The Rise of Ridesharing Platforms

An Uber-assessment of bits and atoms

Magnus Ofstad

Master Thesis Design, Use, Interaction 60 credits

Department of Informatics

UNIVERSITY OF OSLO

05/2017

III

The Rise of Ridesharing platforms: an Uber-assessment of bits and atoms

2017

The Rise of Ridesharing Platforms: an Uber-assessment of bits and atoms

Magnus Ofstad http://www.duo.uio.no/

Trykk: Reprosentralen, Universitetet i Oslo

Abstract

This thesis is a case study of Uber. The research questions are 1. How do ridesharing platforms evolve from startups into large-scale market disrupters? 2. How can ridesharing platforms augment its user base and capture more segments within the transportation business?

Uber’s scaling is traditional. The initial growth consisted of capturing the high end of the market. Later on, Uber moved down market, making ridesharing affordable for the masses. The master stroke was to devise a scalable business model that was replicable in different markets.

More unique is Uber’s managerial approaches. Uber has devised novel strategies to overcome challenges thrown up in front of her. In short, Uber has leveraged its vast user network and partnered-up with outside forces to solve intractable tasks. These task networks have corroborated to turn the tide against the push back from taxi operators and unfriendly regulatory authorities. They have also aided Uber in the competition against other ride sharers. Uber’s decentralized structure consists of multiple innovative task networks. Local teams take initiatives which continuously spawn out new services and products. Simultaneously, Uber’s autonomous car project is a vast innovation network. The goal of the network is to bring the novelty first-to-market.

Above all, the recruitment of drivers is the key to Uber’s scaling. This thesis claims that macro conditions in the labor market make it easy for the app to add new workers. Furthermore, venture funding gave Uber economic leverage to add users at a rapid pace through low prices and subsidized driver earnings.

VII

Preface

Uber is an upstart which has reached a global presence. The venture is in the news almost on a daily basis. Despite Uber’s success, no one has given a full scholarly account of why it has been able to scale the user base exponentially. This thesis is an attempt to uncover the ingredients in this success formula.

During the research process, it was evident that there were and are serious obstacles to Uber’s growth. These barriers arise from how Uber disrupt old practices in society. This thesis shows how Uber has devised fresh managerial approaches to overcome these hurdles. One particular strategy deals for instance with how Uber has solved the emergence of self-driving cars.

The purpose of this thesis is not only to give a glimpse into the specific actions in Uber’s history that proved crucial to its success but also to provide theories that can be used to analyze other upstarts within the sharing economy.

The last point is vital since I struggled with finding an all-encompassing theory to describe all the things that go on in Uber. This thesis, therefore, employs a multidimensional theoretical approach. Many of the theories used her might prove insightful and useful in the study of other cases within the sharing space.

I would like to thank my academic supervisor Professor Ole Hanseth. He gave me suggestions about relevant research papers that inspired me to do this study. In the age of ‘fake news,' I will also compliment great journalists doing a tremendous job. My investigation would have been harder, more time-consuming, and less interesting without their effort.

Uber is a wide case, and an old saying is that ‘the devil is in the details.’ I have tried to get most things right, of course, but if some dates and other minor details are slightly wrong, I will apologize for this in advance. The faults are all mine.

Magnus Ofstad, Oslo, 05/02/2017

VIII

Content

1 Introduction ...... 1 1.1 Introducing ‘ridesharing’ in platform markets ...... 1 1.2 An Uber review ...... 5 1.2.1 Labor ...... 6 1.2.2 Innovation ...... 8 1.2.3 Governance ...... 12 1.3 The ethics of Uber ...... 18 2 Background theory ...... 22 2.1 Brazilianization ...... 22 2.1.1 The first world ...... 22 2.1.2 The third world ...... 24 2.2 Platform economics ...... 25 2.2.1 Multisided platforms ...... 25 2.2.2 Network externalities ...... 26 2.2.3 Early adopters and marquee users ...... 27 2.2.4 Innovation, subsidizing or quality ...... 27 2.3 Generative digital infrastructures ...... 29 2.3.1 Complex adaptive systems ...... 29 2.3.2 Generativity in digital infrastructures ...... 30 2.3.3 Generative contextual conditions ...... 31 2.3.4 The generative replicative pattern mechanism ...... 32 2.3.5 The generative network of networks mechanism ...... 33 2.3.6 The generative innovation mechanism ...... 35 3 Methods ...... 36 3.1 Case selection ...... 36 3.2 Methodology ...... 37 3.3 Data collection ...... 38 3.4 Data analysis ...... 40 4 Findings ...... 42 4.1 How Uber conquered San-Francisco? ...... 42 4.1.1 The creation of an app Leviathan ...... 42

4.1.2 The product as an asset ...... 46 4.1.3 The Users as an asset ...... 48 4.1.4 Flexibility and aggressiveness ...... 49 4.1.5 The launch pad ...... 53 4.2 How Uber scaled exponentially? ...... 54 4.2.1 Revving up the motor with venture capital ...... 55 4.2.2 Establishing the launch playbook ...... 56 4.2.3 Market segmentation and cheaper prices ...... 59 4.2.4 Innovative recruitment and pricing policies ...... 63 4.2.5 Design improvements ...... 66 4.2.6 Technological improvements ...... 70 4.2.7 The fallout from dynamic pricing ...... 73 4.2.8 This is how Uber scaled ...... 74 4.3 Augmentation Obstacles ...... 75 4.3.1 Uber and regulators ...... 76 4.3.2 Are driver’s employees or freelancers? ...... 79 4.3.3 Competition ...... 80 4.3.4 Privacy risk on the Uber platform ...... 86 4.3.5 Uber and women ...... 89 4.3.6 Public relation management ...... 90 4.4 Augmentation strategies ...... 91 4.4.1 The UberPOOL extension ...... 92 4.4.2 New applications ...... 93 4.4.3 Partnering-up ...... 95 4.4.4 Uber Movement ...... 100 4.4.5 Network of networks ...... 101 4.5 What happens when wages goes south? ...... 102 4.5.1 Drivers in the developed world ...... 103 4.5.2 Drivers in the developing world ...... 107 4.5.3 When the boss is an algorithm ...... 111 5 Conclusion ...... 113 Bibliography ...... 119 Appendix A ...... 152

Figure 1 The five biggest companies in the world - ten years apart...... 1 Figure 2: Uber's multisided market...... 26 Figure 3 Framework for network of networks ...... 35 Figure 4 The Uber app, simple technical oversight...... 44 Figure 5 Ride sharers funding in 2016 ...... 56 Figure 6 Growth takes off with non-professional drivers on UberX...... 60 Figure 7 Uber's virtuous cycle ...... 65 Figure 8 Uber - a sleek alternative for executives ...... 67 Figure 9 UberX is dependable, regular, and economical...... 67 Figure 10 Logo change ...... 68 Figure 11 Operation Slog ...... 82 Figure 12 Where Uber drivers reside geographically ...... 103 Figure 13 Uber driver's weekly earnings ...... 104 Figure 14 The successful launch of the Uber platform ...... 113 Figure 15 List of user friendly features ...... 113 Figure 16 Platform economics in the ridesharing space ...... 114 Figure 17 Uber's scaling canvas ...... 115 Figure 18 Resilience in Uber's network ...... 116 Figure 19 Uber as an attractor ...... 117

XII

XIII

1 Introduction

1.1 Introducing ‘ridesharing’ in platform markets

The digitalization of the global economy has heralded in a new gilded age, where a few companies totally dominate in ‘winner-takes-all’ markets. Today, companies like Amazon, Microsoft, Google, Apple, and Facebook are mustering an ever more eye- catching position. These companies have been able to create huge moats around their businesses. Thus, users perceive the costs of switching to other providers as too high.

Market cap 08/01/2006 08/01/2016 1 Exxon Apple $413B $571B 2 GE Google $336B $540B 3 Microsoft Microsoft $245B $441B 4 Gazprom Amazon $244B $364B 5 Citigroup Facebook $240B $357B Source: Bloomberg. Figure 1 The five biggest companies in the world - ten years apart.

Digital enterprises, like the ones mentioned above, seldom bring new physical products to the markets. They are usually peddling intangibles, and their strength lies in being multisided platforms that reorganize the terms on which market participants interact (Cusumano, M. A. & Gawer, A., 2002; Iansiti & Levien, 2004; Gawer, A. & Cusumano, M. A., 2008). Multisided platforms create value primarily by selling access to a marketplace where they enable direct interactions between groups which otherwise would not have traded with each other (Hagiu, A.,2014). Metaphorically speaking, their dominance arises from being positioned as the spider who organizes the network which everyone else has to operate. Alas, the spider profits whenever someone taps one of the strings in the web (Zuboff, 2015). Effectively, positioned as the nexus in an increasingly digital economy, big tech platforms constitute an expanding share of the total world market capitalization. 60 of the world’s largest corporations earned at least half of their revenue from platform markets ten years ago

(Eisenmann, T. R., 2007). Since then, the move to digital preponderance has gathered pace, like figure 1 is showing, and it is reasonable to suggest that the recent development will persist for some years to come.

Digital platforms serve different market functions. Google and Facebook offer search and social media as well as being infrastructures which other platforms run on. Traditionally, Amazon was a marketplace, while Microsoft was a software company, it is not so anymore; through their recent foray into web services, they provide cloud infrastructures which other can use to build even more platforms. Apple is a company mainly driven by innovations in hardware, but through its operative system and its open software development kit, it provides an ecosystem where outsiders can develop and offer apps to users – the Apple app store.

Airbnb and Uber are startups that have grown into platforms enabled by the newly available cloud tools. These two companies are examples of collaborative consumption platforms or ‘sharing platforms’ in short. The term ‘sharing’ has some distinct qualities. Even though Google and Facebook depend on user’s willingness to ‘share’ their personal data, the concept is restricted to the Uber and Airbnb’s of the world since they involve users in “the short term rental activity” of physical assets (Belk, 2014, p. 1597). Thus, we can define sharing startups as businesses which run their ventures on digital platform systems and harness the capabilities of digital infrastructures to facilitate the exchange of physical products or services based on tangible assets (Constantiou, Eaton, & Tuunainen, 2016).

Based on this distinction, i.e. that sharing platforms are based on peer-to-peer transactions taking place both in the digital and the physical realm, we can argue that sharing platforms offer something more than social media, search and other digital platforms. Clearly, it is true that there are some distinct qualities to what sharing platforms offer, but it is also true that they are not very different in how they aim to generate revenue. It is mostly about market intermediations and monetizing user’s efforts. In general, profit driven sharing companies acts as market intermediaries in the supply and demand of services based on tangible assets, such as accommodation (e.g. Airbnb), cab service (e.g. Uber), deliveries (e.g. Postmates), everyday chores (e.g. TaskRabbit). In all of these instances, collaborative consumers must pay fees to the platform owners to take part in the services which the platforms provide.

The digitalization of the global economy has attracted huge interests from academics and media pundits. Right now, it is the rise of sharing platforms that are taking center stage. Current discourse has labeled their emergence differently. ‘Disruptive technologies,' ‘the sharing economy’ and ‘gig work’ seems to be the most prolific terms in use. Whether ‘sharing’ is an appropriate label for most of the activities that take place in this sphere is an interesting topic in and of itself, but it is of even greater importance to discuss what the phenomenon means in a wider sense. The burgeoning phenomenon has transformative economic and social effects in so far as sharing startups have the capacity to displace established businesses and provoke reorganization of work arrangements.

This thesis is about studying the factors behind the rise of one particular type of collaborative consumption platforms, so-called ridesharing platforms, and making sense of what kind of strategies they follow to ensure long-term viability. The central questions are:

1. How do ridesharing platforms evolve from startups into large-scale market disrupters?

2. How can ridesharing platforms augment its user base and capture more segments within the transportation business?

To get an answer to these question; this thesis does a case study on the Uber app. Uber is perhaps best known for having grown to become a worldwide provider of taxi services. The service launched originally as an iPhone application. Hence, users downloaded the app, submitted their credit cards and signed up for the service. When users accessed the app, the interface flashed a map of available cars. By requesting a trip, the nearest available car took off to pick them up. Users paid for the fare by credit card. At the end of every trip, drivers and passengers were obliged to rate each other. When few drivers are around, algorithms send prices up to incentivize more drivers out on the road.

However, Uber have evolved over the years. Today, it is not so easy to describe the service. It still mainly works as we described it above; the only difference is that users can request many types of transportation services through digital devices. Accordingly; this is a study into how Uber have grown into being a global provider of

3 ridesharing services and how it tries to use this position in an attempt to pivot into larger parts of the transportation market.

Let us be frank, despite the fact that Uber has achieved growth on a global scale; it is not the only transportation app that has enjoyed considerable success. New digital transportation ventures are coming to the market in droves, and Uber face increasing competition from startups that have grown into becoming large regional players. Thus, Uber is currently the ridesharing company with the greatest global reach in an increasingly competitive and dynamic business space.

We shall see how Uber has managed to mobilize users onto its platform. Like no other, Uber seem to have developed a playbook that makes it able to scale over and over again. Uber have launched its services in over 576 cities as of April 2017. The starting point for this thesis is that Uber is likely to generate knowledge into how ridesharing companies have capabilities to grow their business exponentially.

We will also see that Uber has encountered pretty severe challenges along the way. In a longer perspective, these problems have grown into becoming obstacles to the augmentation of the user base. They have also played a part in putting Uber’s long term success into question, but cannot bear the responsibilities alone. The emergence of self-driving cars plays an equally important role.

Right now, Uber experience some fault lines in its exponential growth. Like some scholars have noted; being big and dominant in a market are not always enough, many factors can trip up a seemingly fast growing digital marketplace (Hagiu, A. & Rothman, S., 2016). In Uber’s case, much of the problems arise from the fact that its digital solutions also involve physical problems. Since Uber tries to solve inefficiencies in transportation, it leads to resistance from stakeholders with vested interest in upholding the old regime. Further, in the physical realm, geography matters and being dominant on the ground in every corner of the world is quite another thing than being dominant in a one-dimensional digital space. To get people to download an app is hard enough, putting vehicles on the road makes it even harder. In contrast to the companies in figure 1; Uber have not managed to saddle their users with sufficiently high switching cost yet. Moreover, as long as Uber’s technological

4 concept is easily replicable, it cannot shield of its market to the same degree as the other mentioned digital behemoths.

There are multiple reasons to why so many different challenges have been thrown up in front of Uber. One particular reason is that Uber is a trailblazer, and therefore have received the brunt of the attacking fire from forces that don’t like ridesharing. Another more abstract reason is that ridesharing is subject to physical constraints alluded to above. Uber must, therefore, chart a course where digital allowances comply with physical limitations. This duality explains why Uber executives are forced to devise fresh approaches to preserve and bolster its business prospects. It also leads to a whole new set of considerations, which up to now isn’t fully understood neither from an academic nor a business point of view. Hence, this thesis is called The Rise of Ridesharing Platforms: an Uber- assessment of bits and atoms, where bits stand for technology, and atoms stand for objects (mostly people) in the physical realm.

According to Uber, recruiting drivers is the key to scaling (Chen, A., 2016). This thesis frames the adding of drivers in a larger perspective. The primary argument is that Uber is in a recruiting sweet spot. This favorable environment seems to be universal irrespective of geographical and cultural spheres. The main argument of this thesis is that this phenomenon has something to do with a convergence in labor conditions between the first and the third world.

1.2 An Uber review

Uber’s disruptiveness is discussed all over the world. Despite the raging debate, Uber has not triggered much interest from information system researchers; concernment has primarily come from legal scholars, social scientists, and economists. Nevertheless, both information system researchers and economists have studied digital innovation as a phenomenon extensively and this research concern Uber just as much as any other platform. However, what makes sharing platforms distinct, are the involvement of physical assets, and in the case of Uber; it is the cars as economic goods and the drivers as a labor resource. Conditions in the labor market are therefore an additional factor to consider. In mass media, Uber stands out as the bad boy of the sharing apps, and are ruffling the feathers of many opponents eager to shut it down. Lawsuits are something which Uber have gotten used to, and it contributes to making governance a

5 complicated affair for the company. For our purpose, research in the field of labor sociology has more relevance than legal studies, if we want to understand how comfortably Uber mobilize drivers to its platform. It is the other way around; if we want to know what kind of challenges the company faces when administering and growing its user base.

1.2.1 Labor

Scholars have different views on Uber and labor. We can see a formation of three camps.

The Precariat

One camp interprets Uber’s rise more or less in the prism of Guy Standing’s (2011) notion of the Precariat. The Precariat is in Standing’s view an emerging working class who has trouble finding steady jobs. Their members face a life of insecurity, being in and out of short-term jobs that give little meaning to their lives. Scholars in this camp support their view by pointing to recent developments in Western labor markets, characterized by increasing bouts of high unemployment, lack of job security and a rise of workers hired on short-term contracts. The camp worries about what they construe as a devaluation of labor and see Uber as a poster boy for this development. In the eyes of these researchers, Uber’s rise is commensurate with the crisis of work, and without this precondition; its business model would never work. Their general view is that companies like Uber exploit those who are in desperate need of some sustenance (Isaac, E., 2014; Wallsten, 2015; Hill, S., 2015). These scholars argue that companies like Uber must be heavily regulated, if not; the precarious state of modern work will escalate, resulting in a cyberized Downton Abbey with massive wealth inequalities.

Entrepreneurs

In the opposite camp, we find proponents who are considerably more positive both to Uber and the sharing economy in general. Former Secretary of the Treasury in the Clinton administration, Lawrence Summers, portend that ridesharing ventures “bring significant economic, environmental benefits including an increase in employment

6 and a reduction in carbon dioxide emission” (Cannon & Summers, L. H., 2014, p. 1). Obama’s economic advisor, Alan Krueger, emphasizes how Uber have created work opportunities in an anemic recovery (Hall & Krueger, A. B., 2016).

However, no one goes as far as Sundararajan (2016). He portrays ‘gig-workers’ as proto-entrepreneurs and the sharing economy as something which potentially could prove to be a rejuvenation of capitalism itself.

That ridesharing lead to environmental benefits finds some support in research centered upon carpooling (Shaheen, 2016; Santi et al., 2014). Similarly, research by Dillahunt & Malone (2015) offers some support for the argument that ridesharing serves to broaden the cab market and provide opportunities for the unemployed. Whether sharing apps herald in a brighter capitalism; remains to be seen, it is a prophecy more than a research finding.

The splintering precariat

In a polarized debate, there is usually someone taking the middle ground, so also in this case. A study by Schor (2015) looks into what kind of income service providers can expect to earn on different platforms. By comparing income data across several platforms, she concludes that the broad diversity within each entity makes it too difficult to generalize. Despite her mixed findings, she tentatively notes that it seem to be more advantageous for providers when some financial assets are involved in the transaction, as opposed to only the provision of short-term labor.

In our context, the problem with Schor’s study is the absence of Uber in her dataset. A study looking into how drivers for Uber and Lyft perceive their work helps us in filling out what’s missing. The study argues that the driver’s perceptions largely depend on their life-situation. To work part-time for Uber can be a pleasant experience. Satisfied research participants cite the flexibility of the work hours as the main reason. Doing it full time, and as a second job to make ends meet is another story. Drivers in this group were more likely to cite job-stresses, economic worries, the need to work particular hours and other freedom constraints. The authors interpret their findings through a construct which they call a ‘splintering precarity’; the work can be great if you do it as a mean to get flexible supplemental income, but horrible if

7 you do it because there is no other alternative or as a mean to fix economic distress. The drivers face risks which workers in more secure full-time positions never experience, but are also facing brighter prospects than the growing numbers of disenfranchised, who are unable to handle even low-skill jobs (Malin & Chandler, 2016).

Introducing ‘Brazilianization'

This thesis is positioning itself very close to the views offered by those who are critical to Ubers labor economy and finds much sympathy with the notion of a ‘splintering Precariat.' Beyond that; regardless of the disagreements in the ongoing debate, there is some uniformity in the various opinions. All studies aim to interpret how sharing apps affect Western labor markets. The developing world is left out. Going forward with this parochialism might lead us to miss the total picture. In fact, Uber is likely to be very successful in emerging markets. Therefore, we need a theory that can encompass situations all over the world. In the search for a global theory, this thesis has selected Ulrich Beck’s theoretical concept of Brazilianization as an adequate lens to interpret Uber’s mobilization of drivers. Indeed, Beck’s theory can be a useful starting point to understand how drivers all over the globe so readily endorse the app as a work tool.

1.2.2 Innovation

Schumpeter (1994) describes innovation as a process of creative destruction where old businesses get crushed by new, superior and improved business models. Accordingly, we can loosely define innovation as an introduction of something new and more efficient to the marketplace. Hence, we can postulate that Uber is creative destruction at work.

Although Uber has upended the business conditions for taxi operators all over the world, we can characterize it as many types of innovations.

Disruptive innovation

Two criteria make an innovation disruptive. First, the novelty has to originate in a new market or significantly expand a traditional market. Second, the innovation has to

8 render traditional products obsolete and displace previous market leaders (Bower & Christensen, C. M. 1995). Furthermore, small businesses and startups are the most likely candidates for being disrupters (Christensen, C. M., 2013).

Uber should tick all of these criteria comfortably, but in a weird twist; Christensen denounce the company as a disrupter. The app has neither originated in a new market nor first targeted the low end of an existing market. Instead, Uber have exclusively made inroads into an already flourishing taxi industry and mainly attracted existing taxi users (Christensen, C. M., Raynor, & McDonald, 2015).

It is hard to challenge the pope when it comes to having the right beliefs, but it is not hard to understand why a majority of the punditry classifies Uber as a disruptor (Moazed & Johnson, 2016). First of all, Uber is an upstart who has broken through. Secondly, Uber tends to alter local taxi markets where it operates. It pushes prices for all cab services down, steal market share and put pressure on incumbents. As an example, the price of a medallion has plummeted around 50 % from 2013 to 2016 in New York City (Rahel, 2016). Finally, Uber have together with other ridesharing apps contributed to an expanded market, making users out of non-users. Turning to New York City again, as an illustration, Uber now have more cars on the road than taxi operators, and the number of cab trips is increasing by a factor of 4 times per year (Alley, 2016). Damodoran (2016) project that the numbers of ridesharing users in the US will exceed 20 million in 2020, more than a doubling from 2014.

Business process innovation

Innovation in processes is the ability to do more with fewer resources. Cramer and Krueger, A. B. (2016) argue that Uber crushes incumbents in capacity utilization. Thus, Uber drivers spend less time waiting for customers and are 30 – 50 % more likely to have a passenger in the car during business hours. Four factors contribute to the improved productivity. First of all, matching drivers and riders is more efficient through an app than through telephone dispatchers. Secondly, Uber’s growing chauffeur base provides a thicker market, which, in turn, leads to faster matches. Third, not being subjected to inflexible taxi regulations helps, and finally in the extension of this; by drawing upon a more flexible labor force and pricing the services

9 more dynamically, it is possible to match supply more accurately with demand throughout business hours.

We can add many factors to this list. Uber is solving much of the credence good problem associated with taxis. The problem arises whenever it is impossible to ascertain the quality of the service before consumption. A taxi trip can be classified as a type of credence good since users cannot know either the price of the fare beforehand or if the driver is going to choose the best route. The Uber app efficiently handles this problem as it provides the rider with a map where they can select the route and calculate the fare in advance. The standard of the car and the driver is also predetermined (Harding, Kandlikar, & Gulati, 2016).

Market efficiency is one side to the competitiveness of Uber. Another side has to do with organizational simplicity. The taxi industry is vertically fragmented. Taxi operators involve many entities, and the drivers and the vehicle owners are not always the same person. This fragmentation put a low upper band for the scale and scope of taxi operators. Uber have done away with all this fragmentation. It has a direct contractual relationship with its drivers, which are compelled to own their cars. Uber’s command lines are therefore less complex; allowing it to concentrate on adding users (Rogers, B., 2015).

What makes Uber able to deliver services to the market in more efficient ways has undoubtedly something to do with advances in technology. Uber has been able to outsource much of the activities which taxi operators previously had to do in-house, and it has been able to automate much of the operational activities linked to running a cab service. Billing of customers and user-matching are handled by software, while outside contributors take care of the transportation and the infrastructure side of the business. Alas, technology has shed the costs associated with communicating and transacting with a diverse user group (Karjalainen & Gang, 2016 ).

Open innovation

Open innovation means that companies should use internal as well as external ideas to develop new products and services. The basic thinking is that businesses can move forward technological advances faster when they are engaged in knowledge sharing

10 with outside partners. Accordingly, companies should realize that smart people work throughout the field and not within singular companies. Businesses should, therefore, do their best to pair its research with like-minded and friendly companies. The winners are almost certainly not going to be the lone geniuses tinkering from their garages, but those that understand the value of cooperation in an ever more fast-paced world of technological development (Chesbrough, 2006).

When businesses move to an open innovation model, this usually expresses itself in joint partnerships, alliances, mergers and acquisitions, R&D cooperation and so forth. There are indications that Uber have been moving into a more open direction in the last few years and it is especially Uber’s foray into self-driving cars development that underscores this trend.

Introducing platform economics

All the innovative processes referred to above captures important aspects of Uber’s evolution and will certainly be revisited later. Nevertheless, this thesis is going to argue that neither of them is the most salient if we go back to the first stumbling steps of the company. In fact, Uber is not one innovation; it is one of many ridesharing apps, which all are a result of a service bundle innovation. The Uber app work due to a host of novel technologies. The internet; the GPS enabled smartphone; open data and social media. It is the interoperability between different parts of this bundle which allows various components to work together (Miles, 2016).

Uber’s app platform is the significant result of this service bundle innovation. To understand that the platform stands out as the centerpiece of the innovation is of high importance if we want to comprehend how Uber could grow to be one of the world’s largest transportation companies without owning a single vehicle. This understanding rests on the fact that digital platforms have the capability to create a common digital point of reference between different sides of a market. This point of reference can transform itself into a mass market, i.e. becoming the clearing center for a massive number of transactions if it becomes popular among users.

Despite the potential for extracting rent from a wealth of transaction fees; one problem remains. How are platform owners going to get users interested in signing up

11 in the first place? In the case of Uber, drivers will assumedly only sign up if there are a high demand coming from riders, and oppositely, riders will presumably only sign up if there are an adequate supply of drivers.

As an answer to this conundrum, this thesis is turning to a collection of research which we call platform economics here, inspired by Constantiou et al. (2016).

1.2.3 Governance

Uber faces immediate challenges along five dimensions:

1. Driver’s employment status

2. Regulation in various local jurisdictions

3. Fiduciaries

4. Competition

5. Technological unraveling

Employment status

Are driver’s employees or independent contractors? – The problem is that ‘gig- workers’ fits neither category. For ridesharing companies, it is important to classify drivers as independent contractors. Employees cost more. They bring on payroll taxes, benefit expenses, and other social costs. Labor laws also protect employees. Contract Workers are paid only for the service they deliver. Naturally, many drivers see it differently and would have liked to be employed. Employment gives job security and more social benefits. The opposite views between the parties have led to labor disputes, which have ended up in American and British courtrooms.

According to legal scholars, the classification depends on three issues. What kind of purpose the providers on the platform serve for the company, how much control the app companies exert over its service providers and how much permanence there is in the relationship between the parties? Regarding ride sharers, one group of legal scholars side in favor with the drivers. They mainly contend that the purpose of the apps is to run transportation companies and that the drivers play a functional role in making this happen (Prassl & Risak, 2015; Davidov, 2016; Brown, 2016). How

12 passengers are encouraged to rate their drivers might also be considered a form of control (De Stefano, 2016). Work-friendly scholars typically denounce the independence of the drivers with phrases like: “At the end of the day, Uber drivers operate like employees with the unique perk of choosing their own schedule” (Brown, 2016, p. 43).

