of Management support C8 Richardson [42] and King [43] point out that none of the Lack of previous identifcation of legislation C9 key success factors described in the literature is responsible, Badly defned specifcations C10 on its own, for ensuring a project’s success. Tey are all Political, social, economic or legal changes C11 interdependent and require a holistic approach. Groups of success factors and their interactions are of prime importance Project manager Lack of commitment C12 in determining a project’s success or failure [44]. Multivariate Project manager Lack of communication skills C13 statistics methods may be very useful with this purpose. Some Project manager Lack of competence C14 examples applied in the project management complexity are Project manager Lack of vision C15 [4, 45, 46] as well as in the feld of project success [47–49]. Project requirements defciently documented C16 Clusteringmethodshavebeenalsousedinthecontextof Project staf changes C17 exploring complex relationships within the feld of project management, for example, the works [50–52]. Our study uses Project team lack of competence C18 Project team misunderstanding related to customer/user self-organizing maps (SOM) and clustering techniques to C19 fnd patterns in a data set of answers coming from a survey. needs Project team lack of commitment C20 2.3. Self-Organizing Maps and Applications in Project Man- Public opinion opposition to Project C21 agement. SOM is an unsupervised neural network proposed Quality checks badly performed or not performed at all C22 by Kohonen [53] for visual cluster analysis. Te neurons of Extremely new or complex technology C23 the map are located on a regular grid embedded in a low Unexpected events with no efective response possible C24 (usually 2 or 3) dimensional space and associated with the cluster prototypes by the connected weights. In the course of Unrealistic customer expectations C25 the learning process, the neurons compete with each other Wrong number of people assigned to the project C26 through the best-matching principle in such a way that the input is projected to the nearest neuron given a defned set of candidate portfolios. Afer reviewing the literature, we distance metric. Te winner neuron and its neighbors on the can conclude that there is not any former application of SOM map are then adjusted towards the input in proportion with for analyzing patterns of project failure causes. the neighborhood distance; consequently, the neighboring neuronslikelyrepresentsimilarpatternsoftheinputdata 3. Research Methodology space.Duetothedataclusteringandspatializationthrough the topology preserving projection, SOM is widely used As discussed in the Introduction, the ground for this work in the context of visual clustering applications. Despite the is the questionnaire that was designed to gather information unsupervised nature, the applicability of SOM is extended on the perception project managers have of what the success to classifcation tasks by means of a variety of ways, such factors and failure causes are. Afer the information was as neuron labeling method, semisupervised learning, or gathered, a multivariate analysis was performed on the data supervised learning vector quantization (LVQ) [54]. with cluster data mining techniques. SOM is recognized as a useful technique to analyze high- Te sections of the questionnaire considered for this dimensional data sets and understand their hidden relations. study were as follows: It can be used to manage complexity in large data sets [39]. (i) General information on the respondent and typology Nevertheless, there are few cases described in the feld of of projects he/she was involved in: country, type, and project management. Balsera et al. [55] have exposed the size of project. application of SOM to analyze information related to efort (ii) Frequency of diferent causes of project failure, with estimation and sofware projects features. MacDonell [56] 26 multiple-choice questions (from 0–25%, rare or has reported other multidimensional data study visualization improbable occurrence, to 75–100%, always occurs). basedonSOM,identifyinggroupsofdataforsimilarprojects and fnding nonlinear relationships within the explored Te causes of failure are extracted from the existing variables. Naiem et al. [57] have used SOM for visualizing the bibliography and the previous work on the matter. Table 2 Complexity 5

Factors related to organization Factors related to project

·Continuous or dramatic changes to the initial requirements (C2) ·Disagreement or conflict of interest among departments (C4) ·Customer’s requirements inaccurate, incomplete or not defined (C3) ·Deficient management of suppliers and procurement (C7) ·Inaccurate cost estimations (C5) ·Lack of Management support (C8) ·Inaccurate time estimations (C6) ·Lack of previous identification of legislation (C9) ·Project staff changes (C17) -Badly defined specifications (C10) ·Wrong number of people assigned to the project (C26) ·Project requirements deficiently documented (C16) ·Quality checks badly performed or not performed at all (C22) ·Extremely new or complex technology (C23)

FAILURE

Factors related to the project manager and team members Factors related to external environment

·Project manager Lack of commitment (C12) ·Project manager Lack of communication skills (C13) ·Irruption of competitors (C1) ·Project manager Lack of competence (C14) ·Political, social, economic or legal changes (C11) ·Project manager Lack of vision (C15) ·Public opinion opposition to Project (C21) ·Project team lack of competence (C18) ·Unexpected events with no effective response possible (C24) ·Project team misunderstanding related to customer/user needs (C19) ·Unrealistic customer expectations (C25) ·Project team lack of commitment (C20)

Figure 1: Failure causes grouped by category. Te factors are derived from existing bibliography and previous work on the matter. Tey have been classifed according to Belassi and Tukel taxonomy. gathers the identifed factors. Te factors are coded as ��, literature, or even specifc points of consideration, should where � is the number given to the failure cause. belong to at least one group [40]. However, it is not easy Te factors were classifed according to Belassi and to diferentiate in which category to include each of the Tukel taxonomy [40] included in Table 1. Figure 1 depicts diferent factors. Te groups are interrelated. Te authors theclassifcationconsidered.Tefactorsarepresentedun- give some indications to distinguish where to classify them, shorted because they were shufed in order to avoid biases mainly regarding organization and external environment. in the survey. Te frst group includes the factors related to For example, if a customer is from outside the organization, the organization the project belongs to (i.e., factors related he/she should be considered as an external factor (i.e., C25- to top management support). Te second group includes the unrealistic customer expectations). For functional projects, factors related to the project itself and the way the project however, are usually part of the organization, such is managed. Te third one comprises the factors related to as top management. In such cases, factors related to the client the project manager and the team members. Many studies can be grouped under the factors related to the organization. demonstrated the importance of selecting project managers We can fnd correspondences among most of the failure who possess the necessary technical and administrative causes included in our study (Table 2) and the complexity skills for successful project termination, as well as their indicators appointed by Kermanshachi et al. [3]. For example, commitment. Te competences of the team members are also C2 (which is one of the most frequent in accordance with found to be a critical factor. Finally, the factors related to the rank presented in Table 7) is related to several complexity external environment consist of factors which are external to indexes as “level of project design changes derived by RFI” the organization but still have an impact on project success and “magnitude of change orders”. C10, for example, is or failure. Te classifcation can be considered collectively related to “percentage of design completed at the start of exhaustive. Belassi and Tukel state that the four groups construction”. Tis fact reinforces the connection between ofer a comprehensive set in that any factor listed in the failure causes and complexity factors. 6 Complexity

Te recipients were randomly chosen among the mem- Table 3: Countries and number of responses (ordered by number bers of 36 project management groups from LinkedIn net- of responses). When several countries are indicated in the same work.Tequestionnairewasopenfor3monthsand11days row, the number of respondents stands for the number of answers in 2011, in order to obtain a signifcant number of answers. received from each country (i.e., 50 answers from Spain and 50 During that period of time, customized emails were sent answers from the United States). with an invitation to answer the questionnaire to a total of NUMBER OF COUNTRIES 3,668 people. A total of 619 answers were received (16.88%), RESPONDENTS 611 of which were considered for further analysis (the rest Argentina 91 were discarded due to consistency issues). Neither of the Spain; United States 50 questionnaires was partially flled or incomplete, since all felds were marked as mandatory. Previously, in 2010, a pilot Greece 45 survey was conducted with the help of project management Chile 41 experts and practitioners. Tis primary questionnaire was India 40 sent up to 45 people with a response rate of 66.67%. Tose Brazil 30 factors which reached a higher score were the ones fnally Luxembourg 27 selected to confgure the list included in the current study. Mexico 21 Suggestions and comments made by respondents to improve theproposedlistwerealsotakenintoaccount. Uruguay 20 Answers from 63 countries were received, as Table 3 United Arab Emirates 19 shows. Italy 18 Te answers have been grouped frst into 13 diferent Russia 16 geographical zones (more details can be found in [16]), Sweden 14 according to geographical, cultural, historical, and economic South Africa 12 criteria. From those, a total of 6 groups comprise more than 85% of the respondents: AMZ3, EUZ3, EUZ1, AMZ1, ASZ1, Canada 11 and ASZ3. Te other 15% of the responses were grouped in a Germany 8 Saudi Arabia; Te Netherlands; United new category called “Others”. Table 4 summarizes the groups 7 by geographical zone. Kingdom Followingasimilarapproach,53typesofprojectswere Australia; Denmark; Finland; Kuwait 5 present on the answers received. To simplify, they have been Belgium; Paraguay; Slovenia 4 classifed into a total of 17 groups, according to the highest Egypt; France; Israel; Switzerland 3 level of the ISIC Rev. 4 Classifcation [58]. Development aid Philippines; Tanzania 2 projects were listed apart as an independent category. Table 5 presents the number of respondents from each project type Algeria; Austria; Bahrain; Belarus; Botswana; British Virgin Islands; Cameroon; China; and the codifcation. Colombia; Cyprus; Czech Republic; Ghana; Information about the size of the projects the respondents Hong Kong; Hungary; Indonesia; Ireland; 1 are usually involved in was also enquired. Te size of the Japan; Jordan; Nigeria; Pakistan; Peru; Poland; projects is something quite subjective and difcult to compare Qatar; Rwanda; Singapore; Surinam; Tailand; among diferent sectors. For example, considering the project Uganda; Venezuela budget as an indicator of project size, a project considered small in construction could, by contrast, be considered large in information technology projects. In order to avoid these biases, the guideline provided by the University of Each interval factor is defned as presented in Table 6. Te Wisconsin-Madison [59] was used. Te results show that ranking is presented in Table 7. 11.95% of respondents are involved in small projects, 49.26% With the obtained results, a cluster analysis was con- in medium size projects, and 38.79% in large projects. ducted to obtain patterns in the answers data set, grouping Tough the scope of this paper does not cover the a set of data with similar values. Te aim is to classify a set of statistical analysis of the obtained answers, here is a summary simple elements into a number of groups in such a way that ofthemainresults.Overall,boththemostandleastfrequent the elements in the same group are similar or related to one failure causes are presented here. In order to rank each one of another and, at the same time, diferent or unrelated to the the 26 failure causes, a frequency index (FI) was calculated as elements in other groups. follows: Te cluster analysis (also called Unsupervised Classi- 611 4 fcation, Exploratory Data Analysis, Clustering, Numerical [∑�=1 ∑�=1 �� ∗��] FI = , (1) Taxonomy, or Pattern Recognition) is a multivariate statistical 611 technique whose aim is to divide a set of objects into groups where or clusters in such a way that objects in the same cluster are very similar to each other (internal cluster cohesion) and � � � is the number of responses choosing interval , the objects in other groups are diferent (external cluster �� is interval factor �. isolation). Summarizing, it deals with creating data clusters Complexity 7

Table4:Answersgroupedbygeographicalzone.

GEOGRAPHICAL ZONES NUMBER OF COUNTRIES ANALYZED RESPONDENTS AMZ3 Argentina, Brazil, Chile, Colombia, Paraguay, Peru, Surinam, Uruguay and Venezuela 190 EUZ3 Cyprus, Spain, Greece and Italy 114 Germany, Austria, Belgium, Denmark, Finland, France, Ireland, Virgin Islands (UK), EUZ1 86 Luxembourg,UnitedKingdom,TeNetherlands,SwedenandSwitzerland AMZ1 Canada and United States 61 Algeria, Bahrain, Indonesia, Jordan, Pakistan, Qatar, Saudi Arabia, Egypt, United Arab ASZ1 40 Emirates and Kuwait ASZ3 India 40 OTHERS Rest 80

Table 5: Project types and number of responses.

CODE PROJECT TYPE NUMBER OF RESPONDENTS ISIC1 Information and communication 365 ISIC2 Financial and insurance activities 43 ISIC3 Construction 42 ISIC4 Manufacturing 41 ISIC5 Professional, scientifc and technical activities 32 ISIC6 Public administration and defence 12 ISIC7 Mining and quarrying 11 ISIC8 Human health and social work activities 10 ISIC9 Electricity, gas, steam and air conditioning supply 10 ISIC10 Accommodation and food service activities 8 ISIC11 Education 8 ISIC12 Administrative and support service activities 7 ISIC13 Transportation and storage 7 ISIC14 Wholesale and retail trade 6 ISIC15 Real estate activities 4 ISIC16 Development aid 3 ISIC17 Arts, entertainment and recreation 2

Table 6: Intervals and their interval factor. bidimensional) from the input �-dimensional space, which preserves its topology and maintains it. As a result, the Factor (interval Interval Meaning arithmetic mean) network shows the distance between the diferent sets, so thatitcanbeusedasanadequatevisualizationsurfaceto 0–25% Seldom/Unlikely to occur 0,125 display diferent data characteristics as, for instance, their 25–50% Occasional/Likely to occur 0,375 cluster division. Summarizing, SOM Clustering Networks 50–75% Frequent/Very likely to occur 0,625 allow input data clustering and easily visualize the resulting 75–100% Always occurs 0,875 multidimensional data clusters. For the analysis of data collectedinthequestionnaire,SOMToolboxwasusedwith MATLAB [61]. Te methodology used is a two-level approach for partitive clustering, where the data set is frst projected in such a way that each group is homogeneous and diferent using the SOM, and then the SOM is clustered, as described fromtherest.Forthispurpose,manydataanalysistechniques in [62]. Partitive clustering algorithms divide a data set into canbeused.Inthisstudy,theSelf-organizedMaps(SOM) a number of clusters, typically by trying to minimize some technique has been used, a specifc type of neural network criterion or error function. An example of a commonly used [60]. partitive algorithm is the �-means, which minimizes error SOM networks are an excellent tool for exploring and function (2), where � isthenumberofclustersand�� is the analyzing data, which are especially adequate due to their center of cluster �. remarkable visualization properties. Tey create a series of � � �2 prototype vectors which represent the data set and project �=∑ ∑ ��−��� (2) such vectors into a low dimensional network (generally �=1 �∁�� 8 Complexity

Table7:Summaryofresults.

RANK CODE FAILURE CAUSE FREQUENCY INDEX 1 C3 Customer’s requirements inaccurate, incomplete or not defned 0,5992 2 C2 Continuous or dramatic changes to initial requirements 0,5665 3 C6 Inaccurate time estimations 0,5329 4 C16 Project requirements inadequately documented 0,5092 Not or badly defned specifcations at the time the Project Team starts 5C10 0,4859 to work 6 C5 Inaccurate costs estimations 0,4826 7 C8 Lack of Management support 0,4605 8 C25 Unrealistic Customer’s expectations 0,4507 Disagreements or conficts of interest among diferent departments 9C4 0,4478 involved Project Team’s misunderstandings related to Customer/User’s wishes 10 C19 0,4417 or needs 11 C17 Project staf changes 0,4306 12 C22 Quality checks not or badly performed 0,4270 13 C26 Wrong number of people assigned to the project (too many, too few) 0,4135 14 C13 Project Manager’s lack of communication skills 0,4114 15 C7 Inadequate management of suppliers and procurement 0,3828 16 C18 Project Team’s lack of competence 0,3815 17 C15 Project Manager’s lack of vision 0,3746 18 C14 Project Manager’s lack of competence 0,3742 19 C23 Too much complex or new technology 0,3574 20 C20 Project Team’s lack of commitment 0,3537 21 C24 Unexpected events with no efective response possible 0,3292 22 C9 Lack of previous identifcation of relevant rules and legislation 0,3194 23 C12 Project Manager’s lack of commitment 0,3009 24 C11 Political, social, economic or legal changes 0,2662 25 C21 Public opinion opposition to project 0,2428 26 C1 Competitors 0,2261

To select the best one among diferent partitionings, each Te SOM consists of a regular, usually two-dimensional, of these can be evaluated using some kind of validity index. grid of map units. Each unit � is represented by a prototype Several indices have been proposed [63, 64]. In our work, we vector �� =[��1,...,���],where� is input vector dimen- used the Davies-Bouldin index [65], which has been proven sion. Te units are connected to adjacent ones by a neigh- to be suitable for evaluation of �-means partitioning. Tis borhood relation. Te number of map units, which typically index is a function of the ratio of the sum of within-cluster varies from a few dozen up to several thousand, determines scatter to between-cluster separation. According to Davies- the accuracy and generalization capability of the SOM. Dur- Bouldin validity index, the best clustering minimizes (3), ing training, the SOM forms an elastic net that folds onto the which uses �� for within-cluster distance (�(��)) and ��� for “cloud” formed by the input data. Data points lying near each between-cluster distance (�(��,��)). other in the input space are mapped onto nearby map units. � Tus, the SOM can be interpreted as a topology preserving 1 �� (��)+�� (��) ∑max { } (3) mapping from input space onto 2D grid of map units. � �=� ̸ � (� ,�) �=1 �� � � Te SOM is trained iteratively. At each training step, a sample vector � is randomly chosen from the input data set. Te approach used in this paper (clustering the SOM Distances between � and all the prototype vectors are com- rather than clustering the data directly) is depicted in puted. Te best-matching unit (BMU), which is denoted here Figure 2. First, a large set of prototypes (much larger than by �,isthemapunitwithprototypeclosestto�. the expected number of clusters) is formed using SOM. Te � � � � prototypes can be interpreted as “protoclusters” [62], which ��−��� = min {��−���} � � � � � (4) are in the next step combined to form the actual clusters. Each data vector of the original data set belongs to the same cluster Next, the prototype vectors are updated. Te BMU and its as its nearest prototype. topological neighbors are moved closer to the input vector in Complexity 9

N samples M Prototypes C Clusters

Figure 2: Clustering process. the input space. Te update rule for the prototype vector of trying diferent numbers of clusters. One of the most signif- unit � is cant pieces of information provided by this kind of analysis is preciselytheoptimalnumberofclustersoroptimal clustering �� (�+1) =�� (�) +�(�) ℎ�� (�) [�−�� (�)], (5) and how the samples are distributed among the clusters. In order to determine the right number of clusters, the where Davies-Bouldin index (DBI) was used [65] as described above. � is time, Te resulting clusters were analyzed in order to charac- �(�) is the adaptation coefcient, terize the perception of project manager practitioners about ℎ��(�) is neighborhood kernel centered on the winner project failure and hence complexity. An analysis of which unit: factors make each cluster diferent from the rest of the data is performed, as well as what the dependencies between vari- � �2 �� −�� ables are in the clusters. Each cluster was featured fnding the ℎ (�) = (−� � �� ), �� exp 2�2 (�) (6) setoffailurecausesratedoverorunderthemodeforthe global survey and identifying the most remarkable diferences with the other clusters. In order to do that, histograms and � � � � where � and � are positions of neurons and on the SOM radar charts were used. Finally, we studied the population grid. Both �(�) and �(�) decrease monotonically with time. distribution of each cluster regarding geographical areas, To perform the analysis, a fle with the following 30 input projects sector, and projects size. Te analysis is presented by variables was prepared (the columns of the data matrix are means of contingency tables. the variables and each row is a sample): (i) Project size (this categorical variable has been 4. Results 2 encoded as follows: small = 3; medium = 3 =9;large 3 Afer several preliminary trials, a 7×5hexagonal SOM topol- =3 =27). ogy was chosen. Te SOM was trained following the method (ii) 26 failure causes (Table 2). described in the former section. Figure 3 shows the Unifed distance matrix (U-matrix) and 27 component planes, one (iii) Answer ID (encoded from P1 to P611). Tis variable is foreachvariableincludedinthetraining.TeU-matrixis included only for tracing purpose during the valida- a representation where the Euclidean distance between the tion stage, not for training. codebook vectors of neighboring neurons are depicted in a (iv) Country (the countries were grouped into 13 geo- color schema image. In this case, high values are associated graphical areas, taking into account geographical, withredcolorsandlowvalueswithbluecolors.Tisimage economic, historical, and cultural criteria). is used to visualize the data in a high-dimensional space (v) Type of project (they are encoded into 17 activities using a 2D image. High values in the U-matrix represent a derived from the ISIC/CIIU Rev. 4 codes). frontier region between clusters, and low values represent a high degree of similarity among neurons on that region. Each For the training, only the scores of the 26 failure causes component plane shows the values of one variable in each and the project size were used. Te other variables of the map unit. Trough these component planes, we can realize data set (country and type of project) were used just only emerging patterns of data distribution on SOM’s grid and in analyzing the results of the clustering. Te training has detect correlations among variables and the contribution of been performed trying diferent grid dimensions but always each one to the SOM diferentiation only viewing the colored considering hexagonal topology (6 adjacent neighbors) and pattern for each component plane [66]. 10 Complexity

U-matrix Size C1 C2 C3 C4 18 0.28 2 0.7 0.26 0.7 1.5 0.5 16 0.24 0.6 0.6 1 0.22 0.5 0.5 0.4 0.5 14 0.2 0.4 0.4 0.3 d d d d d C5 C6 C7 C8 C9 C10 0.7 0.5 0.5 0.6 0.6 0.6 0.6 0.4 0.5 0.4 0.5 0.5 0.5 0.4 0.3 0.4 0.4 0.3 0.4 0.3 0.2 0.3 d d d d d d C11 C12 C13 C14 C15 C16 0.6 0.35 0.6 0.6 0.6 0.6 0.3 0.4 0.25 0.4 0.4 0.4 0.5 0.2 0.4 0.2 0.2 0.2 0.2 d d d d d d C17 C18 C19 C20 C21 C22 0.6 0.6 0.4 0.5 0.5 0.5 0.5 0.4 0.3 0.4 0.4 0.4 0.4 0.3 0.3 0.2 0.3 0.2 0.3 d d d d d d C23 C24 C25 C26 0.6 0.5 0.45 0.6 0.4 0.5 0.5 0.4 0.35 0.4 0.3 0.4 0.3 0.3 0.25 0.3 d d d d Figure 3: SOM U-matrix and component plane matrices using a 7×5hexagonal SOM.

C2 C25 C9

(a) (b) (c)

Figure 4: Prototypes of the three patterns found in the component planes.

Several similarities among the component planes of some (ii) Pattern II: C25 and C26 (Figure 4(b)) factorscanbefoundinFigure3.Treediferentpatternscan (iii) Pattern III: C9, C12, C13, C14, C15, C18, C20, and C21 be clearly distinguished: (Figure 4(c)) (i) Pattern I: C2, C3, C5, C6, C10, C16, and C17 An example of each component plane has been depicted (Figure 4(a)) in Figure 4 to illustrate each of the patterns. As introduced Complexity 11

Table 8: Summary of patterns interpretation.

Pattern number Focus attention Interpretation Tis pattern of respondents links project complexity mainly with the characteristics of the project and how it is managed: the requirements are Pattern I How the project is managed incomplete or inaccurate, with continuous changes, the specifcations are badly defned and both costs and schedule are inaccurate. Tey consider also that the project staf changes entail a higher complexity. Tey emphasize the infuence of unrealistic customer expectations and a wrong (it is supposed to be insufcient) number of people assigned to the Insufcient number of resources and Pattern II project. It is curious that neither of them is apparently under their unrealistic customer expectations responsibility (under the hypothesis that the resources assigned to the project are a decision made by the organization managers). Project manager and team members skills, Tey have a more personalistic view of the complexity, with the role of the Pattern III competences and commitment project manager and team members being decisive.

