Horizontally integrated or vertically divided?
A comparison of outsourcing strategies at Ericsson and Nokia
”Outsourcing is one of the most misunderstood concepts in use today” Jorma Ollila
Working paper October 2002
Lars Bengtsson and Christian Berggren 2 (22)
PREFACE In this report we try to compare the outsourcing strategies of Nokia and Ericsson, delimited to the production of radio base stations. The results from a preliminary version of this report rendered a great interest in media during summer year 2002. The results and implications have, however, in several cases been over-interpreted and even misinterpreted. We thus feel impelled to elucidate the ambition and status of the report.
This is a working paper which aims at illustrating various choices and discuss their consequences, in order to elaborate questions for our further research. A follow-up study, in which we analyse current outsourcing strategies is in progress. The analysis in this report is delimited to one business area within Nokia and Ericsson, namely the radio base station business. The description of other business areas mainly serves as a background to the comparison.
The working paper mostly deals with the situation and planned activities as they appeared during early spring year 2002. During late spring and summer of 2002, the market situation has dramatically changed for the telecom companies. In order to conclude this report, we have chosen not to include the efforts made to deal with these new conditions.
The descriptions and analysis are based on the picture given by the interviewees and the material we have had access to. We cannot exclude that there are alternative and complementary images of the companies. The rapid changes within the companies have also led to a situation in which the strategies are expressed and interpreted differently in various parts of the organisations. The overall strategies of the corporations have not been available to us. We have thus not analysed the relations between our descriptions and these overall strategies.
Despite these limitations we believe that the questions, decisions and consequences concerning costs and competence, which are elaborated in the report, are of general interest. We hope that the analysis may provide a valuable contribution to the discussion on outsourcing and its effects, which is a topical question in many companies.
Gävle in October 2002
Lars Bengtsson Christian Berggren
University of Gävle University of Linköping 3 (22)
CONTENTS
PREFACE ...... 2
1 BACKGROUND...... 5
1.1 THE FUTURE OF MANUFACTURING: A COMMODITY OR A STRATEGIC COMPETENCE?...... 5 1.2 OUTSOURCING IN THE TELECOM INDUSTRY – ONLY ONE WAY FORWARD? ...... 6 1.3 PURPOSE OF THE REPORT...... 7 2 NOKIA AND OUTSOURCING...... 9
3 ERICSSON AND OUTSOURCING...... 11
4 LEARNING FROM NOKIA OR ERICSSON? COMPARING TWO OUTSOURCING STRATEGIES ...... 14
4.1 TWO ALTERNATIVE STRATEGIES FOR OUTSOURCING ...... 14 4.2 RISKS AND COSTS. AN ANALYSIS OF JUSTIFICATIONS AND OUTCOMES...... 15 4.3 INDUSTRIALIZE WITHOUT VOLUME PRODUCTION? THE PROBLEM OF CORE COMPETENCE 18 4.4 LEARNING FROM NOKIA OR ERICSSON? ...... 19 REFERENCES ...... 21 4 (22) 5 (22)
1 BACKGROUND
1.1 The future of manufacturing: A commodity or a strategic competence? During recent years there has been a strong trend within Swedish industry towards outsourcing of manufacturing and resources supporting manufacturing. An indication is the decreasing level of value-added within the engineering industry, from 31% in 1990 down to 25% in 1998 (VI 1999).
This outsourcing trend signifies a changing view on the significance of manufacturing. Quite a few companies have begun to view manufacturing as an undifferentiated standard activity (e.g. Arnold 2002), a ”commodity” which easily can be sourced from external market actors in the new “network economy”. A representative case of this attitude could be found in Harvard Business Review 2000: ”Does Manufacturing Matter? The short answer is: not much. And that´s a good thing” (Ramaswamy & Rowthorn 2000).
In the research on production and outsourcing, however, several authors have pointed to the problems of large-scale outsourcing. The outsourcing decisions are often justified by cost considerations, but paradoxically it has turned out to be quite difficult to sort out the actual cost effects of outsourcing decisions, not only in the long-term, which is always difficult, but also the short-term outcomes. The costs, ex post, of manufacturing at contract manufacturers, are often different from the quoted price, ex ante. Various transaction expenses, such as the costs of transferring products, equipment and knowledge from the customer firm to the contract manufacturer are left out of the analyses, and so are the costs and potentials of various production learning curves, etc.
Against the ”manufacturing-as-commodity”-perspective, a perspective on manufacturing as a strategic competence within technology-based firms has been formulated (e.g. Brown et al 2000; Pfeffer & Sutton 2000). Whereas the outsourcing trend tends to downgrade the value of manufacturing knowledge and process engineering skills, this perspective points to the need of investing in advanced manufacturing processes and competence development. From this view follows the importance of making the strategic role of manufacturing visible inside and outside the firm. Otherwise valuable resources might be lost as a result of short-term financial considerations. Since outsourcing tends to be an irreversible process, these resources might be impossible to recover. If the role of manufacturing competencies remains neglected and ”invisible”, there could be an increasing gap between actual productivity in the down-graded production departments and the productivity increases which might result from a commitment to invest in strategic production development.
This potential productivity gap has motivated a new research project on the consequences of outsourcing for renewal and learning capabilities in technology-based firms (Bengtsson 2001). A key question concerns the role – if any – of manufacturing for the long-term competitive positions of the studied firms. According to the project proposal, there is a need for an elaboration of the future-oriented and company-unique capabilities, which are or could be realized in advanced manufacturing operations. In previous studies five categories of such potentially strategic competencies have been singled out (Berggren & Bengtsson 2000; Bengtsson & Berggren 2001): 6 (22)
• Operational skills in achieving high levels of quality, productivity, flexibility and delivery precision. The links between these capabilities and the competitiveness of innovative firms have been demonstrated in several studies.