An opposing crowd argues that it not so straightforward, referring to the fact that only a few drivers work full time for one particular app. Most of them are part-timers, and most of them work for more companies than one in a single shift. These scholars will, therefore, consider it fair only to grant employment status in individual cases, where there exists some permanence in the working relationship (Woo & Bales, 2016; Means & Seiner, 2015; Eisenbrey & Mishel, 2016).

Government regulation

Most developed nations regulate taxi operators strictly. The need for monitoring comes from particular features of the business, most notably its low barriers to entry and the ever-present danger of oversupply. There are also issues related to safety, exploitative pricing and transparency. Hence, most cities keep oversight of the business by limiting the number of cab drivers through permits, licenses or medallions; standardizing the pricing of fares; dictating permissible work hours; instituting frameworks for inspection of vehicle and equipment; demanding record- keeping requirements and so forth.

Uber is opposed to nearly all of these regulations. The idea of ridesharing is that everyone should be free to share a car. This stance has pitted Uber against both taxi companies and local governments all over the world, making some of its adversaries take legal and legislative initiatives to stop Uber from operating in their area. Needless to say; whenever Uber gets slammed by negative regulatory rulings, business stands to suffer.

In general, US focused legal scholars agree with Uber that it does not make much sense to regulate ridesharing through the prism of old taxi regulations. Most scholars advocate that courts should adopt a balanced approach. They think that fair competition is a concern, but hard to regulate in a stable legal sense, and it is

13 questionable how much of a role courts should play in regulating the marketplace. The primary obligation of regulatory authorities should be to ensure driver, rider and public safety. Thus, ridesharing companies should commit to do a proper vehicle inspection and do a decent vetting of driver partners. Courts should also demand that ride sharers have commercial insurance policies (Elliott, 2015; Holloway, 2015; Koopman, Mitchell, & Thierer, 2015; Posen, 2015; Ranchordás, 2015).

Uber have been met with significant more push back from regulatory bodies in Western Europe than in America. However, Uber have managed to get their case heard in front of the EUs top judicial authority for market dispute arbitration. To be evaluated according to free market rules is quite another matter than being judged vis- a-vis local taxi regulation. The crux of the problem depends on whether Uber classifies as a transportation company or as a digital company. One legal scholar predicts that the court will favor Uber so that it can go either way (Geradin, 2016). We can anticipate a ruling at the end of 2017.

Fiduciaries

How Uber handles various fiduciary issues, has come under severe criticism. One line of criticism is that Uber is too lax about safety issues. Female passengers are an especially vulnerable target for violent and sexual crime. The main problem according to reporters is that drivers are not sufficiently screened and that Uber circumvents regulatory requirements like vehicle inspections, proper commercial insurance, and other safety standards. A few scholars have been more forgiving. They believe that the problem is largely going to self-correct as Uber reacts to criticism and complies with court rulings (Rogers, B., 2015; Feeney, 2015). Scholars have also refuted the claim that ridesharing is unsafe for female users. A quantitative study conducted in New York City indicates that the presence of Uber in an area leads to less sexual crime (Park, Kim, & Lee, 2016). Another study shows that ridesharing leads to less drunk driving and homicides in cities where ridesharing play an important part in the make-up of the transportation network (Greenwood & Wattal, 2015).

Another line of criticism deals with discrimination. A US study shows that black men get their requested rides canceled more than twice as often as other men and that they have to wait longer to get paired with drivers than white customers (Ge, Knittel,

MacKenzie, & Zoepf, 2016). The possibility to discriminate on the Uber platform comes from the online feedback system; rating system and the fact that users get a picture of the counterparty before the service commence. However, we are taking it far if we blame Uber for having bigoted users. Most people will view racism as an overall societal problem. Some argue that Uber is serving minority communities better than taxi operators (Smart et al., 2015).

Privacy risk is “potential loss of control over personal information, such as when information about you is used without your knowledge or permission” (Featherman & Pavlou, 2003, p. 455). Ongoing criticism about Uber’s misuse of user data is harder to rebut. The risk of losing control over personal information is not likely to go away soon on the Uber platform because the risks stem from features inherent in the technology stack.

The app comes with geolocation and user information exchange. It is the technology which makes riders and drivers hook-up. Simultaneously, these technological capabilities open up for a variety of data misuse. Uber collect vast quantities of sensitive user data, such as credit card information, cell phone numbers, logs of previous trips and information about real-time whereabouts. Katz, V. (2015) points out that Uber collects so much information that it by big data analytics can reach insights that are far more sensitive than what it can get from any single account. In a wider perspective, it opens up the specter of surveillance capitalism, where everyone becomes a target for data, extraction, and analysis (Zuboff, 2015). As almost everything we do, can be traced back to online transactions; “the digital exhaust from those actions creates associations and patterns that may be mined for insight, efficiencies, or more nefarious purposes” (Howard, 2014). The simple fact is that the real-time geolocation of its users provides Uber with surveillance capabilities that might be too tempting not to use.

Competition

Many scholars argue that Uber’s app platform is more about institutional disruption than technological innovation (Isaac, E., 2014; Laurell & Sandström, 2016). In their eyes, the brilliance about Uber has nothing to do with its technology, rather; its

15 competitiveness rests mainly on how it avoids rules that apply to traditional taxi operators. Effectively, by portraying itself as an intermediary, and not a cab service per se, the company can operate in a regulatory void.

Naturally, this creates an unfair competitive situation and is the main reason why so many taxi drivers protest whenever Uber move into a city. Objectively, Uber’s competitiveness is more than a perceived threat for taxi operators. It is real and happening right now. If we take a look at San Francisco, Uber’s startup market, we can see the first-hand fallout from digital disruption. Already in 2014, it was visible that severe price competition had taken its toll, leading both to a diminished market for traditional taxi companies and a sharp drop in income for an ever dwindling number of regular cab drivers (Isaac, E., 2014).

The displacement of incumbents does not mean that Uber is without competition. On the contrary, the competition gets tougher by the day. In the beginning, Uber was alone in offering app-based vehicles-for-hire, but by now; the marketplace is full of ridesharing apps. When we consult the databases of Crunchbase and AngelList, we are getting a tally of over hundred ride-sharing companies who recently have received angel funding.

If we take a look at Uber biggest competitors, we got the following picture. Uber faces national competition from two providers in the US: Lyft (220 cities) and Curb (60 cities). In Southeast-Asia, Uber is encountering increased competition from Grab, who currently have around 75 000 drivers in their network. Similarly, Ola has grown to become a major company in India, operating in over 100 cities with over half a million cars. In Europe, the competition comes from a bunch of individual companies domiciled in various local markets. In China, Didi has beaten Uber to the punch, forcing Uber to leave the market.

Scholars have started to take note of Uber’s growing field of competitors. Some see it as an example of how the disrupter gets disrupted (Elbanna & Newman, 2016). Others have studied how Uber fight tooth and nail with upstarts in emerging markets (Wirtz & Tang, 2016). Furthermore, case studies about other ridesharing apps are starting to come along (Kelestyn, Henfridsson, & Nandhakumar, 2017).

Technological unravelling

Self-driving cars have gotten much attention lately, and Uber are in the midst of it all. Longer term, an introduction of autonomous vehicles is putting significant aspects of Uber’s current business model into question. Longer term, it can do away with drivers, and in that case, who will own the cars? – Uber or someone else? At this point, we do not know if the self-driving technology is going to change the mass market for personal vehicles. At the end of the day, Uber have a vision about a future where most of us do not own cars but use ride-sharing for all our transportation. The idea is not so far-fetched; researchers have already looked into why this scenario could be an advantageous solution in many contexts (Fagnant, Kockelman, & Bansal, 2015).

There is no way to predict the future. For now, it is sufficient to observe that ridesharing startups need to worry by more than growing their user base; they also have to prepare for a future which might require a total rework of their technological and organizational underpinnings.

Uber executives might be right to on their toes in the face of this technological unraveling. The driverless revolution is by some accounts not far away. The span of 5 – 10 years is often mentioned and maybe shorter when it comes to long-distance trucking (ERTRAC, 2015). Uber have reasoned that the move to a driverless future has immediate consequences. It has to participate in the technological race if it wants to stay relevant.

In sum, this creates organizational challenges which new startups do not usually encounter. Autonomous cars are about hardware and software, research and patents, politics and regulatory frameworks.

Introducing generative digital infrastructures

Uber has grown an immensely large business. Despite their tremendous success, Uber managers have to deal with known-unknowns from day to day. How legal and regulatory matters are going to turn out are anyone’s guess. Moreover, it is tough to discern what the fallout of the self-driving car revolution is going to be. On top of all this, Uber executives need to ensure that their business is on a reasonable trajectory;

17 that they eventually will turn profitable and that they are competitive in an increasingly tougher market.

One thing is dealing with what’s knowingly unpredictable; another thing is to handle unknown unknowns. An unknown-unknown is an unpredictable event which is impossible to plan. There is hardly any point to speculate on what an unknown can be, but this matter to a considerable degree for Uber as a business. We need to be aware of the fact that ridesharing by no means is an industry that has matured into the Uber ways of doing it. In the end, users might turn to alternatives that are even more peer- to-peer, and collaborative, such as carpooling and car-sharing, which after all might be harder to monetize. There is also speculation that deep-pocket players like Google and Apple are going to enter the market and provide fierce competition.

This thesis is turning to a construct called generative digital infrastructures to get a handle on these management issues. First of all, the construct helps us understand the various complications arising from governing the Uber app. Second, it provides us with a theory which can help us to understand the approaches Uber have taken to find a way out from under the mess.

1.3 The ethics of Uber

The infamous venture capitalist Peter Thiel called Uber the “most ethically challenged company in Silicon Valley” (Lee, 2014). Thiel is not alone in his views about the morals of Uber. Moral criticism has centered on everything from what we can construe as foul speech in public to deliberate avoidance of rules that govern markets in most developed nations (Lu, A., 2015; Baker, 2014).

Ubers reputation is to ask for forgiveness rather than permission. Naturally, many find this provoking. However, Uber is not alone in this way of doing it; several technological companies are known to operate in this mode (Kirkpatrick, 2011). The ethics of this way of doing business are worth some consideration from a deontological perspective, but it will have no impression on the recipients since many innovators often just contra-punches with self-serving arguments. Travis Kalanick, Uber’s CEO has stated that: “Any rule has the right to become an old rule and be replaced by a new one. Ultimately, all rules have to bend towards people and

18 progress. Cities, states, countries that allow rules to move forward see quicker progress” (Sharma, Mishra, & Arora, 2016).

However, we can have a constructive dialog about the consequences of particular innovations in a utilitarian perspective. Thus, do ridesharing apps lead to beneficial or disadvantageous outcomes for the society as a whole? - Unfortunately, the debate about the impact of the sharing economy is on a polarized glide path between dystopian and utopian views.

As an example of a dystopic view, it does not get more dystopic than Callaway’s (2016) essay about how the sharing economy has turned San Francisco into a city of exploited workers. As an example of a utopian view, it is hard to be rosier than what Sundararajan projects in various speeches, interviews, blogs, op-eds and research papers listed on his personal website http://oz.stern.nyu.edu/. The debate between those that see the glass half full rather than half empty is “useful and worth continuing, but we should be clear that it has no end, and there will be no definite answer” (Kenney & Zysman, 2016, p. 64).

To show that it is not entirely black and white, we can use two other startups as examples. The first one is Juno, an upstart in New York City. It wants to compete with Uber based on being more driver-friendly and socially responsible. The company offers their drivers a stake in the company upon signing up, in effect giving them stock options. The goal is to have 50 % of the company’s equity owned by drivers within ten years. In that case, drivers become something more than freelancers, they become owners and have a say in corporate matters. However, whether this model is going to be the new standard bearer, depend on the reception the company receives in the market. Based on price and other features of the app, it is not much that differentiates it from Uber (Lazzaro, 2016). The company is currently doing around 20 000 fares daily in New York City (Bhuiyan, J., 2016d).

The second company is a French carpooling company called BlaBlaCar. The company was conceived in 2003 and founded in 2006. The service has become a huge success and currently operates in 22 countries with more than 25 million users. BlaBlaCar’s market is long-distance motorists that are susceptible to pick up fellow travelers along the way. Drivers in the network can take advantage of getting

19 companionship and lower traveling costs. For riders, the service has proven to be an alternative to public transport. BlaBlaCar discourages drivers from turning the gig into full-time work, which makes the company stay true to its carpooling mission, and allows it to operate without being challenged by government regulations. Aside from this, the company operates much in the same way as Uber, taking a cut on every fare (Kelestyn et al., 2017).

Contrasted against Uber’s business model, the two companies give us an entirely different trajectory of what the future might bring. The simple fact is that the sharing economy is in its early stages and we have to see more of how it is folding itself out before we pass a final judgement. Currently, about 0.5 % percent of the American workforce is working through an online intermediary, which isn’t particularly significant compared to the total workforce, but it is growing fast, and it certainly will be more important going forward (Katz, L. F. & Krueger, A. B., 2016). In this perspective, the gloom crew raises important questions about the precarious state of app based work. At the same token, the optimists have a good case when they point out that ridesharing is not all about Uber, it is also about sharing apps with a more reasonable approach to its participators. Nevertheless, we can be sure of one thing, if we are not willing to design our future; someone else is going to do it for us. At the end of the day, questions regarding the future state of work are too important, that we can leave it to chance. Governments, therefore, have to be brought in on the action.

Some states have banned Uber. In the long term, this is not likely to be a prudent way to go, it has an aspect of inertia about it, ensuring social harmony in the short term, but forgoing future gains in efficiency, productivity, and economic growth. Arguably, online sharing produces a more efficient use of underutilized assets, where cars are an asset that lends itself naturally to sharing: It is dear to purchase, but stand most of the time parked. In the end, countries with a statist attitude to sharing app innovations are likely to end up poorer.

On the flip side, too put on a laissez-faire stance is not viable either, especially for Western economies. It can lead to income disparities which have adverse effects both in the short and the long term. Economists have in later years grown into being apprehensive about income inequalities; speculating that it can lead to low terminal growth (Stiglitz, 2012).

The most viable alternative seems to be something in the middle. The sharing apps should be able to thrive while the platform owners should be subject to certain reciprocal obligations toward service providers on their platform and society as a whole. Harris and Krueger, A. B. (2015) puts forward one proposal for how such a space can be carved out in the US. The two researchers argue that app companies should offer social security and Medicare payroll taxes for their platform workers, but not things like overtime pay. Indeed, some middle ground like this should not be too hard to find most Western nations. In all likelihood, the majority of us would agree that the regulatory regime as of today is too binary (Hagiu, A. & Biederman, 2015), while a minority seem to be unwilling to consider any compromise (Eisenbrey & Mishel, 2016).

For the time being, the discussion above is pure academics; very few governments have been interested in tackling these issues head on. As an odd man out, the Philippines was the first state to issue rules and regulation which puts a minimum of requirements that ride sharers need to adhere too (Morales, 2015). American states and individual cities around the world have done something very similar, but no other nations have passed laws on a national level. Furthermore, no government has tried to broaden the right of app workers. Kicking the can down the road seems to be the order of the day and is one of the main reasons why the impact of the sharing phenomenon is so hard to discern.

Today, it is fair to say that the ethics of Uber goes from courtroom to courtroom. Sometimes the courts side favorably with Uber, in other cases not. In the meantime, Uber has plans for what’s in store for us. The company expects self-driving cars will end private ownership and that everyone’s drive is going to be an Uber car delivered just in time (Bradshaw, 2014). If things play out the way Uber conceive; it is going to be a revolution that puts the current disruption of the taxi industry in the shadow. Car production is an important part of the manufacturing base in several industrial economies. So, it is highly conceivable that any new development is going to have a significant impact and that it will catch us by surprise; just as it did on disruptive changes in the taxi business.

2 Background theory

Three theories were put forward in the introduction to make sense of Uber as a ridesharing disrupter: Brazilianization, platform economics, and generative digital infrastructures.

2.1 Brazilianization

As a theory to understand the ease of Uber’s driver recruitment, it might seem strange to cite the name of a country, but even though Brazilianization has something to do with the country, it has more to do with an idea. Beck could, in fact, use the term ‘Latin-Americanization’ because he uses examples from all over the continent (Beck, U., 2000).

2.1.1 The first world

So, what is Brazilianization? - In short, Beck claims that countries in Latin-America, “with their high proportion of ‘informal’ multi-activity work, may reflect back the future of the […] western core” (Beck, U., 2000, p. 93). In clear text, this means that working conditions in the West will increasingly resemble labor markets seen throughout developing nations such as Brazil. According to Beck, job market convergence between the first and the third world will have a broad range of repercussions. Primarily, it entails that full employment has entered a state of terminal decline in the developed countries. In the future, only a small part of the Western workforce will be able to depend on full-time paid work and its associated perks: health care, pensions, unemployment benefits and so on. Full employment will be crowded out by alternative work arrangements like the ones that we see a high proportion of in Brazil, such as, self-employment, part-time work, short-term work contracts, contract labor and other conceivable informalities. Moreover, systemic high unemployment is not going to be something that belongs to the excluded inhabitants of Brazilian favelas, but something which will be imported as a constant to the West as well. In total, this process is what Beck describes as the “Brazilianization of the West” (Beck, U., 2000, p. 106).

The notion of Brazilianization is related other parts of Beck’s work. His description of contemporary society as a ‘risk society’ stands out in this respect. According to Beck, the increasingly fragmented state of labor conditions is only one of many risks which have beset the western world as a result of something he calls a transition from the first to the second modernity (Beck, U., 1992). Economic globalization is a central tenet of his thinking, which in his mind represents a shift in power towards capital and away from the bargaining power of national labor unions. Thus, the second modernity is “typified by growing social inequalities, ecological crises, and an increasing individualization of work ... [representing a] shift from the first modernity which took institutional shape in Europe in the post-World War II period and which was centered on paid employment and typified by the standardization of work, full employment, the welfare state, and exploitation of nature” (Kalleberg, 2013, p. 701).

Taken together, Beck’s views about globalization, modernization, and risks produce a dystopic vision. Beck prophesied that under Brazilianization four groups would emerge. (1) “The Columbus class of the global age” as the winners of globalization. (2) “Precarious employment at the top of the skill ladder.” (3) “The working poor” with low or unskilled jobs. (4) “Localized poverty” of the no longer needed (Beck, U., 2000, pp. 106-107).

Many argue that Beck falls very deep in gloominess. It can be viewed as an exaggeration at least on a short-term basis when Beck suggests that the precariousness will travel up the skill ladder and serve to hollow out the middle-class, but regardless; there is no question that he got the vector of his call right. Beck’s call makes him a trailblazer in many respects, in later years several economist and sociologists have more or less copied his views about the rise in income inequality and casual work arrangements (Standing, 2011; Piketty, 2014).

In the United States, “the percentage of workers engaged in alternative work arrangements - defined as temporary help agency workers, on-call workers, contract workers and independent contractors or freelancers - rose from 10.1 percent in February 2005 to 15.8 percent in late 2015” (Katz, L. F. & Krueger, A. B., 2016, p. 1). In the European Union, we see the same trends. Eurostat now projects that over 30 percent of the workforce is in temporary employment. “In the case of Belgium for example, the numbers of hours worked by agency workers rose by over 11 % in 2015

23 alone” (European Political Strategy Centre, 2016, p. 3). In light of this, it could be interesting to investigate how far up the skill ladder precariousness have traveled, but for now, at least; the focus will be on the fact that the overwhelming proportion of insecure work still exist in low-skilled occupations.

Chauffeuring is considered a low-skilled occupation. All that is needed is a driver license. However, for Uber drivers there exists one substantial caveat. Uber demand that the cars ‘in its fleet’ are of a certain standard. Many drivers in the third world are not able to finance such vehicles. In the developed nations, the requirements are usually not insurmountable.

This thesis will posit that the low barriers to entry and a slow recovery from the great recession are part and parcel of why it is so popular to drive for Uber. We know for instance that the great depression led to an explosion in the number of taxi drivers in big American cities (Elliott, 2015). So, macroeconomic distress is part of the story. However, far deeper; this thesis will argue that Uber is in a recruitment sweet spot due to the ongoing Brazilianization of the West. If this is the case, Uber’s recruitment pace will continue unhampered despite cyclical upturns in the economy.

2.1.2 The third world

For our purpose, the comparable framework of Beck’s analysis gives us an entry point to make sense of Uber’s popularity in emerging markets as well. Beck points out that although informal work is prevalent in Latin-America, it has by no means reached a plateau. Despite the fact that Latin-America witnessed stabilization and higher growth in the 1990s, there was no increase in regular work, rather; temporary job arrangements crowded out formalism (Beck, U., 2000). For this reason, we should expect that emerging markets are conducive for Uber’s business model. Thus, in countries where self-employment already is high, Uber should be able to recruit fairly quickly.

In the West, the Uber app is by many viewed as a device which brings workers down in deeper precariousness, but this might be opposite in the third world. In emerging markets, the Uber app might give some order to the reign of informalism, in effect bringing more transparency about prices, payments, and the regularity of salaries. Due

24 to Beck’s comparable analysis, this thesis will, therefore, contend that increased hardship all around is not a primary cause for recruitment ease in emerging markets, but rather the likely benefits drivers are likely to see in doing app based work. Alas, opposite from the first world, recruits in developing nations are more disposed to see the prospect of driving for Uber as a social enhancement. The promotional aspect may also have something to do with third world chauffeurs perceptions about the luxuriousness linked to smartphone usage and Uber’s insistence on using nice cars. Disenfranchised might, therefore, have no chance to join the platform. It could, therefore, be interesting to see what kind of workarounds Uber introduce, if any, to make their recruitment socially fair.

In the third world, app-driving might bring advantages in a wider perspective; if drivers come together around digital platforms, the next step could be a coming together in a social sense, leading to the formation of bodies where drivers find a collective voice to address their needs. Moreover, many of the big metropolises in the third world are witnessing tremendous population growth. Ridesharing can help these cities cope with growing pains in transportation infrastructure and keep a lid on pollution problems. We also know that smartphone usage is on the rise among city dwellers in a growing number of emerging markets (Milian, 2014a).

2.2 Platform economics

Before diving into platform economics, we need to understand why economist views digital platforms as synonymous with markets. “Economists view platforms as special kinds of markets that play the role of facilitators of exchange between different types of consumers that could not otherwise transact with each other” (Gawer, A., 2014, p. 1240). In Uber’s case, we can nod approvingly to this. The Uber app facilitates the use of private transportation services on a scale which clearly surpasses previous forms of analog market organization.

2.2.1 Multisided platforms

Economists refer to platforms as “multi-sided,” since they facilitate “two-sided markets” (Rochet & Tirole, 2003, p. 990). As figure 2 depicts, users on both sides of the market transact with each other and Uber generate income from the market

25 interaction. In Uber’s case, many stakeholders want to have a word in on how the market should function.

Figure 2: Uber's multisided market.

2.2.2 Network externalities

The concept of network effects is central to platform economics. Network effects mean that there is a positive correlation between the number of users and the value it has for the individual user. Two-sided network effects typically characterize sharing platforms. There exists a mutual self-reinforcing feedback loop between the two sides. “With two-sided network effects, the platform’s value to any given user largely depends on the number of users on the network’s other side. Value grows as the platform matches demand from both sides” (Eisenmann, T. R, Parker, G., & Van Alstyne, M. V, 2006, p. 2).

Positive network effects have the power to bring about exponential growth. “Once the number of users who adopt the digital technology reaches a critical point, the value of that platform for potential user’s increases rapidly. This creates positive feedback loops and incentives for their existing users to stay, and others to join become high, all whilst creating less room for competition” (Kelestyn et al., p. 4777). Network

26 effects give platform businesses a combination of offensive and defensive capabilities. At the end of the day, network externalities have the potential to result in ‘winner-take-all’ markets, where the winner in some instances dominates the market to the point of being a monopolist. It is the prospect of windfall profits from this growth trajectory which motivates owners of platform companies to scale in a hurry.

2.2.3 Early adopters and marquee users

What startup platforms should do to get the bandwagon effect of network externalities going is one of the leading questions in platform economics. This research is centered on how platforms are supposed to attract a critical mass of user until the growth reaches a tipping point, alas, before positive network effects kick in and escape velocity occurs in the growth trajectory. Platform economics portray this question as the “chicken and egg problem” (Rochet & Tirole, 2003, p. 990). Consumers have few benefits in participating if there are few producers around, and vice versa, producers, will not see any point in participating if demand is scant.

A classic study in this field stipulates that innovations reach a critical mass when 13.5 % of all users have adopted it. To attract the first users is a difficult affair, and the early adopters can be described as any startups fickle, but vital friends. The first users play the role as risk takers (testers) and opinion leaders (marketers) in the diffusion process. Rave reviews from them can make the innovation take off almost immediately, or oppositely; be deemed dead on arrival (Rogers, E. M., 2010).

Other researchers have done many of the same observations. Eisenmann et al. (2006) suggest that participation of marquee users can serve as an attractor for all types of users. Marquee users spend more, stick around longer, provide feedback and get the word out. Perceptions are everything here. Users tend to follow trusted opinion leaders onto platforms when they hear good things about it and perceive it as the place to be.

2.2.4 Innovation, subsidizing or quality

Sharing platforms have to ensure that users populate both sides of the market and this complicates the recruiting process. Platform economists pick out three scaling

27 strategies that platforms can use. The first strategy is opening up the platform for innovation. Platform economists posit that platforms have the potential to develop faster when creative outsiders are invited to contribute. However, platform owners must take into account that there are both advantages and disadvantages to this strategy. It is positive if openness results in larger scale and more revenue sources (Hagiu, A., 2014). It is negative if it results in a platform that is harder to manage and control. Less control can have fatal consequences; valuable assets might get lost, or even worse; the platform can get swallowed by a third-party which eventually grows dominant within the ecosystem (Eisenmann, T. R., Parker, G., & Van Alstyne, M. V., 2008).

The second strategy is subsidizing one of the two sides of the market (Rochet & Tirole, 2003; Parker, G., & Van Alstyne, M. V., 2005). Hagiu, A. (2014) suggest that most multi-sided platforms subsidize at least one side of the market. In Uber’s case, riders can be recruited faster when fares get priced artificially low. Likewise, drivers can be signed on faster with guarantees of high wages, bonuses, or other perks. The platform owner can, therefore, look at subsidies as an investment to gain a critical mass of users. However, the platform owner also needs to consider the dangers of being too aggressive. Low prices can produce an abundance of demand, while producer subsidies can create a supply glut. Getting subsidizing right is imperative for not ending up with an unbalanced market (Eisenmann, T. R. et al., 2006). Besides that, the financial muscles of the startup decide how much and for how long subsidizing can go on. Platform owners cannot bleed cash forever. Naturally, it also depends on how long it takes for the platform to scale. If the company has considerable financial wiggle room, cross-subsidizing both sides of the market might be an alternative.

Market dynamics chance constantly. To identify these dynamics are important because this can give the platform owner clues about which side of the market to subsidize and information about whether subsidizing is going to work effectively (Rochet & Tirole, 2003; Parker, G. & Van Alstyne, M. V., 2005). To take note of the potential competition is vital. Market dynamics are different when adversaries cannot compete on price versus when the competitors might be willing to engage in price

28 wars. Severe price competition can have binary outcomes. Price competition is, therefore, a game that requires confidence, risk taking, financial muscles, and stamina.

The final strategy is to make value propositions about the superb quality of the service (Eisenmann, T. R. et al., 2008). If subsidizing seems too risky, competing on quality might be an alternative. This strategy usually consists of building up a reputation for providing exclusive services and attracting users through brand recognition. There are different ways to sell quality to users. The term involves both objective and subjective aspects. In objective terms, quality has something to do with performance, reliability, durability, serviceability, and aesthetics. In subjective terms, quality has something to do with capturing the zeitgeist of the day. Fashion makes users perceive some products and services as more valuable than others. In high-end markets, it is often imperative to recognize that consumers want to attach feelings of hope or ambition to the product.