Davies-Bouldin’s index SSE 1.4 200 150 1.2 100

1 50

0 0.8 1 2 3 4 5 6 1 2 3 4 5 6 7×5 Figure6:Sumofsquarederrorsusingahexagonal7×5SOM Figure 5: Davies-Bouldin’s index using a hexagonal SOM � � calculated for each clustering (�-axis: number of clusters/�-axis: calculated for each clustering ( -axis: number of clusters/ -axis: Davies-Bouldin’s index). sums of squared errors).

previously, we can detect correlations among variables by minimization of both parameters. In this case, we have taken just looking at similarities in the component planes [66]. We 5 diferent clusters. can infer combinations of answers from the surveyed project Figure 7 depicts the clustering results taking as reference managers. In this case, a combination of answering patterns the results of the application of the �-means algorithm on can be concluded from each of the three groups. So, the the optimal cluster number according to Davies-Bouldin. Te project managers that rate factor C2 as frequent also rate as SOM plot sample hits (b) represent the number of input frequent factors C3, C5, C6, C10, C16, and C17 (Pattern I). vectorsclassifedbyeachneuron.Terelativenumberof Itisalsothesamewiththeothertwopatterns.Regarding vectors for each neuron is shown via the hexagon’s size and Pattern I, it should be noticed that all the factors considered its color represents the similarity among neurons. belong to the project category, except factor C17 that belongs Pearson’s chi-squared test [67] was calculated to fnd to the organization category. Regarding Pattern II,thereis dependence among the variables and their distribution in an association between the rating of unrealistic customer’s the clusters. Te resulting � values for the contingency tables expectations (C25) and a wrong number of people assigned showninTables9,10,and11werehigherthan0.05,so to the project (C26). Regarding Pattern III,mostofthefactors itcanbeconcludedthatneithertheprojectsizenorthe belong to the project manager and team members (noted in geographical zone or the project type is signifcant for the thefguresasPM&T)category(C12,C13,C14,C15,C18,and groups determined. Nevertheless, the � values obtained for C20). �� are less than 0.05 which implies that the variation of the Table 8 summarizes the interpretation of each pattern. A values of �� through the clusters has statistical signifcance, focus of attention has been remarked for each one and how so we proceeded to the study and categorization of clusters each pattern could be understood. Obviously, interpretation based on these factors with statistical signifcance. Bar charts has a component of subjectivity. were plotted to characterize each cluster (depicted in Figures Next, �-means algorithm was used to build the clusters 8, 9, 10, and 11). Factors have been grouped by their category and the Davies-Bouldin index was considered to determine according to Belassi and Tukel taxonomy [40] in order to the optimum number of clusters. Te Davies-Bouldin index facilitate the interpretation of the information. andthesumsofsquarederrorsdiagramsaredisplayedin Figures 5 and 6, where the �-axis represents the number 5. Discussion of Results of clusters. According to the usual work methodology of this type of neural networks, the best possible clusterization It can be observed that clusters 3 and 5 are associated with the will be the one that reaches a better compromise in the highest rates of the factors for each category, while clusters 12 Complexity

1 1 1 1 1

2 1 1 1 1

2 2 2 4 4

2 2 4 4 4

5 5 5 3 3

5 5 3 3 3

5 5 3 3 3

(a) (b)

Figure 7: Partitioning of the SOM codebooks with 5 clusters (a) and the SOM plot sample hits (b).

Table 9: Contingency table showing the distribution of project size among the clusters.

Cluster 1 Cluster 2 Cluster 3 Cluster 4 Cluster 5 Samples Size Small25141461473 Medium 84 53 68 40 56 301 Large 57 35 61 42 42 237 Samples 166 102 143 88 112 611

0.7 0.7

0.6 0.6

0.5 0.5

0.4 0.4

0.3 0.3

0.2 0.2

0.1 0.1

0 0 C4 C7 C8 C17 C26 C2 C3 C5 C6 C9 C16 C22 C23 C10

Cluster1 Cluster4 Cluster1 Cluster4 Cluster2 Cluster5 Cluster2 Cluster5 Cluster3 Mode Cluster3 Mode Figure 8: Organization factors and clusters bar chart. Factor Figure 9: Project factors and clusters bar chart. Factor importance importance average score of each cluster plotted against the mode average score of each cluster plotted against the mode of the global of the global survey. survey.

1 and 2 are the lowest. In fact, cluster 1 only attributes 3 and 5 rate factors equal to or over the general mode and high importance/frequency to C3 (customer’s requirements clusters 1 and 2 rate factors equal to or below the general inaccurate, incomplete, or not defned). On the other hand, mode, as represented in Figures 8, 9, 10, and 11. An exception cluster 3 rates with high values 19 in 26 factors, and cluster 5 ofthiscanbefoundforC5(inaccuratecostestimations), rates high values 13 in 26. So, the general trend is that clusters where cluster 2 rates higher than the mode. A global rating Complexity 13

Table 10: Contingency table showing the distribution of geographical zone among the clusters (see Table 4 for details about the geographical zones).

Cluster 1 Cluster 2 Cluster 3 Cluster 4 Cluster 5 Samples Zone AMZ1 26 9 11 5 10 61 AMZ3 43 35 48 30 34 190 ASZ1 8 3 12 9 8 40 ASZ3 9 5 13 7 6 40 EUZ1 23 18 17 10 18 86 EUZ3 38 19 22 16 19 114 Others 19 13 20 11 17 80 Samples 166 102 143 88 112 611

Table 11: Contingency table showing the distribution of project type among the clusters (see Table 5 for details about the project type codes).

Cluster 1 Cluster 2 Cluster 3 Cluster 4 Cluster 5 Samples Type ISIC1 100 59 90 49 67 365 ISIC2 14 5 9 3 12 43 ISIC3 12 5 10 8 7 42 ISIC4 7 7 12 8 7 41 ISIC5 10 9 3 6 4 32 ISIC6 1 0 5 4 2 12 ISIC7 1 1 4 3 2 11 ISIC8 3 3 3 0 1 10 ISIC9 6 1 2 1 0 10 ISIC10 1 2 1 1 3 8 ISIC11 2 2 1 0 3 8 ISIC12 3 3 0 1 0 7 ISIC13 2 2 0 1 2 7 ISIC14 3 2 0 1 0 6 ISIC15 0 0 1 1 2 4 ISIC16 0 0 2 1 0 3 ISIC17110002 Samples 166 102 143 88 112 611 overview of each cluster can be drawn from Figure 12, where Table 8) and the clusters. Te most remarkable connections a radar chart was built representing the average rates of each canbefoundwithinPattern I and Cluster 1 (both sharing a clustertothefactorsincludedineachcategory. high importance of incomplete or inaccurate requirements, In a similar way, from Figure 13, we can conclude how the badly defned specifcations, the project staf changes, and each group of factors is rated within each cluster. It can inaccurate costs estimations) and Pattern III and Cluster 3 be observed that all the factors are arranged following the (both remark the importance of project management and same order (project, organization, project manager and team team members skills, competences, and commitment). members, and external factors) with three exceptions: cluster 1 gives PM&T the same importance as the external factors 6. Conclusions (very low in both cases), cluster 4 gives PM&T the same importance as organization factors, and cluster 3 is the most Te analysis performed by clustering techniques has allowed remarkablebecauseitgivesthesameimportancetothe us to conclude that the total number of answers obtained can project, the organization, and the PM&T factors. be grouped into 5 classes of respondents, who behaved dif- An interpretation of each cluster (Table 12) can be in- ferently in analyzing project failure causes and hence project ferred from the presented results, especially Figure 13. In complexity. Tis result is coherent with the conclusions of order to facilitate the understanding, the focus of attention the existing literature on the subject, which claims that of each cluster is remarked. there are no unique criteria and subjectivity is an inherent Finally,wecanalsofndsomeconnectionsbetweenthe characteristic of these assessments. Te behavior of each patterns found in the component planes (summarized in cluster can be understood by means of the bar charts shown. 14 Complexity

Table 12: Summary of clusters interpretation.

Cluster number Focus attention Description Cluster 1 is featured as the pattern of respondents that gives very low infuence over the projects failure to the factors included in the questionnaire. Tey just only attribute some infuence to the project and the organization categories. Te most remarked factors are (highly rated factors that are equal or above the Project factors, specially customer’s general mode): Cluster 1 requirements inaccurate, incomplete (i) C4: Disagreement of confict of interest among departments or not defned. (ii) C17: Project staf changes (iii) C3: Customer’s requirements inaccurate, incomplete or not defned (iv) C5: Inaccurate cost estimations (v) C10: Badly defned specifcations Cluster 2 attributes the highest infuence to the project category, followed by the factors related to the organization and, to a lesser extent, to the project manager and, team members and external categories. Both the project and the organization, All the factors related to the organization are scored with the same importance Cluster 2 remarking the infuence of inaccurate as the general mode. cost estimations. Regarding the project category, all the factors are rated high or medium, excepting C9 (lack of previous identifcation of legislation). Te stress is specially put in C5 (inaccurate cost estimations). Project, organization and, project Cluster 3 is featured as the pattern of respondents that consider the project Cluster 3 manager and team members factors manager and team members factors as the highest value, at the same level as are all considered at the highest value. the project and the organization factors. Cluster 4 is characterized because it associates a medium infuence to project, Project factors mainly, but organization, and project manager and team member almost equally. All the organization and project manager and factors are rated equal to the mode, except three: Cluster 4 team members factors are also very (i) C6: Inaccurate time estimations (rated below the mode) important. (ii) C9: Lack of previous identifcation of legislation (rated above the mode) (iii) C16: Project requirements defciently documented (rated below the mode) Cluster 5 associates a very high infuence on the projects failure to the categories of project and organization. Te main diference with cluster 3 is Cluster 5 Project and organization that cluster 5 attributes less importance to the project manager and team members category.

0.7 0.7

0.6 0.6

0.5 0.5

0.4 0.4

0.3 0.3

0.2 0.2

0.1 0.1

0 0 C12 C13 C14 C15 C18 C19 C20 C1 C11 C21 C24 C25

Cluster1 Cluster4 Cluster1 Cluster4 Cluster2 Cluster5 Cluster2 Cluster5 Cluster3 Mode Cluster3 Mode Figure 10: Project manager and team members factors bar chart. Figure 11: External factors and clusters bar chart. Factor importance Factor importance average score of each cluster plotted against the average score of each cluster plotted against the mode of the global mode of the global survey. survey. Complexity 15

Project Te prevalent order of importance for the factor cate- gories is project, organization, project manager and team members, and external factors. Te most remarkable excep- tion can be found in cluster 3, which attributes the highest infuence also to the project manager and team members category. Te infuence of external factors is in all the cases very low, so it can be remarked that the project managers attribute the complexity to the inner features of the projects External PM&T and the management conditions. On the other hand, the analysis of the component planes has revealed that there are three clear patterns. Te frst one (Pattern I) establishes a close association with the factors included in the project category as factors of failure and hence complexity. However, the third pattern (Pattern III) focuses on the factors associated with the project manager and the team members. Tis is a remarkable fact that shows at least two schools of thought: the one considering the Organization factors inherent to the project and the one that attributes the complexity to the skills and competences of the project Cluster3 Cluster4 managers and the team members. Te second pattern points Cluster5 Cluster1 Cluster2 out a relation between factors C25 and C26, which belong to two diferent categories. Te analysis of the component planes Figure 12: Radar chart representing the average rate for the factors is independent of the cluster analysis, although some similar- of each category. ities can be found between these patterns and the clusters. Te most remarkable are the similarities of Pattern I and Cluster Cluster3 1andPattern III and Cluster 3. Finally, this study provides a multidimensional analysis of the complexity in projects. Some signifcative combinations of factors have been found. A limitation of this study is that only the perception of project managers is considered. Te study may be extended considering other project stakehold- Cluster1 Cluster5 ers. Other limitations are that more than 67% of the answers come from only 10 countries (Argentina, Spain, United States, Greece, Chile, India, Brazil, Luxembourg, Mexico, and Uruguay) and there are a number of countries with a very limited number of answers. In addition, more than 59% of the answers are related to IT projects. Taking into account the precedent limitations detected in the study, results are likely tobebiasedbecauseofcountryandtypeofprojectresponse percentages, and therefore they cannot be generalized. Te Cluster4 Cluster2 results have been exposed per geographical zone and per Project PM&T project type to avoid the referred limitation. Organization External Conflicts of Interest Figure 13: Radar chart representing the average rate of each cluster in each category. Te authors declare that there are no conficts of interest regarding the publication of this paper.

Representing the average of each cluster for each variable Acknowledgments andcomparingittotheglobalmodehaveproventobe Tis work has been subsidized through the Plan of Science, meaningful in characterizing each cluster. Technology and Innovation of the Principality of Asturias Regarding the clusters identifed and based on the dis- (Ref. FC-15-GRUPIN14-132). tribution of samples and the characterization of each one, we can conclude that they are representative, in the sense that each one has its own diferential set of features. It is References remarkable that neither the project size nor the geographical [1]B.Dao,S.Kermanshachi,J.Shane,S.Anderson,andE.Hare, area or the project type is signifcant considering the clusters, “Identifying and Measuring Project Complexity,”in Proceedings soitcanbeconcludedthattheanswersgivenarenotspecifc of the International Conference on Sustainable Design, Engineer- to a particular country or type/size of project. ing and Construction, ICSDEC 2016, pp. 476–482, usa, May 2016. 16 Complexity

[2] K. I. Gidado, “Project complexity: Te focal point of con- [19] M. Bosch-Rekveldt, Y. Jongkind, H. Mooi, H. Bakker, and A. struction production planning,” Construction Management and Verbraeck, “Grasping project complexity in large engineering Economics,vol.14,no.3,pp.213–225,1996. projects: Te TOE (Technical, Organizational and Environmen- [3] S. Kermanshachi, B. Dao, J. Shane, and S. Anderson, “An tal) framework,” International Journal of Project Management, Empirical Study into Identifying Project Complexity Manage- vol.29,no.6,pp.728–739,2011. ment Strategies,” in Proceedings of the International Conference [20]S.Cicmil,T.Cooke-Davies,L.Crawford,andK.Richardson, on Sustainable Design, Engineering and Construction, ICSDEC Exploring the complexity of projects: Implications of complexity 2016, pp. 603–610, usa, May 2016. theory for project management practice,ProjectManagement [4] L.-A. Vidal and F. Marle, “Understanding project complexity: Institute, 2017. implications on project management,” Kybernetes,vol.37,no.8, [21] T. Cooke-Davies, Aspects of complexity: Managing projects in a pp. 1094–1110, 2008. complex world, Project Management Institute, 2011. [5] K. Remington, R. Zolin, and R. Turner, “A model of project [22] C. Ivory and N. Alderman, Can project management learn any- complexity: distinguishing dimensions of complexity from thing from studies of failure in complex systems? Project Manage- severity,” in Proceedings of the 9th International Research Net- ment Institute, 2005. work of Project Management Conference,2009. [23]L.-A.Vidal,F.Marle,andJ.-C.Bocquet,“Measuringproject [6]B.N.Baker,D.C.Murphy,D.Fisher,D.I.Cleland,andW.R. complexity using the Analytic Hierarchy Process,” International King, “Project management handbook,”in Project Management Journal of Project Management,vol.29,no.6,pp.718–727,2011. Handbook New York, pp. 902–919, Van Nostrand Reinhold, 1988. [24]Y.Lu,L.Luo,H.Wang,Y.Le,andQ.Shi,“Measurementmodel [7] D. K. Peterson, C. Kim, J. H. Kim, and T. Tamura, “Te percep- of project complexity for large-scale projects from task and tions of information systems designers from the United States, organization perspective,” International Journal of Project Man- Japan, and Korea on success and failure factors,” International agement,vol.33,no.3,pp.610–622,2015. Journal of Information Management,vol.22,no.6,pp.421–439, [25] A. J. Shenhar, O. Levy, and D. Dvir, “Mapping the dimensions of 2002. project success,” Project Management Journal,vol.28,pp.5–13, [8]T.Huysegoms,M.Snoeck,G.Dedene,A.Goderis,andF. 1997. Stumpe, “A case study on variability management in sofware [26] D. Dvir and T. Lechler, “Plans are nothing, changing plans is product lines: Identifying why real-life projects fail,” Interna- everything: Te impact of changes on project success,” Research tional Journal of Information Systems and Project Management, Policy,vol.33,no.1,pp.1–15,2004. vol. 1, no. 1, pp. 37–48, 2013. [27] T. Cooke-Davies, “Te ‘real’ success factors on projects,” Inter- [9] H. Ernst, “Success factors of new product development: A national Journal of Project Management,vol.20,no.3,pp.185– review of the empirical literature,” International Journal of Man- 190, 2002. agement Reviews,vol.4,no.1,pp.1–40,2002. [28] J. K. Pinto and J. E. Prescott, “Variations in Critical Success Fac- [10] V. R. Montequ´ın, S. C. Fernandez,´ F. O. Fernandez,´ and J. V. tors Over the Stages in the Project Life Cycle,” Journal of Balsera, “Analysis of the Success Factors and Failure Causes Management, vol. 14, no. 1, pp. 5–18, 1988. in Projects,” International Journal of Information Technology Project Management,vol.7,no.1,pp.18–31,2016. [29] D. I. Cleland, “Measuring Success: Te owners viewpoint,” in Proceedings of the 18th Annual Seminar/Symposium,pp.6–12, [11] R. Osei-Kyei and A. P. C. Chan, “Review of studies on the Montreal, Canada, 1986. Critical Success Factors for Public-Private Partnership (PPP) projects from 1990 to 2013,” International Journal of Project [30] L. A. Ika, “Project success as a topic in project management Management,vol.33,no.6,pp.1335–1346,2015. journals,” Project Management Journal,vol.40,no.4,pp.6–19, 2009. [12] J. R. Turner and R. Muller,¨ Te project manager’s leadership style as a success factor on projects: A literature review,Project [31] S. A. Assaf and S. Al-Hejji, “Causes of delay in large construction Management Institute, 2005. projects,” International Journal of Project Management,vol.24, [13] U. Shauchenka, Why Projects Fail. Uladzislau Shauchenka,ISBN no. 4, pp. 349–357, 2006. 0987724207, 2013. [32] A. P. C. Chan, D. Scott, and A. P. L. Chan, “Factors afecting [14]K.G.Cooper,J.M.Lyneis,andB.J.Bryant,“Learningtolearn, the success of a construction project,” Journal of Construction from past to future,” International Journal of Project Manage- Engineering and Management,vol.130,no.1,pp.153–155,2004. ment,vol.20,no.3,pp.213–219,2002. [33] D. W.M. Chan and M. M. Kumaraswamy, “Acomparative study [15] T. Williams, “Assessing and moving on from the dominant of causes of time overruns in Hong Kong construction projects,” project management discourse in the light of project overruns,” International Journal of Project Management,vol.15,no.1,pp. IEEE Transactions on Engineering Management,vol.52,no.4, 55–63, 1997. pp. 497–508, 2005. [34] N.R.Mansfeld,O.O.Ugwu,andT.Doran,“Causesofdelayand [16] S. M. Cousillas Fernandez,´ Valoracion´ y cuantifcacion´ de causas cost overruns in Nigerian construction projects,” International de fracaso y factores de exito´ en proyectos, Universidad de Journal of Project Management,vol.12,no.4,pp.254–260,1994. Oviedo, 2013. [35] M. Sambasivan and Y. W. Soon, “Causes and efects of delays [17] S. M. Cousillas, V.Rodr´ıguez, R. Concepcion,´ and F. Rodr´ıguez, in Malaysian construction industry,” International Journal of “Dentifcation and analysis of causes for project failure and Project Management,vol.25,no.5,pp.517–526,2007. factors for project success in the asturian case,” in Proceedings [36]W.Al-Ahmad,K.Al-Fagih,K.Khanfar,K.Alsamara,S.Abuleil, of the XIV International Congress on Project Engineering,2010. and H. Abu-Salem, “A taxonomy of an IT project failure: root [18] D. Baccarini, “Te concept of project complexity—a review,” causes,” International Management Review,vol.5,p.93,2005. International Journal of Project Management,vol.14,no.4,pp. [37] J. Goedeke, M. Mueller, and O. Pankratz, “Uncovering the 201–204, 1996. Causes of Information System Project Failure,” 2017. Complexity 17

[38] L. J. May, “Major causes of sofware project failures,” CrossTalk, [57] A. Naiem, M. El-Beltagy, and S. Seif, “Project portfolio explo- vol. 11, pp. 9–12, 1998. ration and visualization using self-organizing maps,” in Pro- [39] P. Patanakul, “Managing large-scale IS/IT projects in the public ceedings of the 10th International Conference on Informatics and sector: Problems and causes leading to poor performance,” Te Systems, INFOS 2016, pp. 222–227, egy, May 2016. JournalofHighTechnologyManagementResearch,vol.25,no.1, [58] UNS, International Standard Industrial Classifcation of all pp.21–35,2014. Economic Activities, United Nations Statistical Ofce, New York, [40] W. Belassi and O. I. Tukel, “A new framework for determining NY, USA, 1990. critical success/failure factors in projects,” International Journal [59] PMA - Determine Project Size n.d, March 2018, https://pma of Project Management,vol.14,no.3,pp.141–151,1996. .doit.wisc.edu/size factors.html. [41] C. S. Lim and M. Z. Mohamed, “Criteria of project success: [60] T. Kohonen, “Te Self-Organizing Map,” Proceedings of the An exploratory re-examination,” International Journal of Project IEEE,vol.78,no.9,pp.1464–1480,1990. Management,vol.17,no.4,pp.243–248,1999. [61] J.Vesanto,J.Himberg,E.Alhoniemi,andJ.Parhankangas,“Self- [42] T. Richardson, “Project management pitfalls,” Business Commu- organizing map in Matlab: the SOM Toolbox,” in Proceedings of nications Review,vol.25,p.49,1995. the Matlab DSP conference,vol.99,pp.16-17,1999. [43] I. King, “Te road to continuous improvement: BPR and Project [62] J. Vesanto and E. Alhoniemi, “Clustering of the self-organizing Management,” IIE Solutions,vol.28,pp.22–28,1996. map,” IEEE Transactions on Neural Networks and Learning [44] G. P. Prabhakar, “What is project success: a literature review,” Systems, vol. 11, no. 3, pp. 586–600, 2000. International Journal of Business and Management,vol.3,no.9, [63] J. C. Bezdek and N. R. Pal, “Some new indexes of cluster pp.71–79,2009. validity,” IEEE Transactions on Systems, Man, and Cybernetics, [45] S. M. Qureshi and C. Kang, “Analysing the organizational fac- Part B: Cybernetics, vol. 28, no. 3, pp. 301–315, 1998. tors of project complexity using structural equation modelling,” [64] G. W. Milligan and M. C. Cooper, “An examination of proce- International Journal of Project Management,vol.33,no.1,pp. dures for determining the number of clusters in a data set,” 165–176, 2015. Psychometrika,vol.50,no.2,pp.159–179,1985. [46] S. Shafei-Monfared and K. Jenab, “A novel approach for com- plexity measure analysis in design projects,” Journal of Engineer- [65] D. L. Davies and D. W.Bouldin, “Acluster separation measure,” ing Design,vol.23,no.3,pp.185–194,2012. IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. PAMI-1, no. 2, pp. 224–227, 1978. [47]A.P.C.Chan,D.C.K.Ho,andC.M.Tam,“Designandbuild project success factors: multivariate analysis,” Journal of Con- [66] J. Vesanto, “SOM-based data visualization methods,” Intelligent struction Engineering and Management,vol.127,no.2,pp.93– Data Analysis,vol.3,no.2,pp.111–126,1999. 100, 2001. [67]K.X.Pearson,“Onthecriterionthatagivensystemofdevi- [48] A. J. Shenhar, A. Tishler, D. Dvir, S. Lipovetsky, and T. Lechler, ations from the probable in the case of a correlated system of “Refning the search for project success factors: A multivariate, variables is such that it can be reasonably supposed to have typological approach,” R&D Management,vol.32,no.2,pp.111– arisen from random sampling,” Philosophical Magazine,vol.50, 126, 2002. no.302,pp.157–175,2009. [49]A.Tishler,D.Dvir,A.Shenhar,andS.Lipovetsky,“Identifying critical success factors in defense development projects: A mul- tivariate analysis,” Technological Forecasting & Social Change, vol. 51, no. 2, pp. 151–171, 1996. [50] F. Marle and L.-A. Vidal, “Project risk management processes: Improving coordination using a clustering approach,” Research in Engineering Design,vol.22,no.3,pp.189–206,2011. [51] F.Marle,L.-A.Vidal,andJ.-C.Bocquet,“Interactions-basedrisk clustering methodologies and algorithms for complex project management,” International Journal of Production Economics, vol. 142, no. 2, pp. 225–234, 2013. [52] L. Yang, C. Huang, and K. Wu, “Te association among project manager’s leadership style, teamwork and project success,” International Journal of Project Management,vol.29,no.3,pp. 258–267, 2011. [53] T. Kohonen, “Self-organized formation of topologically correct feature maps,” Biological Cybernetics,vol.43,no.1,pp.59–69, 1982. [54] T. Kohonen, “Learning vector quantization,” in Self-Organizing Maps, pp. 175–189, Springer, 1995. [55] J. V. Balsera, V. R. Montequin, F. O. Fernandez, and C. A. Gonzalez-Fanjul,´ Data Mining Applied to the Improvement of Project Management. Data Mining Applications in Engineering and Medicine, InTech, 2012. [56] S. G. MacDonell, “Visualization and analysis of sofware engi- neering data using self-organizing maps,” in Proceedings of the 2005 International Symposium on Empirical Sofware Engineer- ing, ISESE 2005,pp.115–124,aus,November2005. Hindawi Complexity Volume 2018, Article ID 6058480, 19 pages https://doi.org/10.1155/2018/6058480