• Competence in logistics and sourcing. Manufacturing skills constitute an important basis both for evaluation of suppliers and knowledge exchange with the suppliers.
• Competence in design for manufacture, experiences of participation in cross-functional development and in analyzing and communicating manufacturing consequences of design decisions.
• Competence in industrialization (process engineering and production ramp-up). The ability to achieve a rapid ramp-up from try-out runs of new products to high volume production in a short period of time is of increasing importance for competition in global industries. This competence is based on several production-related capabilities, such as process planning, industrial engineering and materials management, test development and prototype fabrication. A systematical development of product-process flexibility in the production system is a key condition for achieving such a competence in industrialization.
• Competence in environmental management. The experience of structured quality and environmental management efforts in well-run manufacturing units could serve as a model for company-wide environmental management systems, including management of supplier activities.
As long as the potential and actual capabilities of manufacturing are not visible, not understood and not appreciated at corporate levels, it is difficult to evaluate the possible contributions of a firm´s production operation to the future competitiveness of the firm. In this situation efforts to estimate the merits and demerits of outsourcing remain elusive. In financially hard times firms are tempted to find “silver bullet-solutions”, by for example selling off manufacturing divisions, to get rid of visible costs tracked by stock market analysts. In these circumstances it is particularly important to understand the full costs of outsourcing, and the potential role of manufacturing competencies for effective product development, supply strategies and product industrialization in the future.
1.2 Outsourcing in the telecom industry – only one way forward? Within the telecom industry, most cases of outsourcing imply that production activities are transferred to American contract manufacturers, who in the next round relocate these activities to low-wage locations. The far-reaching outsourcing process at Ericsson is well- known. But how is the situation at Nokia? There is an abundance of articles comparing the styling and performance of the two companies´ various mobile phones, whereas studies looking into the manufacturing and production ramp up strategies of the two Nordic firms are in short supply. In Sweden it has sometimes been argued that outsourcing represents the future of the industry, and in this sense Ericsson represents a more ”advanced” case, which soon will be followed by Nokia. This understanding is well received in the EMS community. However, this deterministic perspective of only one way to organize manufacturing in high tech industries is misleading. Maybe Nokia is not at all a ”latecomer” in the outsourcing game, but represents a qualitatively different strategy, and a harbinger of future reorientation, beyond the current fads and fashions. 7 (22)
The CEO of Nokia, Jorma Ollila emphasize that outsourcing, or relocation to low cost sites does not solve any problem by itself. In an interview in Swedish Dagens Nyheter (02-01-25) he argues: ”Outsourcing is one of the most misunderstood concepts in use today. The real challenge is to get all the parts of the business, the entire network to function in an efficient way.” 1997 1998 1999 2000 2001 Nokia Sales (Mkr SEK) 75 900 126 300 170 000 264 200 295 400 Employees 35 500 41 000 51 200 58 700 54 000 Value added 31% 33% 32% 29% 26% Ericsson Sales 167 700 184 400 215 400 273 600 231 800 Employees 100 800 103 700 105 000 101 600 85 200 Value added 33% 33% 29% 20% 10%
Table 1. A comparison of value added at Nokia and Ericsson, corporate level. (Source: Nokia and Ericsson, annual reports 2001. Definitions: Value added = (operating income +costs of personnel) / sales. The figures are only indicative, and sensitive to volume changes as well as accounting principles selected.) The different outsourcing strategies at Nokia and Ericsson are to some extent reflected in the changes of value-added in the two companies from 1997 till 2001. At Ericsson this figure has fallen from 33% to 10%, at Nokia from 31% to 26% in the same period.
1.3 Purpose of the report
This working report is guided by three principle questions:
• What are the outsourcing strategies at Nokia and Ericsson?
• What are the motives for the two different strategies?
• What are the consequences and which strategy is most likely to be successful?
The study is limited to the radio base station business in each company, in particular it is a comparison between Nokia in Oulu and Ericsson in Gävle.
Some parts of the report are based on non-public information provided by various interviewees, mainly during the period January - August 2002. Because of the sensitive nature of the issues the report does not specify the contribution of specific individuals. The comparisons of the Nordic companies constitute the core of the report. There is a certain asymmetry in our access to data in the two companies. As for Ericsson, it has been possible to validate our cost approximations against internal studies of manufacturing costs both inside Ericsson and EMS-firms supplying Ericsson. For confidentiality reasons we cannot publish any precise figures, however. As for Nokia, we have (so far!) not been able to get access to similar company studies. As a result the information from Nokia is more dependent on external validation: ”Do they walk the talk”? So far we have found a satisfying concordance between expressed and revealed preferences concerning outsourcing at Nokia. 8 (22)
An extended version of this English report is available as a working paper in Swedish (Bengtsson & Berggren 2002). 9 (22)
2 NOKIA AND OUTSOURCING Nokia consists of three global divisions: Nokia Networks, Nokia Mobile Phones and Nokia Ventures (as of early 2002). The mandate of Nokia Networks is to develop, build and deliver mobile and fixed networks and IP network infrastructure and related services. In 2001, Nokia Networks represented approximately 24% of Nokia´s net sales. The division employed 22,000 people and seven factories worldwide, four in Finland and three in China.
At Nokia Phones, manufacturing is viewed as a core competence: ”We consider our mobile phone manufacturing to be a core competency and competitive advantage.” (Nokia Form 20- F, 2001: 34.) In contrast to other mobile phone providers, only a minor part, in 2001 approximately 20%, of Nokia´s mobile phones are outsourced to external vendors.