2.3 Generative digital infrastructures

Generativity and information infrastructure are two separate concepts. Some scholars view them exclusively as separable theoretical lenses (Eaton, Elaluf-Calderwood, Sorensen, C. & Yoo, 2015), while others treat them as part and parcel of one construct where generativity serve as a precondition in the evolution of successful digital infrastructures (Henfridsson, O. & Bygstad, 2013). This thesis is taking the latter position and deems it as something which can be called generative digital infrastructures.

2.3.1 Complex adaptive systems

Before treating the two concepts as an integral whole, we need to clarify their meanings. Infrastructure is the underlying structures which make our society function the way it does. The underlying structures consist of various components, such as; roads, rails, airports, ports, electricity, telecommunications, digital assets, water systems, sewerage, waste management and so on and so forth. These components are in themselves distinct infrastructures, which also consists of key elements.

There are different views about how to best describe the properties of digital

29 infrastructures. The most relevant perspective for our purposes is the one who interprets information infrastructures in light of socio-technical issues. Thus, the emphasis on socio-technicality means that digital infrastructures involve everything from users, owners, developers, human resources, organizational power dynamics and technological artifacts to various cultural-, social- and political circumstances (Hanseth & Lyytinen, 2010).

Information system researchers infer that several contemporary digital service systems have grown global in scope and scale and that maintaining them are a complicated socio-technical affair (Tilson & Sørensen, C., 2010; Hanseth & Lyytinen, 2010; Ghazawneh & Henfridsson, O., 2013; Eaton et al., 2015). To characterize digital infrastructures as complex adaptive systems lies near at hand (Hanseth & Lyytinen, 2010). Complex adaptive systems are “systems that involve many components that adapt or learn as they interact” (Holland, 2006, p. 1).

In Uber’s case, we are back to known-unknowns and unknown-unknowns again. Complex adaptive systems sustain themselves through self-organizing mechanisms, which are inherently intricate to govern. We can visualize ourselves the system as multiple nodes in a web; where changes in one node have non-anticipatory effects on other nodes, which, then again, lead to unanticipated ripple effects throughout the system. To rule these systems through commands is hopeless, they have to be nudged along through a gardening approach. Small changes in a complex system can have a massive impact or none at all; it is impossible to know beforehand. A gardening approach to governance is anything but passive. The managers have to play the role as enablers which cultivates the ecosystem in which the organization operates (McChrystal, McChrystal, Collins, Fussell, & Silverman, 2015; Hanseth & Lyytinen, 2010).

2.3.2 Generativity in digital infrastructures

Generativity exists in digital infrastructures when outsiders are allowed to contribute to its further development. Thus, when an infrastructure exhibits a high degree of generativity, the system can be conceptualized as a shell in its initial stages. That is before outsiders start to contribute; combining parts in the shell with other components; making an endless stream of additions and recombination’s, and so forth

30 linking up various technical components to serve the end-purposes of different contributors. Due to user initiated innovations; systems can end up in something entirely different than what the founders originally conceptualized.

2.3.3 Generative contextual conditions

The fault line of the digital age is between open and closed systems (Slaughter, 2017). Generativity is dependent on openness in the digital architecture. The unfinished task of building the Internet is an excellent example of how an information infrastructure has grown as a result of openness and countless contributions from outside users. The internet has organizational and architectural features which make outside contributions possible. More specifically, it is the decentralized control and loosely coupled architecture of the Internet, which allows contributors to recombine new elements with parts of the whole as long as they adhere to the standardized interfaces governing the structure (Henfridsson, O. & Bygstad, 2013). It is the malleability of these structural properties which gives the Internet its generative capabilities. Capabilities that we can call: “a capacity to produce unprompted change driven by large, varied and uncoordinated audiences” (Zittrain, 2006, p. 1980). We can, therefore, summarize that decentralized control and loosely coupled modules are key contextual conditions for making digital infrastructures generative (Henfridsson, O. & Bygstad, 2013).

Researchers have long been aware that architectures win technology wars. “Simply stated, success flows to the company that manages to establish proprietary architectural control over a broad, fast-moving, competitive space “ (Morris & Ferguson, 1992, p. 86). In general, platform-centric ecologies are considered to be effective in establishing these architectural control points. Google and Facebook illustrate the controlling and monopolistic force of digital platforms (Eaton et al., 2015).

Although Uber has grown into being domineering in the taxi-app business; it is by no means almighty. In fact, all taxi markets are hyperlocal; with their long-standing regulatory practices, so it is questionable whether ridesharing naturally lends itself to a global monopolization. There are also question marks related to how much the entire global taxi market is worth (Damodoran, 2016). However, Uber’s investors see

31 the company as something which transcends taxis. They see it as an intermediary who brings all types of logistics under its influence (Gurley, B. 2014b). If we side with the investors and posit that Uber can evolve into something that resembles a global transportation oligopoly, then Uber have to conquer multiple segments of the market.

According to our construct; openness, outside contributions, scaling, and the extension of the platform into new use areas are key ingredients in generative digital infrastructure evolution. We can also posit that mechanisms which bring about such a development can be considered generative (Henfridsson, O. & Bygstad, 2013). Here we are talking about mechanisms in a strictly philosophical sense, meaning, physical causes to natural phenomena. In the case of Uber, we can notice three generative mechanisms: replicative patterns, networks of networks, and innovation. These are all mechanisms which Uber need to cultivate to sprawl into more areas.

2.3.4 The generative replicative pattern mechanism

The replicative pattern mechanism has to do with how sharing platforms deal with various obstacles stemming from physical realities. Ridesharing platforms differ from platforms that only are present in the digital sphere. For Google, it does not matter whether searches happens in Hong Kong or Moscow. A search is a search. For the Apple app store, it does not matter whether apps get downloaded in Africa or Asia. A download is a download. Pure digital ventures scale distances effortlessly, and the whole world is an instant tumbling ground. It is not like that for ride sharers: distance matters. Startups have to conquer markets one by one.

Kelestyn et al. (2017, p. 4776) conclude that all successful ride sharers have developed a playbook for repeated scaling. They “view such rapid global scaling as an organizing logic by which ventures replicate a generic solution to recurring challenges, that are found in expanding a user base across markets; which is usually characterized by slight variables in their conditions.”

Apparently, no sharing company can scale over vast distances if they have to invent the wheel for every market they encounter. Presumably, practice makes perfect. So, learning is part of the scaling process. The general solution must also be open for revisions. Things that worked in one market do not necessarily work in another. The

32 reason for this is that business practices can be dissimilar across regional boundaries (Hofstede & Hofstede, 2001). Markets and cultural differences should, therefore, be analyzed ex-ante. However, when the dynamics of a regional market is figured out, it gets easier to launch within the same sphere.

To introduce the Uber app in new cities are an essential part of the scaling process. It could, however, be interesting to broaden the perspective a bit. Scaling is not only about a one-directional growth path. Healthy growth needs intervals of consolidation. Uber can also deal with consolidation issues in a general recursive manner. Uber’s competitors may, for instance, be different from market to market, but this does not mean that they need to be dealt with all that differently. The same goes for regulators. The rules might be different from city to city, and from nation to nation, but this does not mean that there cannot be replicative answers for how to deal with anticipatory push-back coming from authorities in new cities. Compared to traditional businesses, startups in the digital realm may be more prone to use recursive algorithms when they encounter problems. This way of thinking about problems is usual for software engineers. It is part of their training.

The replicative pattern notion is primarily a top-down scaling mechanism but is dependent on decentralized decision-making authority to work efficiently. Local branches must have the ability to react to unforeseen events and adapt the service to local requirements.

2.3.5 The generative network of networks mechanism

Here we are talking about a collection of mechanisms which center on the fact that Uber is a vast network consisting of users, drivers, local and regional offices, company headquarters, executives in managerial positions and regular employees. Naturally, Uber can also invite outside stakeholders to take part in the network. Investors are obvious candidates. Business partners are another. The network - extended by outsiders or not - is an asset which can be used to strengthen and expand the platform.

Like openness is a precondition for generativity, openness is also encapsulating the logic of networks. Networks are participatory and derive strength from the

33 involvement of the many. Control of information is difficult to achieve in web-like structures, transparency rules. Moreover, networks are autonomous; they encourage self-organization (Slaughter, 2017).

Within a vast network, there are various coalitions of networks working side by side. A team of teams is an example of a network that combines free communication with decentralized decision-making authority. It is about combining adaptability, agility, and cohesion of a small group with the power and resources of a large organization (McChrystal et al., 2015).

Networks self-organize to defend, respond and stabilize. Hence, resilience is central to the idea of networks. Resilience, however, can only be achieved if networks efficiently solve tasks. As a result, task networks often emerge within web-like structures. Task networks are networks which are created to carry out “time-bounded tasks.” We can distinguish between three groups of such networks: (1). “A cooperation network is a linked group of individuals working together to solve prescribed tasks in a prescribed way.” (2) “A collaboration network is a linked group of individuals working together to figure out the best way to solve a prescribed task that itself may evolve over time.” (3). “An innovation network is a linked group of individuals tasked with generating new ideas, processes, and/or products in a service of a prescribed general goal” (Slaughter, 2017, pp. 111-112). A team of teams is a collaboration network.

Networks of networks are a cluster of interdependent networks chained together. Transportation networks are only one of many networks which interact and are dependent on other networks to function (see figure 3). The system coalesces and stabilizes around powerful entities which serve as attractors (Hanseth & Lyytinen, 2010). Followingly, enterprises on the way up can enhance its attractiveness if it has the capabilities to partner up with other organizations. Thus, the more important role an enterprise (Uber in our case) play as a node within the transportation space, the more crucial it get for the functioning of other transportation entities and nodes within other networks in the connected system. There are positive feedback-loops at play here. The interdependent linkages between entities foster cooperation, and cooperation in one field spawns colluding in others. In general, cooperation is a good thing; it means that additional resources are put to the attractor’s disposal more or less

34 for free (Chesbrough, 2006). We can call this cooperative predisposition for the self- organizing invisible hand of interdependent networks.

Source: (Gao, Li, & Havlin, 2014).

Figure 3 Framework for network of networks

2.3.6 The generative innovation mechanism

The innovation mechanism consists of several building blocks. One can look at it as an organic process happening from within. Then, innovation is “a self-reinforcing process by which new products and services are created as infrastructure malleability spawns recombination of resources” (Henfridsson, O. & Bygstad, 2013, p. 918). Another way to look at it is to acknowledge that new services and products also can happen by inviting outsiders to make additional digital applications on top of the platform (Ghazawneh & Henfridsson, O., 2013). To open up the platform for innovation is also a central idea in platform economics, which an alert reader already may have noted.

3 Methods

3.1 Case selection

The idea of doing a case study of Uber happened in a succession of stages. It all started with a general interest in information infrastructures and digital platforms as research fields. A rich research literature accompanies both of these themes. The varying perspectives in the academic papers can be overwhelming but are also one of the features which make it an attractive research field to follow.

Despite all the inquiries that have gone into digital platforms, there are only two studies on the rise of sharing platforms conducted by information system researchers. One study is about how Airbnb evolved from a startup into a sustainable business, and a similar one into the growth of BlaBlaCar (Constantiou et al., 2016; Kelestyn et al., 2017). The small number of studies creates advantages for prospective researchers. By delving into a relatively new research fields, there exist a chance to learn something new and a possibility for making findings that will contribute to insight about a phenomenon which certainly is going to be researched extensively in the coming years. This assessment led to a motivation of doing a case study about one particular sharing venture.

The search for a case quickly led toward Uber. At first glance, Uber can seem like Airbnb’s spitting image, albeit in a different market. Both have scaled their user base extremely fast; both have evolved from startups to global businesses; both are considered market leaders in their respective segments, and primarily for this reason; both are subject to extensive public discourse. It is, therefore, natural to assume that many of the same mechanisms which serve to explain Airbnb’s rapid growth also must apply to Uber.

However, what seem very similar at the surface, changed markedly when the researchers started to look into the collected research material. In comparison with the two other studies undertaken in this field, this thesis is an iteration, which serves to broaden the theoretical fundament from where to understand the rise of sharing platforms.

3.2 Methodology

A case study can be defined “as an intensive study of a single unit with an aim to generalize across a larger set of units” (Gerring, J., 2004, p. 341). The selected case is, “asked to perform a heroic role: to stand for (represent) a population of cases that is often much larger than the case itself” (Seawright & Gerring, J., 2008, p. 294). Thus, case selection is an important part of the research process.

Kelestyn et al. (2017) look at BlaBlaCar as an extreme case. Their overarching aim is to use the exceptionalism of the company’s tremendous scaling to develop new theories. An extreme case is supposed to reveal more information than a typical one. Hence, extreme cases “can be well-suited for getting a point across in an especially dramatic way” (Flyvbjerg, 2006, p. 229).

This thesis is also an extreme case study. In fact, Uber stands out more from the rest of the ridesharing platforms than BlaBlaCar. Primarily, Uber is the only ride sharer who is present on all continents. Uber is, therefore, a case which more than anyone demonstrates what it will say to scale over cultural boundaries and vast geographical distances. Uber has also sprawled in different directions. It is no longer exclusively a ridesharing platform. Today, Uber describe itself as a digital logistics venture.

This thesis employs a multidisciplinary theoretical approach. Theories from sociology, economics, information system research and studies of international relations work in concert to answer our research questions. It is an approach which differs from the BlaBlaCar study, where the three researchers spun out a new theory heavily colored by an information system view of the world (Kelestyn et al., 2017). This thesis is taking advantage of insights reached from this study, but marching through the empirics of Uber, it becomes clear that one-dimensional theories lack sufficient explanatory power. A multipronged theoretical approach serves us better. In this respect, this thesis has more similarities with the Airbnb study, which uses platform economics and information infrastructure theory as explanations to the scaling of the user base (Constantiou et al., 2016).

Case studies often get criticized for being based on subjectively construed narratives. A general line of attack is that the narratives often are colored by the researcher's views beforehand and for this reason has little value as a scientific approach. Like any

37 other case study, this thesis opens itself up for this type of critique. However, the critique does not fit. The researcher felt a considerable degree of cognitive dissonance trying to interpret the different sources collected throughout the study. Finding an all- encompassing theory was hard. Some parts of one theory fitted, but then again other parts did not. Hence, it took a long time to establish theoretical lenses in which to interpret the research object. An eclectic approach was necessary. Many pieces had to fall into place, and that is why the theory employed here are so broadly based.

Singular case studies provide us with rich descriptions of one particular unit within the phenomenon which we are interested in understanding. Naturally, detailed descriptions give us an intimate knowledge of the particular case, and this understanding helps us in making sense of the overall phenomenon. Much of the sense-making is about interpreting what we in normal conditions can construe as being more or less generalizable facts, and oppositely; what we can assume to be particularities which are special for our case. Hence, the objective is not only to uncover generalities, but also to uncover ambiguities which might be hard to explain in general terms. Ambiguities are not problems for researchers per se, but something that contributes to the exhilaration of going in-depth into a case (Flyvbjerg, 2006).

3.3 Data collection

There are two ways for researchers to source data in qualitative studies. One way to go about it is to create data in participation with the research objects. Observation is an example of one method used to collect this type of data. Interviews are another. Another way to go about is to collect data which already exist on the research object.

The first approach is usual when there is little available material on the research object, while the second is more usual when there is an abundance of data. Quite often we see qualitative research that builds on a combination of the two methods. This thesis, however, only collects data that already is out there. In fact, there is so much material produced about Uber that the main challenge in the research process consists of sifting through the weeds – selecting the most relevant data points.

In the collection of existing data, we separate between primary and secondary sources. The difference between them varies depending on the research topic, but in our case,

38 we can posit that; primary sources are data originating from the research object itself, while secondary sources are material on the research object written-up by someone else.

This thesis has collected the following primary sources:

• Artifacts (Uber applications)

• Recorded audiovisual interviews and speeches made by the leadership of the company

• Press briefings (Uber Newsroom)

• Technical information (Uber Engineering)

• Developers feed (Uber Developers)

• Blogs (Uber investors, financial analysis, former employees, driver forums and so forth)

Is there some data that is missing? – Yes, Uber is not a publicly listed company and are very restrictive with releasing user statistics and other quantifiable information about the enterprise.

Secondary sources can complement primary sources. This thesis has collected the following types of secondary sources:

• Research literature (see introduction)

• Commentaries (mass media)

• News (mass media)

We have already reviewed the research literature on Uber. The reason for having done this is threefold. One, it organizes the main points of academic contention around the Uber platform. Two, it offers interesting perspectives which we can make use of in the sense-making. Three, it serves as a backdrop to why this thesis chooses to analyze the material in a certain way.

As an approach to generate reports from mass media, this thesis has made extensive use of the tech news aggregator http://www.techmeme.com. However, the aggregator generates almost exclusively news stories from American media. Europe and the rest of the world are often left out. To cover the fact that Uber is a global company, this

39 thesis venture out in the distributed web to find commentaries and news stories which are non-American. The main reason for doing this is to avoid entrapment by an overwhelmingly American view of the world. This study is also interested to find out about labor and scaling dynamics from a non-American view.

3.4 Data analysis

Collecting data is the first step, using it constructively is the next step. This thesis has used the data to two ends. First, to explain how the Uber app evolved into a successful multi-sided platform with a huge user base. Second, to make sense of what kind of strategies Uber employs to augment the platform in light of all the emerging obstacles alluded to in the introduction. To get a handle on these two sets of inquiries; a review of the gathered data was undertaken to establish a timeline of the company’s development and to pick out the most significant themes vis-à-vis our research questions.

Five themes stand out as significant:

 One, to find out how Uber managed to entice early adopters and efficiently solve the chicken-and-egg problem. The theme gets described under the heading: How Uber conquered San-Francisco?

 Two, to discern the reasons behind the exponential user growth. The theme gets explained under the heading: How Uber scaled exponentially?

 Three, to take heed of all the problems which followed in the wake of Uber’s digital disruption and how these complications lead to obstacles for further growth. The theme gets analyzed under the heading: Augmentation obstacles.

 Four, to find out what kind of strategies Uber employed to overcome the barriers alluded to in theme three. The theme gets evaluated under the heading: Augmentation strategies.

 Five, to understand why Uber is so popular to join for driver partners around the world. The theme falls under the heading: What happens when wages head south?

The basic overview above, make it possible to sort the data into finer granularities. Data that could provide helpful to understand particular points of interests were kept as research material, while sources of little value were left out. Afterwards, the relevant data got sorted into its thematic brackets. For instance, data that could help explain how Uber scaled their user base in the first days of the startup's existence went into one bracket, while data shedding light on obstacles to the augmentation of the user base went into another.

Doing online research comes with particular concerns. It is simply too easy to denote everything that pops up in http://www.techmeme.com as an expression of “knowledge blogging” (Constantiou et al., 2016). In fact, most of the generated material is not blogs at all, but simple news stories. The writers are paid by regular news media and have in most instances a journalistic take. However, there are exceptions, the owners of TechCrunch, are investors in Uber (Appendix A). TechCrunch follows Uber especially close, and it is very hard to find any critical reports emanating from this source. In many cases, TechCrunch serves as an outright spokesman for Uber. Thus, being aware of this fact is vital. Researchers should search for objective facts and try to avoid entrapment in echo chambers that advance industrial interest.

In general, researchers must treat online sources carefully. The validity of a source strengthens when other sources corroborate it, and likewise, weakens if it cannot be backed up by other information (Kjeldstadli, 1999). Good source critic entails that sources have to be put up against each other to get the full picture. That means for example that Uber’s take on things must be triangulated with media punditry and academic research whenever this presents itself as an opportunity.

4 Findings

4.1 How Uber conquered San-Francisco?

The app, which later should take the world by storm, had a modest beginning. The founders never thought of the concept as more than a niche-product, and were, therefore, not interested in putting much personal effort into getting it up and running. So, the founders hold an arm-length oversight on the project while friends and hired help did the actual work.

4.1.1 The creation of an app Leviathan

UberCab, the predecessor to what later became known as Uber, was founded in March 2009 by Garrett Camp and Travis Kalanick. Camp and Kalanick were two grizzled entrepreneurs from the startup scene in Silicon Valley. Both had started successful ventures before, which they had sold on for millions of dollars.

Travis Kalanick Garret Camp

Myths and realities

Uber’s founding myth is that Camp and Kalanick met each other at Le Web Conference in Paris 2008, and talked about how frustrating it was not to get adequate taxi service. It was there and then that they got the idea about changing the taxi industry forever (Kalanick, T., 2010). This romanticized account is not entirely true. Kalanick has later credited Camp for conceptualizing what was going to evolve into Uber (Shontell, A., 2014a; Swisher, 2014). It is true that the two of them discussed

42 several startup ideas in Paris, but it is not true that changing taxi was on top of their minds. Camp’s idea was to create an on-demand limo-service so that he could ride around in style (Kalanick, T., 2011c).

Kalanick’s deep involvement with Uber as a company was something that evolved in stages, and if it had not been for Camp’s constant harping on the idea, he would not have gotten involved at all (Shontell, A., 2014a; Swisher, 2014; Chafkin, M., 2015). So, why was Camp so interested in getting Kalanick involved? - The sources are very silent on this. It might be that conceptualizing a business idea is a social process. Everyone needs a sounding board and a reality checker. Kalanick seems to have been willing to lend his ear to Camp. He might also have appeared enthusiastic about the idea. According to one media report, Camp was impressed by Kalanick’s guts and energy (Swisher, 2014). For whatever reason, the two of them ended up as sparring partners.

At the end of the day, this is what matter, the two of them kept in touch, and with the emergence of the GPS-enabled smartphone and cloud computing tools; they decided to realize Camp’s limousine idea. In August 2009, Kalanick had warmed up so much to the idea that he agreed to seed UberCab with $200K alongside Camp (Appendix A).

Camp makes an app

In the first half of 2009 Camp started to work on an iPhone app that would capture his vision about an on demand limo-service. Above all, he wanted the app to signal exclusivity and elegance (Kalanick, T., 2011c). Camp’s design philosophy was quick product recognition and easy user interaction. The user should immediately see what the product was all about in a tap or in a simple keystroke (Camp, 2016). This philosophy is easy to recognize in how the Uber app looks today, but realizing it, was not done in a jiffy. There were a lot of issues Camp needed to get a handle on in the prototype.

Looked upon from a technical perspective, Camp had to create a triangle of two way real-time online communications between drivers, riders, and the dispatcher unit. He needed visible mapping localization in the user interface. The drivers had to see

43 passenger’s localization and the other way around. Additionally, the app needed user friendly functions both on the rider and the driver side. On the passenger side, it was imperative that potential users could request trips, set pick-up location in the map and pay for the service easily with credit cards. Furthermore, the user should see the car coming towards the pick-up location in the map. On the driver side, the driver should be able to handle requested trips and spot the pick-up location. Behind the scenes, the system had to know which drivers were available at any point in time and a manner of dispatching new orders to the right drivers. On top of all this, the drivers and passengers should be able to rate each other. Deeper down in the technology stack, Camp had to work out several features which were central to the business, such as, how to handle payments; how to manage and store orders; how to price fares and so forth, see figure 4.

Figure 4 The Uber app, simple technical oversight.

The UberCab prototype was a side-project for Camp since he was involved in other projects. A fun fact, which tells how much of an ad-hoc project it was; is that things like method declarations, class names and documentation in the code, was written in Spanish. So, Camp did not code much at all. It was his Spanish-speaking friend, Oscar

Salazar, who did most of the programming (Kosoff, M. & McAlone, N., 2015; Lashinsky, 2015).

Nevertheless, the prototype was ready for testing by the end of 2009. In January 2010, Camp brought Kalanick and Salazar to New York City. They had three cars crisscrossing the lower Manhattan and a few people using the app (Kalanick, T., 2010). Little information about the testing exists, but overall; Camp and Kalanick must have been pleased with the results because afterwards, they were ready to take the next step and commercialize the app.

The basic fact was that neither Camp nor Kalanick could see themselves as running a limo service. So, they both reasoned that they should hire a general manager and some programmers to build the rest of the app. It was supposed to be a low-key affair for both of them (Shontell, A., 2014a; Swisher, 2014).

Key Hiring’s

In February 2010, Ryan Graves was hired as general manager for the company, while Conrad Wheelan was hired two months later as a programmer. Graves reacted to a tweet from Kalanick and got employed after they had e-mailed back and forth a few times (Kalanick, T., 2010). Wheelan was an old friend of Camp. Zalasar was also sticking around, helping out Camp (Shontell, A., 2014a; Lashinsky, 2015).

Graves had just recently given up a well-paying job at General Electric, not finding any meaning being a cog in the corporate machine. He wanted to create something and decided to give starting up a shot. According to Graves, the initial prototype was full of bugs and did not work very well. The first thing he did was, therefore, to outsource the building of the app to an outside company. Simultaneously, he started designing the website for the service and studying the nature of the taxi business, ensuring that the app could handle operational requirements.

The prototype lacked vital functions. There was no user sign-up enrollment, no pricing mechanisms, and no payment system. Following up on these issues became Wheelan’s job. His biggest contribution was to make a pricing algorithm based on passengers' mileage and time spent inside the vehicle (Lashinsky, 2015).

Recruiting drivers

Kalanick and Graves cooperated to recruit drivers. They went only after licensed limousine drivers and tried to make outreaches to limo companies, enticing them to put the whole fleet and their drivers available for the app. It did not succeed (Graves, R., 2010a). Single drivers were their next option. Their main selling points were instant ride requests, never to worry about billing, direct client communication and making a supplement to existing income. Despite good arguments, and the possibility for potential drivers to make additional revenue in a tough business environment, it must have been a hard sell. Nevertheless; one by one, the two of them managed to boot UberCab with ten drivers. However, the drivers were only available on weekends and only when they were free from attending to regular clients (Stone, B., 2017). It was slim pickings, but the user side of the platform was not very sizable either; consisting of around hundred techie friends from Silicon Valley (Shontell, A., 2014a).

Launch

In May 2010, the UberCab app was approved by the Apple app store. The actual launch of the service happened on the 1st of June. Then, one year and two months had passed since UberCab was founded and over two years had gone by since Camp and Kalanick talked about apps, luxury, and limousines in Paris.

Little did the involved parties know about what they had put into motion. Camp had at least reached his goal about riding around in style but didn’t think it would go much further than that (Camp, 2016). Shortly after the launch Graves were named the CEO of the stumbling upstart (Shontell, A., 2014a).

4.1.2 The product as an asset

The UberCab app was about getting instant luxurious rides for a relatively low price. The service main selling points were trust, quality, and ease of use.

The UberCab app

In the beginning, the UberCab app was two apps, one for passengers and one for drivers. According to the UberCab app demo, most of the user functionalities worked

46 like it does today, except for the feature which makes it possible for users to calculate the fares beforehand (UberCab, 2010). There was also less choice available. Cars and prices were standardized at a relatively high level. Still, the idea of ordering a car with a tap on an iPhone application was a novelty that never had been a possibility before. The idea captured the zeitgeist of the time. To order rides on a whim by a handheld device they carried around at all times, was an opportunity many consumers appreciated.

Trust

Sundararajan (2016) calls trust the most valuable capital in the sharing economy and speak of sharing platforms as self-regulatory trust infrastructures. UberCab’s primary goal was to get new users interested in joining through referrals. So, at a very early stage, it became evident for Graves, Kalanick, and Camp that it was imperative to establish trust in the service. The solution became the star-system where users rated each other on a five point scale and were able to give feedback about each other.

Short term, the reasoning behind the star-system was to prod the drivers to do a good job, while the feedback system was supposed to be a way to get into the mindset of the users. Together, these two user exchanges made it theoretically possible to correct driver behavior and take initiatives on user complaints about missing features in the current service (Geidt, 2011).

Longer term, it became an even more important part of ensuring quality, especially; when non-professional drivers joined the platform. Low-rated drivers could be warned to improve their performance and if necessary; be kicked out if the performance did not get any better. The system was meant to work both ways to keep the drivers safe from unstable passengers.