Research Article Measuring the Project Management Complexity: The Case of Information Technology Projects

Rocio Poveda-Bautista ,1 Jose-Antonio Diego-Mas ,2 and Diego Leon-Medina3

1 Institute of Innovation and Knowledge Management (INGENIO) (CSIC-UPV), Universitat Politecnica` de Valencia,` Camino de Vera s/n, 46022 Valencia, Spain 2Institute for Research and Innovation in Bioengineering (I3B), Universitat Politecnica` de Valencia,` Camino de Vera s/n, 46022 Valencia, Spain 3Departamento de Proyectos de Ingenier´ıa, Universitat Politecnica` de Valencia,` Camino de Vera s/n, 46022 Valencia, Spain

Correspondence should be addressed to Rocio Poveda-Bautista; [email protected]

Received 27 October 2017; Accepted 27 February 2018; Published 13 May 2018

Academic Editor: JoseRam´ on´ San Cristobal´ Mateo

Copyright © 2018 Rocio Poveda-Bautista et al. Tis is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Complex projects require specifc project management (PM) competences development. However, while no complex projects have standards that are recognized to guide their management, complex projects do not have guides to deal with their complexity. To lead complex projects to success, this complexity must be measured quantitatively and, in our opinion, project management complexity assessment should be based on existing PM standards. In this work, the main project complexity assessment approaches based on PM standards are analyzed, observing that International Project Management Association (IPMA) approach is the closest to a tool that can be used as a complexity quantitative measurement system. On the other hand, several authors have shown that the inherent complexity of specifc kind of projects must be measured in a particular way. Te main objective of this research is to propose a project management complexity assessment tool for IT projects, providing a Complexity Index that measures the impact that complexity factors inherent to IT projects have under a specifc complexity scenario. Te tool combines the use of complexity factors defned by IPMA approach and the use of complexity factors found in the literature to manage inherent complexity of IT projects. All these factors were validated by expert survey and the tool was applied to a study case.

1. Introduction management systems developed for noncomplex or moder- ately complex projects in complex projects. As Shenhar [5] Although complexity theory was applied to project man- noted, diferent types of projects require diferent managerial agement in the 90s, the discipline of complexity project approaches. management was unofcially launched at the 20th Inter- What does complexity means in project management? national Project Management Association World Congress A complete review has been recently published which sum- in Shanghai [1]. Following Bosch-Rekveldt et al. [2] who marizes the development of this concept and the factors argue that specifc complexities in projects might require spe- that afect project management complexity [6]. Some other cifc competence development, inherent complexity within authors have developed similar work prior to this one [7, 8]. projects must be studied in a particular way. In this sense, it Manystudiesanddefnitionshavebeendevelopedon is worth highlighting what Williams [3] pointed out, which complexity; however, most of these studies focus only on the indicates the increase in the complexity of projects as one of conceptual framework of project complexity and few of such the main causes of project failure. studies have focused on projects complexity measurement In 1991, Bennett [4] already noted the need for an systems. Moving a step beyond the defnition of a conceptual exceptional level of management in complex projects, as well framework of project complexity, the works of some authors as the inadequacy of the implementation of conventional have suggested factors that afect the complexity in projects, 2 Complexity such as Geraldi et al. [7], Bosch-Rekveld et al. [8], and considered uncertainty of objectives and uncertainty of Chapman [9], which are based on a systematic review of the methods used to achieve the projects goals as important fac- literature, Vidal and Marle [10] that proposed a complexity- tors of project complexity. Wood and Ashton [38] proposed driven approach of project management to assist project a similar complete framework. Later Bosch-Rekveld et al. management in decision making defning a complexity- [8] added environmental complexity to Baccarini proposal based criteria, and Ireland et al. [11] that classifed projects giving rise to their TOE (technological, organizational, and (simple, complicated, and complex projects types A, B, and environmental) framework and, in a similar way, Geraldi el al. C) according their complexity and taking into account the [7] highlight structural complexity, uncertainty, and sociopo- systems of systems view. Tese last ones proposed diferent litical elements. Williams [39] complemented this complexity leadership styles and management tools for each type of defnition with other main aspects of complexity: the number project. and interdependence of elements and the uncertainty in goals Tese studies mentioned above have focused on iden- and means. Dunovic´ et al. [40] consider constrains as a third tifying project complexity factors and have built a frame- primary element of a new model of complexity, in addition to work that describes project complexity qualitatively [12, structural complexity and uncertainty. 13]. Despite the fact that several authors have focused on Otherworkssuchas[5,41–43]thatfocustheirresearch measuring project complexity quantitatively [13, 14], besides on structural complexity and uncertainty can be integrated this and in our point of view, these project management in the PMI approach. complexity assessment systems should be based on existing From the perspective of structural complexity and uncer- project management (PM) standards to ensure implemen- tainty, Project Management Institute [44] published Navi- tation of recognized competencies and practices in man- gating Complexity: A Practice Guide as a proposal providing aging complex projects. While no complex projects have guidelines to project managers in order to perform a check standards supported by various bodies of knowledge (BoK), of the status of the project assessment in terms of complexity. Project Management Institute (PMI), International Project Trough a questionnaire for which the answer is afrmative ManagementAssociation(IPMA),andtheAssociationfor or negative a scenario of complexity in which the project is Project Management (APM), complex projects do not yet located can be implied. have a BoK to guide their development. However, some of these standards try to evaluate projects in order to (ii) IPMA Approach. It is based on Crawford-Ishikura Factor assess their complexity and to look at project managers’ Table for Evaluating Roles and focused on measures of competence development in the view of complexity. Tese competence development in complex project management by standards capture the diferent perspectives on complexity the project manager through complexity factors. and encompass factors that contribute to project complexity Te frst standard measurement tool for complexity in considering diferent approaches. project management was developed by the Global Alliance Te objective of this research is to propose a project for Project Performance Standards (GAPPS) whose approach management complexity assessment tool in Information characterises projects based on the management of their Technology (IT) projects based on an adequate existing PM complexity. Te framework developed by GAPPS used a tool standard in order to measure the specifc complexity in called Crawford-Ishikura Factor Table for Evaluating Roles the management of this type of projects. To such end, the (CIFTER). Tis tool is used to diferentiate project manager presentworkisdevelopedfollowingthefollowingsections: roles based on the complexity of managed projects. Te in Section 2, we shall study the main project complexity developmentofsuchstandardwascarriedoutbymembers assessment approaches based on PM standards as well as of the GAPSS [45]. CIFTER identifes seven factors that afect the complexity in IT projects, and the methodology followed complexity project management. to develop a tool for assessing the complexity of IT project As an assessment model of complexity project manage- management will be presented. In Section 3 the proposal ment,IPMAhasdevelopedanimplementationguideforthe of an IT project management complexity assessment tool is assessment criteria, which transfers and adapts the CIFTER presented and in Section 4 this tool is applied to a study case. model to objectively demonstrate the degree of competence Finally, in Section 5, the conclusions and limitation of this of project managers in complexity project management. Such study are shown. adaptation was made under ICRG (IPMA Certifcation and Regulations Guidelines) [15]. Te model suggests ten factors 2. Materials and Methods for the assessment of complexity. Table 1 shows the diferent factors with the description 2.1. Project Complexity Assessment Approaches and the criteria taken into account for the IPMA project Based on PM Standards management complexity assessment. Tis scheme is used to assess the project management (i) PMI Approach. It focuses on structural complexity and complexity. Each indicator is rated according to four levels uncertainty issues. of complexity: very high complexity (4), high complexity (3), Tisapproachisbasedontwomainperspectivesabout low complexity (2), and very low complexity (1). structural complexity proposed by Baccarini [36], organi- On the other hand, the Association for Project Man- zational complexity and technological complexity, and the agement (APM) considers, in the Registered Project Profes- perspective proposed by Turner and Cochrane [37] that sional Candidate Guidance [46], that a project is considered Complexity 3 low limited sequential few parties few, simple few conficts few relations homogeneous low percentage Low Complexity available, known quite transparent quite independent public interest low close, concentrated repetitive approach comparable interest small, easy to handle uniform, well known constant and uniform low, monodimensional few uniform categories much support available few important decisions uni-dimensional, simple many options for actions uni-dimensional, common few and well known relations low potential of opportunities common standards applicable known and proven technology direct, not demanding, uniform high large hidden diverse many diferent many conficts high percentage High Complexity large, demanding numerous parties divergent interests unknown relations distant, distributed large public interest uncertain, changing no support available very interdependent innovative approach numerous, manifold adaptive and variable unknown technology multicutural, unknown large, multidimensional intensive mutual relations many important desicions overlapping, simultaneous limited options for actions indirect, demanding, manifold large potential of opportunities multidimensional, comprehensive few common standards applicable multidimensional, matrix structure many investors and kinds of resources one investor and few kinds of resources large (relative to project of the same kind) low (relative to project of the same kind) Social span Team structure dynamic team structure static team structure Cultural variety Leadersship style Capital investment Financial resources Diversity of context Demand of creativity Number of interfaces many few Structuring of phases Geographic distances Conficting objectives Hierarchical structure Demand for reporting Availability of support Scope for development Mandate and Objective uncertain, vague defned, obvious Application of standards Demand of coordination demanding, elaborate simple, straighforward Number of sub-ordinates many, large control span few, small control span Interested parties, lobbies Potential of opportunities Stakeholder interrelations Description of the criteria Categories of stakeholders Decision-making processes Interests of involved parties Structures to be coordinated numerous structures few structures Demand for communication Signifcance on public agenda Quantity and diversity of staf Interdependence of objectives Number and assessment of results Availability of people, material, etc. Technological degree of innovation Table 1: IPMA Complexity assessment System for IPMA B (IPMA 4-L-C) [15]. Options for action to minimise risks Variety of methods and tools applied Proportion of PM to total project work Risk probability, signifcance of impacts high risk potential, large impact low risk potential, low impact Predictability of risks and opportunities low, uncertain high, quite certain Relations with permanent organisations Transparency of mandate and objectives Criteria (1) Objectives, Assessment of Results (2) Interested Parties, Integration (3) Cultural and social context (4) Degree of innovation, general conditions (5) Project structure, demand for coordination (6) Project organisation (7) Leadership, teamwork, decisions (8) Resources incl. fnance (9) Risk and opportunities (10) PM methods, tools and techniques 4 Complexity

“complex project” if it was highly rated in the following and solutions evolve over time according to the need of the indicators/criteria (not in priority order): objectives, assess- project. Te last ones are those that sequentially follow the ment of results, interested parties, integration, cultural and phases and deliverables of the project. social context, degree of innovation, general conditions, Other studies on IT projects go a step further by doing project structure, demand for coordination, project organi- a root cause analysis and identifying factors which can be zation, leadership, teamwork, decisions resources (including attributed to failure of IT projects [21, 30, 51]. Some of fnance), risks (threats and opportunities), and project man- these factors are characteristic of observed tendencies in agement methods, tools, and techniques. APM provides a project with high extent of complexity. Project managers and project complexity questionnaire to help project managers to researchers have attributed IT projects unsuccessful to the know if they are working on projects considered complex. complexity of such projects [21] and propose the use of agile It may therefore be stated that APM follows an approach organizations and reduce the complexity to achieve success similar to that of IPMA, and its assessment of complexity is in IT projects. performedfollowingthesameprocedure. Afer a systematic review of the literature, Table 2 sum- marizes the factors inherent to the complexity of IT projects, (iii) Te Complex Project Manager Competency Standards specifc to IT sector, and diferent from the complexity factors Approach. Te International Centre for Complex Project used by standard tools to assess any other type of projects. Management of Australian Government develops in 2012 Tese factors have been extracted from studies on project the Complex Project Manager Competency Standards [47]. failure characteristics, abandonment factors, risk factors, and Tis standard defnes a methodology for the assessment of project factors that afect IT development projects. Ten, complexity and classifcation of projects based on their com- these factors were grouped according to the IPMA project plexity and provides tools to categorize projects by their types management complexity assessment criteria that best defne of systems, determine the strategy and appropriate contracts them. for the project, and select competent project managers. As conclusions of the literature review, it is important to Tis approach is not without criticism, since it is a stan- mention that one of the main factors that afects the success dard that, on the one hand, has not satisfactorily established of IT projects is related to the user involvement in project any measure to assess complexity and, on the other, has development. On the other hand, an adequate sponsorship of been used as a requirement for project managers to establish the executive management is also important. Another critical contracts and subcontracts with the Australian government factor is requirements; most projects usually begin with a [1]. clear vision and objectives, but sometimes the requirements of IT projects are based on product iterations; therefore 2.2. Complexity in IT Projects. As we have progressed in it is imperative to increase realistic expectations of project the study of complexity in projects several existing project stakeholders to ensure project success. management standards have been recognizing the need for Resources and skills are key factors in the success of an exceptional level of management in complex projects. the project as well as how new technologies work when In the same way, there is a need for specifc competence applied (sometimes technologies are not mature enough to development in specifc complexities in projects. In this be implemented). sense, several authors have developed measurement methods Several studies confrm that iterative and agile meth- of project complexity taking into account diferent frame- ods (project life cycles) have more success than traditional works in specifcs kinds of projects, such as large engineering approaches. Terefore, the methodology used on project projects [8], large infrastructure projects [48], construction management must be present on any assessment of project projects [49], and design projects [50]. However, there are complexity. no reported researches that focused on the conceptualization of the IT project complexity construct and studies have not 2.3. Methodology. From all the approaches used by the been found to deepen the complexity of the Information diferent recognized standards in project management, IPMA Technology (IT) industry. approachistheclosesttoatoolthatcanbeusedasacomplex- While any industry is exposed to project failure, IT ity quantitative measurement system, since it defnes factors industry shows being more vulnerable to risk and failure than and suggests a measurement scale to measure the degree to other industries. A number of areas related to project risk which these factors afect the management complexity of the management and project failure provide useful study bases project.Whileitispartoftheprojectmanagercertifcation to defne IT project complexity. system, this tool is useful for measuring complexity in Tera are many studies around IT projects; however projects as it attempts to confrm that the project manager is the years of experience of Standish Group developing the capable of managing complex projects. CHAOS report are known. As mentioned in the Standish For the purpose of our study, a framework for IT Group CHAOS Report [20] made on 50000 IT projects, over project management complexity assessment is designed. Tis 20%ofprojectsfailedorwerecancelled.Tisreportshows framework has as baseline the IPMA project management that large projects have less chance of success than small ones. complexity assessment, adding or removing (if necessary) On the other hand, agile or iterative development projects some complexity factors in order to build an assessment have more chance of success in comparison with waterfall or template for IT projects. Te proposed factors to be included incremental projects. Te frst are those whose requirements on the assessment of IT project complexity were extracted Complexity 5

Table 2: IT Projects Complexity factors (literature review).

Group of factors Factors References Clearstatementsofrequirements [16–22] Objectives, requirements Realistic expectations [20, 20–24] and expectations Clear strategic objectives [18–20, 25] Uncertain and changing [16–20, 23] regulatory requirements User involvement [17, 18, 20, 21, 24, 26] Interested Parties, Executive management support [18–20, 23–25] Integration Project Sponsor Committed with [18, 19, 23, 27, 28] project methodology Team motivated by the project [17, 18, 26, 29] HardWorking,focusedstaf [17,18,20,26,29] Near shore/of shore teams Leadership, teamwork, [17, 18, 26, 29] decisions involved Ofshore/near shore teams are familiar with technical and [17, 18, 24, 26, 29] business aspects of project PM methods, tools and Incremental or iterative [24, 30] techniques methodology used in the project Incompetence on using/applying [20, 23] Technology New Technologies [18, 20, 21, 26, 28, 31, 32] Technology IT Management Support [20, 24] TechnologyIlliteracy [18,20,26,28,31,32] Infrastructure, [23, 24, 33–35] Telecommunication Constraints considering the literature review carried out in section before included in the template developed to design the complexity and taking into account the fact that IPMA complexity factors assessment tool (see columns 1 and 2 of Table 2). do not focus exclusively on the scope of IT projects but cover a wider range of projects. 2.3.2. Survey Design. Te objective of the survey was to ask To propose an assessment template in order to build a the experts to identify the main complexity factors that afect tool that measures IT project complexity taking into account IT development projects. Te sections below will describe the inherent complexity of these projects, frst, some IT more in detail the structure of the questions and their complexity factors that are not covered by IPMA assessment objectives. will be added to the baseline IPMA assessment knowledge and, then, this template was validated by experts. Tis Experts Profle Questions. Tese questions were designed tool was called Complexity Index tool because it will allow to know the experts’ profle: industrial sector of the IT measuringthecomplexitylevelofaprojectatonepointunder practitioner, years of experience working on IT projects, and a scenario of concrete project complexity. professional profle. Tis proposal was validated with a survey fulflled by experts. Te selection of the experts was an important issue. Questions Related to Complexity Groups.Questionswere We were looking for IT specialists with deep experience raised about complexity groups, to fnd out those which were working in IT projects: IT chiefs technology ofcer, IT considered by experts as the ones which impacted the most project/program managers, project team members, end users, the complexity of the project. and practitioners with enough expertise and knowledge on IT Te complexity groups were assessed using a 5-point sector. Tese experts were involved in the survey under the Likert type scale. below channels: personal contact and social networks. Questions Related to Complexity Factors within Each Group. 2.3.1. Selection of IT Project Management Complexity Factors. Te experts were asked about the groups in which new We considered all the factors found in the literature as complexity factors were added. Tese are objectives, require- relevant factors in the complexity of IT projects (see Table 2), ments and expectations, interested parties and integration, since all these factors were identifed by several authors as leadership, teamwork and decisions, PM methods, tools and inherent to the complexity of these projects. All of them were techniques, and technology. 6 Complexity

Aerospace, defence & security 5.41%

Technology 40.54% Banking & capital markets 21.62%

Capital projects and infrastructure 2.70% Communications 2.70% Engineering & construction 2.70%

Financial services Other 13.51% 5.41% Industrial manufacturing 2.70% Hospitality & leisure 2.70% Aerospace, defence & security Financial services Banking & capital markets Hospitality & leisure Capital projects and infrastructure Industrial manufacturing Communications Other Engineering & construction Technology

Figure 1: Profle of the respondents. Industrial sector (%).

We considered that the other groups of complexity factors >1 should be part of the tool since these are composed of 5 years 21.62% complexity factors that afect any type of project. 5–10 years Terefore, questions related to complexity factors are, all, 32.43% based on allowing practitioner to rank the factors within their complexity group and discard them if required. Tese 0 years questions were used to fnd out which of these factors 2.70% contributedmosttothecomplexitywithinitsgroup(relative complexity). Te complexity factors were assessed using a 5-point Likert type scale. 1–5 years 13.51% 2.3.3. Survey Results. Of the total number of people that accessed the survey, 13 were not fully completed, so there were 37 responses in the end. Of these 50 responses, it was necessary to eliminate the thirteen partial answers since the 10–15 years 29.73% study must be done with comparable items. 52% of the responses were answered from Spain, 22% Figure 2: Profle of the respondents. Years of experience (%). from Colombia, 6% from United States, and 19% from more than 9 diferent countries. According to the results shown in Figure 3, about 57% of the respondents had management profles. Te remaining Profle of the Respondents. Most of the practitioners are percentage of experts is part of projects with a more technical from technology, banking, and fnancial services sectors (see and business oriented role. Afer screening the profles, we Figure 1). selected all the experts to participate in the survey. According to the results shown in Figure 2, more than 50% of the respondents had more than 10 years of experience Complexity Factors Results. In this part, the specifc sur- working on IT projects, and almost the one-third part vey questions related to complexity factors were analyzed. had at least 5–10 years. Only 1 respondent had 0 years of Trough a brief analysis and taking as an example some experience working on IT projects, since it was an end user, answers from the survey (see Tables 3, 4, and 5), the impact the answer was considered valid for the study. Te results in of the new complexity groups/factors added was studied. terms of level of expertise suggest that the experts were well Table 3 shows the information of the percentages of the qualifed. total of the answers and the average column that was used to Complexity 7

Business specialist Subject matter expert 2.70% 2.70% Chief technology ofcer 2.70% Director 5.41% End user 2.70%

Project manager 29.73%

IT specialist 29.73%

Program manager 10.81%

Other Manager 5.41% 8.11% Business specialist Manager Chief technology ofcer Other Director Program manager End user Project manager IT specialist Subject matter expert

Figure 3: Professional profle of the respondents (%).

Table 3: Complexity groups, importance in IT projects.

Total 1 2 3 4 5 Average Responses 3 2 2 5 25 Objectives, Requirements, Expectations 4.27 37 8.11% 5.41% 5.41% 13.51% 67.57% 1 1 7 12 16 Interested Parties, Integration 4.11 37 2.70% 2.70% 18.92% 32.43% 43.24% 3 3 22 8 1 Cultural and social context 3.03 37 8.11% 8.11% 59.46% 21.62% 2.70% 1 5 12 15 4 Degree of innovation, general conditions 3.43 37 2.70% 13.51% 32.43% 40.54% 10.81% 3 3 7 19 5 Project structure, demand for coordination 3.54 37 8.11% 8.11% 18.92% 51.35% 13.51% 4 3 5 14 11 Project organisation 3.68 37 10.81% 8.11% 13.51% 37.84% 29.73% 3 3 3 15 13 Leadership, teamwork, decisions 3.86 37 8.11% 8.11% 8.11% 40.54% 35.14% 2 3 12 9 11 Resources incl. fnance 3.65 37 5.41% 8.11% 32.43% 24.32% 29.73% 1 2 9 17 8 Risk and opportunities 3.78 37 2.70% 5.41% 24.32% 45.95% 21.62% 3 6 12 10 6 PM methods, tools and techniques 3.27 37 8.11% 16.22% 32.43% 27.03% 16.22% 2 6 13 9 7 Technology 3.35 37 5.41% 16.22% 35.14% 24.32% 18.92% 8 Complexity

Table 4: Factors within Objectives, Requirements and Expectations, relative complexity ranking.