The production and supply structure of Nokia Networks is more complex, but the same basic considerations seem to apply also in this division. Both Nokia Phones and Networks have developed long-standing partnerships with selected suppliers. These are evaluated on a number of factors such as financial resources to increase production capacity, systems of management control, general technological skills and specific skills in quality systems, manufacturing, supply management, etc. On top of these hard issues, the attitudes, values, visions and goals of the potentials suppliers are evaluated (Ali-Yrkkö 2000).
The main site of Nokia Networks´ radio base stations, both development and manufacturing, is located in Oulu, in northern Finland. In addition to research and global product development, the Oulu organization includes prototype fabrication and pilot plants, departments for product industrialization and high volume production of radio base stations for the European market. Close to the Nokia premises a good part of its suppliers are located (even though in a process of relocating to low wage sites). Previously, Nokia owned several of these companies. Similar to Ericsson (but not to the Phone division) Nokia Networks has decreased its degree of vertical integration by selling off complete factories to contract manufacturers. So far, however, the scale of this outsourcing is much more limited. Filter production has been transferred to Filtronics and in 2001, SCI acquired Nokia Networks´ facilities, including personnel, for radio access products in Haukipudas, close to Oulu and the transceiver (TRU) plant in Camberley, UK.
The sourcing strategy of Nokia Networks may be described as ”component oriented”. This means that Nokia maintains strategic components and processes in-house, first and foremost TRU-manufacturing and the majority of final assembly and testing of complete radio base stations. By keeping volume manufacturing of the TRUs, the heart of the radio base station, in-house, Nokia Networks is well placed to prepare production of the next technology generation. Components, which are viewed as non-strategic, are sourced from selected suppliers, normally only one or a few for each component. This include for instance Nokia- specific integrated circuits and radio frequency components, servers and subassemblies such as filters, combiners and power units, mechanical components (e.g. cabinets) and connectors. Some parts of the final assembly, including testing has been transferred to partner companies, but Nokia still tests random samples from these companies. 10 (22)
Capacity suppliers
Product-owing company (OEM): Product Core components, tests, final assembly development and entire system Radio base stations System and component suppliers:
E.g. Integrated circuits, servers, filters,
combiners, power units, cabinets, connectors, etc.
Industrialisation Volume production Phasing out Figure 1. A horizontally integrated model. One company is responsible for selected components during the entire product life cycle, from development to phase-out.
Nokia classifies its suppliers in three categories: suppliers of standard components, suppliers with a specified but limited design responsibility; and system suppliers who are responsible for their on design solutions, but in cooperation with Nokia´s engineers. Linked to this design responsibility the most advanced system suppliers take care of industrialization, volume production and their own suppliers. In theory they are supposed to close their own agreements with these sub suppliers, but in reality there is no clear-cut division of responsibilities here. Similar to Ericsson, Nokia still negotiates supply contracts with many material suppliers servicing the contract manufacturers.
This model means that Nokia and its system suppliers are actively cooperating during the whole product life cycle. The Oulu works have organized separate ”Partnership areas” within Nokia´s premises where selected suppliers (design- as well as manufacturing firms) have their own offices and personnel. This makes it easy to establish close contacts and interaction. Physical co-location also facilitates the implementation a new form of joint development, where design-teams may consist of a design leader from Nokia and project members from partner companies.
According to Jorma Ollila, Nokia should keep control of the entire supply chain for at least the next five years. For Nokia Networks this means thatr it is essential to maintain volume production skills in-house. Intimate production knowledge facilitates interaction with design engineers and supports Oulu´s role as a master architect of new production concepts, which are later transferred to operations in other parts of Nokia´s international production network.
Step by step, Nokia Networks has built three separate supply chains: in Europe, Asia and the Americas. Every new product, however, will first be industrialized and manufactured in high volume production in Oulu, which is the designated master plant, before production starts in the regions outside Europe. Also after this transfer, Oulu keeps its own volume production and continues to support the offshore units. In order to safeguard overall flexibility and high levels of capacity utilization in its own plants, ”capacity suppliers” are utilized. In the future Nokia might shift more production to these ”capacity suppliers. However, Nokia representatives emphasize the importance of keeping sufficient production volume inside its Oulu works, to maintain critical mass and continue in its role as master plant. This means that both outsourcing and insourcing are realistic future options. 11 (22)
3 ERICSSON AND OUTSOURCING
Ten years ago, Ericsson was a vertically integrated company with significant and diversified production resources. The outsourcing process started in the mid 1990s, when ETX the fixed network division, suffering from increasingly difficult business conditions and falling demand started to outsource various simple components, such as cabinets and plastic details. A key issue already at this time concerned Ericssons´s perception of its future core business and core competencies. The answer has been changing over time. Initially, Ericsson regarded mounted circuit boards as part of its own manufacturing business. When the crisis at Swedish Norrköping surfaced, partly as a result of a failed project to develop a new SDH-switch (Synchronous Digital Hierarchy), the “core competence” line was redrawn and the in-house production of advanced PCBs was put to an end. In this case it was no real outsourcing decision, but rather a complete close down of the Norrköping plant, and the elimination of 1800 jobs. In the wake of this crisis, the management of the fixed networks division revised its entire production strategy. It was decided only to keep test platforms in-house, and to outsource basically all other production. The first real outsourcing-case occurred in Östersund, where complete AXE-switches were assembled. This plant and a French sister plant was sold off to Solectron. The decision that was presented as an alternative to a plant closure. The espoused rationale was that the American contract manufacturer would be able to improve capacity utilization and reduce production costs by attracting other customers to the plant. This never materialized, however. As a consequence production efficiencies were never improved in the plant, and the justification of the outsourcing decision collapsed. Recently, Solectron has dismissed all employees in Östersund.