The product

UberCab priced their service 1.7 times higher than regular taxis, but it was 90 % cheaper than what ordinary limo-services could offer (Kalanick, T., 2011c). UberCab sported mostly drivers who were in possession of either Lincoln Town Cars or Cadillac Escalades (Arrington, M., 2010a).

Lincoln Town Car Cadillac Escalade

4.1.3 The Users as an asset

Kalanick and Camp were two well-known characters in Silicon Valley. Above all, they had many friends who were involved both in startup companies and angel investing. This informal network of friends could be called upon to get the word out about UberCab. They should also prove to be an invaluable asset when the fledgling startup needed to raise funds.

Marquee users

Friends of the founders were extremely enthusiastic about the service and gave it rave reviews. Chris Sacca, a well-known venture capitalist, tweeted out: “Rolling in an @ubercab. Eat your heart out Robin Leach” (Swisher, 2014). The founder of TechCrunch, Michael Arrington, encouraged his readers to become UberCab users. His call went as follow: “enjoy superior service and help to break the back of the taxi medallion evil empire” (Arrington, M., 2010a).

Word spreads fast in Silicon Valley. ‘Techies’ wanted to try out the app which renowned users recommended. Apparently, it was not all about helping friends with getting the startup of to a good start. Most of the reviews were heartfelt; many thought it was superb ‘riding in style’ to weekend events by a tap on the cell phone. So, UberCab’s ten drivers got busy very soon. Two weeks after the launch, every driver where making ten trips each day on weekends (Shontell, A., 2014a). Kalanick referred

48 to it as the ‘revenge of the nerds.' Every nerd could now act like sports stars - ‘living it large’ (Kalanick, T., 2011c; Holmes, 2012).

Angel investors

UberCab experienced growing-pains fairly fast. In early autumn 2010, the app started to provide daily service and had trouble recruiting enough drivers to keep up with demand. Graves pleaded with their users to stay patient on the UberCab blog (Graves, R., 2010b). Thus, Kalanick and Camp needed funds to finance more employees and recruit drivers faster. The opportunity came at an open investor forum hosted by their friend Jason Calacanis in October 2010 (Shontell, A., 2014a).

The founders hoped to get $1M out of investors at the forum. In total, they managed to rise $1.25M with First Round Capital as the leading investor (Arrington, M., 2010b). Camp had a friend working in First Round Capital and had actively courted the venture capital firm since the summer (Shontell, A., 2014b). Sacca and Calacanis also invested in the startup. In total, 29 angel investors participated in the funding (Appendix A). They were true fans of the service now. First Round Capital got a member on the board as thanks for their vote of confidence.

4.1.4 Flexibility and aggressiveness

By January 2011, the platform had around hundred drivers available and seven employees (Kalanick, T., 2011a; Lashinsky, 2015). To work on a startup is a chaotic experience. Problems need to be solved as they arise. A small number of people must, therefore, have a flexible approach and work on the issues which seem most pressing. During the autumn of 2010, Kalanick was starting to take an increasingly bigger interest in the venture. Gradually he was going to take the helm of the company. Uber’s aggressive posture is largely a product of his leadership personality.

Topsy turvy in a 10-by-10-foot cubicle

Kalanick and Camp rented a small office space within the premises of another company. The office was about to become cramped when new employees joined. Right after the launch, Graves hired Ryan McKillen as a second engineer. In January 2011, the startup had five software engineers in total, not counting Oscar Salazar who

49 quietly disappeared from the scene. The small team of software engineers had their hands full. The app was full of bugs and did not work properly; sending all limos to one location and incidents of this nature (Kosoff, M. & McAlone, N., 2015; Chafkin, 2015).

The Uber crew, early 2011.

Source: Lashinsky (2015)

In August, Graves brought Austin Geidt on board (Lashinsky, 2015). Geidt had seen a positive tweet about the company from venture capitalist Jason Calacanis and decided to prod Graves for an internship. She was fresh out of college at the time, a former drug addict and didn’t have much going for her in a tough economy. Graves decided to give her a chance because he liked her humor and intensity. Geidt was supposed to be the marketing support for the company and cold call drivers to get them to sign up. However, she soon learned that in the startup world much of the day consisted of handling emergencies. It became, for instance, her job to deal with customer support calls. The calls came mainly in at night because they essentially were running a limo

50 service for party goers. Despite the stress of the job, Geidt was a person who thrived in the chaos of a 10-by-10-foot cubicle and gradually became a valued employee that was involved in every side of the business (Bort, J., 2015; Lashinsky, 2015).

According to Geidt, it was hard to sign-up drivers immediately after she started in the company, but after a few months, she recognized that new recruits turned up by themselves through referrals from existing drivers. She recalls that one of the new drivers had a pink Camaro and that they also had to contemplate saying no to recruits (Chafkin, M., 2015). Apparently, the drivers made respectable earnings in the early days of the startup's existence, and this enticed a steady stream of prospective recruits.

The Kalanick method

Travis Kalanick started out as one of these college dropouts turned entrepreneurs. He started a few software companies in the field of peer-to-peer file sharing in the late 90s and early noughties. The various companies had a precarious existence and when turning 30; he lived at home with his mother (Lagorio-Chafkin, 2013). Kalanick describes himself as a failure looking back on his early career (Kalanick, T., 2011e). However, despite the travails of his earliest business ventures, he was eventually able to sell of a semi-successful startup and become a multi-millionaire, which got him started as an angel investor. Not everyone will call this a failure. In any case, he was well into his thirties by then, and as he self describes it; exhausted by entrepreneurship (Laizos, 2010).

Kalanick’s formative years are a grey matter. We know little about what it is that drives and inspires him. He took an early interest in programming and studied computer science at UCLA (Lagorio-Chafkin, 2013). Beyond facts like this, things get fuzzy. According to some, he shares the philosophy of Ayn Rand and his favorite book is supposedly The Fountainhead (Kosoff, M., 2015). There are similarities between the hero of the book and Kalanick, but it is hard to know how much one should emphasize this. One of his friends from within his circle of Californian tech entrepreneurs characterizes him as generous, extremely ambitious and “smart as hell” (Chafkin, M., 2015).

The media have portrayed Kalanick with different brush strokes. His fiercest critics

51 call him everything from insensitive and arrogant to misogynistic and sexist (Bradley- Carr, P.. 2014; Lacy, S., 2014; Tiku, N., 2014; Lu, A., 2015). More balanced media portraits emphasize that he come across as very salesman like (Anthony, N., 2014; Shontell, A., 2014a), while in other personal interviews he is also depicted as someone who has a knack for thoughtful analysis (Carson, B., 2016d; Bradshaw, 2014). However, the trait that sticks out in the majority of personal interviews is his aggressiveness when it comes to business. He wants to crush his adversaries, and he is pragmatic about how he accomplishes his goals (Lagorio-Chafkin, 2013; Swisher, 2014; Bradshaw, 2014; Chafkin, M., 2015).

There are obviously many layers to Kalanick’s personality. Apparently, his years as tech entrepreneur have colored his attitudes and beliefs. In interviews, he describes entrepreneurship as a very volatile existence, where it is necessary to think fast, stay nimble and solve problems as they arise. Lack of money seems to have been a recurrent problem for him in his first business ventures, which made it necessary to hustle and attract investors to stay afloat (Kalanick, T., 2011e). In this volatile environment, it is easy to see that opportunism, pragmatism, aggressiveness and salesman like abilities must have come in handy. Kalanick seems to possess these personality traits.

UberCab becomes Uber

UberCab’s operations got attention from taxi-operators and local authorities at approximately the same time as it got angel funding. In late October, local authorities decided to act. UberCab got a notice that they immediately had to stop their service until its regulatory status was clarified. The order was severe enough with strong wording about both fines and prosecution (Kolodny, 2010).

The founders of UberCab did not blink. Both Kalanick and Camp had experienced tight situations as entrepreneurs before and were not going to give in. Their supporters, drivers, and investors were also behind the company all the way (Siegler, 2010). Kalanick and Camp’s line of defense was that they were running a matchmaking service and not a cab service per se. UberCab was only a mediator between limo drivers and their potential customers, and everything was fine on the driver side because the chauffeurs were already licensed. To remove any confusion on

52 the part of regulators, they decided to remove ‘Cab’ from the company name (Ha, 2010).

4.1.5 The launch pad

There are three causes behind how the Uber founders managed to grow the Uber platform as a two-sided market in San-Francisco. First of all, the app satisfied consumer taste by combining three ingredients: consumer technology, luxury, and on- demand services. Secondly, the app got great reviews from early adopters. Finally, the Uber startup crew had ‘the right stuff’: they worked hard, were able to recruit drivers, and were good at trouble-shooting and multi-tasking.

We have also uncovered that there were special circumstances at play. Uber got a boost from influential friends that also were in the business of starting up ventures. In total, Uber’s successful launch was the result of a collaborative network between the startup crew, their enthusiastic user-friends, their hard-working drivers and their angel-investor friends. The users were not only helpful getting the word out. Evidently, they were heavy users as well. The high turnover in the number of trips meant that the drivers made sizable earnings and this enabled the recruitment of more service providers through word-of-mouth amongst the drivers. That some users were willing to step in with funding made Uber able to hire engineers that could improve the technology and rid the code of its worst faults.

Despite this fact, there is no denial that the founders were on to something unique. The app had the features which were able to make it a consumer hit. It was not unthinkable that the establishment of the app as a successful venture in San-Francisco could be used as a launch pad to expand the service into other cities.

The problem with regulatory authorities in San-Francisco made Kalanick take a more hands-on approach in dealing with the startup. He was now ready to take the helm of the company, and formally took over the position as CEO at the end of 2010. Graves agreed to step down into the position as manager of operations. It all happened rather amicably according to Graves and Kalanick (Graves, R., 2010c). During the winter Kalanick was on the prowl to get more funding. He now told investors that he had set his sights on New York City (Schonfeld, 2011).

4.2 How Uber scaled exponentially?

Even though Uber is not very forthcoming when it comes to releasing company statistics, we know that the growth in its user base has been formidable. In the summer of 2016, Uber announced that it on an overall basis had “completed more than 2 billion trips on its app” after hitting its first billion 6 months earlier (Tepper, 2016). These numbers must mean that Uber currently is processing over 5 million rides each day. It also seems likely that over half of these rides take place outside of North-America, making markets overseas crucial for Uber’s revenue. The numbers get very impressive when we consider the fact that the company processed around 1 million rides each day at the end of 2014, up from around 120 000 at the end of 2013 (Huet, E., 2014). Taken into account that Uber had completed a little over 10 000 trips in total by February 2011, we do get a grasp of how exponential the growth have been (Arrington, M., 2011).

This growth could not have been possible without drivers. In October 2016, Kalanick revealed that Uber have 40 million drivers who are making at least four trips every month (Lynley, 2016). The number of US drivers has approximately doubled every six months from mid-2012 to the end of 2015. In total, Uber had around 460 000 active drivers in America at the end of 2015, which was up from 160 000 the year before (Hall & Krueger, A. B., 2016). Certainly, Uber had a considerable amount of drivers working globally at the end of 2015 as well, although it is guesswork to say exactly how many. Comments from Kalanick explain how exponential the growth has been. In September 2014, Kalanick boasted that Uber was adding 50 000 drivers a month, up from 20 000 in May the same year (Wagner, K., 2014).

Uber’s user base expansion tracks its larger footprint across the globe. In May 2011 Uber launched in New York City. In the later parts of 2011, Uber expanded to Seattle, Chicago, Boston, District of Columbia and Paris in Europe. In 2012, Uber added eight new cities in the US and London in Britain. From 2013, it was like a dam bursting, by November 2014 Uber opened in a new city every day (Huet, E., 2014), reaching a total of 576 cities as of April 2017.

We are now going to look into how Uber scaled the user base exponentially. There are multipronged causes behind the tremendous growth. The following causes are the most significant ones:

• Venture backed financing.

• Establishing a launch playbook that worked across markets.

• Segmenting the product with cheaper services.

• Subsidizing.

• Design improvements.

• Technological improvements.

4.2.1 Revving up the motor with venture capital

By early 2011, investors could see a venture with enormous potential. Uber was an app that had penetrated and captivated San-Francisco. It was in full swing driving passengers around with a growing fleet. Through their contacts, Kalanick and Camp started the book building process of Uber’s first venture funding round. They got Benchmark Capital, a renowned venture capital firm, to underwrite the book building process. Mid-February investors had committed $11M in total, and Uber closed its first venture round. Benchmark Capital got one member on the Uber board (Arrington, M., 2011).

This way of doing it; first to get funding and then to grow, should become a fixture in the Uber repertoire over the coming years. Eventually, it should become a self- reinforcing process. So, after every successful expansion, Uber got rewarded by more funding, and as the expansion gathered pace, more investors with deeper pockets climbed on board, cheering the company on with more capital injections. To date, Uber has in total raised $11.56B in funding from various investors (Appendix A).

Its two interesting aspects to Uber as a venture based startup. First off, it is worth noting that Uber cultivated its relationship with investors by giving them board memberships. Uber benefited from this in two ways. In addition to contributing with money; venture investors could also give valuable advice on everything from strategic 55 to operational and financial matters. Moreover, by allying itself with deep pocketed investors; Uber created confidence about the soundness of its rapid growth. Involved investors have given the impression that there has never been a question about whether Uber will run out of money.

Source: Bloomberg

Figure 5 Ride sharers funding in 2016

Second off, venture backed financing made it possible to scale fast over vast geographical distances and do specific actions which recruited passengers and drivers in every new market. Compared to other ridesharing ventures, no one has raised the amount of money as Uber. Didi is the only one who even comes close, raising more money than Uber in 2016. That Uber went up against a well-financed adversary in China was one of the explanations for why Uber was not able to dominate there, see figure 5.

4.2.2 Establishing the launch playbook

The funding gave Uber sufficient financial strength to make good on Kalanick’s promises to take on New York. However, the operational roll-out needed organizing and planning to succeed.

Establishing a local organization

Kalanick, Graves, and Geidt spearheaded the New York expansion. The three of them figured out that they should keep it organizational nimble, but they also realized that they could not do it all alone from the west coast. They needed local staff on the ground at all times. Finally, they decided to staff up a New York office with three people. These three individuals were made responsible for recruiting drivers, market the service, and deal with reactions from regulatory authorities and taxi operators. The job for the company headquarters was then to assist in the startup process. This organizational flexibility has become part of Uber’s “launch playbook.” It always starts with three local representatives and grows from there (Milian, M., 2014b).

Market analysis

New York City was not the easiest target Uber could find as their first city to take on outside its startup market. In contrast to San-Francisco, where taxi operators had a reputation for providing bad service, New York was a market that worked fairly well (Schonfeld, 2011). “In Manhattan, a yellow cab is typically available via a street hail within a few minutes. In the outer boroughs, locals have grown accustomed to car services that have taken root in particular neighborhoods and can offer reliable transport to other points within the city” (McCarthy, 2011). Further, the New York market was heavily regulated. Only registered yellow medallion cabs were allowed to pick up street-hailers, while other for hire car services were only allowed to pick-up customers through dispatchers (Rubinstein, 2012). To get hailers was, therefore, out of the question for Uber, but as Kalanick noted; the number of medallions had been stagnant in the city for the last three decades (Kalanick, 2011c).

Despite tight regulation, New York City was a huge market. Before Uber’s move into the city, taxis and other for hire vehicles were making about 1.5 million rides each year. Regulatory authorities had issued about 160 000 taxi-licenses. 60 000 of these were medallions (NYC T&LC). If Uber could only get a piece of this market, it would be off to a good start in the city. Moreover, Kalanick sensed that there was demand since about 1 000 users already had signed up to the service based on rumors that the app was coming to the city (Wortham, 2011).

Kalanick reasoned that Uber could not compete with yellow cabs on Manhattan during business hours, but reckoned that the app could horn in on telephone dispatch taxis in the outer boroughs (McCarthy, 2011). He was also counting on doing business in Manhattan when things went off kilter. It was about being opportunistic, taking customers on events, servicing passengers during taxi shift change, when it was raining and when no other cabbies were near (Kalanick, T., 2011b).

However, Kalanick could not offer his users lower prices. On the contrary, prices would be 1.75 higher than regular taxis, but Kalanick was betting that customers would pay up to get more convenient services (Wortham, 2011). For that to fly, services had to be great from day one.

Operational roll-out

The most important part of rolling out the service was recruiting enough drivers. Just as in San-Francisco, the recruiting effort consisted of many cold-calls, reaching out to professional drivers and trying to get them in for an information meeting or making them sign up directly (Lagorio-Chafkin, 2013). At least; by targeting licensed drivers they did not have to do much explaining of what the job entailed, and they ensured quality in the service from day one. Moreover, on the surface of it, licensed drivers meant that Uber did not have to cross too many regulatory hurdles, insuring a sort of legal compliance from day one.

As in San-Francisco, the main pitch was that drivers were able to generate business in the downtime. They could also use their pre-registered New York users as an additional selling point. Furthermore, Uber would take care of setting up the drivers with iPhones for free, so that they immediately were ready to roll (Kalanick, T., 2011c). Austin Geidt did a tremendous job in this respect, having been through the process once before in San-Francisco (Milian, M., 2014b). In just three months, 100 drivers were recruited (Kalanick, T., 2011b).

After the launch, it was all about creating publicity and getting press coverage. The point was to get referrals going. Moreover, to get referrals going as fast as possible, nothing was better than inviting reporters, celebrities, and investors to a VIP launch party (Shontell, A., 2011). Longer term, company blogs, and social media became

58 instruments to create hype about the service. Just as in San-Francisco, Uber banked on getting rave reviews from early adopters in social media.

Uber’s launch in New York went into the company’s launch playbook which should be refined, improved and repeated over and over again: The main stem of the playbook still consists of the following points:

1. Be adequately funded.

2. Study the new market.

3. Set up an agile local team with entrepreneurial experience.

4. Incite expectations about a city opening through publicity initiatives.

5. Pre-register users.

6. Recruit drivers.

7. Open with fanfare: Have a VIP launch party and get marquee users involved.

8. Do subsequent publicity initiatives to raise awareness and get more users to sign up.

9. Enlist more drivers.

The international launch playbook

Today, Uber have an international launch playbook, compiled by an internal team of around 40 employees (Milian, M., 2014b). As previously mentioned, the main ingredients from the New York move have not changed much. The only difference at the start of the international expansion was that each market had to be studied more intense up front to learn about cultural differences and unfamiliar regulation. Furthermore, the technology had to be tailored to work in different countries. The internal launch team has an updated list over where to expand next based on competition and demand metrics compiled of the various markets (Lu, K., 2016). In theory, Uber is now a global company which is capable of taking on every market.

4.2.3 Market segmentation and cheaper prices

Until the summer of 2013, Uber was all about providing comfortable rides with professional drivers. There is maybe a general misconception that Uber was the first

59 company to do ridesharing with non-professional drivers. In 2012, Lyft and Sidecar were testing the concept in San-Francisco, which was a year before Uber decided to do the same. However, when Uber first made its move into the segment, the growth started to take off, see figure 6. In total, this action alone is the main cause which explains Uber’s extreme growth over the next years. However, the action to go fully for the concept we today associate with ridesharing was something that slowly evolved.

UberTAXI

The Uber leadership spent an extended period mulling over how they should go about widen their product and move down market. There were risks entailed with potentially destroying the luxurious reputation Uber had achieved. They were also apprehensive about what kind of reception such a move would get among regulatory bodies that already were on the warpath. The Uber management was neither sure of the market effects nor what kind of operational model they should employ.

Source: Hall and Krueger (2016)

Figure 6 Growth takes off with non-professional drivers on UberX.

Uber’s reluctance to compromise on quality made it willing to contemplate meshing the Uber app with the traditional way of operating taxis. In April 2012, Uber launched UberTAXI as an experiment in Chicago. Cab drivers were recruited to carry both Uber and regular passengers. The initiative got a positive response from both taxi drivers and riders (Rao, L., 2012a). In September the same year, Uber tried to launch a similar service in New York but was stopped instantly by city officials. In the fall of 2012, Uber met regulatory resistance in Chicago too.

The regulatory limbo concerning the roll-out of UberTAXI is part of the reason why only a few cities have this alternative as a product today. The main reason was that Uber opted for another model.

UberBLACK, UberSUV, and UberX

It was UberX that eventually would transform into a product that everyone could operate, licensed or not. However, Kalanick and the rest of the Uber management team continued to play it slow. Kalanick argued that Uber could keep users who preferred luxury rides while simultaneously capturing lower segments of the market. Upon launch of the extended product, Uber had concluded that it could get taxi drivers to provide rides with hybrids or SUVs. Hybrids could take care of individual users while SUVs could take care of groups.

Both hybrids and SUVs were popular models among taxi drivers, so the supply should theoretically be plentiful. In July 2012, two years after the first Uber car rolled out on the streets of San-Francisco, Uber launched SUVs in Seattle, Boston, Toronto and Chicago, and both options in San-Francisco and New York (Sawers, 2012).

Uber marketed the different services as three choices in the front of the interface. UberX was the cheapest alternative consisting of hybrids, but still, it was around 1.3 costlier than regular taxis. UberBLACK was the name of the old limousine service, mostly comprised of luxury vehicles, costing 35 % more than UberX (Tsotsis, A., 2012a). UberSUV was the costliest of the three options but designed for taking groups up to 6 passengers.

In the beginning, Uber did not have enough drivers to service all their users, so they had to restrict the use of UberX to a smaller group of loyal users. However, Uber had

61 immediate plans to ramp up the service by recruiting more taxi drivers to ride for the app in their downtime (Tsotsis, A., 2012b).

The transformation of UberX into an all-purpose ridesharing product

In August 2012, Lyft launched ride sharing without professional drivers in San Francisco. The service was considerably cheaper than UberX. In fact, users did not pay for the service at all; payments were donations which the users elected to endow at the end of each ride. It was a payment scheme which was supposed to circumvent regulations. The service grew fast, and in October the service had 200 drivers available on the platform (Olanoff, 2012).

Uber was watching the competition closely. In an interview Kalanick commented on the development: “If somebody’s out there and has a competitive advantage in getting supply, that’s a problem. I’m not going to just let that happen without doing something about it” (Lawler, R., 2012b). However, the Uber management had to consider that UberX, for the time being, was making more trips than Lyft in San- Francisco (Gallagher, 2012). Uber was not willing to plunge head first into the low- cost space without thinking about the consequences. Regulators were still breathing down Uber’s neck, and it needed to consider how their driver partners might react to an opening up of the platform. If an opening up of the platform would result in a revolt from their enlisted drivers, the gains stood in danger of being outweighed by losses (Lawler, R., 2012c; Lawler, R., 2013b).

At the beginning of 2013 Uber got some respite from Californian regulators. Kalanick used the opportunity to announce that Uber soon would roll out low-cost ride-sharing in the state (Lawler, R., 2013a). When Lyft started to offer its services in Seattle and Los Angeles, Kalanick felt that Uber could not wait any longer. In April, Uber released a company white paper where it stated that regulation was no longer considered a reason not to move. Uber would start offering ride sharing in states which had given “tacit approval” and would wait for a maximum of 30 days if a competitor already had made a move in a specific market. It was a policy manifesto where Uber signaled that it was going to use its superior position to seriously augment its driver network and fight any digital apps that would challenge its position. It was also a signal that Uber would take advantage of being the first mover in most markets

62 and that it would seek to dominate any market before competitors started to aggregate users (Bhuiyan, J., 2016a).

The point of market segmentation

Inside the digital realm, a one-size-fits-all product can very well work, like for instance Google search. Outside the digital realm, a one-size-fits-all product almost never works. Obviously, Uber’s ability to tailor its product to fit the needs of customers in different segments is one of the causes of its success. In contrast to many of its competitors which overwhelmingly have gone after price sensitive parts of the market; Uber can play in all niches of the market. Like it or not, customers are different.

4.2.4 Innovative recruitment and pricing policies

When Uber transformed UberX, it had an advantage over especially its American competitors. Uber operated in 30 cities against Lyft’s three (Panzarino, 2013). Ergo; the company had already built up expertise in how to recruit driver partners. Flush financing was also there to support the scaling. In total, the company had raised nearly $50M in venture funding before the transformation took place. Later in 2013, it received a whopping $363M in new VC-funding. It was substantially more than what Lyft had raised, and this financial preponderance would more than even out Lyft’s head start. How Uber aggressively put its money to use should prove vital to ensure scalability of the business.

Recruitment tools

In the early days, Uber did its recruiting by reaching out to taxi and limousine drivers. The main selling points were that professional drivers could enhance their income in the downtime by working for a company that emphasized luxury and elegance. Since everyone with a decent car, in theory, now could become a driver on the platform, it could no longer rely on propelling the recruitment by targeting a small audience. The majority of ordinary car owners were also likely to be put off by an emphasis on elegance. Instead, Uber had to construe a message directed to a mass market and entice drivers through more public channels. The company found Facebook ads to be

63 the most effective recruitment tool in concert with referrals and solicitation by existing drivers. Uber toned down its language about quality. The new selling points were freedom and flexibility. It was no pressure; one could drive as much as one liked (Dupre, 2016).

Content drivers were the driving force behind Uber’s referral propelled recruitment. Enthusiastic drivers were not only instrumental in soliciting new recruits but were also ambassadors for the Uber brand. Supposedly, the creation of this supportive user base was for the most part achieved by giving drivers good starting salaries.

Subsidizing driver compensation

Uber put its financial war chest to immediate use. The goal was to spread geographically and scale to be the dominant player in every market. To recruit drivers became now much easier than it was in the company’s more cash-strapped beginning. It was not abnormal that the first drivers in new markets got $25 in hourly guaranteed earnings.

For instance, London’s first UberBLACK driver could not believe what was happening when he made £2 500 every week (Knight, 2016). In San-Francisco, where Lyft and Sidecar were active, Uber pulled out the big guns and offered new UberX drivers $5 000 in guaranteed pay for the first month (Swan, 2014). In Los Angeles, another city where competition against Lyft was heating up, drivers got a $40 hourly guarantee driving Friday and Saturday nights, while additionally getting a $500 sign- up bonus (LA Weekly, 2014). In Europe too, Uber was aggressive, offering their drivers huge bonuses (Dillet, R., 2014). Of course, signing up drivers gets relatively easy when the incentives are this good.

Subsidizing passenger fares

In sum, Uber was subsidizing its fares massively during 2014 and 2015. Some argue that passengers were paying only 40 % of the actual cost of the service (Kaminska, 2016). The recruitment of users happened in much of the same manner as Uber recruited drivers. One tactic was to give away rides to new passengers whenever they launched in new cities. Another was to offer rebates for new travelers in the towns where Uber already was operating.

Free rides and sign-up deals are all well and good, but nothing beats lower prices across the board. Despite starting of as a pricy limo service, Uber understood early on that new users could be enrolled by slashing the cost of fares. In fact, Uber began to test out reduced prices already in 2011 (Rao, L., 2011). The following years Uber was able to cut prices for UberBLACK on a permanent basis in three cities: San Diego, San Francisco and Seattle (Wilhelm & Lawler, R., 2013). After the launch of UberX, the company went all out in January 2014, slashing prices with as much 20 % in 16 out of 24 US cities (Kalanick, T., 2014). In the summer of 2014, Uber became the cheapest alternative in New York City (Wohlsen, 2014). So, the story went on from city to city; prices were subsidized and pared down until Uber was the cheapest alternative (Silverstein, 2014).

Network effects

So, how could Uber push down prices by such a magnitude? Kalanick gives the answer himself. “The answer is in the network effects of our business. More cars and drivers mean better coverage and lower pickup times. Lower pickup times mean better economics for drivers, and thus more drivers and cars. And we want to deliver that value to our riders” (Kalanick, T., 2014). Thus, higher capacity utilization financed lower prices.

Source: Gurley, B. (2014b).