Does not Total 12345 Average apply Responses 1 0 4 15 16 1 Mandate and objective uncertain, vague 4.25 37 2.70% 0.00% 10.81% 40.54% 43.24% 2.70% 1 5 4 17 8 2 Many conficting objectives 3.74 37 2.70% 13.51% 10.81% 45.95% 21.62% 5.41% 0 1 5 19 11 1 Hidden mandate and objectives 4.11 37 0.00% 2.70% 13.51% 51.35% 29.73% 2.70% 0 6 13 9 8 1 Very interdependent objectives 3.53 37 0.00% 16.22% 35.14% 24.32% 21.62% 2.70% Large number of objectives and 1 4 12 13 6 1 3.53 37 multidimensional assessment of results 2.70% 10.81% 32.43% 35.14% 16.22% 2.70% 1 3 1 7 25 0 Unclear requirements 4.41 37 2.70% 8.11% 2.70% 18.92% 67.57% 0.00% 0 2 6 15 11 3 Expectations unlikely to be achieved 4.03 37 0.00% 5.41% 16.22% 40.54% 29.73% 8.11% Strategic Objectives (organizational) uncertain, 0 0 10 18 8 1 3.94 37 vague 0.00% 0.00% 27.03% 48.65% 21.62% 2.70% Uncertain and changing regulatory 2 1 8 12 11 3 3.85 37 Requirements 5.41% 2.70% 21.62% 32.43% 29.73% 8.11%

Table 5: Factors within Technology, relative complexity ranking.

Does not Total 12345 Average apply Reponses Incompetence on using/applying 1 3 6 10 16 1 4.03 37 Technology 2.70% 8.11% 16.22% 27.03% 43.24% 2.70% 1 9 9 15 3 0 Too many new technologies in place 3.27 37 2.70% 24.32% 24.32% 40.54% 8.11% 0.00% 2 1 7 15 12 0 No IT management support 3.92 37 5.41% 2.70% 18.92% 40.54% 32.43% 0.00% 1 4 16 11 5 0 Stakeholders technology illiteracy 3.41 37 2.70% 10.81% 43.24% 29.73% 13.51% 0.00% Many Infrastructure, 1 4 12 17 3 0 3.46 37 Telecommunication Constraints 2.70% 10.81% 32.43% 45.95% 8.11% 0.00% compare between complexity groups. Tese average ratings Table 6: Summary of Questions related to complexity groups. showed the relative importance of each complexity group to the experts. Complexity Average Ten, a survey’s example of the questions related to Objectives, Requirements, Expectations 4.27 complexity groups is shown. Interested Parties, Integration 4.11 Leadership, teamwork, decisions 3.86 Questions (groups of factors). “Please rank each complexity group from 1 (the least complex group of factors) to 5 (the Risk and opportunities 3.78 most complex group of factors) considering its impact on IT Project organization 3.68 projects.” Resources incl. fnance 3.65 Table 6 summarizes the average ratings of each complex- Project structure, demand for coordination 3.54 ity group (rating range from “1, the least complex factor,” to Degree of innovation, general conditions 3.43 “5, the most complex factor”). Technology 3.35 Similar analysis was made within each complexity group in order to classify the factors that make up each group. PM methods, tools and techniques 3.27 A survey’s example of the complexity group “objectives, Culturalandsocialcontext 3.03 requirements, and expectations” is shown in order to clarify theprocedureappliedinthesurveytovalidatetheproposal.

Questions (factors of complexity group “objectives, require- factorfrom1(theleastcomplexfactor)to5(themostcomplex ments, and expectations”). “Please rank each complexity factor) considering its impact on IT projects. Please mark Complexity 9

Table 7: Summary of Questions related to complexity factors within each group.

Order Relative Complexity of Complexity Factors Average (1) Unclear requirements 4.41 (2) Mandate and objective uncertain, vague 4.25 (3) User uncommitted with the project 4.19 (4) Hidden mandate and objectives 4.11 (5) Dispersed team, not focused 4.05 (6) Expectations unlikely to be achieved 4.03 (7) Executive management uncommitted with the project 4.03 (8) Incompetence on using/applying Technology 4.03 (9) Little motivation of the project team 3.95 (10) Strategic Objectives (organizational) uncertain, vague 3.94 (11) No IT management support 3.92 (12) Numerous interested parties and lobbies 3.91 (13) Ofshore/Near shore teams are not familiar with technical and business aspects 3.91 (14) Uncertain and changing regulatory Requirements 3.85 (15) Divergent interest of involved parties 3.78 (16) Many conficting objectives 3.74 (17) Unknown stakeholders interrelations 3.71 (18) Sponsor uncommitted with project methodology 3.64 (19) No assistance to project management available 3.57 (20) Numerous/manifold, variety of methods and tools applied 3.56 (21) Very interdependent objectives 3.53 (22) Large number of objectives and multidimensional assessment of results 3.53 (23) Many Infrastructure, Telecommunication Constraints 3.46 (24) Few common standards applicable 3.44 (25) Stakeholders technology illiteracy 3.41 (26) Many diferent categories of stakeholders 3.36 (27) Dynamic team structure 3.34 (28) High percentage/proportion of PM work from total project work 3.33 (29) Too many new technologies in place 3.27 (30) Many sub-ordinates, large control span 3.19 (31) Totally Iterative methodology used 3.19 (32) Many important decisions in place 3.15 (33) Adaptive and variable leadership style 3.14 (34) Ofshore teams/Near shore teams involved 2.91

‘Doesnotapply’ifyouthinkthatthisfactoritisnotapplicable From the results obtained in the survey it can be con- to IT projects.” cluded that all factor groups and all factors within each group Moreover, the results of the survey obtained for the new should be included in the complexity measurement tool, IT complexity group “Technology” are shown in Table 5 in since the experts thought that they are factors that afect the order to know the relevance of its factors in comparison with complexity of IT projects. factors of other complexity groups. Most of the new IT complexity factors suggested are in the frst half of the relative complexity ranking. As shown in 3. The Complexity Index Tool in IT Project the analysis of survey results, there are no factors considered Management Complexity Assessment: outofthescopeoftheComplexityIndextool.Temaximum Description, Interface, and Functionality value of the responses indicating that this factor does not applywascloseto10%,whichisnotconsideredbythe Tis section describes the tool and the functionality that is researchers sufcient to remove them from the tool. provided. Table7showstogetherallthefactorsstudiedinthesurvey Te tool was designed taking into account all the new and ranked by level of complexity. complexity factors extracted from the literature and validated In Table 7 the new IT complexity factors proposed are by experts. A table with all the factors to be included in the shown in italics. assessment tool is presented as shownin Table 8. 10 Complexity large many hidden diverse the project methodology many diferent many conficts uncertain, vague High Complexity large, demanding numerous parties divergent interests unknown relations distant, distributed uncertain, changing large public interest very interdependent innovative approach numerous structures Unclear requirements unknown technology demanding, elaborate large, multidimensional multicultural, unknown intensive mutual relations overlapping, simultaneous indirect, demanding, manifold User uncommitted wth the project Sponsor uncommitted with project multidimensional, comprehensive Expectations unlikely to be achieved multidimensional, matrix structure Executive management uncommitted to few project project limited sequential few parties methodology few conficts few relations homogeneous few structures defned, obvious available, known Easily achievable Low Complexity quite transparent quite independent public interest low close, concentrated repetitive approach comparable interest small, easy to handle uniform, well known low, monodimensional few uniform categories simple, straightforward uni-dimensional, simple uni-dimensional, common Requirements perfectly clear few and well known relations known and proven technology Sponsor committed with project direct, not demanding, uniform User available and committed to the Executive management committed to the Table 8: Complexity Index tool Draf. Social span methodology Requirements Cultural variety User Involvement Diversity of context Demand of creativity Number of interfaces Realistic Expectations Structuring of phases Geographic distances Conficting objectives Hierarchical structure Demand for reporting Scope for development Mandate and Objective defned, obvious uncertain, vague Demand of coordination Interested parties, lobbies Stakeholder interrelations Description of the criteria Categories of stakeholders Interests of involved parties Structures to be coordinated Demand for communication Signifcance on public agenda Interdependence of objectives Executive Management Support Clear Statement of Requirements Number and assessment of results Uncertain and changing regulatory Technological degree of innovation Project Sponsor committed with project Relations with permanent organisations Transparency of mandate and objectives Clear Strategic Objectives (organizational) Criteria (1) Objectives, Requirements and Expectations (2) Interested Parties, Integration (3) Cultural and social context (4) Degree of innovation, general conditions (5) Project structure, demand for coordination (6) Project organisation Complexity 11 high Many low, uncertain Dispersed team high percentage Little motivation High Complexity project chain links uncertain, changing aspects of the project no support available numerous, manifold adaptive and variable dynamic team structure many, large control span many important decisions limited options for actions No IT management support large potential of opportunities high risk potential, large impact Stakeholders technology illiteracy Totally Iterative methodology used few common standards applicable Too many new technologies in place Technological Incompetence in any of the Teams unfamiliar with business/Technical Ofshore teams/Near shore teams involved literacy low Few teams few, simple Focused team low percentage Domestic teams Low Complexity available, known Highly Motivated project chain links high, quite certain static team structure constant and uniform few, small control span much support available few important decisions many options for actions Full IT Management support Well known technologies used low risk potential, low impact low potential of opportunities common standards applicable Stakeholders technology Technological competence in all of the Totally Incremental Methodology used Good know how in ofshore/near shore one investor and few kinds of resources many investors and kinds of resources low (relative to project of the same kind) large (relative to project of the same kind) Table 8: Continued. project Technology Constraints in the project Team structure Leadership style New Technologies Capital investment Financial resources Technology Illiteracy Availability of support IT Management Support Application of standards Number of sub-ordinates Potential of opportunities Description of the criteria Decision-making processes adWrig FocusedHard-Working, Staf Team motivated by the project Quantity and diversity of staf Incompetence on using/applying Infrastructure, Telecommunication Near shore/Ofshore teams involved Availability of people, material, etc. with technical and business aspects of Options for action to minimise risks Variety of methods and tools applied Ofshore/Near shore teams are familiar Proportion of PM to total project work Risk probability, signifcance of impacts Predictability of risks and opportunities Incremental or iterative methodology used Criteria (7) Leadership, teamwork, decisions (8) Resources incl. fnance (9) Risk and opportunities (10) PM methods, tools and techniques (11) Technology 12 Complexity

Table 8 gathers “low complexity” and “high complexity” (7) User input bar: another option to slip within the rank values for each factor. Please note that “high complexity” of complexity measure value is the one used to formulate the survey questions (see (8) Read only: graph of the complexity group. Tables 4 and 5). Te next step to develop was how to really measure the 3.2. Assessment Result. Te bottom part of the interface’s tool complexity based on the items shown in Table 8. IPMA shows a graph which provides to the user the result of the complexity assessment considers a project complex if the assessment (Complexity Index score), advising fnally if the measure of complexity reaches a complexity level of 62,5%. project under evaluation is complex or the practitioner has Following IPMA framework, complexity is measured skills to drive complex projects. against that of similar projects in the singular professional Figure 6 is an example of the assessment result. environment of the project manager, scoring each complexity At the same time that the user changes the values of the factor. IPMA claims that a project can be considered complex assessment of any complexity group, the graph will refect the when the average of the assessments’ score of the 10 groups of new Complexity Index value. complexity factors that make up the general assessment tool is higher than 2.5. Tis value was chosen because it is between 4. Study Case low complexity (2) and high complexity (3) (see Section 2.1). Tus, the minimum score obtained for the evaluation of In order to validate the tool, it was applied to assess the any project considered complex should be 25. Tis supposes complexity of an IT project. 62.5% of the maximum value of the complexity measurement. Tis study case is divided into two parts; the frst one Tis maximum value would be 40 if all the factors that is the description of the taxonomy of the IT project with contribute to the project’s complexity are assessed with the the objective of understanding what is the starting point of highest score (4) [52]. the assessment. Te duration of the project was two years; Terefore, the Complexity Index tool is based on the same thus two scenarios were considered: 2015 scenario and 2016 score to defne the Complexity Index. Te new tool is focusing scenario (slight diferences between these will be recognized on the measure of the complexity of IT projects including new in Section 4.1.1). specifc factors in calculation of the Complexity Index. Te second part is the application of the Complexity Complexity Index tool is based on the complexity groups Index tool to the project in 2015 and 2016 status. Te and factors validated by the experts as a result of the survey assessment was performed by the project manager of this (see Table 8). project during these 2 years, who led this project towards Te interface of the designed tool is shown in Figure 4. success despite the complex environment. Te template proposes assessing the complexity groups according to the 4 defned levels of complexity, considering 4.1. Project Overview. Te project analyzed in the study the level of complexity of the factors that make up each group. case is a sofware development project in a banking sector Te sum of the normalized values of each complexity group company, with maintenance and support operations. Te provides the fnal score of complexity of the project under project could be described as a consolidation layer of infor- evaluation. mation from external systems, which will adjust the data to a Te next section will describe the functionality of the tool predefned standard format, in order to report risk measures for the complexity group assessment. to downstream systems. Te project is part of a wholesale banking system and 3.1. Complexity Group Assessment Functionality. Te example customers are all over the world. Diferent suppliers are shown in Figure 5 shows how a user of the tool can measure subcontracted to handle diferent aspects of the project. the complexity of a particular complexity group. Te project organization chart is shown in Figure 7. Te numbers in Figure 5 colored in green indicate the felds described below: Locations. Human resources and the three main suppliers of the project were allocated in 5 countries. Project management Field type: meaning was located in United Kingdom. (1) Read only: complexity group description Stakeholders. From the point of view of the main stakeholders (2) Read only: criteria (complexity factors) within the oftheproject,wecoulddescribeitsinfrastructureasa complexity group synthesis of 36 upstream systems and 4 support systems that (3) Read only: 4 levels of complexity from very low to very provide reference data and another 10 downstream systems to high which risk measures need to be reported. On the other hand, (4) Read only: description of what represents a very low there were audit systems reviewing the project; therefore, value some ad hoc audit teams asked for new requirements. It is important to mention that each system is an external (5) Read only: description of what represents a very high team, with its own organization chart. In many cases, these value organizations are working with ofshore teams. Terefore, (6) User input feld: user feld to rank complexity of a communication matrix between stakeholders is not easy to group build. Complexity 13

Figure 4: Complexity Index tool interface.

Figure 5: Complexity group functional description.

Figure 8 provides a brief overview of the project rela- (iv) Slow learning curve due to the amount of components tionships and dependencies within stakeholders’ information of the project fows.

Constraints 4.1.1. Diferences between 2015 and 2016. In2016,somedown- stream systems were integrated into the project; therefore, (i) Communication issues: very complex communica- deliverables were more complex. By the end of 2015, there tion matrix wasaninitiativetomovesometasksoftheprojectto (ii) Cross-national cultural and legal diferences ofshore teams. Aferwards, there was a transition phase and (iii) High rotation of team members knowledgetransfertoworkwiththem. 14 Complexity

Assessment: project is not complex enough/PM does not require adequately drive dynamic structures of the teams that were complex project management competences built depending on the level of requirements. Figure 10 shows the 2015 assessment results. Complexity Index score was 56,82%. Terefore, according to the defnition of the Complexity Index tool, the project cannot be considered complex. As a reminder, the minimum 1 Complex, 25.00 Simple, 75.00 defned is 62,5%. Assessment 4.2.2. 2016 Project Complexity Assessment Results. Figure 11 shows the result of IT project manager’s assessment of the project in 2016. 0 20 40 60 80 100 According to the project manager’s assessment, we can (%) observe the following.

Figure 6: Complexity Index tool assessment result graph. VeryHighComplexityGroups.Tegroupsassessedwithavery high level of complexity were the same as those in the 2015 project complexity assessment. 4.2. Complexity Index Methodology: A Case of Study Applied to a Sofware Project. Te Complexity Index tool was applied High Complexity Groups.Tegroupsassessedwithavery to the study case assessing the complexity of this IT project. In high level of complexity were the same as those in the 2015 this section, the profle of the project manager was explained project complexity assessment. However, there was one new and then the results obtained comparing 2015 assessment group under this complexity level: “objectives, requirements, with 2016 assessment were exposed. and expectations”; since the project expands in scope, with Te project manager who assessed the complexity of the the inclusion of downstream systems as part of the project, it project has more than 15 years working on IT projects. He was more difcult to manage this group in 2016. More stake- has Ph.D. in chemistry and is PMI-certifed. He is an IT holders also meant that project expectations were slightly professional with consulting experience and he worked as changed. On the other hand, strategic expectations about industry leader across public and private sectors, including changes in scope were increasing pressure on the project by managing multimillion budgets and complex program orga- top management. nizations. He was contacted by email and how the Complexity Figure 12 shows the 2016 assessment results. Index tool works but without any interference on performing Complexity Index score was 63,64%; therefore project can the assessment was explained to him. His responses were be considered as complex. impartial about the project complexity. Subsequently, the results were presented to the project manager so that he may express his qualitative opinion on 4.2.1. 2015 Project Complexity Assessment Results. Figure 9 the complexity of the project and, thus, analyze the level of shows the result of IT project manager’s assessment of the agreement with the results obtained from the implementation project in 2015. of the tool. Te project manager expressed his agreement According to the project manager’s assessment, we can with these results in general terms, since he acknowledged an observe the following. increase in complexity in 2016 that forced him to implement complex project management competences. Very High Complexity Groups. Most of the complexity high- lightedbytheexpertisin“projectorganization”duetothe 5. Discussion and Limitations number of interfaces the project had, the communication demand, and the hierarchical structure of the project and the Te present work describes a new tool, based on IPMA teams. approach, to assess IT project management complexity in In addition, the other group that contributed to the an efcient and reliable way. It includes complexity factors complexity was the “cultural and social context.” As men- adapted to this particular industrial sector since it is consid- tioned in the project description, the great diversity of the ered that the development of projects in this sector is more project context, together with the cultural variety of the teams prone to failure than in other sectors, due to the complexity involved, local teams and near shore and ofshore teams of its projects. Te tool combines the use of complexity factors working together, as well as geographical distances of teams, defned by IPMA approach to measure complexity of any made this group very complex. kind of project and revised following a systematic literature review and the use of new complexity factors found in the High Complexity Groups. Two groups had high complexity; literature to manage inherent complexity of IT projects. Te the frst one was “interested parties, integration” of the use of all these factors were validated by IT experts by means project, which highlighted the variety of stakeholders and of a questionnaire. Te experts were chosen according to their interrelationships that were present in the project. experience and knowledge of IT industrial sector. Te use Te second group was “leadership, teamwork, and deci- ofIPMAapproachcanbejustifedbyitsabilitytoobtain sions” that refected the adaptive leadership style required to quantitative scores of project management complexity. Complexity 15

Report Report IT program manager

Business Report staf, Management

Ofshore production Report PMs Analysis team Production support Quality assurance Development team Infrastructure team support team Team team Levels 1, 2 Report

Vendor manager Report Vendor manager Vendor manager Vendor manager Vendor manager Vendor manager

Project manager 1 … Technology 1 team Technology 2 team Technology 3 team … Project manager 6

Project managers 15 business analysts 8 members 8 members 8 members 8 members Up to 40 developers/analysts

Analysis Support Infrastructure Quality, testing Support

Development

Figure 7: Project organization chart.

U/S U/S U/S U/S U/S U/S U/S U/S U/S U/S U/S U/S U/S U/S U/S U/S U/S U/S Upstream systems (product U/S U/S U/S U/S U/S U/S U/S U/S U/S U/S U/S U/S U/S U/S U/S U/S U/S U/S U/S data source) Audit systems/ S/S Support systems (other data Ad hoc S/S sources) audit teams Project S/S

S/S Technical hardware, sofware Downstream Reporting and D/S D/S D/S D/S D/S teams/ systems (data RR/S regulatory specialists target) systems D/S D/S D/S D/S D/S

Figure 8: Data fow and external stakeholders.

Te tool allows obtaining a Complexity Index that mea- underlying project manager conception of how complex the sures the level of global impact that the complexity factors studycaseprojectis.Ofallthenewspecifccomplexityfactors inherent to IT projects have on the project, under a specifc for IT projects, added to the IPMA framework, “project complexity scenario. For its validation the tool was applied to organization,” “cultural and social context,” “interested par- a sofware development project in a banking sector company. ties, integration,” “leadership, teamwork, and decisions,” and With the implementation of the tool to the case study it “objectives, requirements, and expectations” are those that canbeobservedthat,byincreasingtheassessmentofsome have contributed the most to a greater increase in complexity. factors towards greater complexity, the Complexity Index Tese factors represent, overall, a 63.64% contribution to increased and the overall percentage in which the project the complexity of the project against 62.5% of the minimum could be considered complex could be measured. index in which a project is considered complex. On the other hand, the project manager in charge of Regarding the results obtained by the scores of the the project acknowledged greater complexity in 2016 and projectcomplexityassessment,eachofthem,obtainedduring expressed, through a satisfaction survey with the tool, his diferent periods of the same project, showed how high their agreement that the rate of increase in the percentage refected complexity degree was with respect to the other complexity the increase in complexity that the project had sufered. scenario. It allowed comparing project complexity in 2015 Teexpertalsoshowedagreementandsatisfactionwith with project complexity in 2015. the project complexity assessment process. In addition, he Basedonthereviewoftheliteratureandthefndingsof acknowledged that, in this second year, he had had to manage the present study, we can conclude that it is not so important project situations in complex contexts due to a greater for an organization to measure all particular factors in a number of downstream systems that were integrated into project complexity assessment system since it may become a the project, increasing deliverables complexity and including difcult process; by contrast, it is relevant for any organization ofshore teams. to know key complexity areas (groups of complexity) and Te weighting of complexity groups of factors provides their metrics (factors) as well as their corresponding weights some important insights into the overall philosophy and that directly contribute to the complexity of the project. 16 Complexity

Figure 9: 2015 project assessment with Complexity Index tool.

Assessment: project is not complex enough/PM does not require this case would consist of one-to-one calculation of the complex project management competences Complexity Index of each project within a portfolio. Tereby, the focus is placed on the most complex projects or the most complex areas and main project complexity sources. Tese are the ones where more complexity related project management competences are needed. Tese assessments 1 Complex, 56.82 Simple, 43.18 should be done in the most complex scenario of each project

Assessment so that the cost/time of implementing it in a project portfolio is not excessive and the comparison is conclusive. In this sense, a risk-beneft analysis in the evaluation of the project portfolio could be used as additional information source 0 20 40 60 80 100 when implementing the Complexity Index tool in project (%) portfolio. In this way, the assessment of each project within a portfolio could be carried out only in cases where the Figure 10: Study case Complexity Index score for 2015. overall risk (technological-commercial)/beneft (economic- strategic) balance is below a threshold value since the rest of the projects could be discarded. Projects that have not On the other hand, an organization would beneft from passed the risk-beneft analysis flter would obtain a high the use of the Complexity Index tool by applying it in project expected Complexity Index (the higher the risk, the greater portfolio prioritization. To implement the methodology in the complexity of the project), but their evaluation would no Complexity 17

Figure 11: 2016 project assessment with Complexity Index tool.