Within the Radio division, ERA, comprising systems as well as terminals (phones) there was initially an entirely different attitude towards outsourcing. As late as 1998, manufacturing and testing were perceived as critical business competencies in this division, and actually plants were transferred from ETX to ERA. This view was to change rapidly. Already in 2000, the systems and networks business, still highly profitable, adopted the very outsourcing-strategy formulated at the ETX division and criticized by the radio people only two years earlier. According to this production strategy all volume production should be phased-out and transferred to contract manufacturers. The centre for development and design of Ericsson´s radio base stations is located in Kista, whereas the main production site used to be Gävle, supported by the Kumla plant. Recently Ericsson had a base station plant in Visby, too. But in 2000 as a result of the new outsourcing strategy, the Visby plant was sold out to Flextronics. Previous outsourcing decisions within ERA had been component-oriented, contracting out computer, signal cards and subassemblies like cabinets, cables and power supply units to companies such as Emerson, Flextronics and Sanmina. With the transfer of Visby, Flextronics acquired sole responsibility for the complete TRU-production to the first generation GSM- systems (RBS 200) for the rest of its product life. And in the following year, the entire mobile phone production was suddenly transferred to Flextronics, who rapidly started to close factories in Sweden and other high-cost countries and move production to low-wage locations.
According to the new production strategy, the critical core competencies within Ericsson will be product design, test development and industrialization of new products both strategic components such as radio modules and of complete products such as radio base stations, including final assembly and testing. So far, the strategy has been implemented for two GSM- products, and will be fully implemented for all 3G products. 12 (22)
” Industrialization” refers to the process of preparing products for production (including necessary design modifications), the ramp-up of production from prototype fabrication and try-out series to full industrial scale, and the design and fine-tuning of the supply chain from order to delivery and customer hand-over. Reduction of unit costs and lead-time are the over- riding goals of the industrialization process. The success is strongly dependent on achieving product and process stability. One key target is to increase process yield (i.e., to reduce the ratio of products requiring rework after testing); another is to reduce through-put time by shortening the test time, which initially may be quite long (several hours).
When the process is regarded as stable, the baton is passed to the contract manufacturers, who will now be responsible for volume production and the continued efforts to increase process yield and lower costs. Product and process stability are no unambiguous concepts, however. Within Ericsson, the cut-off point (final take over) between ERA and the contract manufacturers normally equals a process yield of 60%. This is still far from efficient volume production. If the yield is raised from 60% to 90%, production capacity will theoretically increase by 50% and unit costs will drop 33%. This means that a lot of process improvement and to some extent also product modifications need to be carried out after this cut-off point is reached. Actual studies of changes in production costs at Ericsson plants as well as external plants indicate strongly positive learning curves during several years. The lead-time needed to reach high yield levels decides when the new product can be produced profitably. In the Ericsson strategy of 2000, contract manufacturers are responsible for this decisive process (and to some extent product) refinement.
Previously Gävle was a master plant for global radio base station production within Ericsson. The new strategy means that the most important mandate of Gävle and its 1,400 employees (May 2002) now is to be responsible for the industrialization process, in close cooperation with the design departments in Kista. After several steps and iterations of product and process development, production ramp-up starts, and the plant focuses on reducing test times and increase process yield. As noted above, the interface between industrialization and volume production is by no means cast in concrete. Within Gävle there are strong arguments for an extended mandate, to run new products in high volumes during a short period of time, in order to verify the process and safeguard its own competence (in the figure below this is called NPI- volume, i.e. volume production during New Product Introduction). An unresolved issue here is the cost for investing in high-capacity production that will only be used for a strictly limited period.
The Ericsson way of sourcing could be described as a vertically divided model. During the end phase of the industrialization process, the selection of contract manufacturers is finalized. A transfer project is set up to convey product and process knowledge and production technology including the dedicated test systems to the selected EMS-company. The Ericsson model of industrialization and supplier cooperation is presented in Figure 2. 13 (22)
?
Contract manufacturer (EMS): Product Strategic components, tests and
development , Volume complete station After sales Radio base (OEM): , TPI and repair
stations NPI System and component suppliers: company assembly system E.g. Computer, signal cards
Product Power units, climate, mechanics Strategic entirecomponents tests, final
Industrialisation Volume production Phasing out Transfer Figure 2. A vertically divided model. The product-owning company takes care of industrialization of strategic components; contract manufacturers are in charge of basically all high volume production.
The strategy is justified in several ways. A general, ideological argument is the notion of Ericsson as a ”knowledge company” not involved in manufacturing, which other companies are said to perform more efficiently. A more down-to-earth argument is the expectation that outsourcing will reduce costs. With the deepening crisis in the telecom sector this has become the decisive argument. The argument concerns both reduction of own assets to free capital and to reduce product costs. In this report we focus the latter. It is assumed, for example, that the global contract manufacturers, enjoying high volumes and a broader customer base, will benefit from more favorable materials prices. Other arguments supporting the Ericsson strategy refer to increased flexibility, reduced risk exposure and possibilities to expand output with limited capital investments. 14 (22)
4 LEARNING FROM NOKIA OR ERICSSON? COMPARING TWO
OUTSOURCING STRATEGIES
4.1 Two alternative strategies for outsourcing Both Ericsson and Nokia are outsourcing production, citing the same general reasons: the need to focus on core competencies and team up with global partners, to enhance flexibility and reaction speed, etc. As noted above, however, there is a significant difference in the degree of outsourcing and in the speed of change. De-integration of production operations has proceeded much faster at Ericsson than at Nokia. Furthermore, there is a difference in the actual mode of outsourcing.