Figure 7 Uber's virtuous cycle

In city after city, Uber implemented its competitive pricing model: subsidizing drivers and passengers. Higher user density provided the network effects. When Uber used its financial muscles to buy more drivers and riders, the chauffeurs enjoyed less downtime, and the consumers benefited from faster pick-up times. Less driver downtime and less passenger wait time created higher utilization of resources and increased transaction based revenues; allowing Uber to reduce prices even further, which, in turn, created more demand and more supply, ultimately leading to an even thicker market (see figure 7).

According to Uber insiders, an important tipping point comes when pick-up times becomes less than 5 minutes. Then, Uber become the consumer’s default choice for any transportation, which increases use cases dramatically (Rothman, S., 2016). Primarily, it gives users the choice to opt out of car ownership and opens up for a scenario where ridesharing explicitly serves as an extension of mass transit systems (King, Peters, & Daus, 2012).

4.2.5 Design improvements

Kalanick has compared Uber to an entity where bits and atoms meet up. Bits are the technology that propels the business forward. Atoms are a representation of the people which make it all happen (Kalanick, T., 2013). People are not only involved as users, but also as employees working on developing efficient and user friendly software. Technological development goes hand in hand with Uber's growth strategies. So, as Uber has grown, engineers within the company has adapted and enhanced the design of the app based on input from users, general learning, and technological advancements. Specific adaptations have been necessary to support international growth.

Design changes

The Uber app has gone through several design phases that reveal some of the attitudes and goals of the company. In the first stages, the company was keen to portray itself as an expression of luxurious convenience, with a Lincoln Town Car and an iPhone featured prominently on the webpage (McAlone, N., 2016). Uber’s emphasis on

66 luxury topped out in 2013 when it worked with the graphical design company Noise 13 to give the app an executive look for business leaders (See figure 8).

Source: Noise 13

Figure 8 Uber - a sleek alternative for executives

After the transformation of UberX into ridesharing for everyone, Uber got increasingly more interested in marketing itself as more folksy. Slowly, the website and app got more colors, and the business executives were replaced by images of regular people (see figure 9).

Source: Jacinthe Busson http://uxtimeline.com/uber.html

Figure 9 UberX is dependable, regular, and economical.

The use of more colors and more friendly images were the starts of a rebranding operation which began in 2015 and lasted for a whole year. An in-house design team did the rebranding, and Kalanick himself was strongly involved. The team settled on branding Uber around five pillars: grounded, populist, inspiring, highly evolved, and elevated. The design team figured that the app and the company website should reflect the fact that Uber now was a global company consisting of a plethora of local markets. Uber was, therefore, given a new logo consisting of a shifting set of color pallets depending on which part of the world the app was accessed. The new logo, which has gotten some critic for being totally gibberish, is supposed to reflect Kalanick’s vision of Uber as bits & atoms, see figure 10 (Kalanick, T., 2016a; Prigg, M. & Liberatore, 2016).

Source: Jacinthe Busson http://uxtimeline.com/uber.html

Figure 10 Logo change

Device compatibility

When Uber first launched, it was available exclusively on the iPhone. The app was released on Android first half a year later (Rusli, 2010). The story repeated itself when Uber released a new version of its app in 2012; it first came on the iPhone, six months later on Android (McMullen, B., 2012). That Uber focused on making its app native on iPhone first has obviously something to do with the fact that Apple at this point had more users, especially in the US. Today, most users are on Android devices, and Uber do not favor one over another any longer when it comes to updates.

In September 2013, Uber also launched on BlackBerry and Windows (Summers, N., 2013). Needless to say; to be compatible across all devices are essential to reach out to a maximum of users. Furthermore, Uber have always allowed users to request trips

68 through SMS messaging. This opportunity has been beneficial for user adaptation in some emerging markets where there is less smartphone usage. However, smartphones are on the rise in less developed nations as well, so Uber have reasoned that it do not need to invest so much in this service (Milian, M., 2014a).

Mobile payments options

Cash free transactions are one of Uber’s appeals. It is convenient for riders and much safer for drivers (Feeney, 2015). When Uber launched, riders were required to setup a credit card on their account. Indeed, this is as hassle-free as it can get, but as mobile payments system is starting to take hold; Uber has been eager to integrate these solutions as well. In May 2013, Uber made it possible to pay with Google Wallet for users on Android (McMullen, B., 2013a). Later the same year, Uber partnered up with PayPal to deliver an additional way to deliver mobile payment options for users. Kalanick reasoned that PayPal could help Uber in gaining user adaptation in countries where the use of credit-cards was scarce (Tam, 2013). Still, different markets have different needs, and in India, neither credit cards nor PayPal would suffice, so in November 2014 Uber teamed up with Paytm to introduce payment through debit card, net banking or credit cards using a Paytm Wallet (Akshay, 2014).

To split the bill is not straightforward on credit card transaction. Mid-summer 2013 Uber did something with this as well, allowing riders to split the credit card bill on multi-passenger fares. This feature helped especially passengers who travelled with the pricier alternatives. An UberSUV split six ways could give an 83 % discount and take the price down to public transport levels (McMullen, B., 2013b).

Today, there is also an option pay with a debit card. Moreover, as Uber expands in emerging markets, the local teams adopt cash payments when it is necessary (Bright, J., 2016a).

Internationalization of code

At the end of 2011, Uber started its global expansion with Paris as its first city. Kalanick wrote in a blogpost that every city was “unique in its transportation pain points, its density, its transportation alternatives, regulation, even its transportation culture” (Kalanick, T., 2011f).

The venture into Paris was done head over heels without the necessary customization of the app: translations were not correct, the mobile payments system was in dollars and distances were done in miles, not kilometers. After the launch, everything in the code needed correcting (Radhakrishnan, 2014). Moreover, it became apparent to Uber’s engineering team that the front end of the system needed a more flexible design. It was too strenuous if Uber engineers in the company headquarter were required to tweak every little bit of code. It was better if the backend system would allow for small changes done by local offices.

The Uber engineering team, therefore, had to create a translation workflow that was consistent with the values of the brand. Next, they had to figure out a way to build tools to communicate system-wide changes (Lu, K., 2016). The solution was to get engineering, product, and design out of the way, and let the system be self-serviceable for every local team, so, they could do the kind of customization they needed locally (Radhakrishnan, 2014). Designing for flexibility allowed local teams to adapt their product and service offerings to meet the needs of each market.

Team of teams

Designing for flexibility allowed Uber to sprawl into a team of teams structure. For Uber headquarters, it is eyes on, and hands of - every team are giving wide ranging freedoms. Team of teams enables every local office to take innovative initiatives. It is practical too, for instance, the concept of black cars has no meaning of luxury in France. In France, UberBLACK is, therefore, called UberBERLINER. In Britain, sport utility vehicles are an unknown entity. UberSUV is, therefore, called UberXL.

4.2.6 Technological improvements

Kalanick likes to put his love of math across as one of his personal traits (Kalanick, T., 2011c). He underscores that Uber is a high-tech company that is involved in solving intractable problems. In 2012, Kalanick staffed up Uber’s engineering team with employees that went beyond traditional software development. Kalanick called them his math team, and their task was twofold: first, to crunch data and then to use the insight from the analysis to improve the algorithms of particular components in the technology stack (Fisher, 2013). The engineers had many algorithms to work on:

70 the dispatch algorithm, the fare estimator, the surge pricing algorithm, the heat map algorithm and the estimated time-to-arrival (ETA) algorithm, just to mention some of the most important ones. Uber organize the implementation of new code through a multiple of micro services, which is an approach that consists of building independently deployable software systems in small modules (Reinhold, 2016).

The ETA algorithm, the driver heat map algorithm, and the surge pricing algorithm have been long standing darlings for Kalanick (Lacy, S., 2011). The reason is evident; short pick-up time is something which drives network effects in hyperlocal markets. Together, all of these three algorithms are supposed to ensure less than five minute pick-up time. It is tough to discern from the outside how this kind of algorithms work because they are proprietary and Uber insiders are obviously obliged to keep their lips sealed. Some information is revealed, though, by the blessing of the Uber engineering team.

The ETA algorithm

In the early days, Uber used Open Source Routing Machine and overlaid it with big data analytics of similar trips stored in databases. So, it was all about picking out the best subset of data organized in tree structures (Elias, 2014). In general, this produced more accurate ETA than using a single routing resource. However, the engineering team had problems with so called “cold starts,” meaning, that there was no statistics in newly opened cities (Nguyen, T., 2015).

In 2014, the Uber engineering team began to build an in-house ETA-algorithm from the ground up. To make accurate predictions about pick-up times is basically about graph algorithms. Generally speaking, most students of computer science know about the shortest path problem. However, Uber’s problem was more complicated, because the engineers needed to update the weighted nodes in the modelled graphs with real- time traffic data. The Uber engineers found out that it took too long and expended too much computing power to update all the edges in the chart with real time data. So, the solution became to break down the graph into small cells. By dividing up the chart, the updates could happen in a parallel cell by cell independent of each other. Exactly, what kind of algorithms Uber uses to crunch the shortest path and time-estimation is hard to know. Apparently, this is something which they do not want to reveal, but it is

71 likely to consist of some combination of big data analytics of empirical data and some shortest path logic to the nearest available vehicle (Carson, B., 2016a; Nguyen, T., 2015). In any case, it is hard to make the ETA’s completely accurate in all situations. Uber still seem to struggle with getting accurate estimates at the edges of the graph, i.e. at the outskirts of every local market to put it in layman terms (Carson, B., 2016a).

The heat map algorithm

The driver heat map algorithm is an algorithm which shows where there are many customers and few drivers in specific cells in a graph. Kalanick pondered early on whether he should release it to its drivers. The problem was that if drivers knew where surges in demand were happening, all drivers would rush to these grids, leaving other parts of the city underserved (Kalanick, T., 2011c). In the autumn of 2014, Kalanick threw his caution to the wind and released the feature. In several markets, the supply was now so high that heated areas fairly fast would become saturated, leading to limited surges.

Today, many US drivers use the SherpaShare app that works as an adjacent and fully functional service alongside the Uber app. SherpaShare is allowed to access the Uber driver API (Singer & Isaac, M., 2015).

The surge pricing algorithm

Kalanick boasts that Uber is taking on the travelling salesman problem and that his staff gets “to ‘NP Complete’ before you get your cup of Joe in the morning” (Graves, R., 2010c). A problem of NP completeness, though, is an issue with no optimal solution, so if we try to bring his bombastic down to earth a bit, he means something of the following. When all these drivers and passengers are distributed in no-matching numbers, during the day, and throughout various city districts, how can it then be brought together so that supply matches demand through business hours and districts.

The solution is to adjust the prices to get drivers out on the street when demand is high, and vice versa, make them stay away when demand is low. Kalanick describes the for-hire-vehicle business as one with abnormal spurts in demand. These demand spurts were by his calculus driven by three main “accelerants”: nightlife, special events, and weather (Kalanick, T., 2011b). Kalanick figured that dynamic pricing

72 could take care of supply problems in demand spurts, meaning; prices could be raised artificially by algorithms that factored in demand spikes beforehand (Lacy, S., 2011).

The dynamic pricing algorithm was quite crude in the beginning - just a 2x multiplier for New Year’s 2010. In October 2011 Uber stepped up its game on Halloween night (Joyce, 2012). The algorithm was now a bit more refined as prices were set to vary between one and two times the normal price. Uber notified users beforehand that surge pricing was in effect with having a Halloween pumpkin appearing in the user interface (Kalanick, T., 2011d). On New Year’s Eve 2011, Uber went all out and let the algorithm run without caps. The users were now getting a stronger notification and had to press an ‘ok’ tab if they agreed to use the higher priced service. However, Uber had laid out a trap. When users kept hitting the ‘ok’ button, prices would increase even further in the next iterative update of the system (Joyce, 2012). The trap is no longer part of the algorithm, but the rest is generally intact.

So today, how does the surge pricing algorithm work exactly? - Well, according to sources from within Uber back in 2012; the prices gets higher when expected pick-up time increases over a certain level in specific grids on the map (Narducci, 2012). It is likely that it still work like this; in reality, it is an extension of the heat map algorithm.

It is worth noting that these three algorithms have a common denominator. It is all about picking up traffic data in individual cells in a graph. Based on this data – various calculi is performed.

4.2.7 The fallout from dynamic pricing

Dynamic pricing is maybe the most offensive part of Uber’s customer treatment. On several occasions, the practice has led to protest storms. Despite the unpopularity, Uber have never wavered. Spokespeople for the company claims that dynamic pricing is an important instrument to achieve one important goal: that users never should go without being serviced. Surge pricing incentivizes drivers to work and reduces the wait for customers willing to pay (Gurley, B., 2014a; Chen, M. K. & Sheldon, 2015).

Surge pricing debacles

Many users felt they were ‘taken for a ride’ trying to get home from New Year

73 festivities in 2011 when there were no caps keeping prices in check. Users reacted extremely negative, and several journalists reported the uproar. One user tweeted out in outrage that he was charged $125 for a 2 miles ride and another did a major shout- out on social media about a 5x surcharge (Agrawal, 2012; Mulligan, 2012).

In total Uber received 97 complaints on the incident. Kalanick admitted no wrongdoing and stated that all in all the experiment had gone well. He announced that users could expect that surge pricing would be a permanent fixture in the future (Goode, 2012). New Year's Eve 2011 should, therefore, prove to be one of many New Year Eve’s debacles yet to come (Kovach, 2014; McNeal, 2016). The uproar about dynamic pricing has been especially strong in situations of natural disaster and other emergencies, which many users find unethical and as an example of price gouging (Bradley-Carr, P., 2012; Ries & Ryall, 2014). Uber have therefore decided to close- out surge pricing under these conditions (Prigg, M., 2014). The issue is still contentious, though, because it can be hard for engineers in control of an information system to keep tabs on all the calamities that go around in the world (Revesz, 2016).

In 2015, three researchers reversed engineered Uber’s opaque algorithm and tested out which effects it had on market dynamics in San-Francisco and New York. The researchers concluded that Uber’s surge pricing has a minuscule effect on supply and a large negative impact on demand. Furthermore, they found that the algorithm was extremely volatile since it did new calculations on various grids in 5 minute intervals. Small changes in the user’s geolocation could, therefore, affect the pricing of fares in an unfair and non-transparent way. They also found that users within an area could get different price calculations for no reason what so ever. When the researchers reported this problem to Uber, the company had to admit that it was a bug in their system causing the failure (Chen, L., Mislove, & Wilson, 2015).

4.2.8 This is how Uber scaled

The fallout from dynamic pricing is a taste of how unpopular Uber is in many corners. In most cases, the protests against the company come from more natural adversaries and not their users. The push back from adversaries has hampered Uber’s growth. It has also proven to be obstacles when Uber attempt to augment the user base from the current level. We will deal with these issues in the next chapter.

In short, there are three simple explanations to Uber’s exceptional growth. Uber was funded to the brim and used the money to scale fast in every city by subsidizing both sides of the market. Fares were low, and drivers got artificially high earnings. To expand the number of potential service providers by opening up for non-professional drivers was a crucial decision. It brought the venture out of the niche concentrated around luxury and paved the way for the masses.

Deeper down, organizational nimbleness and a modular technological architecture played a vital role. Uber found a method which allowed it to cross geographical boundaries by applying a recursive launch playbook. Hence, the agile team that got Uber started in San-Francisco sprawled into many flexible teams all over the world. The team of teams structure enabled organizational decentralization, which meant that every team got the freedom to act upon local requirements. The technological architecture played a role in making the team of teams structure work. Many independent deployable modules in so-called micro services allowed for independent local initiatives. The way Uber perfected various algorithms to bring available cars to their users; made the service more efficient.

4.3 Augmentation Obstacles

Slowly, from 2014 and onwards, Uber have steadily tried to use their position as ‘everyone private driver’ to pivot into other areas of transportation. The business now revolves around everything that is movable from one point to another. Thus, Uber “aspire to make transportation as reliable as running water, everywhere and for everyone” (Kalanick, T., 2016c). Uber’s slogan is now simply ‘get there.'

We have seen that Uber has conquered the globe with its digital bits. The atoms have been harder to tackle. As we became acquainted with in the chapter 'An Uber-review,' Uber have encountered many governance challenges. Additionally, some new challenges have arisen on the horizon. One reporter gives this account on Uber’s problems: “Uber’s combative streak served it well in the beginning when its goals were to decimate the traditional taxi industry and barnstorm its way across the globe, gobbling up market share in its relentless quest to become everyone’s de facto mode of transportation. But now the company’s chickens are coming home to roost, and Uber appears to be at a loss for what to do” (Hawkins, A. J., 2017c). So, what is the

75 company’s chicken? – Uber’s biggest problem in the long term boils down to its crumbling relationship with both drivers and users. A lack of trust is spreading through the network, making some tasks which require cooperation harder to achieve. To explain this view, we need to go through some of the obstacles in a systematic manner.

4.3.1 Uber and regulators

Uber operate with relative ease today in the Anglo-American sphere. Elsewhere, the picture is more mixed.

Anglo-American regulators

In general, Uber did not have to function long in a city before regulatory authorities would take a closer look at the company’s activities. As previously mentioned, Uber got served with a cease-and-desist order in San-Francisco just a few months after its launch. The charges directed against Uber were that it operated an unlicensed taxi service and limousine dispatch. Uber found a loophole claiming it was not a cab service, just a mediator, and that all their drivers were licensed limousine drivers (Kolodny, 2010).

The San-Francisco strife was just the opening birth-pang for what was to become a long and arduous struggle with regulators. In nearly all cities where Uber launched, some push back from the authorities followed. The charges brought against Uber were not always the same. In Chicago, it was about vehicle inspection and driver insurance (Rao, L., 2012b). In Boston, it was about the GPS technology not being approved for business purposes (Lawler, R., 2012a). In New York, it was about credit card payments, which was not an approved procedure (Tiku, N., 2012). In Washington D.C, it was about payment procedures, which according to regulators needed to be settled beforehand if Uber wanted to use professional limousine drivers (Eldon, 2012). If we move forward in time, Uber is still meeting resistance. In 2016, Uber suspended services in Austin Texas, when it could not comply with city demands of doing fingerprint background checks of the drivers (Dockterman, 2016).

Uber’s answer to this onslaught has been to carry on, go into dialog with authorities, find loopholes and muddle through somehow. When that does not work; Uber take on a more confrontational stance. The company has reasoned that there is power in numbers. Uber have a task network at hand which is eligible for mobilizing since users are required to provide an email address upon signing up for the service. The users can, therefore, easily be solicited to campaign for the company, petitioning local authorities and so forth. Uber have been able to generate tens of thousands of signatures in specific campaigns (Helderman, 2014).

Users willing to speak out for the company served to legitimize Uber as a service. However, as the regulatory resistance against Uber steadily mounted in lockstep with its expansion, Uber managers realized that popular support was not always enough. In short, it became a virtue of necessity to lawyer up, bolster public relation management and influence governmental bodies more discretely. So far, lobbyist and lawyers have proven to be an effective instrument when it comes to keeping a lid on bad press, fighting back in litigation cases and working out compromises with local authorities (Weise, 2015).

Several pundits have uncovered a replicative pattern to how Uber behave towards opposing regulatory forces (Public Citizen, 2016; Weise, 2015). It goes something like this: (1) Uber launches in a city regardless of existing regulations, (2) scale the user base rapidly, (3) rally users to pressure local officials, and (4) unleash well- connected lobbyist to influence regulatory bodies. If this is not working: (5) try to change the laws by influencing lawmakers through grassroots-campaigning, political donations and lobbying behind closed doors.

Uber’s tactics have paid off. In September 2013, California’s state wide regulatory body unanimously voted to create a new category for app based ride sharers called Transportation Network Companies (TNC). It was the first ruling recognizing ridesharing as an individual entity to be regulated based on its own merits. Under terms of the ruling, TNC’s had to comply with a minimum set of requirements regarding insurance, vehicle inspections, driver training programs, a zero tolerance policy on drugs and alcohol, and criminal background checks. TNC drivers also had to obtain licenses from California Public Utility Commission (CPUC, 2013). Uber was already in compliance with most of the requirements.

In the following months and years, most American states followed suit and legalized TNC’s under more or less the same framework as California. The same goes for most nations in the Anglo-Saxon world. Uber became legal in London in 2015. In Australia, Uber became legal in nearly all provinces from 2016 and onwards. Canada has been a struggle for Uber but is now considered legit in most provinces.

European regulators

Uber have been less successful in influencing regulators in Europe. An overwhelming number of nations have moved to ban ridesharing with non-professional drivers. The violations of license obligations are considered unlawful in regards to safety and fair competition.

We can only speculate on why Uber have been unsuccessful in swaying European regulators. It can have something to do with the strength of the push back Uber have encountered. Taxi drivers have in many cities protested forcefully and efficiently. Another explanation may have to do with body politics. Unionized labor is in contrast to Anglo-American nations, more of a power factor in Europe. Thus, the political power of Uber lobbyist was no match for European taxi unions. Naturally, political philosophy is also playing a role. Europeans put less trust in free markets than Anglo- Americans. The public is, therefore, looking more favorably on upholding regulations.

As previously mentioned, Uber have not given up on legitimizing its business model. EU’s top judicial authority might side with Uber. European nations may also come to display various interests. Uber have its European headquarter in the Netherlands and BlaBlaCar is a French success story. These factors may sway French and Dutch attitudes.

Attitudes in emerging markets

It is harder to find a red thread traversing the emerging markets. In south-East Asia, ridesharing is banned in South-Korea, operates with difficulty in Thailand and Hong Kong, but is more or less legal in the rest of the region. In the Middle-East and North- Africa, Uber operates despite facing widespread regulatory resistance. Cairo is, for instance, supposed to be Uber’s fastest growing city as of 2016 (Bhuiyan, J., 2016b).

In Latin-America, Uber is storming ahead in a massive expansion (Newcomer, E., 2016c).

By the looks of it, regulations do not seem to be the same concern in emerging markets as it has in developed nations. Isolated incidences of trouble are of course not welcoming news in the Uber headquarters, but far worse; is favoritism of national champions in markets where Uber are trailing the competition. Russia, India, and China come to mind. We return to this later.

4.3.2 Are driver’s employees or freelancers?

Legal scholars have characterized it as the million dollar question (Woo & Bales, 2016). The question is a legal grey area, but a paramount one for Uber. Uber will have problems making profits if the courts give drivers employment status.

In 2013 a group of Californian drivers filed a class action lawsuit against Uber claiming that the venture had misclassified them as independent contractors. The drivers demanded repayment of tips which Uber had collected on their behalf and that Uber pays for some of their operational expenses (Bowe, 2013). Uber tried to have the plaintiff's claims rejected by the court. However, the court rejected Uber’s motion. The case was valid and deserved its day in court (Shapiro, 2015).

Woo & Bales (2016) gives an excellent summary of the two sides to the question. Uber lawyers argue that drivers have considerable independence and that their relationship with the company is limited. Drivers can work as much as they like and they can decline to take on specific rides. Uber cannot fire drivers but only terminate user accounts when violations are apparent. Reciprocally, the drivers can end the relationship at any moment by turning of the app.

The plaintiffs contest this view. Drivers go through a hiring process which includes background checks, city knowledge exams, vehicle inspections and interviews. Uber also have significant control over how they conduct trips. Uber manuals emphasize that all drivers behave professionally and that drivers should not reject trips as a general rule. User feedback monitoring and the star rating system are strong control mechanisms. If drivers do not live up to conventional standards, there can be reactions ranging from friendly reminders to user account termination. To have its user

79 accounts terminated is in the driver’s opinion equal to being fired. Moreover, they argue that Uber control what they earn through the prepaid fares.

Uber argue that this is a misunderstanding. User feedback and manuals provided by the company must be viewed as useful advice and not as actual requirements. Uber also dispute that it has any control over pay because drivers can negotiate higher fares with riders by turning of the app before the trip is complete and choose to work when they know that the pricing algorithms spikes. However, the drivers call this nonsense; arguing that redoing the prepaid agreement is only possible in theory and not in real- life.

The legal wrangling back and forth could obviously go either way. Uber have, therefore, been eager to settle the case. In April 2016, Uber succeeded in reaching a settlement where it admitted the driver’s $100 million in retribution for confiscated tips. The company also promised to pave the way for unionizing among the drivers, and that it will open up for collective bargaining on driver issues. The drivers, though, kept their status as independent contractors. On this point, Uber will never budge (Kalanick, T., 2016b).

However, the issue is not going away entirely. In Britain, two Uber drivers sued Uber on much of the same terms; claiming that they were misclassified as freelancers and entitled to minimum wage, paid holidays and average benefits. In the autumn of 2015, London courts sided with the drivers. Uber have appealed the decision, and the case is still in limbo.

4.3.3 Competition

When Uber transformed UberX, its goal was clear. It was not going to let competitors get the upper hand in any market. In short, Uber have devised a recursive pattern to how it tries to solve competitive challenges. If no competitors were around, Uber would use the first mover advantage to scale fast and dominate in order not to leave any new market entrants the chance of horning in on the market. If there already were other upstarts active in the market. Uber would use its financial muscles to engage in price wars until the competitors capitulated. The standing operative was that all

80 measures could be used to crush adversaries. If all else failed, Uber would consider tactical pull-backs and make deals to consolidate its market position.

‘Slogging’ Lyft

In the United States, Uber was a latecomer to ridesharing with non-professional drivers. Lyft was the first mover, when Uber first started to move, though, it used its financial muscles to gain the upper hand quickly. Another factor working in Uber’s favor was that it already had established its service with professional drivers in the major cities. Uber was, therefore, able to use its teams of teams’ structure to get the ball moving much faster than Lyft. In a couple of years, Uber was able to dominate ridesharing across America.

Despite Uber’s dominance, it has never been able to deny Lyft a substantial share of the market. In response to Lyft’s marketplace stamina, and other startups coming to the market, Uber managers devised a collaboration network to step up the aggressiveness against all of its competitors. During 2014, the competitors were about to learn what this meant.

In 2014 Uber had to issue a public apology for using unacceptable competitive methods. Evidently, Uber had targeted a competitor in New York City with prank orders, sending their drivers to dead pick-up spots. Simultaneously, Uber operatives poached its increasingly more frustrated drivers (D'Orazio, 2014). Uber’s apology, though, was nothing more than window dressing. The same year, Uber’s competitive strategy leaked out to the media, and it turned out that the tactics used in New York City were part of a nationwide campaign; called ‘Operation Slog.' The idea behind Operation Slog was to give Uber an arm-length's distance to a whole range of subversive actions; constructed to deal major blows to its competitors (Newton, 2014).

Figure 11 is sketching out how Uber intended to slog its competitors, represented here as Lyft. On the right side of the figure; Uber contracted out its dirty tricks to an outside provider, company X, to cover its tracks. Company X employed and equipped the Sloggers with burner phones, credit cards and instructed them to keep silent about the activities. Simultaneously, Uber employees trained the Sloggers in how they

Figure 11 Operation Slog

should go about gathering intelligence about Lyft’s business and recruit their drivers, i.e. tempting them with perks and sign-on bonuses. The Sloggers used the credit cards to create dummy Lyft accounts and the burner phones to request Lyft rides from various pick up points. Well positioned in the car of the enemy, the Sloggers could provide Uber with intelligence about Lyft’s drivers and a substantial number of poached drivers. On the left side of the figure; Uber mobilized its network of drivers and staff to hustle for the company; which in reality meant swamping Lyft with fake orders and cancelling them.

Uber has neither issued any apologies nor indicated that it would end the program. The negative publicity does not seem to matter. On the contrary, in a disturbing turn of events, it looks like Uber competitive strategies are about to spread to other actors in the industry. One reporter has suggested that Lyft has copied many of Uber's dirty tricks and redirecting them back at Uber (Carson, B., 2016c).

‘Out-Ubered’ in China

In February 2014, Uber officially launched in Shanghai (Wu, 2014). The Chinese market was an enticing treat with high smartphone usage, a growing e-commerce

82 sector and a huge number of potential customers. However, China is a difficult market for outsiders to operate in, irrespective of the business segment. Chinese authorities keep a watchful eye on foreigners and favor local producers.