Assessment: project is Complex/PM requires assessment of several projects across diferent sectors, since complex project management competences it has been used in project manager certifcation systems for many years in order to demonstrate their ability to lead complex projects. Tis validation was frstly performed Complex, Simple, through a systematic review of the literature and, secondly, 1 63.64 36.36 through an expert survey in IT project management. Tis

Assessment surveyonlyaimedtovalidatetheadequacyofallthese factors (IPMA approach factors and new factors found in the literature for IT projects) to fnd out whether they should be 0 20 40 60 80 100 included in a tool that measures the complexity of IT projects. (%) Terefore, the purpose of the survey used in the methodology was not to fnd a statistical signifcance of the results but a Figure 12: Study case Complexity Index score for 2016. threshold value from which each factor should be included in the tool or, below which, it should not be included. Tus, the statistical analysis performed to process the results of longer be necessary since they would have been discarded, the survey is merely descriptive and uses the mean value as thus reducing the cost/time of the implementation of the threshold value. methodology. Te main objective of the methodology used for the Conflicts of Interest development of the tool was the validation of an already verifed approach in the professional use of the complexity Terearenoconfictsofinterest. 18 Complexity

References [20] Standish Group, Te Chaos Report, M. A. Dennis, Ed., Standish Group International, Inc., 2015. [1] S. J. Whitty and H. Maylor, “And then came complex project [21] J. P. Murray, “Reducing IT project complexity,” Information management (revised),” International Journal of Project Man- Strategy: Te Executive’s Journal,vol.16,no.3,pp.30–38,2000. agement,vol.27,no.3,pp.304–310,2009. [22] K. Lyytinen, R. Hirschheim, and R. Hirschheim Paper, “Lyyti- [2] C.Bosch-Rekveldt,M.Mooi,H.Verbraeck,A.Sjoer,E.Wolsing, nenKandHirschheimRPaper1987.pdf,”Oxford Surveys in and B. Gulden, “Mapping project managers competences to Information Technology,vol.4,pp.257–309,1987. project complexity,” in IPMA 23rd WorldCongress, Research Track Human Side of Projects in Modern Business,2009. [23] K. Ewusi-Mensah, Sofware development failures anatomy of abandoned projects, 2003. [3] T. Williams, “Assessing and moving on from the dominant project management discourse in the light of project overruns,” [24] R. T. Nakatsu and C. L. Iacovou, “Acomparative study of impor- IEEE Transactions on Engineering Management,vol.52,no.4, tant risk factors involved in ofshore and domestic outsourcing pp. 497–508, 2005. of sofware development projects: A two-panel Delphi study,” Information and Management,vol.46,no.1,pp.57–68,2009. [4] J. Bennett, International Construction Project Management: General Teory and Practice,Oxford,1991. [25] S. L. Jarvenpaa and B. Ives, Executive Involvement and Partic- ipation in Te Management of Information Technology,vol.15, [5]A.J.Shenhar,“Onesizedoesnotftallprojects:exploring Management Information Systems Research Center, 1991. classical contingency domains,” Management Science,vol.47, no. 3, pp. 394–414, 2001. [26] L.Wallace,M.Keil,andA.Rai,“Understandingsofwareproject risk: A cluster analysis,” Information and Management,vol.42, [6]A.Bakhshi,J.Ireland,andV.Gorod,“Clarifyingtheproject no.1,pp.115–125,2004. complexity construct: Past, present and future,” International Journal of Project Management,vol.34,no.7,pp.1199–1213,2016. [27] B. Bahli and S. Rivard, “Te information technology out- sourcing risk: a transaction cost and agency theory-based [7]J.Geraldi,H.Maylor,andT.Williams,“Now,let’smakeitreally perspective,” Journal of Information Technology,vol.18,no.3, complex (complicated): A systematic review of the complexities pp. 211–221, 2003. of projects,” International Journal of Operations and Production Management,vol.31,no.9,pp.966–990,2011. [28] H. Taylor, “Critical risks in outsourced it projects: Te intractable and the unforeseen,” Communications of the ACM, [8]A.Bosch-Rekveld,M.Jongkind,Y.Mooi,H.Bakker,andH. vol.49,no.11,pp.75–79,2006. Verbraeck, “Grasping project complexity in large engineering projects: Te TOE (Technical, Organizational and Environmen- [29] F. P. Brooks, “Essence and accidents of sofware engineering,” tal) framework,” International Journal of Project Management, Computer (Long. Beach. Calif),vol.20,no.4,pp.10–19,1987. vol. 29, no. 6, pp. 728–739, 2011. [30] K. M. Whitney and C. B. Daniels, “Te root cause of failure [9] R. J. Chapman, “Aframework for examining the dimensions and in complex it projects: complexity itself,” Procedia Computer characteristics of complexity inherent within rail megaprojects,” Science,vol.20,pp.325–330,2013. International Journal of Project Management,vol.34,no.6,pp. [31] M. J. Earl, “Te Risk of Outsourcing IT,” MIT Sloan Management 937–956, 2016. Review,vol.37,no.3,pp.26–32,1996. [10] L. Vidal and F. Marle, “Understanding project complexity: [32] C. F. Kemerer and G. L. Sosa, “Systems development risks implications on project management,” Kybernetes,vol.37,no.8, in strategic information systems,” Information and Sofware pp. 1094–1110, 2008. Technology,vol.33,no.3,pp.212–223,1991. [11] V. Ireland, B. Rapaport, and A. Omarova, “Addressing wicked [33] B. W. Boehm, “Sofware risk management: principles and problems in a range of project types,” Procedia Comput. Sci,vol. practices,” IEEE,vol.8,no.1,pp.32–41,1991. 12, pp. 49–55, 2012. [34] C. A. Brown, “Te opt-in/opt-out feature in a multi-stage [12] D. Owens, J. Ahn, J. Shane, J. Strong, and K. C. Gransberg, delphi method study,” International Journal of Social Research “Defning Complex Project Management of Large U.S,” Trans- Methodology, vol. 10, no. 2, pp. 135–144, 2007. portation Projects, vol. 17, no. 2, pp. 170–188, 2012. [35] R. Kliem, “Managing the risks of ofshore It development [13]B.XiaandA.P.C.Chan,“Measuringcomplexityforbuilding projects,” Information Systems Management,vol.21,no.3,pp. projects: A Delphi study,” Engineering, Construction and Archi- 22–27, 2004. tectural Management,vol.19,no.1,pp.7–24,2012. [36] D. Baccarini, “Te concept of project complexity a review,” [14] S. Sinha, B. Kumar, and A. Tomson, “Complexity measure- International Journal of Project Management,vol.14,no.4,pp. ment of a project activity,” International Journal of Industrial and 201–204, 1996. Systems Engineering,vol.8,no.4,pp.432–448,2011. [37] J. R. Turner and R. A. Cochrane, “Goals-and-methods matrix: [15] I. AEIPRO, NCB-Bases para la Competencia en Direccion´ de coping with projects with ill defned goals and/or methods of Proyectos, 2006. achieving them,” International Journal of Project Management, [16] R. N. Charette, Sofware Engineering Risk Analysis and Manage- vol.11,no.2,pp.93–102,1993. ment, Intertext Publications, New York, NY, USA, 1989. [38] H. L. Wood and P. Ashton, “Te factors of project complexity,” [17] S. Sakthivel, “Managing risk in ofshore systems development,” in Proceedings of the CIB World Building Congress,pp.69–80, Communications of the ACM,vol.50,no.4,pp.69–75,2007. Salford, UK, 2010. [18] R. Schmidt, K. Lyytinen, M. Keil, and P. Cule, “Identifying [39] T. M. Williams, “Te need for new paradigms for complex sofware project risks: An international Delphi study,” Journal of projects,” International Journal of Project Management,vol.17, Management Information Systems,vol.17,no.4,pp.5–36,2001. no.5,pp.269–273,1999. [19] M. H. A. Tafi, “Risks factors associated with ofshore IT [40] K. A. Dunovic,´ I. B. Radujkovic,´ and M. Skreb,ˇ “Towards a new outsourcing,” IndustrialManagement&DataSystems,vol.105, model of complexitythe case,” Procedia-Social and Behavioral no. 5, pp. 549–560, 2013. Sciences, vol. 119, pp. 730–738, 2014. Complexity 19

[41] P. Austin, S. Newton, A. Steele, and J. Waskett, “Modelling and managing project complexity,” International Journal of Project Management, vol. 20, no. 3, pp. 191–198, 2002. [42] M. V. Tatikonda and S. R. Rosenthal, “Technology novelty, project complexity, and product development project execution success: A deeper look at task uncertainty in product innova- tion,” Heating, Piping, and Air Conditioning,vol.72,no.2,pp. 74–87, 2000. [43] T. Little, “Context-adaptive agility: Managing complexity and uncertainty,” IEEE Sofware,vol.22,no.3,pp.28–35,2005. [44] Project Management Institute, “Navigating Complexity: A Practice Guide,” 2014. [45] A. Aitken and L. Crawford, “A study of project categorisation based on project management complexity,” in Proceedings of the International Research Network for Organizing by Projects Conference (IRNOP VIII), Brighton, United Kingdom, 2007. [46] S. Seeman, “River Protection Project (RPP) Project Manage- ment Plan,” Association for Project Management Registered Project Professional – RPP Candidate Guidance RPP – the standard for extraordinary project professionals from the Asso- ciation for Project Management 2 APM Registered Project Professional – RPP Candidate Guidan, 2015. [47] Australian Government, Complex Project Manager Competency Standards,vol.1,2012. [48] D. Lessard, V. Sakhrani, and R. Miller, “House of Project Complexity—understanding complexity in large infrastructure projects,” Engineering Project Organization Journal,vol.4,no.4, pp.170–192,2014. [49] R. M. Lebcir and J. Choudrie, “A Dynamic Model of the Efects of Project Complexity on Time to Complete Construction Projects,” International Journal of Innovation, Management and Technology,vol.2,no.6,p.477,2011. [50]K.Shafei-MonfaredandS.Jenab,“Anovelapproachfor complexity measure analysis in design projects,” Journal of Engineering Design,vol.23,no.3,pp.185–194,2012. [51] T. O. A. Lehtinen, M. V. Mantyl¨ a,¨ J. Vanhanen, J. Itkonen, and C. Lassenius, “Perceived causes of sofware project failures - An analysis of their relationships,” Information and Sofware Technology,vol.56,no.6,pp.623–643,2014. [52] J. Alba, D. Cron, M. El Ouarti, G. Pugliese, W. Schmehr, and L. Simonelli, “IPMA International Certifcation and Regulations for the Assessment of Individuals in Project, Programme and Portfolio Management,” in Proceedings of the International Project Management Association, Zurich, Switzerland, 2016. Hindawi Complexity Volume 2018, Article ID 3190251, 17 pages https://doi.org/10.1155/2018/3190251

Research Article Strategies for Managing the Structural and Dynamic Consequences of Project Complexity

Serghei Floricel ,1 Sorin Piperca,2 and Richard Tee3

1 Department of Management and Technology, School of Management (ESG), UniversiteduQu´ ebec´ aMontr` eal´ (UQAM), Montreal, QC, Canada 2Department of Management, Birkbeck, University of London, London, UK 3Management Department, Libera Universita` Internazionale degli Studi Sociali Guido Carli (LUISS), Rome, Italy

Correspondence should be addressed to Serghei Floricel; [email protected]

Received 27 October 2017; Accepted 5 March 2018; Published 13 May 2018

Academic Editor: JoseRam´ on´ San Cristobal´ Mateo

Copyright © 2018 Serghei Floricel et al. Tis is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

We propose a theoretical framework that highlights the most important consequences of complexity for the form and evolution of projects and use it to develop a typology of project complexity. Tis framework also enables us to deepen the understanding of how knowledge production and fexibility strategies enable project participants to address complexity. Based on this understanding, we advance a number of propositions regarding the strategies that can be most efective for diferent categories of complexity. Tese results contribute to the integration of various strands in the research on project complexity and provide a roadmap for further research on the strategies for addressing it.

1. Introduction of project complexity? Second, what strategies can be used to manage them efectively? While we know that complexity afects project efectiveness, Given its impact, the project management feld has made it is unclear what strategies can be used to manage difer- signifcant eforts to understand the nature and consequences ent forms of complexity. Tis paper aims to advance our of project complexity and to develop approaches for manag- understanding of these strategies. From infrastructure con- ing it [5]. One source of inspiration was the fundamental lit- struction to developing biotechnology or sofware products, erature on complexity. Concepts such as heterogeneity, emer- complexity has been blamed for causing unexpected events, gence, or chaos provided a fresh perspective on the dominant late changes, additional costs, and delays and for afecting approaches for planning and managing projects [6, 7]. How- project performance and value creation [1–4]. One example ever, while these abstract concepts ofered a fertile ground for of how complexity afects projects is the launch-day failure criticizing the foundations of the discipline, they proved to of the luggage handling system at Heathrow Terminal 5. be less useful as a basis for developing concrete practices that While the immediate source of problems was a sofware bug, would help address complexity. Terefore, another approach during late execution stages, on-site interference between was to investigate managers’ perceptions in order to map contractors produced delays that eventually required sched- and measure the concrete factors that appear to increase ule compression that, in turn, prevented an adequate testing project complexity [8–11]. Some researchers went even fur- of the system. Tis example points out to the multiplicity of ther and attempted to identify the strategies that managers factors and interactions involved in a project as the source of deem appropriate for addressing each factor [12]. However, its complexity and to consequences concerning the form the the large number of identifed factors makes it difcult project will take as well as its evolution in time. However, in to provide theoretical justifcations and empirical evidence spite of this basic understanding of complexity, two key issues that goes beyond managers’ opinions on what factors are remain open. First, what are the most important dimensions important and what strategies are efective [13]. Tere is even 2 Complexity a risk that such approaches only collect “rationalized myths” particular its heterogeneity and interactions, and, hence, to [14] circulating in the organizational feld of the project predicting and controlling their consequences [17, 18]. Te management discipline. intrinsic viewpoint is held by philosophers and scientists who Our attempt to address these issues relies on two pre- assume a heterogeneous world, featuring diverse compo- mises. First, the most fruitful conceptual strategy for under- nents with multiple aspects and properties, which, therefore, standing complexity and how it can be managed consists of interact in multiple ways to produce unpredictable forms bridging abstract theories with problematic situations and and dynamics [19–21]. For example, research on biological practices encountered in projects. We believe that the essen- systems highlights the indeterminacy and multiplicity of tial aspects of project complexity can be captured by a few material properties and of their interactions, while research generic dimensions, which can be connected to concrete on artifcial systems emphasizes the number and diversity of factors observable in various types of projects. Second, we parts or components and their intricate interoperation [22– assume that each form of complexity is best addressed by 24]. Project researchers sharing this view tend to emphasize specifc strategies out of the variety of available ones. Some of the number and diversity of relevant factors, such as the these strategies create an improved representation of the number of diferent stakeholders or disciplines involved in a project and of its environment, helping participants to under- project, as well as their interrelations [8, 16, 25]. stand the specifc complexity they face and to prevent its ef- In turn, the cognitive viewpointarguesthatperceived fects, while others aim to increase the fexibility of the project complexity is an artifact of our imperfect representations. in order to address the unexpected consequences of complex- A frst reason for such perceptions is the imperfect cor- ity [15, 16]. respondence of representations with reality. Our perceptual Based on these premises, this paper makes three contri- abilities, even when they are enhanced by detection, magni- butions. First, we integrate various strands in the research on fcation, and measurement capabilities, overlook, regularize, project complexity into a model that highlights its most con- and approximate many aspects of the world [26]. Engineering sequential dimensions for projects and use these to develop and other conventions for representing objects increase the a typology of complexity. Second, we use this framework distance even more by overlooking additional details [27]. In to develop a parsimonious categorization of knowledge pro- turn, knowledge representations, such as scientifc abstrac- duction and fexibility strategies for addressing project com- tions or engineering formulas, necessarily simplify the prop- plexity. Tird, we derive a number of propositions regarding erties of natural and artifcial objects [28]. In their quest for the strategies that are most efective with respect to various ideal parsimony, symmetry, and generality, they forego irreg- forms of complexity. In our view, these contributions provide ularities in shape, structure, or texture, along with secondary a roadmap for subsequent empirical research on project com- forces and interactions [29]. In turn, computer scientists plexity and the strategies for addressing it. By increasing the highlight what they call computational complexity [30, 31]. feasibility of empirical corroboration, our results also open Namely, they point out that increasing the correspondence new perspectives for creating more efective project manage- with reality also raises the computational power required for ment practices. operating with resulting representations [2, 32, 33]. From this Te paper proceeds as follows. In the following section, point of view, the complexity of an object corresponds to Section 2, we disentangle the literature on project complexity the difculty of extracting regularities that can simplify its in three themes: antecedents, mechanisms, and consequen- description, along with the volume of computations required ces. In Section 3, we detail our theoretical framework with for retrieving its form from this condensed representation respect to complexity and the strategies for addressing it [34–36]. and propose a series of propositions that connect these ele- A second key theme of complexity is related to what ments. A discussion and conclusion section summarizes the we term mechanisms, which, given particular antecedents, argument and suggests ways to carry this research further. produce unpredictable consequences or overwhelm cognitive abilities. Existing research has focused on two aspects: the 2. Theoretical Context emergence of systemic properties and their evolution over time. Te emergence stream investigates how the aggregation Te project management literature and the fundamental of component entities produces systemic properties that can- contributions that inform it use the term complexity to cover not be explained, let alone anticipated, by only considering three distinct topics: a series of antecedents, such as the mul- the properties of these components, even when their interac- tiplicityofrelevantfactors;aseriesofmechanisms,suchas tion propensities are taken into account [19]. Tis irreducible emergence or self-organization; and a series of consequences character of higher-level properties becomes apparent at all such as unpredictability and lack or control. Figure 1 orga- successive levels of organization, from molecules with respect nizes these aspects of complexity and identifes the key di- to atoms and cells with respect to molecules [37] to organiza- mensions of each aspect. tions with respect to the individuals composing them [38]. A In terms of antecedents, some researchers see complexity frst dimension of emergence is the “upward” nonadditivity as an intrinsic property of reality, irreducible even if our of component aggregation [39, 40]. Proposed mechanisms knowledge were perfect. Others argue that complexity per- start with the breaking of symmetry with respect to physical ceptions are a product of the cognitive limitations of human laws in aggregate systems such as molecules [41]. For higher and organizational actors and of their knowledge and infor- levels, the focus is on the nature and pattern of interactions mational tools with regard to representing this reality, in or couplings between components [42, 43]. For example, Complexity 3

Antecedents Mechanisms Consequences Intrinsic properties Emergence Structural (i) Heterogeneity (i) Upward nonadditivity (i) Unpredictable form (ii) Multiple interactions (ii) Downward conditioning (ii) Unexpected properties

Imperfect representations Unfolding Dynamic (i) Correspondence (i) Destabilizing engines (i) Irregular change patterns (ii) Computational power (ii) Stabilizing feedback (ii) Unexpected events

Figure 1: Disentangling the literature on project complexity. research on processes occurring in materials and living In terms of consequences, the literature also sets apart a organisms turned its attention to the role of supramolecular, structural and a dynamic aspect [81]. Te structural kind of noncovalent ties in their complex self-organizing or adaptive consequences refers to the unpredictable form of the project: properties [44]. Research on artifcial objects stresses the role organizational and contractual structure, technical or archi- of secondary interactions between components in shaping tectural solutions, and as-built artifacts. For example, during their form [45]. Sociomateriality researchers suggest that development, project concepts go through several iterations, emergent properties of materials and artifacts infuence the which push the end result far away from the initial vision agency, practices, and relations between human actors [46, [82, 83]. Some of this unpredictability stems from underlying 47]. In turn, the properties of social networks and organiza- material connections between entities [49] and the “rich tions, including projects, emerge from this tissue of practices indeterminacy and magic of matter” ([84]: 91), but also from and relations and the objects with which they are intertwined, a range of multilevel infuences on projects, the interplay of by means of constant translation and maintenance eforts as their “human side” [85] with the formalizing, political and well as unintended routinization [48, 49]. A second dimen- economic forces in organizations [86], industrial sectors, and sion of emergence mechanisms includes the downward condi- broader institutional and social systems. tioning of collectives of components by higher-level systems A second kind of structural consequence focuses on unex- [50, 51], or mutual infuences between levels [52]. Tis kind pected properties. While such properties may appear to be of hierarchical interlevel infuences or multilevel nestedness minordeviations,becausetheyarediscoveredquitelateinthe underlies the complexity of biological [53–55], communica- project, when prior decisions and actions are more difcult to tion [56], and social systems [57]. Research on projects has reverse, they are even more likely to cause major crises or con- also turned its attention towards this kind of infuences by ficts around the required corrective changes [87]. For exam- recognizing that projects, teams, and artifacts are nested in ple, advances in execution uncover unexpected interferences broader technical systems, organizations, interorganizational between project subsystems or between the project and soil networks, and institutional frameworks [58–62]. conditions, adjacent structures, natural habitats, neighboring In turn, unfolding mechanisms address processes through communities, and various regulations [88, 89]. Tese may, in which, in time, complexity drives the systems of interest along turn, interfere with the access on site of contractors for sub- multiple pathways in unpredictable ways. Tis theme has sequent activities, accentuate conficts of interests between also been addressed from two perspectives. Te frst focuses participants, and amplify personal animosities [90]. on destabilizing engines that bring about sudden, radical Other researchers emphasize dynamic consequences for transformations or unpredictable, chaotic change patterns the project and its environment. A frst dynamic efect is irreg- [63]. Researchers seek to characterize the causal mechanisms, ular change. Management scholars have paid special attention such as nonlinear interactions or path-dependent processes to dynamics in organizations and their environments, as well that amplify small variations in initial conditions [64, 65], asthechallengestheypose[79,91–95].Someresearcherseven and attempt to understand the conditions, such as number of argue that organizing is an ongoing activity, struggling to interacting factors [66], that activate these mechanisms. Teir reconnect a world of intersecting event strands, in which sta- insights inspired project management researchers in address- bility is just a cognitive artifact or a fragile result of recurring ing various dynamics from catastrophic artifact failures to the processes [96–98]. Project scholars also pay attention to the confictual unraveling of teams and interorganizational net- dynamics of projects and their environments, in particular to works [16, 67, 68]. A second perspective focuses on stabilizing chaotic change and the special kind of uncertainties it creates feedback that tames destabilizing tendencies but still produces [99, 100]. Tey single out the consequences of unpredictable unpredictable evolution [69–71]. Fundamental research in evolution in user needs and markets, as well as in technolog- self-organization uses simulations to show how order emer- ical and regulatory environments [82]. ges out of chaos [72]. In turn, organization scholars study me- A second type of dynamic outcomes consists of unex- chanisms such as structuring from repeated interactions pected events. Tese are emphasized by scholars who argue between actors [73, 74] or lifecycles in which stabilization that social systems having an inward, self-referential orien- followsturbulentperiods[75–77].Otherspayattentionto tation, such as projects attuned to their goals and operating processes that maintain highly dynamic patterns, such as high plans, are subject to perceptual discontinuities when dealing velocity and exponential growth [78–80]. with environments whose complexity is higher than their 4 Complexity internal communicational complexity [101, 102]. Te man- or downward, we propose four types of unfolding processes, agement literature highlights the challenges caused by unex- with specifc dynamic consequences for projects. Because all pected events [103], for example, jolts [104], attributes their of these consequences have unexpected or surprising aspects, emergence to a complex “causal texture” of the environment, participants will have to respond afer the fact rather than and suggests that they lead to a sudden “increase the area of anticipating specifc dynamic paths. Terefore, we assume relevant uncertainty” [105]. Project research also singles out that efective strategies for addressing dynamic consequences unexpected events occurring inside and outside the project rely on cultivating project fexibility and propose, for each of as a key source of trouble [106, 107]. Similar to unexpected the four types of unfolding, a specifc type of appropriate interactions, their surprising and late occurrence amplifes fexibility. their impact on projects [108]. Tis model makes two strong assumptions, namely, that In the next section, we use this background to develop a each type of consequence stems directly from only one type parsimonious set of dimensions and a typology to character- of mechanism and, moreover, is mitigated by only one type ize project complexity, and then we use it to understand the of strategy. In the following subsections, we detail the model strategies that can be used to manage complexity and derive by dividing the discussion along these two key infuence aseriesofpropositionsaboutthemostefectivestrategiesfor pathways and further explaining these and other assumptions managing each type of complexity. we make.