The different production networks that are formed as a result of idiosyncratic ways of teaming up with external manufacturing partners could be described in terms of configuration and coordination (Shi & Gregory 1998). Configuration refers to the geographical structure and focus of the various locations. Theoretically there are two types of configuration, product focussed or process focused (Hayes & Schemmer 1978). Product focus means that each production unit is responsible for a specific product or component. Process focus refers to a network divided according to process types and process steps. Coordination, the other key concept, refers to the modes of managing and controlling the various units in the network.
The sourcing strategy of Nokia Networks has resulted in a product-focused production network, akin to the component-based supply model common in the automotive industry (Corswant & Fredriksson 2002). Graphically, a horizontally sliced but vertically integrated model represents the product life cycle in this network (see Fig 1). A key aspect in the model is the fact that Nokia maintains responsibility in-house for strategic components during their entire life cycle: design and engineering, industrialization, volume production, after sales and phase out. Furthermore, Nokia Networks takes care of the final testing of key components, base station assembly and delivery. The responsibility for so-called non-strategic components is allocated to several different system suppliers. The model as a whole presupposes an active collaboration between Nokia and its selected suppliers during the entire product life span.
The production network shaped by the sourcing strategy of Ericsson Radio is of a composite type. In parts of the network there is a product focus. This involves the suppliers of non-core components, such as Emerson (power supply) or SCI-Sanmina (cabinets). The new strategy implemented since 2000, where volume production is transferred to contract manufacturers, had added a process-focused system of relations. As Figure 2 illustrates, this network can be represented as a model, which is both horizontally sliced and vertically divided. In this model, outsourcing also encompasses production of strategic components, final assembly, test and installation. Most of the components, as well as the testing system and the product as a whole, are developed and industrialized by Ericsson, however. For the non-core components, system suppliers are responsible also for design (collaborating with Ericsson), and production. In these cases, the responsibility for industrialization is carried out mainly by the suppliers.
A key issue in both types of production network concerns coordination and management of the interfaces between components, both in the design and industrialization phases. How do Nokia and Ericsson engineers collaborate with their counterparts at the system suppliers? The ambition of Nokia is to create cooperative interaction in its joint design areas at the Oulu 15 (22) works. There is also an ambition to devise common IT-enabled development processes on a global level. So far this has not been accomplished (and maybe never will).
In the Ericsson network, the Swedish company retains responsibility for design and industrialization of most components, but also strives to develop new forms of supplier collaboration. As a result of its outsourcing of volume production, another interface issue is added to its supply chain management. Both Nokia and Ericsson are coordinating their component manufacturers, but on top of that Ericsson has to coordinate and integrate a technologically highly complex process chain. The interface problem is aggravated by the different orientation of Ericsson and its contract manufacturers. Product industrialization at Ericsson is focused on design adaptation in order to improve manufacturability, not on the production process per se. As a result, the production systems formed at Ericsson may diverge from the standardized processes at the contract manufacturers.
4.2 Risks and costs. An analysis of justifications and outcomes.
There are various motives behind outsourcing (The Outsourcing Institute 2001). Cost considerations is a key factor driving outsourcing decisions. Neither Nokia nor Ericsson would outsource production if they expected this would bring higher costs to the customer. Another motive is related to strategic focus, and the interpretation of future core competencies. A third factor concerns risk exposure and flexibility in the highly turbulent telecom sector. According to Ericsson management, the EMS-companies can level their production across several customers. By outsourcing production to these firms Ericsson will reduce the risk of being caught with over-capacity in a negative business cycle with falling demand.
The large contract manufacturers are not unconditionally flexible, however, but demands compensation if real demand falls short of booked capacity. Alternatively Ericsson foots the bill for their production investments. The expressed intention at Ericsson is that contract manufacturers should be able to increase their output with 30% within 2-4 weeks, or alternatively, to cease production (-100%) within the same time. The actual flexibility level as of today, and the price Ericsson has to pay for this flexibility is not disclosed.
In general presentations such as the annual report Nokia, too, emphasizes that sourcing from third party suppliers increase flexibility and reaction speed (Nokia Form 20-F 2001: 34). In interviews, a considerable doubt is expressed if increased outsourcing really enhances flexibility. Interviewed managers point out that almost all the EMS-firms are subject to the same business cycle, which means that several of these companies, for example Solectron, is now under heavy financial pressure. Another important consideration at Nokia is the risk of being too dependent on international manufacturers whose strategies sometimes are unconvincing. One instance, mentioned by Nokia, is the tendency by several manufacturers, including Flextronics, to evolve into conglomerates. ”They now look similar to our structure 10 years ago. What is their competitive advantage when they are manufacturing ´everything´?”. At the end of the day, Nokia managers argue, the issue in risk management is not outsourcing but forecasting, the capability to analyze future trends and adapt the capacity levels accordingly. A possible objection to this reasoning is the question if reliable forecasting is possible in the deregulated telecom industry. It could be argued that the key question is which way of designing and controlling the production network yields the highest level of real (not to confuse with contractual) flexibility to the lowest cost. An investigation of this important problem falls outside the scope of this report. 16 (22)
What is the possible cost advantage of outsourcing production? The opinions at Nokia and Ericsson differ. Ex ante estimates demonstrating cost reductions are not difficult to find. It is harder to identify rigorous ex post analyses, which try to factor in all relevant parameters. Some follow up studies at Ericsson indicate cost savings, others zero or negative outcomes. The contract manufacturers regularly present positive figures. An ex post analysis at Ericsson of the transfer of a radio base product, which included the cost of the transfer process, yielded a negative result. In the research literature too, it is a moot question if outsourcing really results in cost savings, although many authors take this for granted, and several conflicting studies have been presented. In this report it is not possible to present any precise figures. Instead we outline the main cost factors affecting the production of a radio base station and estimate their relative importance. The project related costs of industrialization and transfer to third party suppliers are not included. Schematically the total unit cost could be divided into the factors presented in Figure 3.