When Uber decided to launch in China, ridesharing was no newcomer in the country. 30 Chinese startups were already active, and the competition between them was tough. Like almost everywhere, Uber met resistance from authorities but was still able to scale in the country, taking some advantage of the fragmented state of Chinese startups. In the start of 2015, Uber looked to have the advantage over local competitors. Then, all of a sudden, the Chinese consolidated and began to strike back. A company called Didi emerged on the scene as Uber’s toughest competitor to date. Didi was the result of a merger between two of the largest Chinese ridesharing companies, and the merged entity became in one strike bigger than Uber. Subsequently, Didi attracted massive funding, and the competition with Uber was on. Through a year and a half, Uber and Didi burned through several billion dollars of cash in subsidizing prices and wages for drivers, trying to achieve network effects and dominate the market (Stone, B. & Chen, L. Y., 2016).

In the end, Didi was the most successful. Uber came to a point when it decided that it was too costly to keep on fighting and that it was better to call for a truce (Kalanick, 2016c). In an agreement between Uber and Didi, Uber agreed to leave China against gaining a 20 % stake in Didi and a promise of a $1Billion quid pro quo investment from Didi into Uber (Abkowitz & Carew, 2016). The withdrawal rested upon the mere fact that Uber never was able to scale as fast as Didi. Before the merger, Didi was operating in 400 cities, while Uber never managed to get over 100 (Alba, 2015; Stone, B. & Chen, L. Y., 2016).

There are mainly two explanations to why Uber met its Uber-man in Didi. (1). When Didi emerged on the scene, it was the market leader in many megacities. It gave Didi first-mover advantage. (2). Didi had better knowledge about the Chinese marketplace which entailed some benefits. First of all, it was able to create alliances with governmental bodies. Biased authorities threw up roadblocks for Uber, but not for Didi. Secondly, Didi had a better understanding of customer’s needs which made it able to offer a stronger product mix than Uber. Thirdly, Didi was more successful in recruiting a diversity of drivers. It started with professional taxi drivers and worked its

83 way down to non-professional ridesharing drivers. Uber commenced on the bottom and had to recruit more drivers to equal out the professionals who were more disposed to drive full-time. Finally, Didi was able to launch popular carpooling services on some predefined routes with a high number of fixed commuters. Uber was never able to venture into the same segment of the market (Wirtz & Tang, 2016).

‘Even Steven’ in India

Losing out in China, Uber have put its main focus on India and South-East Asia. India is now Uber’s largest market after the US, but just as in China, Uber is up against a formidable adversary. Uber are currently trailing Ola by around 200 000 drivers.

The Uber headquarters reckon that it have bigger chances of succeeding in India. Ola does not have the same funding as Didi, and cannot subsidize and burn through cash to the same degree. Price competition between the two should, therefore, be more benign. Regulatory authorities also seem to be fairer, treating the two more or less alike, but this could change of course. Besides that, Ola has many of the same advantages as Didi. Ola has Intimate knowledge of the marketplace. It is agile and innovative. Always willing to test out novel services, and try out new businesses processes; like for instance opening up for cash payments, which Uber soon copied (Rai, 2016).

UberIndia is equally innovative. Many of the newest business processes which Uber have come up with since 2015 have originated in India. Local projects include online bookings, car ordering without doing app downloads, an alert button for riders in distress, Uber Hire (where you can order a driver for a whole day), and the UberFLEET app (for car-leasing firms) (Russel, J., 2017a).

Uber might surpass Ola in revenue fairly soon considering how much money the company is investing. However, it is not very likely that Ola will crumble under the pressure of Uber. Ola is having success with its Micro line, which is the cheapest alternative on the market as of 2017. Micro cars are compact hatchbacks like Hyundai Eon’s and so on. Ola is also able to make profits out of its premium service, a segment which Uber have trouble mastering in the country (Bhuiyan, J., 2017a).

The question is whether Uber is willing to consider operating as a duopoly in the country. India is certainly big enough to sustain two large companies. Nevertheless, the fight for supremacy is very hard between the two. In an ironic twist, Uber is accusing Ola of ‘slogging’ and have sued the company for unfair competitive practices (Hawkins, A. J., 2016a).

Going nowhere in Russia

In Russia, Uber have trouble scaling the user base. Uber is a distant third in size, hopelessly behind Yandel. In 2015, Uber operated with 3 000 cars in Moscow against 15 000 Yandex vehicles (Khrennikov, I., 2015). After the Russian invasion of Crimea, the civil war in Donbas, and the subsequent Western sanctions, things have taken a turn for the worst for Uber. From 2017, Kremlin levied special taxes on Uber while leaving Yandex untouched. The result is that drivers are leaving the Uber platform (Khrennikov, I., 2017). So, government favoritism of local champions is something Uber have to struggle with in some markets overseas.

‘Winner-takes-all’ markets?

The empirical evidence above suggests that ridesharing is not a winner-takes-all market in a global sense. Furthermore, there is no agreement between pundits about whether sharing is ripe for wide competition or whether the markets naturally fluctuates to being domineered by a few players.

On the one hand, ridesharing has no natural moats. Ridesharing apps are relatively easy to develop and require small amounts of capital investments. It depends on freelance service providers with no special skills or special loyalty. Moreover, users can, if they feel for it, download new apps in a matter of seconds. Winner-takes-all markets usually do not develop in markets where there are relatively few barriers to entry (Evans & Schmalensee, 2016).

On the other hand, digital platform markets have tremendous network effects which usually points to monopolization. Many of Uber’s investors claim that in the long run, it is room for only one in every market. When one player gets dominant, it gets hard to push it from the throne because it would require too much of a cash burn for

85 challengers. That explains Uber aggressiveness when it moves into a market and explains why the Uber headquarters decided to pull out of China (Levinson, 2015).

Consolidation of fiefdoms

The merger between UberChina and Didi signals that established ride sharers now have reached a scale where they want to solidify their market position. Not surprisingly maybe; since new startups continue to attract venture capital. Following, we see that cross-ownership and cooperation grow within the sector. In 2015, Didi, Lyft, and Grab formed what became known as the anti-Uber alliance. The partnership was a construct devised by Softbank, which is a major investor in all of them. Naturally, Softbank had no interest in seeing the three companies challenge each other in their home markets (Lunden & Buhr, S., 2015).

Further, we see that other big businesses in the tech and car sector invest significant stakes in ridesharing companies. Apple has invested in Didi (Chen, L. Y., Webb, & Oster, 2016). General Motors have invested in Lyft (Newcomer, E., 2016a). Google and Toyota have invested in Uber (Wilhelm, 2013; Newcomer, E., 2016b). Additionally, the biggest ridesharing companies have cross-invested in each other. Uber and Didi collude, as mentioned above. Didi have also invested in Lyft and Grab (Russell, J., 2015c; Buhr, S., 2015a)

Damodoran (2016) notes that this consolidation looks like how mafias are establishing fiefdoms to make their business thrive without any disruptions. He also put forward the thesis that ridesharing lack of natural moats is one of the main reasons to why Uber make such a huge bet on developing self-driving cars. It is an attempt to get more control of the infrastructure and rise barriers to entry by making the business more capital intensive.

4.3.4 Privacy risk on the Uber platform

As previously mentioned, privacy risks are not going to go away anytime soon on the Uber platform. It has to do with properties in the technology stack, and maybe more alarmingly; the lack of self-control which the company exhibits when it comes to surveilling their users.

Risks in the online reputation system

Trust is one of the central properties in creating efficient marketplaces. Parties need assurances that the other will stick to agreements and refrain from trickery. Thus, trust propels markets to efficiency. Efficient markets are characterized by high turnover of goods and small spreads between bids and ask prices. Most sharing platforms solve the trust quandary by establishing online feedback systems as a way to create trust between trading parties. “The basic idea is to let parties rate each other, for example after the completion of a transaction, and use the aggregated ratings about a given party to derive a trust or reputation score, which can assist other parties in deciding whether or not to transact with that party in the future. A natural side effect is that it also provides an incentive for good behavior, and therefore tends to have a positive effect on market quality” (Jøsang, Ismail, & Boyd, 2007, p. 618). The most optimistic see sharing platform as trust infrastructures with the ability to self-regulate (Sundararajan, 2016).

Unfortunately, there is also a negative aspect of online trust systems. On the Uber platform, it has the potential for creating tensions between drivers and riders. First of all, reputation matter for drivers. It is a well-known fact that Uber drivers get kicked of the platform if their rating falls under a certain grade. It is part of Uber’s algorithmic management of driver partners. Naturally, this management can create huge anxieties for drivers on the precipices from being kicked out from the platform. Then, it is maybe not surprising that drivers have been known to beg and threaten passengers to avoid getting more bad ratings. For travelers, it is best to keep the high score as their default rating to prevent any retribution from aggravated and perhaps unstable drivers (Kane, 2015). However, bad ratings will always occur. Uber recognize that 1 % of all trips ends with the lowest driver ratings. Usually, the cause is heated arguments and outright fights between the drivers and the passengers (Cook, 2015).

Secondly, Uber create dummy numbers between drivers and riders so the two of them can communicate before and after the ride. Drivers can reach the rider for up to half an hour after the trip has ended, but the rider can get to the driver indefinitely. On the positive side, it is a useful feature since passengers can claim lost items through the channel. On the downside, it is simultaneously a channel which riders have used to

87 harass drivers long after the trip has ended (Bhuiyan, J., 2015). For drivers, this is maybe more of a nuisance than a privacy risk unless it is the start of something more serious.

Finally, and this is where the two first points end up in getting serious from a privacy perspective. Users reveal extensive information to other parties on the Uber platform. Moreover, the driver is likely to know where passengers live or work because that is where most trips end. If all this information leads to unwanted attention, where drivers stalk and harass passengers, it is clearly a privacy risk; and naturally; it can also lead to violent and sexual crime. It is the properties of the online feedback system, which many argue make Uber unsafe for female passengers (Nuzzi, 2014; Paul, 2016).

Surveilling users

Uber have something the Uber insider’s calls ‘God view.' The view is an online map where Uber staff can see trips users do in real time. When Uber opened in Chicago, the God view was put on display for all party goers to see under the launch party. Peter Sims, a relatively high profile venture capitalist, and Uber user were mightily surprised when he got text messages about his real-time whereabouts from friends he knew from Chicago. They were watching his movements in real time during the launch festivities and having fun with it (Hill, K., 2014). Sims contemplated suing Uber for privacy breach, but in the end decided not to. He is still not very pleased about the ethics of the company (Sims, 2017).

The God view seem to be hard not to look at for Uber managers. Notable tech journalists have also experienced to get their privacy breached in much of the same manner as Sims (Hern, 2014). Incidents like this have led users to delete of the app and politicians to blast the company for loose handling of user data (Isaac, M., 2014; Holson, 2014). More formally, Uber have also received complaints about the company’s privacy policies at the Federal Trade Commission (EPIC Complaint, 2015). Uber settled this complaint by paying $20M (Levin, 2016).

In response, Uber have tightened up their privacy policy (Tassi, 2015). However, the question remains about how much the new policy have changed things. Later reports

88 show that snooping into personal accounts has been a widespread practice among Uber employees (Conger, K., 2016b). The simple fact is that Uber have the technical ability to surveil users permanently with the app running in the background on the smartphone and Uber is a business which lives and breads on knowing their users. The more it knows, the better. Being able to predict and model users movements are capabilities which the Uber leadership cherish. It is part of Uber math and Uber engineering (Graves, R., 2010c). Very early on, Uber engineers bragged about the ability to perform big data analytics. Uber is among other things able to analyze what the purpose of singular rides are, be it prostitution, be it one-night-stands or other things which it finds amusing to know (Voytek, 2012; Pagliery, 2014).

Hence, to know their user's normal travel patterns are precious for Uber. It can make the algorithms predict with higher accuracy where individual users are going the next time they use the Uber app and make it allocate driver resources more efficiently. According to Uber’s privacy policy, the company now states that it will restrict the tracking to five minutes after a trip has ended (Conger, K., 2016a).

Regardless, new updates in Uber’s software and new software products are continuously pushing the boundaries for how much surveillance the company thinks is permissible. The release of the UberFLEET app in 2016 underscores this fact. According to its specifications, it gives leasing firms the ability to micro-manage drivers through Uber’s technology tracking devices. Leasing partners can check individual driver’s statistics like fares, cash collected, hours online, and login/logout timestamps, trip details, incentives earned, and live mapping of drivers that are online.

The UberFLEET app shows that surveillance is part of the company’s business culture.

4.3.5 Uber and women

Susan Fowler’s tell all blog entry about the work environment inside Uber has only served to uncover something which has been there all along, namely, that Uber have a woman problem. In her blog entry, she describes a misogynist atmosphere of discrimination, and where she experienced episodes of sexual harassment (Fowler, 2017).

Fowler’s revelations have been followed up by a damning report in The New York Times. The report is damning not only because it confirms Fowler’s assertions about an aggressive and sexist workplace culture, but also because it builds on independent interviews with over 30 former and current employees in the company. Furthermore, the report states that three female employees have sued Uber based on sexual harassment and that more women are waiting in the wings considering following suit (Isaac, M., 2017).

Uber CEO Travis Kalanick have a hard time to admit that Uber have a systemic problem regarding the treatment of female employees (Anand, 2017). However, Kalanick insists that Uber will get to the root of the problem and have promised an independent inquiry into the matter (Wagner, K., 2017). Uber board members have announced that the investigation will be transparent and the result revealed to the public. The final goal is to improve the workplace culture (Etherington, D., 2017).

Thus, Uber is in serious damage control mode. However, the question is whether the damage already is done from a public relations perspective. The Fowler case is one in a long series of incidents which have painted the company as insensitive to female issues. Mass media have over the years reported numerous stories about sexual assaults and rape committed by Uber drivers. Uber have tried to calm female users potential anxieties by releasing internal company statistics. Internal numbers show that only 1 out of every 3.3 million trips end in a sexual aggravation of some sort (Bort, J., 2016). It is questionable whether female users will find solace in these odds.

Uber have pledged to recruit 1 million female drivers by 2020. This promise is in danger if not Uber sort out the mess. The problem is both internal and external. The organizational culture needs to be cleaned up, and the drivers need to go through a more arduous vetting process. The star-rating system has apparently not been enough to exclude criminals on the platform.

4.3.6 Public relation management

We have seen that the bond between the Uber managers and the drivers were quite strong from the founding and up to 2014 - 2015. The Uber management was able to mobilize the drivers to perform several tasks. They were mobilized to pressure

90 regulatory officials and hustle against Lyft. These task networks were dependent upon common interests between the leaders and the workers. Both parties wanted Uber to succeed. Uber drivers wanted to continue making money chauffeuring, while the manager's motivations were profits, prestige and the possibility of being promoted within the company hierarchy.

In later years, there are signs that the interdependence of the nodes within Uber’s network has started to crumble. Uber have shown disinterest when drivers have complained about wage cuts. The seriousness of letting the drivers unionize, which was part of the settlement in the employment status lawsuit, is also brought into question since Uber has taken drivers in Seattle to court over the issue (Hawkins, A. J., 2017b). Neither does it help when Kalanick is shown berating an Uber driver on newsreels (Newcomer, E., 2017). In January 2017, Lyft surpassed Uber in the number of downloads on the Apple App store for the first time (Hawkins, A. J., 2017a).

Uber have named the year 2017 for the year of the driver. It is on to a bad start if it wants to follow through on this slogan. In all likelihood, Uber need to cultivate its network better and avoid public relations scandals to ensure continued success. We will now turn to the strategies which Uber have devised to find a way forward.

4.4 Augmentation strategies

In general, Uber uses four strategies to surpass the obstacles we have mentioned. The first strategy has to do with innovating new services on its app. The second strategy has to do with innovating new logistic services on the back of its vast user network. The third strategy has to do with inviting outside developers to make add-on services to its platform. The last strategy has to do with partnering up with outside forces to advance the launch of self-driving cars. The common denominator of these strategies consists of openness and cooperation with the outside world. To give the drivers more services to drive for, can also improve their earnings and rectify Uber’s crumbling relationship with its drivers.

4.4.1 The UberPOOL extension

In late summer 2014, Uber launched carpooling services in San-Francisco. The service simply entails that drivers can pick up more riders going the same route and letting the users split the bill. Uber advertised that this could bring fares down 40 % for individual travelers. The drawbacks for carpooling passengers are that they must accept small detours and some waiting time around new pick-ups. Uber called the service UberPOOL and offered it as a select button alongside all the other services in the app (Lawler, R., 2014a).

The lower price points are making carpooling popular. In April 2015, Uber announced that UberPOOL had completed 1 million rides in San Francisco, Paris, New York, and Los Angeles (DeAmicis, C., 2015a). From this date on, the service has continued to scale and is available in a little over 31 cities around the world as of April 2017. Not surprisingly; the service is very popular in India since it makes ridesharing attainable for customers that cannot afford pricier options. Uber’s Indian headquarters campaigns for the service under the slogan #SwitchtoPool, and sees it as the way to the future (Jain, A., 2016).

Competition in the carpooling sphere has already gotten quite hefty. Large ridesharing competitors like Lyft, Didi, Grab and Ola are not the only ones who are active in the space. Apps that solely mimic mass transit solutions are also coming to the market.

When a vehicle makes successive pick-ups along a lane in a cyclic manner, it resembles public transportation more than cab-services. Several transportation startups in America have taken advantage of this fact. On roads with heavy repetitive user traffic, startups are scrambling to establish so-called ‘smart routes.’ Smart riding means that passengers can hop on to carpooling services which run in shuttle traffic. Lyft is in proportion to its size more aggressive in this area than Uber, but then again; Uber is punching its weight around as well (Constine, J., 2015b; Hawkins, A. J., 2017e).

The emergence of smart routes has brought carpooling in direct competition with public transportation. Mass transit has one distinct advantage versus ridesharing companies. They are often heavy subsidized by taxpayers. Still, greater flexibility makes carpooling a viable business alternative. In localities, where public transport is

92 poorly developed, ridesharing has an opportunity to gain market share. Ride sharers can also strike up deals with individual companies and shuttle employees who are living close to each other to and from work. In some American cities, this service is popular among employees since they get tax deductions from it (Hawkins, A. J., 2016c).

Local American politicians have also opened their eyes for the flexibility of carpooling. Many cities are struggling with ongoing budget cuts. In these situations, it is tempting to compete out mass transit routes with little capacity utilization. In Florida, a municipality has started to subsidize UberPOOL to operate parts of the mass transit system. It is a tacit understanding between the two parties that this replaces inefficient public transport services. More municipalities are also looking into this alternative. They see that Uber and Lyft in some instances can offer commuters better services and reduce the costs for the public (Brustein, 2016; Lazo, 2016).

So, it is no longer only apps versus taxis; now it is also ridesharing versus public transport. The ridesharing companies argue that carpooling is good for the environment. The main argument is that it replaces private car ownership, and thereby leads to less traffic and less Co2 emissions (Kalanick, T., 2015). This reason is also partly why many see carpooling as a solution to congestion problems of the growing megacities in the emerging markets. Some take the opposite view, arguing that it leads to more traffic overall. If carpoolers mainly replace mass transit, it is hard to see that it will lead to fewer cars being on the road (Bershidsky, 2016).

4.4.2 New applications

Uber’s CEO Travis Kalanick argued very early on that Uber is a logistics company, not a cab service (Bradshaw, 2014). From experimenting with logistics in 2014, Uber started to develop it as services in earnest from 2015. UberRUSH and UberEATS are the two most important ones.

UberRUSH

In 2014, Uber tried out package deliveries with bike couriers in New York. A year later it tested the hauling of cargo in Hong-Kong (Lawler, R., 2014a; Russell, J.,

2015a). The experimentation led to the launch of UberRUSH in October 2015, a delivery service which took care of both smaller and larger items (DeAmicis, C., 2015b). As of April 2017, the service is available in New York City, Chicago, and San-Francisco. The service is structured a bit differently than the regular Uber app. The UberRUSH service is more business to business and is integrated into individual merchant’s check-out processes. Uber look at it as more of a distributed infrastructure service than something which is suitable for a two-sided market structuration (Droege, 2015). Naturally, Uber’s network of drivers is taking care of the deliveries.

Digital delivery services are a highly competitive business space in America. Postmates is the biggest platform. It is especially significant in California where it operates in big and small cities all over the state. However, food stands for 75 % of Postmates deliveries (Buhr, S., 2015b). Uber have, therefore, reasoned that it could challenge Postmates better around a specially tailored platform for the sole purpose of delivering food.

UberEATS

In fall 2014, Uber launched UberFRESH as an experiment in California. Back then, it was an additional choice on the Uber application (McGregor, 2014). Approximately one year later Uber tested out food delivery on a standalone app in Toronto (Hempel, 2015). In March 2016, Uber took the experimentation to the next phase, and UberEATS launched as a standalone app in Chicago, Houston, Los Angeles and San Francisco (Perez, 2016).

The service works pretty much as the Uber app. The only difference is that there is an additional layer of service providers. UberEATS offers pre-arranged lunch and dinner menus from different partner restaurants. It seems to be a popular extra venue to make money for restaurants. In Stockholm, as an example, UberEATS offer menus from over 100 restaurants. Uber’s drivers, or couriers on bikes and scooters, take care of the deliveries.

The service has scaled tremendously considering that it only has been around for one year. As of April 2017, the service is available in 71 cities and exists on every continent. Obviously, Uber have one advantage when it comes to scaling food-

94 delivery geographically. The company has already an established distribution network consisting of its driver partners.

For drivers, the extension of driving opportunities is a good thing. It means that they have a greater pool of demand. Currently, the drivers are guaranteed a minimum pay of $10 - $20 per hour (RideSharing Guy, 2016). Dynamic pricing is also a fixture in the UberEATS platform (Hawkins, A. J., 2016d). The competition in the field is harsh though. Several startups are active in the sphere: Postmates, Caviar, Sprig, Munchery, GrubHub, Seamless, DoorDash and more (Carson, B., 2016b).

4.4.3 Partnering-up

In short, Uber need to partner-up with others to advance many of its new goals. These targets consist mainly of developing autonomous cars and transforming Uber into a logistics company.

Opening up the platform

To open up the platform for outside developers is an alternative when a corporation wishes to create new services. In essence, there are two ways to open a platform: the first, is giving outsiders access to use parts of the code base in the platform to make complementary services tailored around the platform, the second is giving up control over vital areas of the platform itself (Boudreau, 2010). In Uber headquarters, the second alternative has never been an alternative. Uber want to have full control over its technology stack and do not want innovation that can alter the product it wants to sell. For Uber’s owners, it is all about, ‘what is it in it for them.'

In 2013, Google Ventures made a significant investment in Uber. The following year, Uber was allowed to integrate its service as a traveling alternative in Google Maps (Etherington, D., 2014). Later the same year, Uber mulled openly upon whether they should grant outside developers access to its platform. The idea was to entice third- parties to put Uber trip-request buttons in their applications. Like one reporter in close contact with Uber speculated: “colonizing the mobile web with Uber buttons could pay big dividends, [ just think about what kind of business opportunities this could bring,] maybe Yelp, Foursquare, or OpenTable could let you book a one-tap ride to a

95 restaurant. Travel apps like HotelTonight could offer an instant Uber to your bed” (Constine, J., 2014).

The practical way to grant access to the platform is to publicize an application programming interface (API). It makes it possible for third-party developers to integrate services from the platform into outside software. This sort of software integration is very common in the development of platform ecosystems. When it happens, an ecosystem emerges where there exist a core unit (the platform) that remain more or less or less fixed and several modules (applications) loosely attached to it in the periphery that is allowed to vary over time (Baldwin & Woodard, 2009). Accordingly, digital platforms can be defined as “the extensible codebase of a software-based system that provides core functionality shared by the modules that interoperate with it and the interface which they operate” (Tiwana, Konsynski, & Bush, 2010, p. 675). Thus, the interface is the layer which makes the core parts link- up by hook and crook with the peripheral parts.

The Uber API came out in autumn 2014 (Lawler, R., 2014c). Uber also established an affiliate program (Uber Developers) as a resource for third-party developers. In the beginning, trip request buttons were the only thing developers were allowed to integrate with third-party apps. Today, the access points have sprawled into four use areas. One, where developers can enable the whole user experience of riding with Uber in third-party apps, which covers everything from getting fare estimates to making payments at the end of the trip. A second, where third-parties can hook themselves into user’s account on the Uber platform. The point is to deliver services to users while they are riding. These services can be anything really, but the most popular ones are entertainment of various sorts. The Uber users must give third- parties the permission to access their accounts. A third, where developers can integrate same day UberRUSH deliveries in the third party check out processes and access real time updates of where the deliveries are on route to its destination point. Lastly one for drivers, where developers can make add-ons to the Uber driver side application. Typical add-ons are trip data, economic management software, and other services which enhance the driver experience. The drivers must give third parties permission if they want to access their driver data.

The Uber developer community has not grown exceptionally large. So, there are not so many companies that have integrated their services with Uber. Maybe, that is why Uber have put up an incentive scheme to get developers interested (Constine, J., 2015a). A few companies have installed Uber trip requests into their apps. The most known are United Airline, Hyatt Hotels & Resorts, Microsoft Calendar, Tripadvisor, OpenTable, and Starbucks. It can help Uber doing more business among travelers and restaurant goers, but as of now, it is hard to discern whether this contributes to revenue in a meaningful way. By enabling third parties to access users while they are travelling, Uber is hoping that it can make users time in an Uber vehicle into a content marketplace, but not much have happened so far (Russell, J., 2017b). We know for instance that Washington Post is hooked up as a news service and Otto Radio as a podcast. The same has happened with the UberRUSH. Only a few shops and restaurants are involved. The driver side API has not seen too much integration either, but the integration it has seen might be important overall seen from the driver’s point of view. Stride is an app which helps drivers estimate their taxes. SherpaShare is an app which gives drivers tips on where demand is high and where surge pricing is likely to be in effect. It also gives drivers analysis of their driving habits in the sense of where and when they earned the most in the past. Activehours is an app which able drivers to get paid-as-they-go from Uber.

To integrate applications to the east and west might be a thing of the past. It is a cumbersome and drawn out process. In the end, it can lead to an app fatigue among users. The emergence of chatbots solves the disintegration of the various platforms. Chatbots are designated app’s that integrates the APIs of the most popular on-demand services that currently is in use. So, these bots make requesting different services easy for consumers. In this respect, Uber’s deal with Facebook Messenger may be the most profitable way to populate the ways that users can request different types of Uber rides while not being physically on the Uber app itself (Buhr S. & Constine, J, 2015).

Autonomous cars

Uber’s CEO Travis Kalanick stated in the summer of 2014 that self-driving cars were the way to the future. He lamented that the “reason Uber could be expensive is you're paying for the other dude in the car. When there is no other dude in the car, the cost of taking an Uber anywhere is cheaper. Even on a road trip” (Carlson, 2014). In 2015, 97

Uber signaled that it was going to be in the forefront in the development of self- driving cars. It represents a major shift for the company. Developing self-driving cars is a massive technological undertaking and something entirely different than the software development a small number of people did in Uber’s infancy. Regardless of whether Uber is going to succeed in this matter or not, it signals that this is a company coming of age; it is a company which no longer intends to restrict itself to tasks which consist solely of software engineering. By this action, it is taking on a huge engineering task; something that is going to involve many people both from within and outside the company. Alas, to make this fly, Uber need to hire highly qualified personnel, partner up with research institutions, and a wide variety of industrial actors, regulators, and public officials. Such a development will create dependencies with outside stakeholders, and do away with the free-wheeling and spontaneous management style of the past.

Political alliances

There will be no revolution in autonomous cars without getting lawmakers behind the effort. Rules and regulation must be put in place to make room for how to handle problematic issues and second order effects that might arise from using the new technology (Scoble, 2016). Early in 2016, several ridesharing, car and technology companies decided to form a group which would go into dialog with authorities to advance the introduction of self-driving cars. On 26 of April 2016, Uber, Ford, Google, Lyft, and Volvo established the Self-driving Coalition for Safer Streets.