3. Theoretical Framework and Propositions 3.1. Pathway and Strategies for Structural Consequences. Te endpoint of this pathway, structural consequences, is con- Our theoretical framework uses the same dimensions of com- cerned with the unexpected aspects of project form, con- plexity proposed in Figure 1, by transforming them into vari- sidered as an (end) state, namely, a set of properties. While ables and types of complexity. We then use them to develop accepting that processes leading to this state occur in time, a more selective model about the infuences between these emergence mechanisms, as discussed in the review section, complexity variables. Our approach can be termed mecha- do not emphasize the temporal aspects of phenomena, but nismic [109, 110], because we suppose that antecedent factors particular conjunctions of antecedent properties that produce will trigger alternative mechanisms, namely, bits of processes emergent properties, namely, forms with unexpected aspects. that are potentially present in the relevant systems, but are To address these phenomena we sought inspiration in the activated only in certain conditions and not in others. Tis debate between the intrinsic and representational viewpoints enables us to tie structural and dynamic consequences to on the nature of complexity. Building on phenomenological a given set of antecedents. Te overall model relating these views on projects (Cicmil et al. 2006), we could assume that, aspects is represented in Figure 2. given a current level of knowledge, both sides of the debate Te model includes essentially two infuence pathways, contribute to our understanding of complexity perceptions, ending, respectively, with structural and dynamic consequen- as experienced by project participants. From this perspective, ces. We assume that both pathways start with the antecedents the aspects emphasized by each side of the debate imply two discussed above, namely, intrinsic properties and representa- dimensions of complexity antecedents. tion imperfections. Te upper path, ending in structural con- Te frst dimension is suggested by researchers who view sequences, starts with the infuence of these conditions on the complexity as an intrinsic property of reality. We interpret emergence mechanisms present in the project. We argue that, this to mean that project participants, given the knowledge depending on the combination of intrinsic conditions and that they deem available to them, do not see any chance representation imperfections that prevails in the project, four of completely understanding and mastering the relevant types of emergence processes will be observed in projects. phenomena. Consequently, at some residual level, complex- Moreover, each type of emergence process will result in a par- ity is an inseparable property of the perceived world. With ticular kind of structural consequence. Further, we assume regard to the antecedents related to this dimension we distin- that the strategies for minimizing these structural conse- guish heterogeneity, namely, the infnite variety of elements quences essentially refer to knowledge production capabil- and aspects present in the word, from the multiplicity of inter- ities, which enable project participants to understand and actions between these factors. In other words, the context master the respective emergence processes. of some projects may be dominated by the multiplicity of Te second pathway, ending in dynamic consequences, relevant factors while others by the multiple interactions be- also begins with a direct infuence from antecedents, which in tween the various factors. We further assume that the hetero- this case refer exclusively to intrinsic properties and in a more geneity end of this dimension is more likely to give promi- specifc way. Namely, we argue that particular combinations nence, from the participants’ perspective, to upward nonad- of relevant factor diversity and interaction number and non- ditivity mechanisms, namely, to conjunction processes that linearity increase the chances to trigger either destabilizing or will converge towards overall confgurations with unpre- stabilizing unfolding mechanisms. A second infuence from dictable properties (project form). Te interactions end of antecedents proceeds indirectly, through emergence mecha- this dimension is more likely to give prominence to down- nisms, in particular whether the prevailing processes involve ward conditioning mechanisms, because participants will upward nonadditivity or downward conditioning. Depend- perceive the mostly hidden interactions as inexorably driving ing on whether antecedents trigger stabilizing or destabilizing the aggregate towards some stable state, which, in turn, mechanisms and interlevel emergence is primarily upward starts to subsume and condition the behavior of converging Complexity 5

Knowledge strategies

Emergence mechanisms Structural consequences Antecedents

Unfolding mechanisms Dynamic consequences

Flexibility strategies

Figure 2: Integrative model of complexity infuences and strategies for diminishing them. elements. However, the multiplicity of interactions and “ver- typical of market aspects, in which additional customer tical” infuences, downward as well as interlevel, means that needs, market segments, competitors, products, and strategic the aggregate system will have many unexpected properties. moves,aswellasahostofotherfactors,arelikelytoemergeas Te second dimension, inspired by those who see com- signifcant in most projects. plexity as a consequence of imperfect representations, focuses Because such a context continuously reveals new relevant on the nature of these imperfections, as perceived by project aspects,itisimportanttodetectandbringthemtobearon participants who attempt to use their perceptions, knowledge, project shaping activities as early and as minutely as possible. and modeling capabilities to anticipate and infuence the Terefore, we believe that distributed learning about various form and properties of aggregates. Tis dimension sets apart aspects involved in the project [114, 115] through role diferen- a weak representation correspondence with reality from an tiation as well as background and external network diversity insufcient computation power to anticipate consequences. [116] and by bottom-up decision making [117] is likely to On the low correspondence end, the critical issue is that be more efective than attempts to centralize learning and currently available knowledge does not include all relevant decision making. Closeness to the “feld” and specialization factors or interactions or mistakenly considers some of them in detecting the given aspect are likely to reduce the distance negligible. As a result, project participants do not take these between representation and reality for participants, while aspects into account when shaping the project and are bottom-up decentralized selection will speed up and make bound to discover them while implementing or operating the more efective the emergence of the fnal form. Concrete project. On the weak computation power end, the key issue examples include the approaches used by companies such as is that, even with good correspondence, or perhaps because 3M and Google. For example, Google encourages individual participants attempt to increase it, they cannot work out all learning by employees and uses employee panels to pre- the consequences of contributing factors, in terms of aggre- dict market demand and democratic voting procedures to gate form and properties. decide on innovation projects [118]. Information systems that Based on these two dimensions, we identify four types of support a “complex, distributed, and evolving knowledge- complexity contexts and propose a corresponding emergence base” and an “unstructurable, dynamically changing process processes for each of them. Table 1 presents the four types, of deliberations and tradeofs” are likely to enable such repre- suggests the projects and project aspects in which they are sentation strategies ([119]: 206). Tis translates into the fol- most likely to be observed, their most likely structural conse- lowing proposition. quences, and suggests the most appropriate representation- producing strategies that help project participants to increase Proposition 1. In a high heterogeneity and low correspondence their anticipation and infuence over aggregates, such as context, strategies that build a distributed learning capability the project system. Inspired by organization theory, project are most likely to be efective in preventing the “endless compli- management, and engineering design literature, we believe cation” structural consequences of complexity. that these strategies reduce the distance and computation difculties, given the specifc combination of heterogeneity, Magic Field. Tissecondtypeofstructuralcomplexitycon- interactions, and representation shortcomings afecting the text is prominent in settings with multiple interactions that project [111–113]. are inadequately captured by existing knowledge. Tese interactions drive the project to converge into an overall Endless Complication. Tis frst type of structural complexity confguration, but interactions that are unaccounted or mis- results in a context of high intrinsic heterogeneity and low represented, such as previously negligible secondary interac- representation correspondence. Te aggregate project form tions that become a problem in physical artifacts designed is unstable and its fnal confguration cannot be predicted for larger scales or performance levels [45, 120], are likely because of the constant discovery of new relevant factors. We to manifest themselves through aggregate properties that consider that this situation is typical of innovation projects, participants cannot control adequately. As a result, project in which knowledge derived from previous projects is not artifacts cannot sustain the required constancy of operation, entirely pertinent and may induce participants to overlook and corrective interventions do not have the intended efect. some key factors or to underestimate their role. Tis is also Tis kind of context is likely to be observed in infrastructure 6 Complexity

Table 1: Types of emergence processes and structural consequences as a function of antecedents, together with the most efective strategies for dealing with them.

Intrinsic properties Heterogeneity Multiple interactions Endless complication Magic feld Innovation project, market aspects CoPS & infrastructure, organization aspects Correspondence Consequence: unpredictable form Consequence: uncontrollable properties Strategy: distributed learning Strategy: heedful connectedness Representation Irreduciblevariety Intractablemess Sofware project, technical aspects Biotech project, institutional aspects Computation Consequence: slow-converging form Consequence: unexpected properties Strategy: simulation through representation Strategy: massive trial and error

construction projects, such as airports or power plants, or in communications are most likely to be efective in preventing the projects that develop “complex products and systems” (CoPS) “magic feld” kind of structural consequences of complexity. such as aircraf or military systems [58, 62]. As larger scale, higher performance, or additional constraints, such as build- Irreducible Variety. Tis third type of structural complexity ing new structures in a crammed site surrounded by continu- contextbecomesprominentwhenthenumberofrelevantfac- ously operating systems, are imposed upon such projects, pre- tors is important but existing representations cannot reduce viously unknown or negligible interactions, including those them to a more parsimonious set of essential properties with with the foundations of adjacent buildings or with nearby which participants can operate. As a result, participants are ecosystems, start to have signifcant consequences. Because of likely to concentrate selectively on some of these factors at multiple interactions, these consequences are likely to propa- some moments and on others at another time. Tis precludes gate in unpredictable ways to other parts of the project and to them from converging on a stable project form. Tis kind of subsequent activities. Te organizational aspects of projects context is likely to be observed in sofware projects in which are also likely to face a similar complexity context, because the the high number of concrete factors cannot be adequately multifaceted relations between participating actors, subunits, captured with most architectural modeling methods, leading and corporate entities, as well as with the broader environ- to an endless succession of beta prototypes before the release ments, are likely to bring new interactions to prominence in of a still imperfectly stable product. A particular case could various projects. be ICT solutions developed for various implementation con- Teobviousstrategyismappingallinteractionsinthe ditions, such as countries, sectors, or types of organizations, project and using pathway analysis to identify and address all of which impose diferent needs and constraints, which the systemic risks that they may cause [121, 122]. However, are difcult to compress into a unique parsimonious set of these methods suppose that interactions are known or can requirements. More generally, this is also the case for the be imagined, while the real danger in this context is that technical aspects of projects, in which taking into account the project participants “cannot anticipate all the possible inter- impact of so many factors (dimensions and other properties) action of the inevitable failures” ([123]: 11). Even when they makes designing them a highly iterative endeavor [132, 133]. encounter unexpected interactions, it is likely that these will In this context, the problem is accounting for the variety be revealed indirectly and to participants that, organization- of factors in a way that is not afected by the low computa- ally, may be responsible only for some of the interacting tional power. We believe that the early virtual representation factors. Terefore, an adequate reaction of project partici- and simulation of the project and its behavior constitute pants depends on their ability to detect these signs [124] and the most efective strategy. Representations may range from update their routine communicational ties in ways that pencil and paper sketches to multidimensional digital proto- account for the newly discovered interactions [125]. Building typing using CAD or Building Information Modeling (BIM) alertness, mindfulness, and heedfulness to make sense of new tools and to preliminary embodiment of artifact materials interactions and restructure communicative ties accordingly and form. Even when powerful sofware tools are used, the islikelytobethemostefectivestrategy[68,126,127]. efcacy of this strategy does not come from their added Organizationally, it relies on building strong communication computational power, but from the way representational ties [128] and routines for rearranging these ties [129] as well outputs boost participants’ individual and collective cognitive as strong sensemaking and integration capabilities [130, 131]. abilities[134].Evenanimperfectoutputthatsimulatesthe Tese arguments are synthesized in the following proposi- real behavior of project aggregates enables participants to rely tion. on their pattern-recognition abilities to uncover diverse fac- tors and creates a boundary object that enables participants Proposition 2. In a context with numerous interactions to contribute their varied perspectives in an integrated way and low representation correspondence, strategies that build [135–138]. All this is likely to speed up the iterative process of a capability for mindful integration and restructuring of convergence towards a relatively stable form. Stigliani and Complexity 7

Ravasi’s[139] study of the design frm Continuum provides an Tis set of propositions assumes that one type of context example of such strategy, by highlighting the way project par- will be present, because of prevailing antecedents, and sug- ticipants use sequences of varying sketches and assemblages gests a single strategy that is likely to be efective in this con- of material objects to converge towards a client-oriented nar- text. Strategies are also mutually exclusive to some extent. For rative about the object they will design. Tese considerations example, the simulation through representation strategy sup- arerefectedinthefollowingproposition. poses a rather long sequence of representations that gradually approximate the project form, while the massive trial and Proposition 3. In a context with high heterogeneity and low error strategy calls for arriving as fast as possible at a material computational power, strategies that build a capability for instantiation of project form. However, Table 1 suggests that aggregate representation of form and behavior are most likely diferent antecedents may prevail with respect to diferent to be efective in preventing the “irreducible variety” kind of aspects of the project, such as market, technical, organiza- structural consequences of complexity. tional, or regulatory; to various subsystems, such as hardware versus sofware; and even to diferent stages, such as defni- Intractable Mess. Te fourth and fnal type of context is more tion versus design. To the extent that these portions of the likely to occur when the number of interactions between rel- project can be treated separately, diferent strategies could evant factors makes it impossible to compute their joint con- coexistinthesameproject.Tiscouldbeperhapsthecom- sequences based on existing representations. As a result, par- mon situation in major projects and programs, or in mega- ticipants base their predictions only on a manageable subset, projects. One example could be a complex product and sys- which is likely to cause unexpected properties of the aggre- temproject,suchastheF35militaryaircraf.An“endless gate, as evidenced by the “black swans” problem in statistics complication” context may prevail during the early defnition [140] and the difculty of creating high-reliability technical stage, because of the endless, shifing, and contradictory systems [68, 123, 141]. One example is biotech projects, in client requirements, an “irreducible variety” context may pre- which the large number of levels and interactions in living vail during the design phase, and a “magic feld” context beings can hardly be considered at the same time in order may prevail during testing and early exploitation stages, when to predict the properties of drug candidates [142] or even many unexpected properties and interactions emerge to to replicate successful experiments [143]. Tis leads to unex- causeaccidentsanddelaytheefectiveuseoftheartifact.Te pected properties such as a lack of therapeutic efect and recommended strategies could be emphasized, in parallel, undesirable side efects. It is also likely that the institutional withrespecttovariousaspects,andinsequence,invarious aspects of most projects will be characterized by a similar stages of the project. complexity context. Te large number of social interactions underlying this aspect, even as they are captured in regula- 3.2. Pathway and Strategies for Dynamic Consequences. Te tions, makes it difcult to work out their consequences for endpoint of this pathway is a pattern of change that is the project and will likely produce undesirable interference unpredictable or unexpected. In this case, the relative timing and conficting interests. and sequence of events, rather than the conjunction of prop- Because in this context interactions are so numerous that erties, are the center of attention. As illustrated in Figure 2, no amount of computation will account for all of them, the we suggest that these consequences result from the activa- most efective strategy is to let material reality itself work tion of unfolding mechanisms, under the joint infuence of out the consequences of aggregation and then test this result antecedent conditions and of prevalent emergence mecha- against required properties. In other words, the most efective nisms. Contributions addressing dynamic processes suggest strategy is trial and error based on concrete objects [144]. that particular conditions may favor destabilizing mecha- For example, in spite of many scientifc advances and scores nisms while others, stabilizing ones. For example, Dooley of new computational methods, pharmaceutical and biotech- and Van De Ven [66] argue that a system deterministically nology projects rely increasingly on massive trial and error infuenced by a small number of variables produces a periodic instead of rational design based on representations [145, 146]. pattern if variables interact linearly. Yet, nonlinear interac- Te efectiveness of this process can be improved by varying tions between a small number of variables produce a dynamic the reliance on preexisting knowledge, radicalness, precision, termed “chaos,” whose path is unpredictable but which and number of trials and of iterations [147, 148]. Moreover, follows a predictable pattern of change. Likewise, a system strategies could rely on serendipity to increase variation in infuenced by many variables produces a pattern termed object forms and to induce accidents in their operation, with “white noise,” meaning totally random, if variables act inde- the goal of exploring areas outside what is normally expected pendently from each other. If their actions are constrained [149] and seeking a sort of “falsifcation” [150] for any conclu- by interactions, the consequence is “colored noise,” such as sions. Tese arguments are synthesized in the following pro- “pinknoise”—arandompatternthattendstoreverseitstrend position. with low frequency—or “brown noise”—a random pattern with path-dependent tendencies. Tis suggests that intrinsic Proposition 4. In a context with numerous interactions and contextual aspects, such as heterogeneity and interactions, low computational power, strategies that build a capability for are also responsible for generating various types of dynamic testing concrete objects in real conditions are most likely to be patterns, with varying levels of predictability. efective in preventing the “intractable mess” kind of structural Like in the case of structural consequences, we used these consequences of complexity. contributions as a source of inspiration but adopted again the 8 Complexity pointofviewofparticipants’livedexperience.Weassume overlooking many details among their multiple aspects and that dynamic patterns are considered from the point of view interactions. Tis helps distinguish diferent kinds of sur- of a system at some mid-range level of aggregation, which prises for project participants. A further consequence of the is of interest to project participants. Moreover, the evolution unpredictability and unexpectedness assumption is our of this system is infuenced by the aggregation of lower-level assumption that the most efective mitigating strategies re- elements as well as by constraints from a higher-level system. volve around cultivating a particular kind of fexibility that For example, if participants are interested in the dynamics enables participants to respond ex post to dynamic conse- of the environment of their project, the market or industry quences. is the system of interest, organizations such as frms and In order to understand what kind of fexibility is most projects are the elements, and the broader societal system is efective for the various unfolding processes and dynamic the higher-level constraining system. If the system of interest consequences, we use the two dimensions introduced above is the project artifact (building, infrastructure, product, etc.), to set apart four typical dynamics. Figure 3 presents each of elements are the needs, technologies, components, materials, them as a function of the emergence direction and, respec- and activities that converge to shape it, and the overarching tively, unfolding mechanisms it involves. Each resulting system is formed by the broader technical, organizational, quadrant includes, on the lef, a schematic description of how and natural environments (technical networks, built sur- these mechanisms interact; in the center, the name of the roundings, landscape and soil, frms, communities, etc.) that dynamic and its most important consequence; and on the incorporate the artifact. In terms of antecedents, we consider right, the strategy most likely to be efective and a depiction only intrinsic properties. of how the strategy addresses the respective dynamic conse- We further assume that the dynamics of the focal system, quence. as perceived by project participants, will be infuenced by two dimensions. Te frst dimension is related to the nature Efervescence. Te dynamic depicted in the lower lef box of unfolding mechanisms that are activated through a direct designates irregular evolution of the focal system (yellow line) infuence from intrinsic complexity antecedents. In partic- which results from nonadditive interactions between lower- ular, we distinguish mechanisms that are primarily destabi- level entities. We assume that heterogeneous lower-level enti- lizing from those that are primarily stabilizing. As explained ties autonomously initiate changes, but these changes interact above, certain intrinsic antecedents, namely, combinations of in multiple ways with those initiated by other entities. Te factor heterogeneity and interactions, may favor destabilizing situation becomes problematic when several actions concur unfolding mechanisms, while other combinations may favor to create a local trend that, due to interactions, becomes a stabilizing mechanisms. As a corollary, we assume that con- path-dependent tendency strong enough to erupt into and texts in which destabilizing mechanisms are prevailing imply destabilize the next level of aggregation, which is the focal that the focal system has a relatively low inertia, while con- system. Tis kind of intrusion from what can be seen as texts dominated by stabilizing mechanisms involve systems insignifcant lower-level details will be perceived by project with high inertia. Te second dimension is related to the in- participants as coming out of nowhere. Required unfold- direct infuence of the intrinsic context, occurring through ing conditions are similar to Brownian motion of lower- emergence mechanisms, in particular through the prevailing level entities, which produce dynamic patterns that can be “vertical” interactions involving the focal system and the termed brown noise or random walk. However, what trans- other two relevant systems. Hence, we distinguish situations forms occasional spikes into important changes are upward in which emergence processes involve upward nonadditiv- nonadditivity mechanisms, for example, nonlinear interac- ity infuences from the lower-level system from situations tions between the elements involved in the spike. Tis kind involving downward conditioning from higher-level systems. of phenomena can be observed in project environments such As a corollary, we assume that systems in contexts dominated as markets, where several independent competitive actions, by upward nonadditivity are more easily decomposable, while which in themselves may have insignifcant consequences, systems in contexts characterized by downward conditioning create, through nonlinear interactions, a major trend that are not easily decomposable, because of the constraints that unexpectedly transforms the competitive context. With the higher-level system imposes upon them. regard to project artifacts, this kind of situation, as well as the In examining the impact of these two dimensions and continual disruption it induces, can happen in the develop- the strategies for addressing their consequences, we start with ment stage of a material project and throughout the lifecycle the premise that complex dynamics are problematic because of sofware projects. In both cases, elements are characterized of their unpredictability and unexpectedness; their specifc by low inertia and high subsystem separability, which enables dynamic consequences are of the “unknown unknowns” the required frequency and autonomy of changes. kind. In contrast to our treatment of structural consequences, Te consequence of efervescence is a continuous disrup- we also assume that participants consider the sources of dy- tion of the system of interest. An example could be an IT namicconsequencestobeintrinsic,ratherthanimperfect project with physical infrastructure components jointly representations. We account for representation issues only by developed by several public transportation authorities oper- assuming that participants consider the mid-range system as ating in the same major urban area. Teir numerous their task environment [151] and focus their attention on it, demands, many of which arrive quite late in the process, trig- leaving out some areas of the higher-level system, examin- gered by what other participants have asked or by new under- ing lower-level components with a coarse resolution, and standing of technical possibilities, generate debates among Complexity 9

Unfolding Emergence Destabilizing (low inertia) Stabilizing (high inertia) Discontinuity Amplification

Strategy: iteratioiterationn Strategy: options Downward conditioning (low separability)

Consequence:nsequence: irrelevanceirrelevance Consequence: insufciencyinsuf

EEffervescenceffervescence AAccelerationcceleration

Strategy: agility Strategy: modularity Upward nonadditivity (high separability)

Consequence: disruption Consequence: obsolescence

Figure 3: Types of dynamic complexity, as a function of prevailing mechanisms, and strategies for dealing with them. Note.Tethree horizontal lines on the lef side of each cell in the table represent, respectively, from top to bottom, the macro level, the meso-level (uses orange color to highlight the fact that it represents the context deemed relevant by project participants: task environment, etc.), and the micro level. Te red arrows suggest the direction of interlevel infuences. Te pictures in the middle part of each cell rely on alterations of the Endless Column (or Column of the Infnite) by Constantin Brancusˆ ¸i (completed in 1938, Targuˆ Jiu, Romania) to represent the four patterns of dynamic complexity. Strategies on the right side of each cell suggest ways of organizing project activities, namely, of connecting them simultaneously as well as intertemporally, which are likely to be most efective for dealing with each type of dynamic complexity. Colored lines represent strands of activities and the extent to which these change direction in time.