Profit
Value added
Standard material
Specific material
Figure 3. Schematically drawing of total unit cost
Purchased materials constitute the singularly most important cost factor, approximating 70%-80% of total unit cost. It is often assumed that the large EMS-companies enjoy a cost advantage here. At a consolidated level the purchasing levels of Flextronics, for example, are much higher than those of Ericsson. Consequently, it is assumed that contract manufacturers will benefit from lower component prices. This scale advantage is only relevant for a minor part of the materials, however. The majority of all components are designed and specified by the customer company (Nokia or Ericsson) for products. This means that the materials costs are already decided in the contracts which Nokia or Ericsson negotiate with these sub- suppliers. Thus, the purchasing discount enjoyed by contract manufacturers applies only to a minor part, approximately 20%-30%, of total materials costs. So even if the cost of standard materials could be lowered by 25%, which is a very optimistic assumption, total product cost would only be affected by 4-5 percentage points. By coordinating all its purchasing across the group, Nokia argues that they can reach very favorable agreements. The large Nokia Phone division supports Nokia Networks, and as a result ”nobody has better purchasing prices than the Nokia group”.
Ericsson expects the contract manufacturers to obtain advantageous price levels by relocating to low cost countries and this is factored into its supply agreements. In this context it is important not to confuse economies of scale and economies of location. Relocation of TRU- manufacturing, for example, does not by itself reduce the price of purchased materials, which 17 (22) is delivered from several different locations, (and also supply the original Ericsson plant). It might be argued that the international contract manufacturers could be in a stronger position to entice their materials supplier to relocate and in this way build a more comprehensive base of materials suppliers in low cost locations. The advantage of possible price reduction needs to be weighed against the potential loss of supply chain control as a result of the outsourcing process. Follow-up studies in 2002 comparing Ericsson plants and selected supplier sites indicate that so far costs are approximately similar.
The costs of value-added are in the region of 20-25%. The principal components are personnel costs and capital costs for equipment and buildings. According to Nokia, the product-specific testing systems are the most expensive portion of the manufacturing equipment. The telecom manufacturer designs this equipment so the contract manufacturers enjoys no advantage of scale. Another production component is the SMD-lines (surface mounting devices). This machinery commands the same prices worldwide, and low cost countries have no advantage. Contract manufacturers may achieve a higher capacity utilization by combining orders from several customers. Nokia´s strategy is to keep its own high volume lines as busy as possible and shift excess order to its “capacity suppliers”.
Personnel costs are per definition lower in low cost countries, at least the labor cost component. Management costs are normally not lower, frequently the opposite is true. In specific cases (not referring to TRUs), Flextronics has offered to reduce labor costs by half by moving from Sweden to Poland. The wage components when producing radio base station is insignificant, however, normally lower than 10% of total unit cost. In an estimate of total outsourcing costs, this modest advantage needs to be balanced against the costs of transferring knowledge, equipment and products to the contract manufacturers. The consultant firm Booz Allen Hamilton has estimated these costs to be in the range of 5%-10% of total product costs (Jackson et al., 2001); but arguably this level is highly dependent on actual production volumes.
Furthermore, some of the personnel costs consist of administration and services, not directly influenced by outsourcing or location.
Finally, the contract manufacturers have to make their own profit on the outsourced products.
Taking all these factors together the cost advantages of outsourcing tend to be marginal at best, calculated as the cost of the final product. Reduced prices for a minor part of materials purchases, lower labor costs and maybe somewhat higher capacity utilization on the one hand, increased expenditures in terms of transfer costs, extended and more complicated industrialization and supplier margins on the other. For Ericsson an estimate by the authors indicates a 5-10% reduction of total product cost, not including transfer costs, etc. The estimate has been verified by cross-checking company internal studies. These studies have demonstrated that the real costs of manufacturing at the EMS-firms tend to exceed contracted prices, when items such as capital costs paid for by Ericsson, general administrative expenses, etc. are corrected for. Follow-up studies have also pointed out that the calculations preceding the outsourcing decisions tend to neglect the consequences for Ericsson units which continue manufacturing some of the otherwise outsourced product. Because of reduced volume these units will suffer from a diseconomy of scale, and units costs will increase, sometimes so much that the entire justification of involving contract manufacturers falls apart. Generally, both the ex ante and the ex post estimates performed by Ericsson and its consultants lack a dynamic perceptive, i.e. considerations of how and if learning effects and improvement potentials in its own plants might offset the 5-10% cost difference in the static perspective. 18 (22)
Managers at Nokia Networks does not recognise the cost advantage in outsourcing all volume production. Moreover they argue, “you cannot only look at costs, you have to look at the effects on revenue, specifically on TTM” (Time to Market). If in the final analysis outsourcing would reduce production costs, but at the same time imply delays in bringing products to the market, and consequently, to lost revenues, the bottom line would be a loss for the company.
Also at Ericsson, the risks that outsourcing could negatively affect speed in the process of bringing new products to the market would be well known. So far such considerations have had no visible effects on the outsourcing decisions. Maybe they should. The experiences of Flextronics taking over the entire phone manufacturing operations from Ericsson in 2001 do not support hypothesis of outsourcing and contract manufacturing as a way to enhance speed and responsiveness in the supply chain.