The coalition intentions were to influence lawmakers and regulatory authorities in the US. The coalition argued that self-driving cars would “make America's roadways safer and less congested." Accordingly, the best solution was to: “have one clear set of federal standards, and the Coalition will work with policymakers to find the right solutions that will facilitate the deployment of self-driving vehicles" (Hawkins, A. J., 2016b).

The industry is not the only ones who see opportunities in the new technology. Local politicians are also competing to get industrial actors to set up research facilities in their states (Grandoni, 2015).

R&D partnerships

In the winter of 2015, Uber launched a research center in Pittsburgh, Pennsylvania. The Uber Advanced Technologies Center in Pittsburgh was the result of a strategic partnership with Carnegie Mellon University and National Robotics Engineering Center. The three partners stated that they intended to cooperate on primarily the areas of mapping and vehicle safety and autonomy technology (Biggs, 2015). Later the same year, Uber struck up another partnership with the University of Arizona. The goal of this partnership was to develop the optics system needed for making autonomous cars work (Brandom, 2015).

During 2015 Uber did key hirings of technical personnel to further the advance of the project. In total it was over hundreds of engineers, robotic experts, car mechanics, and so on and so forth (Chafkin, M., 2016).

Acquisitions

Uber have made three acquisitions according to Crunchbase. All of them have been done to advance Uber efforts in the development of autonomous vehicles. On the 18th of August 2016, Uber acquired the self-driving truck startup Otto for $680M in cash (Dillet, R., 2016). Otto consisted of many defectors from the Google self-driving car project and represented a major boost for Uber. It had developed proprietary LIDAR sensors. A component which makes the car orientate itself on the road. The head of Otto was made the leader of Uber’s self-driving car division (Chafkin, M., 2016).

The Otto acquisition paired up well with the purchase of the mapping company deCarta which was acquired a year before (Mangalindan, J. P., 2015). “Combining Otto’s technology with Uber’s massive push in mapping makes Uber stronger than ever when it comes to putting self-driving cars on the road. Uber will also be able to use its enormous user base to tap into ride data and improve its mapping solution “ (Dillet, R., 2016).

At the end of 2016, Uber acquired another startup Geometric Intelligence. The startup specializes in artificial intelligence and will help Uber enhance their mapping technology.

Joint ventures

To acquire Otto was not the only announcement Uber did on the 18th of August 2016. The same day, Uber announced a joint venture with Volvo. The two companies agreed in principal to join their knowledge to develop self-driving cars. Primarily, Uber will fit Volvo cars with all its software, and bring the two of them closer to the goal (Bhuiyan, J., 2016c). In February 2017, Uber struck up a similar partnership with Mercedes Benz (Davies, 2017).

Testing strategy

Uber have tested self-driving cars in Pittsburgh, San-Francisco, and Phoenix. Otto also has done long-distance test runs. The tests have not always been condoned by the authorities beforehand (Hawkins, A. J., 2016e). Following, Uber is living up to its adage that it is better to ask forgiveness than permission. According to some pundits, Uber’s aggressive testing have paid off. It receives dividends by acquiring massive amounts of data from the numerous test runs. Uber is in some quarters considered to be ahead of Google in the competitive race, which mostly has concentrated on doing test runs in laboratory environments (Barr, 2016; Chafkin, M., 2016). The competition between Uber and Google is heating up. Google accuses Uber of stealing industrial secrets (Hawkins, A. J., 2017d).

Other news reporters are not so impressed by Uber’s self-driving car progress. According to inside company material, every test run needs an intervention from the driver for every mile it travels. The self-driving cars also have one kind of ‘bad experience’ from every other mile it travels. A ‘bad experience’ is Uber’s jargon for “how many times the robot car abruptly decelerated, braked or jerked around” (Bhuiyan, J., 2017b). Needless to say; the self-driving car research has some children diseases which need addressing before we can start to talk about an imminent breakthrough.

4.4.4 Uber Movement

In January 2017 Uber launched the ‘Uber Movement’ website. In short, it is a compilation of massive traffic data that Uber have gathered over the years and some transportation studies the company has done. The point of making the data public was

100 to share it with government agencies, helping local transportation authorities in their planning work. In the start, the data was made available for the Washington metro area, Boston, Manila, and Sydney, but Uber signaled that more cities were soon to be released (Dwozkin & Siddiqui, 2017).

People who want to access Uber’s data need to sign up for the service. Moreover, Uber decides whether it want to grant access or not. On the Uber Movement website, it says that the data is relevant for city officials, planners, and policymakers. Uber states that it want to open the databank to the general public as well, but have not announced releases yet.

4.4.5 Network of networks

In reality, Uber have always tried to tap outside resources to grow. From early on, venture capitalists got board posts. Since then, the board has been broadened by representatives with valuable industrial expertise. Uber is also seeking support from outside sources to solve governance problems. In the charges about sexism in the workplace, Ariana Huffington, a female board member with no financial strings attached to the company, is handling the accusations in public. Kalanick is also on the look-out for a deputy to help him out with internal responsibilities.

So, a clear picture emerges on how Uber tries to tackle various augmentation obstacles. Uber partners up with outside forces not only to advance its goals but also as a way of seeking protection.

A network of network effect is creeping onto the scene as a general evolutionary trend in Uber’s governance. Uber strike up partnerships in east and west. Politically, it wants to share data with authorities, and thereby grow into being considered a transportation utility. To have political backing can also help Uber to defend against charges of data mishandling. Industrially, it cooperates with other relevant companies to develop autonomous cars. Digitally, it wants to gain partnerships and sprawl into other business segments with the aid of outside developers. Simultaneously, it further its innovative processes by spinning out new services based on the vast network of local offices. In total, these are all examples of innovative and corroborative networks that are set up to solve specific tasks.

101

How American political authorities are starting to cooperate with Uber to develop transport solutions is also a telling sign. It shows that many public officials have changed attitude. Politicians are not considering Uber as a total nuisance any longer. In many states, Uber is part of the solution to transport policies.

The ultimate result of all these partnerships is that the position of the company solidifies and its power grows.

4.5 What happens when wages goes south?

Uber is in a popularity race, and then again it is not. It wants to look attractive so it can entice drivers to sign up and carry passengers, but simultaneously; it cannot give drivers more incentives than what is sensible from a business perspective. Uber is a broker. Its main goal is not to ensure service providers top earnings, but to increase the numbers of transactions on its platform. Transaction fees equal Uber’s revenue stream. Uber executives have figured out that more transactions mean higher revenues, even though the rate of every fare depreciates.

The Uber math work as follows: The first drivers deserve high guaranteed wages, but when a market first is created, the salaries are exclusively coming out of the prices which the riders are willing to pay. Uber knows that when it lower prices, more users are willing to buy transport, which, in turn, leads to a bigger market and a better capacity utilization on the driver side. Due to less idleness; drivers earn more despite lower fares. More demand also means that more drivers can be added onto the platform until supply matches demand more accurately. If Uber slashes prices yet again, the virtuous cycle starts all over. Uber portray this as a win-win situation for the drivers (Gurley, B., 2014b; Knight, 2016).

The combination of lower prices and more transactions are a double-edged sword for drivers. Yes, drivers can earn more money when business increases, but it also means that they need to complete more trips to earn the same as before. In fact, the Uber math does not seem to work so well for drivers as its dynamic logic portend. When Uber succeeds in chasing the number of transactions higher, what we see across the board is that driver earnings fall and that the time spent behind the wheel to earn a

102 decent living goes up. It is a wage dynamic that is universal for saturated markets both in the developed and in the non-developed world.

4.5.1 Drivers in the developed world

Uber offer standardized work, but it seems that the earnings level vary somewhat between different regions.

The United States

The American economy has created about 5 million new jobs since the great recession. There has been little job creation in fly-over-America. The new jobs have overwhelmingly been set up on the two coasts in the major metropolitan areas. The net gains in jobs are also comparable to the rise in alternative work arrangements (Katz & Krueger, 2016). Some pundits have also pointed out that the new jobs to a significant part have gone to non-whites, which is natural since minorities and immigrants mostly live in urban areas (Porter, 2016; Isaac, M., 2017). Uber mirrors what’s going on in the American labor market. So far, ridesharing is a city phenomenon, see figure 12.

Source: Hall and Krueger (2016)

Figure 12 Where Uber drivers reside geographically platform. 60 % quit within two years, and only around 30 % of the drivers state that Uber is their only source of income (Hall & Krueger, A. B., 2016). The majority are working under 15 hours a week, earning less than $300 weekly (see figure 13).

103

Source: SherpaShare

Figure 13 Uber driver's weekly earnings

So, most look at driving for Uber as a way of making some money on the side or as a bridge to fill out wages from another part-time position. Regardless; earnings matter for those that work part-time or full-time for Uber.

Decent wages matter for Uber as well. It gets easier to add service providers to the platform when the work is seen as financially worthwhile. Uber executives have, therefore, been eager to make major shout-outs on significant driver earnings. However, the media has ridiculed the company on its most outrageous wage projections (Griswold, A., 2014). Just recently, Uber had to pay $20 million to Federal Trade Commission to settle claims that it falsely advertised how much drivers would make (Carson, B., 2017).

According to U.S. Bureau of Labor Statistics, average hourly earnings is around $26. Uber drivers are in general earning well below this level. Hourly earnings are almost never crossing the $20 mark (SherpaShare, 2015; Hall & Krueger, A. B., 2016). Some reporters have pointed out that drivers income in some states is close to the minimum wage (O'Donovan & Singer-Vine, 2016).

104

Looking at it from a historical perspective, it sounds like drivers earnings have been stair-stepping down in lockstep with Uber’s price cuts. As a tradition, Uber have slashed prices on fares every January 2014, 2015 and 2016. Some of these price cuts have been steep. The one in 2014 was for instance 20 %. It does not look as the drivers have been able to make up the gap by increased capacity utilization. A majority of the drivers will refer to the fact that it is hard to make more than two trips per hour regardless of user demand (Mangalindan, J. P., 2014; Russell, J., 2015b; Bates, 2016). Furthermore, Uber have raised transaction fees in many cities. In the beginning, Uber never charged more than 20 % in commission. Now, it charges 25 % in areas where supply are sufficient (Huet, E., 2015).

However, it looks like Uber has taken a pause for the cause, and have started to realize that its limits to how low prices can go. There has, therefore, been no price cuts from Uber in the winter of 2017 (Griswold, A., 2017). When wages head so far south as today, Uber might have to consider that it is not only competing for labor with fellow transportation companies but also other low-paying occupations in fast food processing or retail. After all, these companies offer workers holidays, health benefits and unemployment insurance.

Then again, how surprising is Uber’s race to the wage bottom? - Transport have never been a high-paying line of work, so why should ridesharing be any different.

Europe

Uber launched in London in the autumn of 2012. The app has taken the city by storm, and during the second half of 2015, the number of Uber drivers surpassed the number of black cabs (Knight, 2016). Simultaneously, it was also evident that Uber provided much cheaper services than taxi operators (Telegraph Staff, 2015).

In a labor perspective, London mirrors big cities in America. Today, average driver pay is £16 an hour according to Uber, which is a bit better than the national average (Chalabi, 2013). However, many drivers beg to differ with Uber’s wage claims. According to two drivers that have sued Uber on labor issues, the net earnings are well below a minimum wage off £7.5 (Telegraph Reporters, 2016).

London drivers tell the same story as the Americans. When Uber launched everything

105 was superb, and drivers were guaranteed £25 per hour. Steadily though, after Uber pushed the market to liquidity with lower prices and more drivers, wages went on a downward trajectory. Nowadays, most drivers argue that driving for the app are tedious and unprofitable (Knight, 2016).

Western Europe tells a somewhat different story than what we can deduct from the Anglo- American experience. In Western Europe, ridesharing has never taken off to the same degree. Non-professional drivers are banned from doing cab services in the four largest Euro-countries. Thus, Uber is still operating in these countries, but with difficulty.

Paris serves as an excellent example of how the situation currently is. Uber launched in the city already in 2011. The service was fairly popular and grew healthily. In the beginning, Uber worked exclusively with professional and licensed chauffeurs. The trouble started when Uber decided to sign up cheaper unlicensed drivers. It sparked immediate protest from taxi operators and discontent from regulatory authorities. In 2015, the government banned Uber from using unlicensed drivers (Lomas, 2014).

Recently, the French government has also threatened to restrict the number of professional licenses, which up to now have been relatively easy to attain. However, this action is widely unpopular among present and aspiring Uber drivers. Uber is a popular way to earn income among immigrant youths in the Parisian suburbs. Drivers are on average earning €20 per hour, and Uber creates a low-entry job-opportunity for many in this group (Chassany, 2016).

On the flip side, we can argue that restrictive French regulation is a blessing in disguise for those that have established themselves on the platform. In stopping ridesharing from taking off; Uber have not been able to push the market to saturation, and this have enabled earnings to hold up.

Australia

Uber operates more freely in Australia than in Europe. The last Australian report about personal travel habits shows that ridesharing has contributed to an expanded market of vehicle-for-hire transportation (Visentin, 2017). So, taxi operators have not seen their market shrink dramatically. The Australian experience is how everyone

106 likes the market to work. Australia is about to break the record for consecutive 27 years of economic growth, and it is not surprising that Australians can afford to complement their use of public transport and private car use with various kinds of taxi services. It is an expression of a higher living standard. Naturally, Uber and other ridesharing services are still expanding in the country, and the recruitment of new drivers and the wage level seem to be unproblematic. Ridesharing is also legal in almost all Australian regions. Australia is, therefore, something of an outlier in the OECD area.

4.5.2 Drivers in the developing world

It is hard to cover all emerging markets in just a few pages, but we can make a sampling of cities to see whether there are some similarities. Uber started its expansion into emerging markets from the fall of 2013 and onwards.

Sub-Saharan Africa

Lagos is maybe the most interesting city in Africa from a ridesharing growth perspective. Officially, the Nigerian business capital has the most congested traffic in the world, and the population grows at a breakneck pace (Taylor, 2012). Uber premiered its service mid-summer 2014, and 2 000 drivers are already working regularly on the app (Onu & Alake, 2016). In Nigeria, Uber is considered a high end product. The chauffeurs cannot drive vehicles older than a 2006 model. Hyundai Accent, Toyota Corolla, and Kia Rio are popular cars among Uber drivers (Agunbiade, 2016). Even though Uber drivers use nice cars by Nigerian standards, the fares are 25% cheaper overall than a ride with a ramshackle car from one of the taxi operators. So, it is the smartphone usage which contributes to the luxury status, and not the prices. As smartphone usage increases, the bigger Uber’s potential market becomes (Adeleke, 2016).

Uber’s vehicle requirements make it difficult for aspiring drivers to qualify. The local Uber management is also putting quite a lot of effort into training and instructing the drivers it recruits. Due to the quality standards, most drivers are leasing the cars they use from an intermediary situated between them and Uber. Just recently, Uber have

107 arranged deals with car dealerships and financial institutions to make cars more affordable for their drivers (Onu & Alake, 2016).

In general, Uber seem to be something which many Nigerian wants to join. Normally, a Nigerian Uber driver makes good middle class wages, which is about $320 monthly (Jari, 2016; Adewunmi, 2011). Leasing and operational costs eat into this revenue, but still, wages are decent by Nigerian standards.

Drivers face an entirely different situation in Nairobi. The Kenyan capital is called the Silicon Savannah due to its vibrant startup scene and a large number of tech companies. In total, twelve ridesharing startups are active in the city (Bright, j., 2016b). Uber launched in January 2015 and was an instant success in the city. The number of drivers available on the app has already reached a thousand drivers (Finnan, 2016). However, due to the high number of startups, the price competition is severe. Little Cab; Uber’s biggest Kenyan competitor, is just as, if not more, successful (Wesangula, 2016). Mid-summer 2016 Uber slashed prices with 35% to be relevant in the competitive environment (Wahito, 2016).

The price cuts did not go down well with drivers. In fact, the price competition makes it hard to earn a living for drivers on all platforms. What happened next, though, is interesting from a labor perspective. As a response to lower wages, Kenyan ridesharing drivers have organized themselves in the Kenya Digital Taxi Association. In the fall of 2016, the members of the union went on strike (Kuo, 2016).

Uber did not budge, and in the end, the strike ended with no gains for the drivers. In the following months, the situation seemed to calm down as the market went into an unexpected expansive phase (Atolagbe, 2016). In March 2017, the conflict between platform owners and drivers were on again. Drivers were now complaining that they did not get rising fuel costs reimbursed from the app companies. The strike seems more organized than the one last year and Uber have been forced to go into collective bargaining with the union. Further; the government participates as an arbiter between the parties (Mungai, 2017).

If we travel further down south from Nairobi, we eventually come to Johannesburg, where Uber started its African expansion in August 2013. Today, Johannesburg is a well functional market from an Uber point of view. Uber have logged over 1 million

108 trips in the city so far. The Uber app offers several types of vehicles for hire: luxurious cars, ordinary sedans, and vans. UberEATS is also up and running.

Uber sees no shortages of prospective recruits joining even though the market is fairly saturated, In fact, Uber have a waiting list for drivers who want to join (Rawlins, 2016). This waiting list has emerged despite the fact that many of the drivers complain about salaries (De Greef, 2016). According to Uber, a driver can expect to earn between 1200 - 2800 US dollars each month, which is above average in the country The problem for drivers in Johannesburg seems to be the same as in Lagos and Nairobi; they do not have control over the means of production. Thus, most drivers are leasing the car from an intermediary, and have to split the gross earnings between Uber, gas, and the car-owner (Van Zyl, 2017).

Uber is aware of the problem and tries to mitigate the situation by making beneficial car lease agreements available for its top service chauffeurs. Going forward, it also wishes to prioritize car owners in the recruitment process (Rawlins, 2016).

India

Uber launched its services in Bangalore India at about the same time as it opened in Johannesburg (Prithivi, 2013). Uber’s growth is on a tear in the country. It operates in 30 cities and has around 400 000 drivers on the platform, as of 2017. Uber plans to have 1 million drivers by 2018 (Rai, 2016).

From Uber’s point of view, the Indian market is not quite as open as in Africa. Every vehicle used by Uber must have been approved by the state for commercial purposes and carry so called commercial license plates. Uber have, therefore, incentivized owners of commercial vehicles to attach their cars to the Uber platform by advertising good earnings prospects and offering attractive sign-on bonuses. On the looks of it, to own a fleet of cars under Uber auspices was tremendously profitable at the start, but the profitability for fleet owners decreased dramatically during 2015 (Roushan Ali, 2015).

From a labor perspective, the Indian story resembles what happens in Sub-Saharan Africa. There is a huge difference between what drivers can expect to earn based on whether they are car owners or not. It looks like hired drivers work after an incentive

109 scheme offered by the car owners, and that the owner and the driver split the net profit 50/50 (Pradhan, 2016). According to Uber, Indian drivers earn 20 - 40 US dollars on a daily basis (Jain, A., 2017). Thus, for drivers that lease their cars, we must expect that earnings are in the lower band of Uber’s calculus, which is about average for Indian urbanites (Singh, 2017).

As Uber and Ola have added cars to their platform, the markets have reached liquidity in many cities, which then again have made it possible for Uber and other app companies to cut incentives offered to car owners and drivers. Naturally, service providers on the platforms are alarmed about this development. Many drivers feel duped by the app companies. They complain that wages are far from what the advertising says. Today, many drivers are frustrated and angry, which, in turn, has led to spontaneous strikes in several Indian cities. Indian drivers demand that app companies reduce the fees associated with transacting on the platforms (Ram & Kazmin, 2017; Singh, 2017).

The division between car owners and workers is a problem for how Uber’s business model works in India. It is simply too many entities which have to split the profits in a low margin business, and this contributes to labor tensions. To counter this, Uber have turned to the same solution as in South-Africa, and have started its own leasing business of commercially approved vehicles together with an Indian financial company (Verma, 2015). By cutting out the outside intermediary, Uber think it can offer drivers a better deal overall. Another factor is also important; when drivers are pinned down in leasing agreements, Uber gains greater control of its drivers. This control can be crucial if it wants to avoid violent strikes and other labor induced disruptions to its business (Ram & Kazmin, 2017).

Maybe, the situation gets a bit different if we look at an economy that is considered more advanced but still seen as an emerging market.

Mexico City

Mexico City is a tale of two cities for Uber drivers; it can be the worst of times, it can be the best of times. Uber moved into Mexico City in 2013. Currently, it has around 50 000 drivers in the city (Newcomer, E., 2016c). A good thing about Mexico City is

110 that City authorities are positive to Uber and have legalized the service (De Haldevang, 2015). Uber chauffeurs can, therefore, operate without having to fear any repercussions from the police. The bad thing is that taxi operators are organized and militant. Uber drivers are, therefore, subject to violent attacks by taxi chauffeurs (Babb, 2015).

In general, Uber drivers have nicer cars than what Mexican taxi operators use. Uber do not compete on price but are considered more reliable and safe for users (Associated Press, 2015). When Uber opened in the city, drivers could expect to make a decent living, beating out middle class wages by a healthy margin (Schwietert Collazo, 2014). Recently, it looks like earnings have come down a bit, but drivers can still expect to earn middle class wages (Garcia Soriano, 2016).

In contrast to Africa and India, Most Mexican drivers are vehicle owners. Ownership makes driving on the Uber platform more profitable. Aside from this, the relationship between drivers and Uber is not without skirmishes. In 2016, Uber drivers went on strike protesting the introduction of UberPOOL due to its lower price. Drivers also wanted to limit the numbers of chauffeurs available on the platform (Mosquera, 2016).

4.5.3 When the boss is an algorithm

Uber executives admit that the magic of their math is counterintuitive. How can drivers come out on top when fares constantly get cut? The magic resides on the win- win logic of network effects. Users get charged less, but drivers earn more because they get more to do (Knight, 2016).

However, the magic does not seem to work for drivers. Uber’s dynamic theories run up against simple arithmetic’s. App driving is not yet a seamless conveyor belt of continuous pick-ups and drop-offs. Drivers still need to wait a while for new trip requests and drive to reach the new passenger's pick-up point.

In any case, the magic work for Uber. Pay more, recruit more. Pay less, recruit more. Anything works, and this speaks volumes about the recruitment sweet spot which Uber experiences. The ‘Brazilianization of the West’ and the informalism of the

111 developing world means that many workers have few alternatives. Recruitment shows, therefore, no signs of abating.

Uber’s labor administration is a form of Janus-faced algorithmic management. On the left side of the face; Uber promotes the drivers as freelancers who can do as they wish. On the right side of the face; strong control is exerted through various algorithmic manifestations. Algorithms determine everything that goes on: prices, payments, where to drive, and how to drive through the rating system. Furthermore, Uber uses psychological tricks to keep the drivers involved: gamifying the driver experience, sending drivers virtually on to the next pick-up before the current trip is over, incentivize the drivers through encouraging messages and so on (Scheiber, 2017). This Janus-faced dynamic tilts the power relationship to Uber’s advantage.

In the West, when wages head south, drivers end up in precarious situations. ‘To hustle’ for Uber, became a term among US drivers whenever the app was in trouble against regulatory bodies and competitors. The worker's mobilization in defense of the app served to legitimize Uber in parts of the opinion. This relationship seems more fragile as of now. It is harder for the drivers to see that it has common interests with the Uber headquarters when wages are lousy.

In the developing world, app driving entails less of a downward spiral in work conditions. Precarious work is something which people have grown accustomed too. Obviously, drivers want to earn decently here as well, but drivers seem to have no illusions about the benevolence of the digital platforms. We have seen that drivers in Kenya and India have organized to protect their rights. Due to the social circumstances in many of these countries; we cannot exclusively judge Uber as a force which pushes informalism. Uber is also creating job opportunities and a chance to unionize. When Uber reaches a level where it has become a large employer, drivers know that they may find colleagues in online driver forums and Facebook groups. Through these fora, drivers can plan collective actions and so forth.

The schism between emerging markets and the developed world makes it hard to construct a simple take on how to view Uber as a force in the labor market.

112

5 Conclusion

Marquee users played a major role in Uber’s successful launch. This fact is nothing new according to earlier research. However, in Uber’s case, they did more than what is usual, namely that they also contributed with money. This thesis is, therefore, interpreting the successful launch of the Uber app as a process undertaken by a corraborative network. In the start, the central players in the network were the founders, a startup crew of fewer than ten people, and the angel investing marquee users. After a while, a growing crowd of enthusiastic drivers was brought in on the action, see figure14.

Figure 14 The successful launch of the Uber platform

This study has not focused much on the superiority of the technology as a success factor, even though it obviously played a humongous role. The reason for this is that it is covered so well in other research (see introduction). In figure 15 we cite the most crucial parts which over the years made the app popular with users.

 A user can order a ride with the tap of an app

 The fare is estimated beforehand

 Multiple choices in services and price

 The user can track the driver as he arrives at his location

 Transparent and flexible payment options

 Priced below taxi fares, and less wait time

Figure 15 List of user friendly features

113

We have seen that platform economics played a significant role in the scaling of the Uber platform. Uber get full score on all the metrics: quality, subsidizing and an opening up of the platform to outside developers. Figure 15 shows that platform economics have played a role for the scaling of other ridesharing companies as well. However, Uber and Didi extinguish itself by having used all techniques to the fullest.

Companies Quality Subsidizing Open platform Uber High High Yes (2014) Didi High High Yes (2016) Lyft Low Medium Yes (2016) Ola Medium Medium Yes (2015) Grab Medium Low No

Figure 16 Platform economics in the ridesharing space

Platform economics seem, therefore, to fit well with the understanding of how ridesharing platforms scale. However, when we studied the exponential scaling in detail, we became aware of facts that revealed that it was more than economics going on. Figure 16 give us a picture of all the factors that went into Uber’s exponential scaling. There is no point revisit them all in detail. At this stage, it is vital to acknowledge that the theory of generative digital infrastructures come in as a handy framework. Organizational decentralization and a modular technological architecture were an essential generative contextual condition. Micro services as an innovation mechanism enabled individual teams to adapt existing services to local requirements. In essence, the team of teams organizational structure created a vast collaborative network, where local teams got the backing they needed from a large corporation. We could see that the team of teams organization begun to yield results when innovations started to take place locally. UberIndia stands out as a salient example by its championing of cash payments and carpooling. The India headquarters has also launched services which it is the sole originator and user of so far, such as UberFLEET and UberHIRE. At the end of the day, the innovation mechanism provided by micro services and the team of teams structure became a “self-reinforcing

114 process by which new products and services are created as infrastructure malleability spawns recombination of resources” (Henfridsson, O. & Bygstad, 2013, p. 918).

Uber’s scaling canvas The main Scalable Market Subsidizing Design Technological partners business segmentation improvements improvements model

Investors The UberBLACK Lower prices A capacity to The ETA launch UBERSUV rebrand itself algorithm Drivers Playbook UberTAXI More UberX… demand Device The heat map (The compatibility algorithm replicative ‘Everyone’s Higher pattern private driver.' density/ Payment The surge mechanism) Capacity flexibility pricing Utilization algorithm Micro More drivers services/ Decentralized Lower prices structure Spawns: Team of User choice Lower Team of Enable more teams ETA teams/ efficient Enable local services, the adaptation and goal: no more innovations than 5. Min. wait time.

Figure 17 Uber's scaling canvas The generative replicative pattern mechanism was also vital. The launch playbook created a scalable business model. It was also responsible for creating the organizational backbone for the teams of teams structure. Uber headquarters was instrumental in establishing local offices and assisted them in launching the app, but afterwards, it more or less handed over the keys to the local team. This decentralization was necessary because the Uber headquarters have no chance of controlling all the 576 cities in Ubers current portfolio.