them and opposition from the technical development team, Discontinuity. Te dynamic depicted in the upper lef corner but some generate sufcient impetus to be included among ofFigure3showsasurprisingchangeinthefocalsystem the requirements of the project, which may perhaps prompt attributabletothedownwardconditioningbythehigher-level a restructuring of the technical architecture, which in turn system. Tis kind of change appears when new interactions mayinducenewdemandsandsoon.However,thesame between the components of the overarching system arise out- characteristics that enable this surprising dynamic are also side the area which is usually considered relevant by project likely to enable the fexibility strategy that we term agility, participants. Tis creates what Emery and Trist [105] call a namely, dividing the project into small separate strands, each turbulent feld, which generates irregular dynamics in the of which traces a relevant aspect through frequent small itera- higher-level system. Because of the signifcant conditioning tions and a continual restructuring of ties between strands impact that these changes have on the focal system, the latter [152–154]. Flexibility results from maintaining the project on is perceived as sustaining a series of sudden radical changes the edge of chaos [155], which precludes its coagulation into a or jolts. Such change is unpredictable because it originates stable form that can no longer trace the emerging changes. In outside the task environment that project participants nor- practical terms, this strategy favors subdividing the project mally monitor. In project environments, such as markets into a very large number of parts and work packages and and industries, this kind of changes occurs when previously minimizing the direct and indirect impact that activities unrelated areas in the global economy interact to impact and decisions concerning these subsystems have on adjacent a given sector. Examples include the recent impact of the subsystems and subsequent activities. Tis kind of approach subprime crisis in the United States on, say, the economy of is assumed by agile project management strategies such as Greece (via its debt) or the real estate market in Spain. For Scrum. Tese arguments are summarized in the following project artifacts, this kind of change may arrive when political proposition. or economic interests, perhaps hidden, combine to force the client to signifcantly change project requirements. Tis Proposition 5. Te presence of destabilizing unfolding and occurs frequently in projects, such as movies, videogames, upward nonadditivity produces a dynamic of efervescence, and high performance CoPS, in which external infuences whose consequence of disruption is most likely to be efectively vary considerably, but systemic constraints boost the inter- addressed by strategies that increase the agility of the project. actions between elements [156]. 10 Complexity

Because of these interactions, the answer cannot be piece- separate adaptation of various parts of the project. Terefore, mealchangeintheproject.Infact,theformtheprojecthad we argue that the most efective strategy for this context is prior to a jolt is likely to become irrelevant in its entirety afer preparing real options [160] via small proactive investments the jolt. In this case, we argue that the efective strategy is that open the possibility of later adding markets or capacities preparing for deep, hierarchically driven iterations [1, 82, 157], or even open entirely new projects in a relatively short essentially getting ready to restart the project several times, time and with reduced investment [161, 162]. For example, a keeping only the learning from previous iterations. Tis power plant project may purchase land for a second unit and strategy calls for reducing intertemporal inertia in the form even complete site preparation activities and be ready for a of irreversible commitments and sunk costs. Tus, projects second unit in case demand grows beyond anticipated level. focus efort on essential elements, whose development pro- Likewise, a university may initially only migrate the human vides the shortest path to an integrated solution [158]. Tis resource management system towards the new information strategy reduces commitment by avoiding the additional ef- platform, but the platform may include from the beginning fort involved in cultivating fexibility through modular archi- the option to add other systems. Options to abandon or tectures and interfaces. In addition, its swif implementation delay the project or parts of it may help address downward may signal decisiveness and help to structure the fuid post- trends. Tese considerations are summed up in the following discontinuity context of the meso-level system. Tese argu- proposition. ments are synthesized in the following proposition. Proposition 7. Te presence of stabilizing unfolding and Proposition 6. Te presence of destabilizing unfolding and downward conditioning produces a dynamic of amplifcation, downward conditioning produces a dynamic of discontinuity, whose consequence of insufciency is most likely to be efectively whose consequence of irrelevance is most likely to be efectively addressed by strategies that prepare options for the project. addressed by strategies that prepare the project for deep iterations. Acceleration. Lastly, the dynamic depicted in the lower right Amplifcation. Te dynamic depicted in the upper right quadrant of Figure 3 is likely to be present in a context corner of Figure 3 occurs in the presence of stabilizing mech- of upward nonadditivity and stabilizing unfolding. Tis anisms involving vertical interactions between the focal sys- dynamicoccurswhenpositivefeedbackmechanismsinvolv- tem and higher levels systems. Tese self-organizing or struc- ing the meso-level system and its components increase the turing processes produce important and continuous change pace of change in the system while accentuating inertia andimposeastronginertiaonthisdynamic,butpositive through path dependency. With respect to project environ- feedback between levels also makes it hard to predict the end ments, this kind of interaction can be observed between state of the system; small variations are likely to be amplifed industrial sectors and frms. A notorious example is high into major diferences. In the case of project environments, velocity sectors [79], which maintain an accelerated pace of such as markets, this kind of dynamic can be observed change, such as that captured by Moore’s Law. Micro level when vertical interactions with the societal level produce entities, such as frms, perceive and eventually take for signifcant growth. For example, sectors proposing a series granted this pace and synchronize their internal processes, of innovative technologies, particularly those providing a such as new technology and product development, with it. In new kind of infrastructure, may induce society to redirect resources towards these sectors. In turn, these resources doing so, they collectively reproduce the pace [78]. If the enable further investment in innovation, which increases meso-level system is an artifact, similar self-reinforcing acce- product functionality, performance, and reliability, which in- leration or pacing may occur between the rhythm of acti- duces new swaths of society to pour resources into the sector. vity or change at the system level and the actions of various A current example of such processes is the growing “smart” participants involved in its development. sector of economy, involving smart phones, TVs, home appli- Te challenge of acceleration is the continuing advance- ances, houses, cities, and governments. Despite the continuity ment of relevant knowledge and the constant diversifcation of trends, amplifcation of small deviations makes it difcult of required skills, which threatens to make solutions and to predict where dynamic processes will bring the sector decisions obsolete. Project participants may cope by pacing [159]. In the case of artifacts, similar interactions between the activities and capability renewal on the rhythm of meso-level project and its clients may lead to growing demand or, on the advancements [155, 163]. However, in spite of the relative pre- contrary, to a downward spiral of mistrust. dictability of trends and pace, acceleration makes it difcult to Te typical consequence of this dynamic for a project is follow changes. As Bergvall-Kareborn˚ and Howcrof ([164]: insufciency, for example, in terms of project scope and 425) put it, “changes ripple through and accentuate ongoing capacity, or, in case of a downward trend, overcapacity. In trends and developments.” For project participants, compo- thecaseofinsufciency,astrategybasedoniterationsmay nent interactions and nonadditive processes may unfold too increase project vulnerability because of their parsimonious rapidly to be understood in a timely manner; the project may nature.Tecontinuityofchangeandtheinertiaimposedon escape their control and run away towards an unpredictable the meso-level by the downward conditioning mechanisms outcome [165]. Trying to follow all intersecting strands with maintain the relevance of signifcant portions of the project, an agile strategy is likely to overwhelm the cognitive and which makes starting all over unneeded. Downward con- adaptive capacities of actors and organizations, even if these ditioning is also likely to boost interactions and prevent are maintained at the edge of chaos. Complexity 11

A strategy with better chances of addressing this complex Likewise, a context of hypercompetition [91] may involve dynamic would be to separate the project into independent efervescence regarding markets preferences and competitor parts, allowing participants to develop and update primarily moves and acceleration with respect to the advancement of the specialized knowledge required for the part in which they technological frontiers. Projects are likely to respond with a are involved and to only track and respond to a subset of combination of strategies. Yet, many requirements of alter- overall changes. Te unfolding inertia created by feedback native strategies are incompatible; some, such as modularity, between the meso-level system and lower-level entities and require intense front-end preparation, while others, such as upward nonadditivity processes enables a relatively continu- iteration, require fast action. Terefore, projects are more ous dynamic, which provides a chance for durable partitions likely to be efective if they are able to cultivate a sort of of this kind. In essence, this strategy cultivates modularity ambidexterity [179], which enables the coexistence of various [166], splitting the project into semiautonomous modules strategies, within the same organization. interacting through limited and well-defned interfaces that contain most changes locally and regularize the infuences 4. Discussion and Conclusions between modules [167, 168]. Te resulting fexibility is twofold. On the one hand, modularity enables responding Complexity has a major impact on projects, through execu- to changes that afect each module without afecting the tion failures, changes, delays, additional costs, dissatisfaction, others, and, on the other hand, it simplifes the less frequent and accidents [180]. Te project management literature has replacement or rearrangement of modules [169, 170]. Despite depicted this impact as resulting from a variety of factors, imposingwhatmayappearintheshorttermasanarchi- which either act independently or interact in unclear ways. tectural straightjacket, evidence suggests that modularity Our theorizing suggests that this impact results from a enables faster as well as less disruptive change in the long conjunction of antecedents and mechanisms to produce two term [171]. With respect to artifacts and their development kinds of consequences. On the one hand, intrinsic prop- activities, modularization requires a partition that minimizes erties and representation shortcomings combine to induce the fows of information, energy, and so on between modules, deviations of project form and behavior from the expected as well as the number of design iterations that would cross stable confguration or controllable variation. On the other interfaces, perhaps isolating the subsystems that appear hand, emergence and unfolding processes combine to induce more likely to be impacted by future change [172–174]. For deviations from the anticipated patterns of activity and project organizations or networks, a similar structure enables change, expected to be regular or predictable. Our theorizing knowledge specialization and minimizes interactions across efort resulted in four categories for each of these two kinds unit or frm boundaries, perhaps following the fault lines of of consequences. While relying on abstract notions derived technical architectures. However, depending on the nature of from fundamental research on complexity, this efort creates artifacts, particularly of the interactions between their parts, a conceptual framework with a moderate level of abstraction. modularization also requires an overarching “system inte- Toourknowledge,thisisoneofthefewattemptstoadopt grator” unit or frm [175]. Modularization can be extended such a strategy in project management research on project into project environments, such as a markets, via the proac- complexity. tive standardization of technical architectures and interfaces Tis framework contributes to the integration of the two [176], and the development of alliances that would impose streams of research that addressed project complexity from a particular architecture, platform, or “stack” as a de facto generic and highly abstract perspectives and, respectively, standard [177, 178]. Tese arguments are summarized in the from a practical viewpoint by relying on managers’ opinions. following proposition. Because of the profusion of concepts advanced by these streams, most attempts at integration had, so far, to rely on Proposition 8. Te presence of stabilizing unfolding and a quasi-bibliometric approach to inventory and classify the upward nonadditivity produces a dynamic of acceleration, terms used to discuss project complexity [6, 181]. We believe whose consequence of obsolescence is most likely to be efectively that our original, theory-driven integration approach and addressed by strategies that increase the modularity of the the moderate level of abstraction of the resulting framework project. provide two benefts. On the one hand, they open the black box of complexity and help ground the understand- Because they may involve diferent levels, it is possible ing of complexity factors in fundamental notions such as to observe in the same project a combination of dynamics emergence. Tis enabled the creation of categories with a showninthesamecolumninFigure3.Forexample,ongoing deeper theoretical meaning, tied to essential aspects and efervescence would be punctuated by relatively less frequent relevant consequences of project complexity, rather than to major discontinuities. Likewise a dynamic of acceleration commonsense notions such as technology and organization. inside a sector may be combined with a dynamic of amplifca- On the other hand, the categories we propose remain easy tion with respect to a broader society. Also, diferent aspects to connect to concrete aspects of projects. Terefore, they of the project may be subject to dynamics pertaining to dif- can help researchers make sense of the rich set of factors ferent columns. For example, the market environment may be identifed by the practical stream of complexity research and subject to a dynamic of amplifcation, while the institutional guide them towards understanding how each factor becomes environment may face discontinuities caused by accession a complexity antecedent or mechanism trigger. Of course, any to power of politicians sharing radically diferent ideologies. increase in abstraction may result in losing some empirically 12 Complexity derived richness. Terefore, further research should rely on and could inspire the development of new complexity man- case studies and grounded theorizing to map back concrete agement practices. complexity factors upon the more general categories included Our results also provide insights that may contribute to in our framework. other felds of research. First of all, they could help advance A second important contribution of this paper is theo- the more abstract conceptualizations of complexity, in par- rizingaboutthestrategiesthatcanbeusedtopreventorto ticular by helping identify commonalities and connections mitigate the impact of complexity. Contributions regarding between concepts used by various researchers. Also, taken strategies applied specifcally to address project complexity together, the results of our theorizing efort portray the are even less frequent than those discussing the nature and becoming of complex projects as unpredictable and uncon- sources of complexity. Our contribution includes, on the trollable emergence taking place in a context of surprising one hand, theorizing the knowledge- and representation- events and irregular unfolding. Tis could contribute to the producing strategies that are efective in preventing the struc- emerging research on organizing as a process or an event sys- tural consequences of complexity. On the other hand, we the- tem. Besides, results on representational and organizational orized strategies for organizing project activities, namely, by strategies for addressing complexity can inform other funda- connecting them simultaneously as well as intertemporally, mental theories of collaboration and organization. to mitigate the dynamic consequences of complexity. Once again,foreachofthesetwotypesofstrategies,wepropose Conflicts of Interest four categories with mid-range of abstraction. Tis level of abstraction enables the creation of meaningful categories for Te authors declare that there are no conficts of interest re- a host of concrete project management practices observed latedtothispaper. across a variety of domains. Further research could inventory such practices and connect them to the categories of our References framework. Te fundamental grounding of this framework would guide researchers towards understanding how these [1] A. J. Shenhar, “One size does not ft all projects: exploring classi- practices block or mitigate detrimental emergence processes cal contingency domains,” Management Science,vol.47,no.3, or respond to irregular or unexpected dynamics. Tese results pp.394–414,2001. could, in turn, inform research on projects as organizations [2] D. P. Tegarden, S. D. Sheetz, and D. E. Monarchi, “A sofware and networks, in particular in what is related to their complexity model of object-oriented systems,” Decision Support fexibility and response capacity [90], along with the research Systems,vol.13,no.3-4,pp.241–262,1995. on knowledge production in projects [26], and even on [3] T. M. Williams, “Te need for new paradigms for complex knowledge management and interproject transfer in project- projects,” International Journal of Project Management,vol.17, based organizations [115]. no.5,pp.269–273,1999. A third contribution of this article is using the theoretical [4] T.Katsila,G.A.Spyroulias,G.P.Patrinos,andM.-T.Matsoukas, framework to develop propositions about the most appro- “Computational approaches in target identifcation and drug discovery,” Computational and Structural Biotechnology Journal, priate strategies for addressing each kind of project com- vol. 14, pp. 177–184, 2016. plexity. With some notable exceptions [15, 114], past research [5]J.Geraldi,H.Maylor,andT.Williams,“Now,let’smakeitreally hadproducedfewtheoreticalmodelsaboutthestrategiesthat complex (complicated): A systematic review of the complexities are efective in addressing various kinds of complexity and of projects,” International Journal of Operations and Production even fewer results that have been validated in a variety of Management, vol. 31, no. 9, pp. 966–990, 2011. empirical settings. Terefore future empirical research could [6]J.Bakhshi,V.Ireland,andA.Gorod,“Clarifyingtheproject corroborate these propositions both through longitudinal complexity construct: Past, present and future,” International case studies that would investigate how various strategies Journal of Project Management,vol.34,no.7,pp.1199–1213,2016. work, perhaps in diferent project stages, as well as through [7]T.Cooke-Davies,S.Cicmil,L.Crawford,andK.Richardson, cross-sectional studies that could rely on diferences sug- “We’re not in Kansas anymore, Toto: Mapping the strange gested by our framework with regard to the nature of com- landscape of complexity theory, and its relationship to project plexity in various industries to determine whether the pro- mangement,” IEEE Engineering Management Review,vol.36,no. posed strategies are indeed more efective in addressing 2, pp. 5–21, 2008. various categories of complexity. One particular stream of [8]M.Bosch-Rekveldt,Y.Jongkind,H.Mooi,H.Bakker,andA. research could focus on deepening our understanding of the Verbraeck, “Grasping project complexity in large engineering interrelations of the various types of complexity and strategies projects: Te TOE (Technical, Organizational and Environmen- inaproject.Inadditiontograspingthepolyphonyofcom- tal) framework,” International Journal of Project Management, vol.29,no.6,pp.728–739,2011. plexity and the mutual exclusivity of strategies in a project, [9] R. J. Chapman, “Aframework for examining the dimensions and this stream could lead to the development of new research characteristics of complexity inherent within rail megaprojects,” programs on the management of project paradoxes [182] International Journal of Project Management,vol.34,no.6,pp. and the ambidexterity of project organizations [179, 183]. In 937–956, 2016. practical terms, these results may suggest what practices or [10] A.T.Nguyen,L.D.Nguyen,L.Le-Hoai,andC.N.Dang,“Quan- combinations of practices are best suited for addressing the tifying the complexity of transportation projects using the fuzzy specifc forms of complexity that afect a variety of domains, analytic hierarchy process,” International Journal of Project such as sofware, biotech, and infrastructure construction, Management,vol.33,no.6,pp.1364–1376,2015. Complexity 13

[11] L.-A. Vidal, F. Marle, and J.-C. Bocquet, “Measuring project [31] A. M. Turing, “On Computable Numbers, with an Application complexity using the Analytic Hierarchy Process,” International to the Entscheidungsproblem,” Proceedings of the London Math- Journal of Project Management,vol.29,no.6,pp.718–727,2011. ematical Society. Second Series,vol.42,no.3,pp.230–265,1936. [12] B. Dao, S. Kermanshachi, J. Shane, and S. Anderson, “Project [32] L. Biggiero, “Sources of complexity in human systems,” Nonlin- Complexity Assessment and Management Tool,” in Proceedings ear Dynamics, Psychology, and Life Sciences, vol. 5, no. 1, p. 19, of the International Conference on Sustainable Design, Engineer- 2001. ing and Construction, ICSDEC 2016, pp. 491–496, usa, May 2016. [33]P.F.Katina,C.B.Keating,andR.M.Jaradat,“Systemrequire- [13] W. Xia and G. Lee, “Complexity of information systems devel- ments engineering in complex situations,” Requirements Engi- opment projects: Conceptualization and measurement develop- neering,vol.19,no.1,pp.45–62,2014. ment,” Journal of Management Information Systems,vol.22,no. [34] A. N. Kolmogorov, “Tree approaches to the quantitative defni- 1,pp.45–83,2005. tion of information,” International Journal of Computer Mathe- [14] J. W. Meyer and B. Rowan, “Institutionalized organizations: matics. Section A. Programming Teory and Methods. Section B. formal structure as myth and ceremony,” American Journal of Computational Methods,vol.2,pp.157–168,1968. Sociology,vol.83,no.2,pp.340–363,1977. [35] B. Goertzel, “Measuring static complexity,” International Jour- [15] S. Floricel, J. L. Michela, and S. Piperca, “Complexity, uncertain- nal of Mathematics and Mathematical Sciences,vol.15,no.1,pp. ty-reduction strategies, and project performance,” International 161–174, 1992. Journal of Project Management,vol.34,no.7,pp.1360–1383, [36] M. Gell-Mann and S. Lloyd, “Information measures, efective 2016. complexity, and total information,” Complexity,vol.2,no.1,pp. [16] H. Maylor and N. Turner, “Understand, reduce, respond: project 44–52, 1996. complexity management theory and practice,” International [37] U. Sauer, M. Heinemann, and N. Zamboni, “Getting closer to Journal of Operations and Production Management,vol.37,no. the whole picture,” Science,vol.316,no.5824,pp.550-551,2007. 8, pp. 1076–1093, 2017. [38] N. Crossley, Towards Relational Sociology, Routledge, London, [17] H.-J. Rheinberger, “Experimental complexity in biology: Some UK, 2011. epistemological and historical remarks,” Philosophy of Science, [39] R. K. Sawyer, “Te mechanisms of emergence,” Philosophy of the vol.64,no.4,pp.S245–S254,1997. Social Sciences,vol.34,no.2,pp.260–282,2004. [18] S. J. Whitty and H. Maylor, “And then came complex project management (revised),” International Journal of Project Man- [40] W. C. Wimsatt, “Aggregate, composed, and evolved systems: agement,vol.27,no.3,pp.304–310,2009. Reductionistic heuristics as means to more holistic theories,” Biology and Philosophy,vol.21,no.5,pp.667–702,2006. [19] M. Bunge, Treatise on Basic Philosophy: Ontology II: A World of Systems,Dordrecht,Holland:D.Reidel,1979. [41] P. W. Andersen, “More is diferent,” Science,vol.177,no.4047, pp. 393–396, 1972. [20] D. Rind, “Complexity and climate,” Science,vol.284,no.5411, pp. 105–107, 1999. [42] C. Alexander, Notes on the Synthesis of Form,HarvardUniver- sity Press, Cambridge, MA, 1964. [21] E. Mayr, “Biology in the twenty-frst century,” Bioscience,vol.50, no. 10, p. 895, 2000. [43] P. Anderson, “Complexity theory and organization science,” Organization Science,vol.10,no.3,pp.216–232,1999. [22] M. DeLanda, Digital Tectonics,N.Leach,D.Turnbull,andC. Williams, Eds., John Wiley & Sons, Chichester, 2004. [44] J. M. Lehn, “Toward self-organization and complex matter,” Science,vol.295,no.5564,pp.2400–2403,2002. [23] G. Weng, U. S. Bhalla, and R. Iyengar, “Complexity in biological signaling systems,” Science,vol.284,no.5411,pp.92–96,1999. [45] G. Simondon, Du mode d’existence des objets techniques,Aubier, Paris, 1989. [24] A. Menges, “Material computation: Higher integration in mor- phogenetic design,” Architectural Design,vol.82,no.2,pp.14–21, [46] K. Barad, “Posthumanist performativity: Toward an under- 2012. standing of how matter comes to matter,” Signs,vol.28,no.3, pp.801–831,2003. [25]A.vanMarrewijk,S.R.Clegg,T.S.Pitsis,andM.Veenswijk, “Managing public-private megaprojects: Paradoxes, complexity, [47] W. J. Orlikowski, “Sociomaterial practices: Exploring technol- and project design,” International Journal of Project Manage- ogy at work,” Organization Studies,vol.28,no.9,pp.1435–1448, ment,vol.26,no.6,pp.591–600,2008. 2007. [26] S.Floricel,J.L.Michela,M.George,andL.withBonneau,Refn- [48] M. Callon, “Some elements of a sociology of translation: ing the Knowledge Production Plan: Knowledge Representations Domestication of the scallops and the fshermen of St Brieuc In Innovation Projects, Project Management Institute, Newton Bay,” Te Sociological Review,vol.32,no.1,pp.196–233,1984. Square, PA, 2011. [49]S.Floricel,C.Bonneau,M.Aubry,andV.Sergi,“Extending [27] K. Henderson, “Te Political Career of a Prototype: Visual project management research: Insights from social theories,” Representation in Design Engineering,” Social Problems,vol.42, International Journal of Project Management,vol.32,no.7,pp. no. 2, pp. 274–299, 1995. 1091–1107, 2014. [28] R. Kline, “Science and engineering theory in the invention and [50] H. A. Simon, Te Sciences of the Artifcial, MIT Press, Cam- development of the induction motor,” Technology and Culture, bridge, MA, 2nd edition, 1981. vol. 28, no. 2, pp. 283–313, 1987. [51] Y. Bar-Yam, “A mathematical theory of strong emergence using [29] P. Nightingale, “A cognitive model of innovation,” Research multiscale variety,” Complexity,vol.9,no.6,pp.15–24,2004. Policy,vol.27,no.7,pp.689–709,1998. [52] K. M. Kontopoulos, Te logics of social structure, Cambridge [30] M. Blum, “A machine-independent theory of the complexity of University Press, Cambridge, UK, 1993. recursive functions,” Journal of the ACM,vol.14,pp.322–336, [53] C. Adami, “What is complexity?” BioEssays,vol.24,no.12,pp. 1967. 1085–1094, 2002. 14 Complexity