4.3 Industrialize without volume production? The problem of core competence
The Ericsson strategy for outsourcing spells out three technical core competencies for the company in the future (in addition to these there are of course also core competencies in the commercials operations): product design, design of test equipment and industrialization of new products. By maintaining industrialization of key components and new systems in-house, Ericsson seeks to handle the important interdependencies between design and manufacturing strategy (cf. Swink 1999). This presupposes that Ericsson also in the future is able to keep a critical mass of manufacturing knowledge as a basis for carrying out industrialization. In the units recently designated for industrialization this is still the case, because of their earlier experience of volume production, but this knowledge may have a short shell-life.
For the future there are theoretically two alternatives at Ericsson. The first implies a systematic way of organizing collaboration and knowledge exchange with external partners during the entire development process, from design to production ramp-up. Currently the EMS-firms are brought into this process very late. Managers and engineers at Ericsson think of themselves more as teachers than as pupils in these relations, but this attitude would have to change. This first alternative is no easy route. Recently published research has demonstrated the difficulties for a customer firm with no production of its own to absorb and make good use of a supplier’s manufacturing skills (Veugelers & Cassiman 1999). The second alternative is to keep enough production operations within Ericsson so as to maintain a critical knowledge mass. This alternative is congruent with theoretical analyses of core competencies as systemic, emphasizing the critical links between product and process knowledge. The espoused strategy of outsourcing all volume production, and bringing contract manufacturers into the development process only at a late stage, makes it less than clear how Ericsson will manage these interdependencies.
At Nokia, too, the production and outsourcing strategy is justified by arguments concerning core competence, and the relationship between critical competence elements. Thus, one rationale for keeping volume production in-house, apart from cost and revenue considerations, is the importance of maintaining a high skill level in manufacturing, and be able to feedback experiences directly from mass production to designers and product engineers. Another competence argument refers to the close interrelations between production and industrialization skills, and to the need for Oulu as a master plant to devise comprehensive production concept, which can be transferred to off-shore production sites. Furthermore, 19 (22)
Nokia managers view advanced production knowledge as an important element in the management of the entire supply chain.
At the Ericsson industrialization unit in Gävle, there is an awareness of the importance of manufacturing knowledge. Said by a member of the management team: ”Today we still have the knowledge and skills needed to set-up efficient, high-quality (volume) production. However, we don´t know how the outsourcing will affect our future industrialization skills. And maybe it will be hard to assess and select suppliers when we don’t have the relevant production skills in-house any more.”
To preserve and update a sufficient base of manufacturing knowledge, it is now proposed that Gävle continue volume production of each new product for a short period after the official industrialization process is concluded. (In figure 2 this is referred to by the acronym NPI- volume). Such ”NPI-volumes” could both act as capacity regulators and as a strategic source of production knowledge.
This solution, however, begs the question how to synchronize new production processes at Ericsson and the corresponding processes at the contract manufacturers. The risk of a double industrialization process is still very real. There is a potential conflict between an emphasis on investing in resources for maintaining in-company industrialization skills and an emphasis on exploiting the skills of the contract manufacturers in arranging standardized manufacturing processes. Given the outsourcing strategy the dilemma is: Should Ericsson let go of new products at an earlier stage in the industrialization process, and risk harming the informal knowledge exchange between product and process engineers needed to increase production yield? Or would it be a better option invest more resources in the internal industrialization process, including some high volume production, but incur possibly substantial costs for maintaining knowledge and capacity and still not solve the problem of a “double industrialization”?
4.4 Learning from Nokia or Ericsson?
Summing up, it could first be stated that both Nokia and Ericsson argue that out-sourcing and cooperation with third party suppliers are necessary, but secondly, that their specific strategies differ in important aspects.
• Ericsson views far-reaching outsourcing and sell-off of production as a key measure to increase flexibility and reduce risk exposure. At Nokia, it is argued that forecasting is more important than outsourcing, citing Cisco as a “teaching case”. Cisco prided itself of being a virtual company with practically no manufacturing of its own. Nevertheless, when the telecom boom ran out of juice in early 2000, the Californian corporation suddenly had to write of 2 billion USD in inventories. “If there are excess stocks somewhere in the delivery chain it´s a question about uncompetitiveness of the whole chain”, is the Nokia argument.
• Expected cost reductions are the key driver for outsourcing decisions at Ericsson. If these expectations come true, and to what extent, is open to argument. Interviewed in early 2002, managers at Nokia Networks did not see any cost advantages by outsourcing more production. Looking at the revenue side, it was pointed out that too much outsourcing could result in product delays and lost sales. 20 (22)
• The modes of outsourcing at the two Nordic companies differ significantly. Ericsson applies a vertically divided model, where volume production in principle is to be carried out by contract manufacturers. The Oulu way of outsourcing could be described as a horizontally integrated model, where Nokia is still responsible for strategic components and the final product during its entire product life.
The introduction to this report referred to arguments popular among consultants and the EMS- firms that outsourcing represents the future of the industry. From this perspective Nokia is construed as a laggard compared to Ericsson. This report supports the opposite view that there is no one best way of organizing development and production in turbulent high-tech industries. Nokia represents a healthy reminder that there is a different way of achieving both cost efficiency and speed. The argument about forecasting as a key success factor is not entirely convincing, however. Rather, the strategic challenge is how to increase the responsiveness and flexibility of the production system in its entirety without incurring uncompetitive costs.
Integrating production and development competencies is a long-term strategy at Nokia. One key aspect is the decision to maintain responsibility for design, production and logistics of strategic components and products in Oulu. This reduces the extent of organizational interfaces and simplifies product transfers after industrialization. But some difficult interfaces do remain, for example in relation to the company´s capacity suppliers. At the same time, selected suppliers are made responsible for their parts including design for manufacture already in the design phase. This implies a high dependency on suppliers and their input and presupposes effective ways of coordination and control by the product owner. The Nokia strategy furthermore does not preempt future production and supply decisions. By keeping volume manufacturing in its own plants both insourcing and outsourcing are viable future options.