[54] P. Kohl and D. Noble, “Systems biology and the virtual physio- [75] C. E. Helfat and M. A. Peteraf, “Te dynamic resource-based logical human,” Molecular Systems Biology,vol.5,no.1,pp.1–6, view: Capability lifecycles,” Strategic Management Journal,vol. 2009. 24, no. 10, pp. 997–1010, 2003. [55] D. W.McShea, “Te hierarchical structure of organisms: A scale [76] S. Klepper, “Industry life cycles,” Industrial and Corporate and documentation of a trend in the maximum,” Paleobiology, Change,vol.6,no.1,pp.145–181,1997. vol.27,no.2,pp.405–423,2001. [77] R. M. Kramer, “Trust and distrust in organizations: emerging [56] O. Hanseth, E. Monteiro, and M. Hatling, “Developing informa- perspectives, enduring questions,” Annual Review of Psychology, tion infrastructure: Te tension between standardization and vol. 50, pp. 569–598, 1999. fexibility,” Science, Technology & Human Values,vol.21,no.4, [78] W.C.BognerandP.S.Barr,“MakingSenseinHypercompetitive pp.407–426,1996. Environments: A Cognitive Explanation for the Persistence of [57] K. J. Klein, F. Dansereau, and R. J. Hall, “Levels issues in theory High Velocity Competition,” Organization Science,vol.11,no.2, development, data collection, and analysis,” Academy of Man- pp.212–226,2000. agement Review (AMR),vol.19,no.2,pp.195–229,1994. [79] K. M. Eisenhardt, “Making fast strategic decisions in high-velo- [58] S. Brusoni, A. Prencipe, and K. Pavitt, “Knowledge specializa- city environments,” Academy of Management Journal (AMJ), tion, organizational coupling, and the boundaries of the frm: vol.32,no.3,pp.543–576,1989. Why do frms know more than they make?” Administrative [80] S. Floricel and D. Dougherty, “Where do games of innovation Science Quarterly,vol.46,no.4,pp.597–621,2001. come from? Explaining the persistence of dynamic innovation [59] R. Drazin, M. A. Glynn, and R. K. Kazanjian, “Multilevel theo- patterns,” International Journal of Innovation Management,vol. rizing about creativity in organizations: A sensemaking per- 11, no. 1, pp. 65–92, 2007. spective,” Academy of Management Review (AMR),vol.24,no. 2, pp. 286–307, 1999. [81] H. Benbya and B. McKelvey, “Using coevolutionary and com- plexity theories to improve IS alignment: A multi-level ap- [60] M. Engwall, “No project is an island: Linking projects to history proach,” Journal of Information Technology,vol.21,no.4,pp. and context,” Research Policy,vol.32,no.5,pp.789–808,2003. 284–298, 2006. [61] D. H. Gobeli, H. F. Koenig, and I. Bechinger, “Managing Con- [82] A. MacCormack, R. Verganti, and M. Iansiti, “Developing prod- fict in Sofware Development Teams: A Multilevel Analysis,” ucts on “Internet time”: Te anatomy of a fexible development Journal of Product Innovation Management,vol.15,no.5,pp. process,” Management Science,vol.47,no.1,pp.133–150,2001. 423–435, 1998. [83] R. Miller and D. R. Lessard, “Te Strategic Management of [62] R. Miller, M. Hobday, T. Leroux-demers, and X. Olleros, “Inno- Large Engineering Projects: Shaping Institutions, Risks, and vation in complex systems industries: Te case of fight simu- Governance,” Cambridge, MA: MIT Press. lation,” Industrial and Corporate Change,vol.4,no.2,pp.363– 400, 1995. [84] S. Kwinter, “Te computational fallacy,” Tresholds,vol.26,pp. 90–92, 2003. [63] R. Tom, “Stabilitestructurelleetmorphogen´ ese,”` Poetics,vol. 10,no.C,pp.7–19,1974. [85] R. House, Te Human Side of Project Management, Addison- [64] W.B. Arthur, “Competing technologies, increasing returns, and Wesley, Reading, Mass, 1988. lock-in by historical events,” Te Economic Journal,vol.99,no. [86] M. Aubry, R. Muller,¨ B. Hobbs, and T. Blomquist, “Project man- 394, p. 116, 1989. agement ofces in transition,” International Journal of Project [65] G. M. Whitesides and R. F. Ismagilov, “Complexity in chem- Management, vol. 28, no. 8, pp. 766–778, 2010. istry,” Science,vol.284,no.5411,pp.89–92,1999. [87] A. Davies, D. Gann, and T. Douglas, “Innovation in megapro- [66] K. J. Dooley and A. H. Van De Ven, “Explaining Complex jects: Systems integration at London Heathrow terminal 5,” Organizational Dynamics,” Organization Science,vol.10,no.3, California Management Review,vol.51,no.2,pp.101–125,2009. pp. 358–372, 1999. [88] T. Y. Hsieh, S. T. Lu, and C. H. Wu, “Statistical analysis of causes [67] S. Floricel and S. Piperca, “Project management between will for change orders in metropolitan public works,” International and representation,” Project Management Journal,vol.47,no.3, Journal of Project Management, vol. 22, no. 8, pp. 679–686. pp. 124–138, 2016. [89] M. Sun and X. Meng, “Taxonomy for change causes and ef- [68] F. C. Saunders, “Toward high reliability project organizing in fects in construction projects,” International Journal of Project safety-critical projects,” Project Management Journal,vol.46, Management,vol.27,no.6,pp.560–572,2009. no. 3, pp. 25–35, 2015. [90] S. Floricel, S. Piperca, and M. Banik, Increasing Project Flexibil- [69] I. Prigogine, Te End of Certainty, Free Press, New York, NY, ity: Te Response Capacity of Complex Projects,ProjectManage- USA, 1997. ment Institute, Newton Square, PA, 2011. [70] H. R. Maturana and F. Varela, Autopoiesis and Cognition: the [91] R. A. D’Aveni, Hypercompetition: Managing the Dynamics of Realization of the Living, Reidel, Dordrecht, 1980. Strategic Maneuvering, Free Press, New York, NY, USA, 1994. [71] N. Luhmann, Social systems, Stanford University Press, 1995. [92] A. D. Meyer, V. Gaba, and K. A. Colwell, “Organizing far from [72] S. Wolfram, “Cellular automata as models of complexity,” equilibrium: Nonlinear change in organizational felds,” Orga- Nature,vol.311,no.5985,pp.419–424,1984. nization Science, vol. 16, no. 5, pp. 456–473, 2005. [73] J. S. Coleman, “Foundations for a theory of collective deci- [93] D. Polley, “Turbulence in Organizations: New Metaphors for sions.,” American Journal of Sociology,vol.71,no.6,pp.615–627, Organizational Research,” Organization Science,vol.8,no.5,pp. 1966. 445–457, 1997. [74] M. S. Feldman and B. T. Pentland, “Reconceptualizing orga- [94] J. Sydow, G. Schreyogg,¨ and J. Koch, “Organizational path nizational routines as a source of fexibility and change,” dependence: Opening the black box,” Academy of Management Administrative Science Quarterly,vol.48,no.1,pp.94–118,2003. Review (AMR),vol.34,no.4,pp.689–709,2009. Complexity 15

[95] D. J. Teece, G. Pisano, and A. Shuen, “Dynamic capabilities and [116] W. M. Cohen and D. A. Levinthal, “Absorptive capacity: A new strategic management,” Strategic Management Journal,vol.18, perspective of learning and innovation,” Administrative Science no.7,pp.509–533,1997. Quarterly,vol.35,no.1,pp.128–152,1990. [96] H. Tsoukas and R. Chia, “On organizational becoming: rethink- [117] R. A. Burgelman, “Intraorganizational Ecology of Strategy Mak- ing organizational change,” Organization Science,vol.13,no.5, ing and Organizational Adaptation: Teory and Field Research,” pp. 567–582, 2002. Organization Science,vol.2,no.3,pp.239–262,1991. [97] T. Hernes, Understanding Organization as Process: Teory for a [118]B.IyerandT.H.Davenport,“ReverseengineeringGoogles Tangled World,Routledge,NewYork,NY,USA,2008. innovation machine,” Harvard Business Review,vol.86,no.4, [98] T. Hernes and E. Weik, “Organization as process: Drawing a pp. 58–134, 2008. line between endogenous and exogenous views,” Scandinavian [119] M. L. Markus, A. Majchrzak, and L. Gasser, “Adesign theory for Journal of Management,vol.23,no.3,pp.251–264,2007. systems that support emergent knowledge processes,” MIS [99] S. Collyer and C. M. J. Warren, “Project management approach- Quarterly: Management Information Systems,vol.26,no.3,pp. es for dynamic environments,” International Journal of Project 179–212, 2002. Management,vol.27,no.4,pp.355–364,2009. [120]R.F.Hirsh,Technology and Transformation in the American [100] S. C. Sommer and C. H. Loch, “Selectionism and learning in Electric Utility Industry, Cambridge University Press, Cam- projects with complexity and unforeseeable uncertainty,” Man- bridge, UK, 1989. agement Science, vol. 50, no. 10, pp. 1334–1347, 2004. [121]T.A.CarboneandD.D.Tippett,“Projectriskmanagement [101] W. R. Ashby, “Requisite Variety and its implications for the using the project risk fmea,” EMJ - Engineering Management control of complex systems,” Cybernetica (Namur),vol.1,no.2, Journal,vol.16,no.4,pp.28–35,2004. pp.83–99,1958. [122] A. Qazi, J. Quigley, A. Dickson, and K. Kirytopoulos, “Project [102] N. Luhmann, Risk: A sociological theory, Aldine De Gruyter, Complexity and Risk Management (ProCRiM): Towards mod- New York, NY, USA, 1993. elling project complexity driven risk paths in construction projects,” International Journal of Project Management,vol.34, [103] F. P.Morgeson, T. R. Mitchell, and D. Liu, “Event system theory: no.7,pp.1183–1198,2016. An event-oriented approach to the organizational sciences,” Academy of Management Review (AMR),vol.40,no.4,pp.515– [123] C. Perrow, Normal accidents: Living with high risk technologies, 537, 2015. BasicBooks,NewYork,NY,USA,1984. [124]L.A.Kappelman,R.McKeeman,andL.Zhang,“Earlywarning [104] A. D. Meyer, “Adapting to environmental jolts.,” Administrative signs of IT project failure: Te dominant dozen,” Information Science Quarterly,vol.27,no.4,pp.515–537,1982. Systems Management,vol.23,no.4,pp.31–36,2006. [105] F. E. Emery and E. L. Trist, “Te causal texture of organizational [125] R. M. Henderson and K. B. Clark, “Architectural innovation: environments,” Human Relations,vol.18,no.1,pp.21–32,1965. Te reconfguration of existing product technologies and the [106] A. Soderholm, “Project management of unexpected events,” failure of established frms,” Administrative Science Quarterly, International Journal of Project Management,vol.26,no.1,pp. vol.35,pp.9–30,1990. 80–86, 2008. [126] K. E. Weick and K. H. Roberts, “Collective mind: Heedful [107] S. Piperca and S. Floricel, “A typology of unexpected events in interrelations on fight decks,” Adminstrative Science Quarterly, complex projects,” International Journal of Managing Projects in vol. 38, pp. 357–381, 1993. Business,vol.5,no.2,pp.248–265,2012. [127]K.E.Weick,K.M.Sutclife,andD.Obstfeld,“Organizingfor [108] S. Floricel and R. Miller, “Strategizing for anticipated risks and high reliability: Processes of collective mindfulness,” Crisis turbulence in large-scale engineering projects,” International management,vol.3,no.1,pp.81–123,2008. Journal of Project Management,vol.19,no.8,pp.445–455,2001. [128] M. T. Hansen, “Te search-transfer problem: the role of weak [109] P. Hedstrom¨ and R. Swedberg, “Social Mechanisms,” Acta ties in sharing knowledge across organization subunits,” Admin- Stereologica,vol.39,no.3,pp.281–308,1996. istrative Science Quarterly, vol. 44, no. 1, pp. 82–111, 1999. [110] G. F. Davis and C. Marquis, “Prospects for organization theory [129] M. Zollo and S. G. Winter, “Deliberate learning and the evo- in the early twenty-frst century: Institutional felds and mech- lution of dynamic capabilities,” Organization Science,vol.13,no. anisms,” Organization Science,vol.16,no.4,pp.332–343,2005. 3,pp.339–351,2002. [111] C. H. Loch and C. Terwiesch, “Communication and uncertainty [130] A. O’Sullivan, “Why tense, unstable, and diverse relations are in concurrent engineering,” Management Science,vol.44,no.8, inherent in co-designing with suppliers: An aerospace case pp.1032–1048,1998. study,” Industrial and Corporate Change,vol.15,no.2,pp.221– [112] S. K. Chandrasegaran, K. Ramani, R. D. Sriram et al., “Te 250, 2006. evolution, challenges, and future of knowledge representation [131] K. E. Weick, K. M. Sutclife, and D. Obstfeld, “Organizing and in product design systems,” Computer-Aided Design,vol.45,no. the process of sensemaking,” Organization Science,vol.16,no. 2, pp. 204–228, 2013. 4, pp. 409–421, 2005. [113] W. Visser, “Designing as construction of representations: A [132] V. Krishnan, S. D. Eppinger, and D. E. Whitney, “A model- dynamic viewpoint in cognitive design research,” Human–Com- based framework to overlap product development activities,” puter Interaction,vol.21,no.1,pp.103–152,2006. Management Science,vol.43,no.4,pp.437–451,1997. [114] T. Ahern, B. Leavy, and P. J. Byrne, “Complex project man- [133]S.D.Eppinger,D.E.Whitney,R.P.Smith,andD.A.Gebala,“A agement as complex problem solving: A distributed knowledge model-based method for organizing tasks in product develop- management perspective,” International Journal of Project Man- ment,” Research in Engineering Design,vol.6,no.1,pp.1–13, agement,vol.32,no.8,pp.1371–1381,2014. 1994. [115] A. Prencipe and F. Tell, “Inter-project learning: Processes and [134] D. Nicolini, J. Mengis, and J. Swan, “Understanding the role of outcomes of knowledge codifcation in project-based frms,” objects in cross-disciplinary collaboration,” Organization Sci- Research Policy,vol.30,no.9,pp.1373–1394,2001. ence,vol.23,no.3,pp.612–629,2012. 16 Complexity

[135] S. Tomke and T. Fujimoto, “Efect of ‘front-loading’ problem- [155] S. L. Brown and K. M. Eisenhardt, “Te art of continuous solving on product development performance,” Journal of Prod- change: linking complexity theory and time-paced evolution uct Innovation Management,vol.17,no.2,pp.128–142,2000. in relentlessly shifing organizations,” Administrative Science [136] P. R. Carlile, “A pragmatic view of knowledge and boundaries: Quarterly,vol.42,no.1,pp.1–34,1997. Boundary objects in new product development,” Organization [156] D. A. Levinthal, “Adaptation on rugged landscapes,” Manage- Science,vol.13,no.4,pp.442–456,2002. ment Science,vol.43,no.7,pp.934–950,1997. [137] B. Ewenstein and J. Whyte, “Knowledge practices in design: Te [157] J. Whyte, A. Stasis, and C. Lindkvist, “Managing change in the role of visual representations as ’Epistemic objects’,” Organiza- delivery of complex projects: Confguration management, asset tion Studies,vol.30,no.1,pp.7–30,2009. information and ’big data’,” International Journal of Project [138] Y. Liu, S. van Nederveen, and M. Hertogh, “Understanding Management,vol.34,no.2,pp.339–351,2016. efects of BIM on collaborative design and constructionAn [158] N. O. E. Olsson, “Management of fexibility in projects,” Inter- empirical study in China,” International Journal of Project Man- national Journal of Project Management,vol.24,no.1,pp.66–74, agement, vol. 35, no. 4, pp. 686–698, 2017. 2006. [139] I. Stigliani and D. Ravasi, “Organizing thoughts and connecting [159] R. G. Lord, J. E. Dinh, and E. L. Hofman, “Aquantum approach brains: Material practices and the transition from individual to time and organizational change,” Academy of Management to group-level prospective sensemaking,” Academy of Manage- Review (AMR), vol. 40, no. 2, pp. 263–290, 2015. ment Journal (AMJ),vol.55,no.5,pp.1232–1259,2012. [160] L. Trigeorgis, Real options: Managerial fexibility and strategy in [140] N. N. Taleb, “Black swans and the domains of statistics,” Te resource allocation, MIT press, Boston, MA, 1996. American Statistician, vol. 61, no. 3, pp. 198–200, 2007. [161] M. J. Garvin and D. N. Ford, “Real options in infrastructure [141] N.Leveson,N.Dulac,K.Marais,andJ.Carroll,“Movingbeyond projects: Teory, practice and prospects,” Engineering Project Normal Accidents and High Reliability Organizations: A Sys- Organization Journal,vol.2,no.1-2,pp.97–108,2012. tems Approach to Safety in Complex Systems,” Organization [162] R. G. Fichman, M. Keil, and A. Tiwana, ““Options thinking” in Studies,vol.30,no.2-3,pp.227–249,2009. IT project management,” California Management Review,vol. [142]G.Sliwoski,S.Kothiwale,J.Meiler,andE.W.LoweJr.,“Com- 47, no. 2, pp. 74–96, 2005. putational methods in drug discovery,” Pharmacological Re- views,vol.66,no.1,pp.334–395,2014. [163] P. Klarner and S. Raisch, “Move to the beat-rhythms of change and frm performance,” Academy of Management Journal [143] C. G. Begley and L. M. Ellis, “Drug development: Raise stand- (AMJ), vol. 56, no. 1, pp. 160–184, 2013. ards for preclinical cancer research,” Nature,vol.483,no.7391, pp. 531–533, 2012. [164] B. Bergvall-Kareborn˚ and D. Howcrof, “Persistent problems and practices in information systems development: A study of [144]W.G.Vincenti,What Engineers Know and How Tey Know It, mobile applications development and distribution,” Information John Hopkins University Press, Baltimore, 1990. Systems Journal,vol.24,no.5,pp.425–444,2014. [145] P. Nightingale and P. Martin, “Te myth of the biotech revolu- [165] C. B. Perrow, “Complexity, catastrophe, and modularity,” Socio- tion,” Trends in Biotechnology,vol.22,no.11,pp.564–569,2004. logical Inquiry,vol.78,no.2,pp.162–173,2008. [146] S. Mandal, M. Moudgil, and S. K. Mandal, “Rational drug design,” European Journal of Pharmacology,vol.625,no.1-3,pp. [166] H. A. Simon, “Te architecture of complexity,” Proceedings of the 90–100, 2009. American Philosophical Society,vol.106,no.6,pp.467–482. [147]G.GavettiandD.Levinthal,“Lookingforwardandlooking [167]R.SanchezandJ.T.Mahoney,“Modularity,fexibility,and backward: cognitive and experiential search,” Administrative knowledge management in product and organization design,” Science Quarterly,vol.45,no.1,pp.113–137,2000. Strategic Management Journal,vol.17,no.2,pp.63–76,1996. [148] L. Fleming and O. Sorenson, “Science as a map in technological [168] N. P. Suh, “Complexity in Engineering,” CIRP Annals - Manu- search,” Strategic Management Journal,vol.25,pp.909–925, facturing Technology,vol.54,no.2,pp.46–63,2005. 2004. [169] K. Ulrich, “Te role of product architecture in the manufactur- [149] R. D. Austin, L. Devin, and E. E. Sullivan, “Accidental inno- ing frm,” Research Policy,vol.24,no.3,pp.419–440,1995. vation: Supporting valuable unpredictability in the creative [170] M. A. Schilling, “Toward a general modular systems theory and process,” Organization Science, vol. 23, no. 5, pp. 1505–1522, its application to interfrm product modularity,” Academy of 2012. Management Review (AMR),vol.25,no.2,pp.312–334,2000. [150] K. R. Popper, Te logic of scientifc discovery, Hutchinson and [171] R. N. Langlois and P.L. Robertson, “Networks and innovation in Co., Ltd., London, 1959. a modular system: lessons from the microcomputer and stereo [151] G. G. Dess and D. W.Beard, “Dimensions of organizational task component industries,” Research Policy,vol.21,no.4,pp.297– environments,” Administrative Science Quarterly,vol.29,no.1, 313, 1992. pp.52–73,1984. [172] J. B. Dahmus, J. P. Gonzalez-Zugasti, and K. N. Otto, “Modular [152]E.C.Conforto,D.C.Amaral,S.L.daSilva,A.DiFelippo,and product architecture,” Design Studies,vol.22,no.5,pp.409–424, D. S. L. Kamikawachi, “Te agility construct on project man- 2001. agement theory,” International Journal of Project Management, [173] K. Holtt¨ a-Otto,¨ N. Chiriac, D. Lysy, and E. S. Suh, “Architectural vol. 34, no. 4, pp. 660–674, 2016. decomposition: Te role of granularity and decomposition [153] J. A. Highsmith, Agile project management: Creating innovative viewpoint,” Advances in Product Family and Product Platform products, Addison-Wesley, Upper Saddle River, NJ, 2nd edition, Design: Methods and Applications, pp. 221–243, 2014. 2010. [174]M.MagnussonandM.Pasche,“Acontingency-basedapproach [154]P.SerradorandJ.K.Pinto,“DoesAgilework?-Aquantitative to the use of product platforms and modules in new product analysis of agile project success,” International Journal of Project development,” Journal of Product Innovation Management,vol. Management,vol.33,no.5,pp.1040–1051,2015. 31, no. 3, pp. 434–450, 2014. Complexity 17

[175] S. Brusoni and A. Prencipe, “Unpacking the black box of modu- larity: technologies, products and organizations,” Industrial and Corporate Change,vol.10,no.1,pp.179–205,2001. [176] M. Wadczak, T. B. Meriem, B. Radier et al., “Standardizing a reference model and autonomic network architectures for the self-managing future internet,” IEEE Network,vol.25,no.6,pp. 50–56, 2011. [177] M. Kenney and B. Pon, “Structuring the smartphone industry: Is the mobile Internet OS platform the key?” Journal of Industry, Competition and Trade, vol. 11, no. 3, pp. 239–261, 2011. [178] A. Gawer and M. A. Cusumano, “Industry platforms and eco- system innovation,” Journal of Product Innovation Management, vol. 31, no. 3, pp. 417–433, 2014. [179] C. A. O’Reilly and M. L. Tushman, “Ambidexterity as a dynamic capability: resolving the innovator’s dilemma,” Research in Organizational Behavior,vol.28,pp.185–206,2008. [180] B. Flyvbjerg, M. Garbuio, and D. Lovallo, “Deception in large infrastructure projects: Two models for explaining and preventing executive disaster,” California Management Review, vol.51,no.2,pp.3–193,2009. [181] M. Padalkar and S. Gopinath, “Are complexity and uncer- tainty distinct concepts in project management? A taxonomical examination from literature,” International Journal of Project Management, vol. 34, no. 4, pp. 688–700, 2016. [182]W.SmithandM.Lewis,“Towardatheoryofparadox:Ady- namic equilibrium model of organizing,” Academy of Manage- ment Review (AMR),vol.36,no.2,pp.381–403,2011. [183] P.Boumgarden, J. Nickerson, and T. R. Zenger, “Sailing into the wind: Exploring the relationships among ambidexterity, vacil- lation, and organizational performance,” Strategic Management Journal,vol.33,no.6,pp.587–610,2012.