During the 1990s, the Nordic area emerged as an internationally leading telecom region, on the back of advanced user, operators and manufacturers. Cooperation and competition between user-developers and manufacturer-developers, and between manufacturers played an important role for achieving this outstanding position (Berggren & Laestadius 2002). Competition between Ericsson and Nokia has been cut-throat, but in strategic standardization issues three has also been a productive cooperation. In this period of unprecedented crisis in the international telecom industry, it might be an advantage for both companies if competition in the market-place could be combined with more openness and exchange in the manufacturing area, to safeguard Nordic production competencies and a competitive production base. 21 (22)
REFERENCES
Interviews
Anders Wennberg, Manager Sourcing, Ericsson Radio Systems, Gävle Bo Westerberg, Vice President, Corporate Sourcing, Ericsson, Kista Gunnar Herdin, Manager, NPI and TPI projects, Ericsson Radio Systems, Gävle Hemmi Piirainen, General Manager, Manufacturing Partnering, Product Operations, Nokia Networks, Oulu Jan Hedlund, Swedish Metal Workers´ Union, Kista Jyrki Ali-Yrkkö, ETLA, Helsinki Kari Sairo, The Finnish Metal Workers' Union, Helsinki Karin Swärd-Hertel, Manager Business Development, Ericsson Radio Systems, Gävle Magnus Bergqvist, Manager Technology, Ericsson Radio Systems, Gävle Raimo Lovio, Researcher, VTT Technical Research Centre of Finland Mats Lundin, General Manager, RBS Supply, Ericsson Radio Systems, Kista Pekka Pesonen, Senior Manager, Strategic Planning, Nokia Mobile Phones, Operations, Logistics and Sourcing Risto Lehtilahti, Nokia Networks, Oulu Tommy Nejderyd, Manager Main Planning, Ericsson Radio Systems, Gävle Åke Svenmarck, The Swedish Association of Graduate Engineers (CF), Kista
Literature
Ali-Yrkkö, J et al. (2000). Nokia – a big company in a small country. Helsingfors: Taloustieto Oy. Ali-Yrkkö, J. (2001). Nokia´s Networks - gaining competitiveness from co-operation. Helsingfors: Taloustieto Oy. Arnold, U. (2000). New dimensions of outsourcing: a combination of transaction cost economics and the core competencies concept. European Journal of Purchasing & Supply Management, 6, 23-29 Bengtsson, L. & Berggren, C. (2002). Horisontellt kopplad eller vertikalt klippt? Två motsatta strategier för outsourcing. En jämförelse av radiobasverksamheten inom Nokia och Ericsson. Working paper. Gävle: Högskolan i Gävle. Bengtsson, L. (2001). Outsourcing av produktion och dess betydelse för innovations- och förnyelseförmågan. Projektbeskrivning till Vinnova. Gävle: Högskolan i Gävle. Bengtsson, L & Berggren, C. (2001). Produktionens förändrade roll. Mager klickfunktion eller kunskapsfabrik? I Backlund, T et al. (red.). Lärdilemman. Lund: Studentlitteratur. Bengtsson, L. Niss, C. Stjernstöm, S. & Westin, S. (2002). The Challenge of Maintaining New Product Manufacturability when Outsourcing Volume Manufacturing. Paper to be presented at CINet2002, Espoo, September. 22 (22)
Berggren, C. & Bengtsson, L. (2000). Kunskapsfabriken. Stockholm: Svenska Metallindustriarbetarförbundet. Berggren, C. and S. Laestadius. (2002). Co-development and Composite Clusters - the Secular Strength of Nordic Telecommunications. Accepted for publication by Industrial and Corporate Change, no. 5. Brown, S., Lamming, R., Bessant, J. & Jones, P. (2000). Strategic Operations Management. Oxford: Butterworth-Heinemann. Corswant, F. & Fredriksson, P. (2002). Sourcing trends in the car industry. A survey of car manufacturers’ and suppliers’ strategies and relations. International Journal of Operations & Production Management, Vol. 22, No. 7, pp. 741-758. Ericsson. Annual report (2001) and Ericsson (2000), Form 20-F, access: www.ericsson.com Hayes, R.H. & Schemmer, R.W. (1978). How Should You Organize Manufacturing. Harvard Business Review, 56 (19), pp. 106-118. Häikiö, M. (2002). Nokia. The Inside Story. Financial Times. Prentice Hall. Jackson, T., Iloranta, K. & McKenzie, S. (2001). Profits of Perils? The Bottom Line on Outsourcing. Booz Allen Hamilton. Inc. Nokia. Annual report (2000) and Nokia (2001), Form 20-F, access: www.nokia.com Pfeffer, J. & Sutton, R.I. (2000). The Knowing-Doing Gap. Boston: Harvard Business School Press. Ramaswamy, R. & Rowthorn, R. (2000). Does Manufacturing Matter? Harvard Business Review, November–December, 32. Shi, Y. & Gregory, M. (1998). International manufacturing Networks – to develop global competitive capabilities. Journal of Operations Management, 16 (2,3), pp. 195-214. Swink, M. (1999). Threats to new product manufacturability and the effects of development team integration processes. Journal of Operations Management, 17, 691–709. The Outsourcing Institute. (2001). Executive Survey. The Outsourcing Institute´s Annual Survey of Outsourcing End Users. http://www.outsourcing.com. Access 2001-02-09. Veugelers, R. & Cassiman, B. (1999). Make or buy in innovation strategies: evidence from Belgian manufacturing firms. Research Policy, 28, p. 63-80. VI (1999). Nya industriella system. OH-material. Stockholm: Sveriges Verkstadsindustrier.