115

The augmentation obstacles illustrate the saliency of viewing digital infrastructures as socio-technical constructs. There are unfriendly stakeholders in the physical realm which create the most severe barriers to further growth. It is part of why this thesis is called an Uber-assessment of bits and atoms. The atoms are by Kalanick’s words the people that make it happen, or as in our case, those that try to prevent it from happening. The idea about digital infrastructures as complex adaptive systems is pertinent. Uber is an enterprise that has to navigate through a vast web of friendly and unfriendly stakeholders. However, Uber adapt and learn as it goes on. Somehow, it muddles through by the help of partners within its user network. In short, this study discerns these partnerships as various task networks which rush in to defend the system. Sifting through the weeds of all the different socio-technical issues, we found that the replicative pattern mechanism is not only about scaling. The mechanism was also active as an instrument in dealing with regulators and competitors. We can, therefore, construe it as a consolidation tool as well. Task networks were active in carrying out these consolidation efforts. In total, we get a picture as described below in figure 18.

Figure 18 Resilience in Uber's network

In total, the generative replicative pattern mechanism is very close to something which Henfridsson, O. and Bygstad (2013, p. 918) would call for an adoption mechanism. Here it is both a scaling mechanism and an adoption mechanism, which consists of making investments in new markets and cultivating the existent user network. We can, therefore, draw the same conclusion. The replicative pattern mechanism is a “self-reinforcing process by which more users adopt the infrastructure as more resources invested increase the usefulness of the infrastructure.”

116

Uber’s augmentation strategy is mainly about seeking to partner-up with outsiders. The growing cooperation between Uber and other entities are a result of three different causes. First of all, Uber need partners if it is going to succeed in becoming a logistics company. Secondly, Uber’s growing user base makes it an attractive partner for other entities that want access to its user network. Thirdly, the emergence of autonomous cars shake up the transportation business and make enterprises within the sector scramble for new positions. Partnerships are a method to preserve and enhance its standing in anticipation of the new disruption.

There are feedback loops and self-reinforcing processes at play here. As Uber grows, its influence increases both politically and economically, and as its influence grows, the more attractive it gets for partners, and as more partners coalesce around Uber, the more its influence and power increases, which, in turn, make it even more attractive as a partner. In total, we got a picture where Uber serves the role as an attractor which defends and develops a transportation network in upheaval due to the emergence of new business practices (see figure 19).

Overall, the network of networks mechanism seems to fit well as a backdrop for how Uber try to surpass obstacles and bolster its business prospects. Uber have created various innovation networks which is working side by side to generate new ideas, processes, and products. These innovative task networks are increasingly consisting of cooperation with outside partners.

Figure 19 Uber as an attractor

117

In total, the generative network of networks mechanism is something which Henfridsson and Bygstad (2013, p. 918) would call for a scaling mechanism. Here it is primarily an augmentation mechanism which also is about scaling in a wider sense. We can, therefore, draw the same conclusion. The network of networks mechanism is a “self-reinforcing process by which an infrastructure expands its reach as it attracts new partners by creating incentives for collaboration.”

This thesis put forward Brazilianization as an adequate lense to view the recruitment of Uber drvers. Beck’s call that Western labor markets would start to resemble third- world precariousness has become a reality. This precariousness play to Uber’s advantage. There is no shortage of drivers even though wages have headed south. Right now, Uber is growing at a rapid pace in emerging markets. The high number of informal work arrangements in the third world make these countries fertile recruitment grounds for Uber. Surprisingly, anecdotes in this study suggest that drivers in less developed countries have been more prone to strike and make collective demands against Uber. In the West, labor protests among Uber drivers have primarily expressed itself through the court system.

118

Bibliography

Abkowitz, A., & Carew, R. (2016, August 1). Uber Sells China Operations to Didi Chuxing. The Wall Street Journal. Retrieved from http://www.wsj.com/articles/china-s-didi-chuxing-to-acquire-rival-uber-s- chinese-operations-1470024403

Adeleke, D. (2016, June 8). Uber, the sharing economy, and the future of transportation in Lagos. TechPoint. Retrieved from https://techpoint.ng/2016/06/08/uber-sharing-economy-future-transportation- lagos/

Adewunmi, F. (2011, October 2). Nigeria’s middle-class: How we live, and what we want from life. How We Made It In Africa. Retrieved from https://www.howwemadeitinafrica.com/nigerias-middle-class-how-we-live- and-what-we-want-from-life/

Agrawal, R. (2012, Janaury 3). Car service Uber passes economics test, fails marketing. Venture Beat. Retrieved from http://venturebeat.com/2012/01/03/car-service-uber-passes-economics-test- fails-marketing/

Agunbiade, T. (2016, August 8). 12 things you need to know about driving an Uber in Nigeria. TechCabal. Retrieved from http://techcabal.com/2016/08/08/12- things-you-need-to-know-about-driving-an-uber-in-nigeria/

Akshay. (2014, Novemebr 12). New payment options make Uber available to 600 million Indians.[Uber Newsroom]. Retrieved from https://newsroom.uber.com/india/new-payment-options-make-uber-available- to-600-million-indians/

Alba, D. (2015, June 30). The Chinese Ride-hailing Startup that's out-ubering Uber. Wired. Retrieved from https://www.wired.com/2015/06/the-chinese-ridehailing-startup-thats-out-ubering-uber/

Alley, J. K. (2016). The Impact of Uber Technologies on the New York City Transportation Industry. The university of Arkansas.

Anand, P. (2017, February 24). Uber Women To CEO Travis Kalanick: We Have A Systemic Problem. BuzzFeed. Retrieved from https://www.buzzfeed.com/priya/uber-women-to-ceo-travis-kalanick-we-have- a-systemic-problem?utm_term=.rhnoGROKWe#.mjOqGOBEox

Anthony, A. (2014, December 21). Travis Kalanick: Uber-capitalist who wants to have the world in the back of his cabs. Guardian. Retrieved from

119

https://www.theguardian.com/theobserver/2014/dec/21/travis-kalanick-ubercab-app-observer-proifile

Arrington, M. (2010a, August 31). What If UberCab Pulls An Airbnb? Taxi Business Could (Finally) Get Some Disruption. TechCrunch. Retrieved from https://techcrunch.com/2010/08/31/what-if-ubercab-pulls-an-airbnb-taxi- business-could-finally-get-some-disruption/

Arrington, M. (2010b, October 15). UberCab Closes Uber Angel Round. TechCunch. Retrieved from https://techcrunch.com/2010/10/15/ubercab-closes-uber-angel- round/

Arrington, M. (2011, February 4). Huge Vote Of Confidence: Uber Raises $11 Million From Benchmark Capital. TechCrunch. Retrieved from https://techcrunch.com/2011/02/14/huge-vote-of-confidence-uber-raises-11- million-from-benchmark-capital/

Associated Press. (2015, July 30). Uber drivers attacked outside Mexico City airport as taxi drivers demonstrate. Guardian. Retrieved from https://www.theguardian.com/technology/2015/jul/30/uber-drivers-attacked- outside-mexico-city-airport-as-taxi-drivers-demonstrate

Atolagbe, M. (2016, November 15). According to Uber, its drivers in Nairobi are making more money following the price cut. TechCabal. Retrieved from http://techcabal.com/2016/11/15/according-to-uber-its-drivers-in-nairobi-are- making-more-money-following-the-price-cut/

Babb, F. (2015, July 30). Uber protests in Mexico turn violent. Venture Beat. Retrieved from http://venturebeat.com/2015/07/30/uber-protests-in-mexico- turn-violent/

Baker, D. (2014, May 27). Don’t buy the “sharing economy” hype: Airbnb and Uber are facilitating rip-offs. Guardian. Retrieved from http://www.theguardian.com/commentisfree/2014/may/27/airbnb-uber-taxes- regulation (Visited Jan. 04, 2017)

Baldwin, C., & Woodard, J. (2009). Platforms, markets and innovation. In A. Gawer, ed. The Architecture of Platforms: A Unified View. Edward Elgar, Cheltenham, UK, pp. 19-44.

Barr, A. (2016, September 12). Google's Self-Driving Car Project Is Losing Out to Rivals. Bloomberg. Retrieved from https://www.bloomberg.com/news/articles/2016-09-12/google-car-project- loses-leaders-and-advantage-as-rivals-gain

120

Bates, D. (2016, February 1). Uber Cuts Prices—and Kneecaps Drivers. The Daily Beast. Retrieved from http://www.thedailybeast.com/articles/2016/02/01/uber- cuts-prices-and-kneecaps-drivers.html

Beck, U. (1992). Risk Society: Towards a New Modernity. London: Sage.

Beck, U. (2000). The Brave New World of Work. Cambridge: Cambridge University Press.

Belk, R. (2014). You are what you can access: Sharing and collaborative consumption online. Journal of Business Research, 67(8), pp. 1595-1600.

Bershidsky, L. (2016, May 25). Uber's Carpools Are No Match for Public Transit. Bloomberg. Retrieved from https://www.bloomberg.com/view/articles/2016- 05-25/uber-s-carpools-are-no-match-for-public-transit

Bhuiyan, J. (2015, February 10). Men Are Using Uber’s Lost-And-Found Feature To Harass Female Drivers. BuzzFeed. Retrieved from https://www.buzzfeed.com/johanabhuiyan/faced-with-harassment-female- uber-drivers-often-left-to-fend?utm_term=.tvpgBQ5wdl#.pxbAE3NVoY

Bhuiyan, J. (2016a, April 13). Three years ago, Uber X was born with a policy manifesto. Recode. Retrieved from http://www.recode.net/2016/4/13/11586110/three-years-ago-uber-x-was-born- with-a-policy-manifesto

Bhuiyan, J. (2016b, April 22). Uber explodes in Egypt, while EasyTaxi stutters in the Middle East. Recode. Retrieved from http://www.recode.net/2016/4/22/11586316/uber-careem-middle-east-ride-hail

Bhuiyan, J. (2016c, August 18). Uber and Volvo commit $300 million to developing autonomous cars together. Recode. Retrieved from https://www.recode.net/2016/8/18/12533058/uber-volvo-self-driving-otto- pittsburgh

Bhuiyan, J. (2016d, November 7). Even Juno — known as the driver-friendly ride- hail service — has self-driving ambitions. Recode. Retrieved from http://www.recode.net/2016/11/7/13524760/juno-self-driving-cars-drivers- uber

Bhuiyan, J. (2017a, March 7). Ola's New Service Quickly Gaining Ground on Uber India. Recode. Retrieved from https://www.recode.net/2016/3/17/11587066/olas-new-service-quickly- gaining-ground-on-uber-india

Bhuiyan, J. (2017b, March 17). Inside Uber’s self-driving car mess. Recode. Retrieved from https://www.recode.net/2017/3/24/14737438/uber-self- driving-turmoil-otto-travis-kalanick-civil-wa

121

Biggs, J. (2015, February 2). Uber Opening Robotics Research Facility In Pittsburgh To Build Self-Driving Cars. TechCrunch. Retrieved from https://techcrunch.com/2015/02/02/uber-opening-robotics-research-facility-in- pittsburgh-to-build-self-driving-cars/

Bort, J. (2015, December 3). From drug addiction to Uber intern to powerful Uber exec at age 30, Austin Geidt's life is already the stuff of Valley legend. Business Insider. Retrieved from http://www.businessinsider.com/ubers- austin-geidts-insane-life-2015-12?r=US&IR=T&IR=T

Bort, J. (2016, March 6). Uber says it's had 'fewer than 170 complaints' of sexual assaults involving rides in 3 years. Business Insider. Retrieved from http://www.businessinsider.com/uber-shares-data-on-sexual-assult- complaints-2016-3?r=US&IR=T&IR=T

Boudreau, K. (2010). Open platform strategies and innovation: Granting access vs. devolving control. Management Science, 56(10), pp. 1849-1872.

Bowe, R. (2013, August 20). Lawsuit alleges Uber unfairly withholds tips from drivers. SF Bay Guardian. Retrieved from http://uberlawsuit.com/Bay.pdf

Bower, J. L., & Christensen, C. M. (1995). Disruptive technologies: catching the wave. Harvard Business Review Video, pp. 506-20.

Bradley-Carr, P. (2012, October 31). As NY floods, "Robin Hood" Uber robs from the rich and... Nope, that's about it. Pando. Retrieved from https://pando.com/2012/10/31/assholes-shrug

Bradley-Carr, P. (2014, November 20). On stage at Goldman Sachs event, Uber CEO compared his company's woes to Ferguson. Pando. Retrieved from https://pando.com/2014/11/20/on-stage-at-goldman-sachs-event-travis- kalanick-compared-his-companys-woes-to-ferguson/

Bradshaw, T. (2014, May 9). Lunch with the FT: Travis Kalanick. Financial Times. Retrieved from https://www.ft.com/content/9b83cbe8-d5da-11e3-83b2- 00144feabdc0 (Visited January 05, 2017)

Brandom, R. (2015, August 25). Uber will partner with University of Arizona for self- driving car research. The Verge. Retrieved from http://www.theverge.com/2015/8/25/9207229/uber-university-of-arizona- tucson-autonomous-self-driving-cars

Bright, J. (2016a, June 2). As it expands in Africa, Uber adapts to local markets and adopts cash payments. TechCrunch. Retrieved from https://techcrunch.com/2016/06/02/as-it-expands-in-africa-uber-adapts-to- local-markets-and-adopts-cash-payments/

122

Bright, J. (2016b, July 31). In Kenya, Safaricom’s Little Cab app goes head to head with Uber. TechCrunch. Retrieved from https://techcrunch.com/2016/07/31/in-kenya-safaricoms-little-cab-app-goes- head-to-head-with-uber/

Brown, G. E. (2016). An Uberdilemma: Employees and Independent Contractors in the Sharing Economy. Md. L. Rev. Endnotes, 75, 15.

Brustein, J. (2016, August 15). Uber and Lyft Want to Replace Public Buses. Bloomberg. Retrieved from https://www.bloomberg.com/news/articles/2016- 08-15/uber-and-lyft-want-to-replace-public-buses

Buhr, S. (2015a, September 16). China’s Didi Kuaidi Put $100M Into Lyft, Inks Ridesharing Alliance To Rival Uber. TechCrunch. Retrieved from https://techcrunch.com/2015/09/16/ubers-rivals-didi-kuadi-and-lyft-form- international-ridesharing-partnership/

Buhr, S. (2015b, October 14). Uber Takes On Postmates With UberRUSH, An On- Demand Delivery Service. TechCrunch. Retrieved from https://techcrunch.com/2015/10/14/uber-takes-on-postmates-with-uberrush-to- deliver-all-the-retail-things-to-you/

Buhr, S., & Constine, J. (2015, December 16). You Can Now Order Ubers (And Soon Lyfts) In Facebook Messenger To Prove You’re On Your Way. TechCrunch. Retrieved from https://techcrunch.com/2015/12/16/facebook-messenger- transportation/

Callaway, A. (2016, January 1). Apploitation in a city of instaserfs: How the “sharing economy” has turned San Francisco into a dystopia for the working class. Monitor. Retrieved from https://www.policyalternatives.ca/publications/monitor/apploitation

Camp, G. (2016, September 23). How I came up with Uber - the startup mentality with Garrett Camp - Zeitgeist 2016. Youtube [Video file]. Retrieved from https://www.youtube.com/watch?v=_vxk4Z1u7RE

Cannon, S., & Summers, L. H. (2014). How Uber and the Sharing Economy Can Win Over Regulators. Harvard business review, 13.

Carlson, N. (2014, May 28). Uber Is Planning For A World Without Drivers — Just A Self-Driving Fleet. Business Insider. Retrieved from http://www.businessinsider.com/uber-is-planning-for-a-world-without-drivers- -just-a-self-driving-fleet-2014-5?r=US&IR=T

Carson, B. (2016a, January 3). Uber's CEO explains why the app's arrival estimates are 'almost always' wrong. Business Insider. Retrieved from

123

http://www.businessinsider.com/travis-kalanick-explains-why-uber-cars-take- longer-to-arrive-2016-1?r=US&IR=T&IR=T

Carson, B. (2016b, March 2). Uber's GrubHub killer is finally in the US — here's the inside story on its big bet on food. Business Insider. Retrieved from http://www.businessinsider.com/why-uber-launched-uber-eats-2016- 3?r=US&IR=T&IR=T

Carson, B. (2016c, July 7). Using tactics it once slammed, Lyft is poaching Uber drivers to build a competing luxury line. Business Insider. Retrieved from http://nordic.businessinsider.com/lyft-driver-recruiting-tactics-criticized-2016- 6?r=DE&IR=T

Carson, B. (2016d, August 18). Travis Kalanick on Uber's bet on self-driving cars: 'I can't be wrong'. Business Insider. Retrieved from http://www.businessinsider.com/travis-kalanick-interview-on-self-driving- cars-future-driver-jobs-2016-8?op=1%3fr=US&IR=T&IR=T

Carson, B. (2017, January 20). Uber to pay $20 million to FTC to settle claims that it exaggerated how much drivers would make. Business Insider. Retrieved from http://nordic.businessinsider.com/uber-to-pay-20-million-to-ftc-over-driver- earning-claims-2017-1?r=US&IR=T

Chafkin, M. (2015, September 15). What Makes Uber Run. Fast Company. Retrieved from https://www.fastcompany.com/3050250/what-makes-uber-run

Chafkin, M. (2016, August 18). Uber’s First Self-Driving Fleet Arrives in Pittsburgh This Month. Bloomberg. Retrieved from https://www.bloomberg.com/news/features/2016-08-18/uber-s-first-self- driving-fleet-arrives-in-pittsburgh-this-month-is06r7on

Chalabi, M. (2013, December 12). UK median weekly pay is £517 - but who earns that? Guardian. Retrieved from https://www.theguardian.com/news/datablog/2013/dec/12/uk-median-weekly- pay-is-517-but-who-earns-that

Chassany, A. (2016, March 7). Uber in France: A route out of the banlieues. Financial Times. Retrieved from http://macaudailytimes.com.mo/files/pdf2016/FT-2511-2016-03-07.pdf

Chen, A. (2016, January 25). Uber’s virtuous cycle. Geographic density, hyperlocal marketplaces, and why drivers are key. @andrewchen [Newsletter]. Retrieved from http://www.andrewchen.co/ubers-virtuous-cycle-5-important-reads- about-uber/

Chen, L. Y., Webb, A., & Oster, S. (2016, May 13). Apple Backs Didi With $1 Billion in Blow to Uber in China. Bloomberg. Retrieved from

124

https://www.bloomberg.com/news/articles/2016-05-13/apple-invests-1-billion- in-uber-s-china-competitor-didi

Chen, L., Mislove, A., & Wilson, C. (2015). Peeking beneath the hood of Uber. In Proceedings of the 2015 ACM Conference on Internet Measurement Conference (pp. 495-508). ACM.

Chen, M. K., & Sheldon, M. (2015). Dynamic pricing in a labor market: Surge pricing and flexible work on the Uber platform. Mimeo, UCLA.

Chesbrough, H. W. (2006). Open innovation: The new imperative for creating and profiting from technology. Harvard Business Press.

Christensen, C. M. (2013). The innovator's dilemma: when new technologies cause great firms to fail. Harvard Business Review Press.

Christensen, C. M., Raynor, M. E., & McDonald, R. (2015). What Is Disruptive Innovation? Harward Business Review.

Conger, K. (2016a, November 28). Uber begins background collection of rider location data. TechCrunch. Retrieved from https://techcrunch.com/2016/11/28/uber-background-location-data-collection/

Conger, K. (2016b, December 13). Uber snooping lawsuit exposes data and security practices. TechCrunch. Retrieved from https://techcrunch.com/2016/12/13/uber-snooping-lawsuit-exposes-data-and- security-practices/

Constantiou, I., Eaton, B., & Tuunainen, V. K. (2016). The Evolution of a Sharing Platform into a Sustainable Business. In System Sciences (HICSS), 2016 49th Hawaii International Conference on (pp. 1297-1306). IEEE.

Constine, J. (2014, August 12). Uber Is About To Launch An API. TechCrunch. Retrieved from https://techcrunch.com/2014/08/12/uber-api/

Constine, J. (2015a, June 30). Uber’s API Affiliate Program Pays Devs $5 For Each User They Refer. TechCrunch. Retrieved from https://techcrunch.com/2015/06/30/uber-pays-to-be-everywhere

Constine, J. (2015b, August 24). Uber Tests Bus-Style Discounted “Smart Routes.” TechCrunch. Retrieved from https://techcrunch.com/2015/08/24/uber-smart- routes/

Cook, J. (2015, February 11). Uber's internal charts show how its driver-rating system actually works. Business Insider. Retrieved from http://www.businessinsider.com/leaked-charts-show-how-ubers-driver-rating- system-works-2015-2?r=US&IR=T&IR=T

125

CPUC. (2013, September 19). DECISION ADOPTING RULES AND REGULATIONS TO PROTECT PUBLIC SAFETY WHILE ALLOWING NEW ENTRANTS TO THE TRANSPORTATION INDUSTRY. Retrieved from http://docs.cpuc.ca.gov/PublishedDocs/Published/G000/M077/K112/7711228 5.PDF

Cramer, J., & Krueger, A. B. (2016). Disruptive change in the taxi business: The case of Uber. The American Economic Review, 106(5), pp. 177-182.

Cusumano, M. A., & Gawer, A. (2002). The elements of platform leadership. MIT Sloan Management Review, 43(3), 51-+.

Damodoran, A. (2016, August 17). The Ride Sharing Business: Is a Bar Mitzvah moment approaching? Retrieved from http://aswathdamodaran.blogspot.no/2016/08/the-ride-sharing-business-is- bar.html

Davidov, G. (2016). The Status of Uber Drivers: A Purposive Approach. Retrieved from Spanish Labour Law and Employment Relations Journal (2017), Forthcoming; Hebrew University of Jerusalem Legal Research Paper No. 17- 7.

Davies, A. (2017, February 1). Uber's Mercedes alliance is a clever new route to self- driving dominance. Wired. Retrieved from https://www.wired.com/2017/02/ubers-mercedes-alliance-clever-new-route- self-driving-dominance/

De Greef, K. (2016, May 27). We're working long hours for lousy pay: Uber drivers. Times Live. Retrieved from http://www.timeslive.co.za/local/2016/05/27/Were-working-long-hours-for- lousy-pay-Uber-drivers

De Haldevang, M. (2015, July 15). Mexico City unveils first regulation on Uber in Latin America. Reuters. Retrieved from http://www.reuters.com/article/us- mexico-uber-idUSKCN0PP2SU20150716

De Stefano, V. (2016). Introduction: Crowdsourcing, the Gig-Economy and the Law. Retrieved from Comparative Labor Law & Policy Journal, Vol. 37, No. 3, 2016; Bocconi Legal Studies Research Paper.

DeAmicis, C. (2015a, April 16). Uber Says There Have Been Millions of Trips on UberPool, Its Carpool Option. Recode. Retrieved from https://www.recode.net/2015/4/16/11561556/uber-says-that-millions-of- people-have-taken-uberpool-its-carpool

DeAmicis, C. (2015b, October 14). Uber Expands Its Same-Day Delivery Service: 'It’s No Longer an Experiment'. Recode. Retrieved from

126

https://www.recode.net/2015/10/14/11619548/uber-gets-serious-about- delivery-its-no-longer-an-experiment

Dillahunt, T. R., & Malone, A. R. (2015). The promise of the sharing economy among disadvantaged communities. In Proceedings of the 33rd Annual ACM Conference on Human Factors in Computing Systems (pp. 2285-2294). ACM.

Dillet, R. (2014, February 28). Uber Wants To Take Over The French Market With Aggressive Bonus Tactics. TechCrunch. Retrieved from https://techcrunch.com/2014/02/28/uber-wants-to-take-over-the-french- market-by-using-dumping-tactics/

Dillet, R. (2016, August 18). Uber acquires Otto to lead Uber’s self-driving car effort. TechCrunch. Retrieved from https://techcrunch.com/2016/08/18/uber- acquires-otto-to-lead-ubers-self-driving-car-effort-report-says/

Dockterman, E. (2016, May 8). Uber and Lyft Are Leaving Austin After Losing Background Check Vote. Fortune. Retrieved from http://fortune.com/2016/05/08/uber-lyft-leaving-austin/

D'Orazio, D. (2014, Janaury 24). Uber employees spammed competing car service with fake orders. The Verge. Retrieved from http://www.theverge.com/2014/1/24/5342582/uber-employees-spammed- competing-car-service-with-fake-orders

Droege, J. (2015, October 14). UberRUSH – Now Open for Business in Chicago, NYC, and SF. [Uber Newsroom]. Retrieved from https://newsroom.uber.com/rush-open-for-business/

Dupre, E. (2016, August 29). How Uber Won The Rideshare Wars and What Comes Next. DMN. Retrieved from http://www.dmnews.com/marketing- strategy/how-uber-won-the-rideshare-wars-and-what-comes- next/article/518825/

Dwozkin, E., & Siddiqui, F. (2017, January 8). Uber is finally releasing a data trove that officials say will make driving better for everyone. The Washington Post. Retrieved from https://www.washingtonpost.com/news/the- switch/wp/2017/01/08/uber/?utm_term

Eaton, B., Elaluf-Calderwood, S., Sorensen, C., & Yoo, Y. (2015). Distributed tuning of boundary resources: the case of Apple's iOS service system. Mis Quarterly, 39(1), pp. 217-243.

Eisenbrey, R., & Mishel, L. (2016). Uber business model does not justify a new ‘independent worker’category. Employment Policy Institute.

Eisenmann, T. R., Parker, G., & Van Alstyne, M. W. (2006). Strategies for two-sided markets. Harvard business review, 84(10), 92.

127

Eisenmann, T. R. (2007). Managing Proprietary and Shared Platforms: A Life-Cycle View. Harvard Business School.

Eisenmann, T. R., Parker, G., & Van Alstyne, M. W. (2008). Opening Platforms: How, When and Why? This paper has been published under the same title as Chapter 6 in Platforms, Markets & Innovation (ed. Gawer, 2009) pp 131-162; Harvard Business School Entrepreneurial Management Working Paper No. 09-030.

Elbanna, A., & Newman, M. (2016). Disrupt the Disruptor: Rethinking'Disruption'in Digital Innovation. MCIS 2016 Proceedings.

Eldon, E. (2012, January 14). The DC Taxi Commissioner’s Attacks On Uber Have Gotten Even More Ridiculous. TechCrunch. Retrieved from https://techcrunch.com/2012/01/14/ubercommissioner/

Elias, J. (2014, August 4). These Delivery Startups Are Racing To Build The Best ETA Engine On The Block. Fast Company. Retrieved from https://www.fastcompany.com/3033913/these-delivery-startups-are-racing-to- build-the-best-eta-engine-on-the-block?show_rev_content

Elliott, R. E. (2015). Sharing App or Regulation Hackney: Defining Uber Technologies. Inc. J. Corp. L., 41, 727.

EPIC Complaint. (2015, June 22). In the Matter of Uber Technologies, Inc. Federal Trade commission. Retrieved from https://epic.org/privacy/internet/ftc/uber/Complaint.pdf

ERTRAC. (2015, June 21). Automated Driving Roadmap. Retrieved from http://www.ertrac.org/uploads/documentsearch/id38/ERTRAC_Automated- Driving-2015.pdf

Etherington, D. (2014, May 6). Google Maps On Mobile Gets Uber Integration And More. TechCrunch. Retrieved from https://techcrunch.com/2014/05/06/google-maps-on-mobile-gets-uber- integration-and-more/

Etherington, D. (2017, March 21). Uber details early steps to change culture, and ongoing COO search. TechCrunch. Retrieved from https://techcrunch.com/2017/03/21/uber-details-early-steps-to-change-culture- and-ongoing-coo-search/

European Political Strategy Centre. (2016, June 10). The Future of Work: Skills and Resilience for a World of Change. Issue 10. Retrieved from https://ec.europa.eu/epsc/publications/strategic-notes/future-work_en

Evans, D. S., & Schmalensee, R. (2016). Matchmakers: The New Economics of Multisided Platforms. Harvard Business Review Press.

128

Fagnant, D. J., Kockelman, K. M., & Bansal, P. (2015). Operations of shared autonomous vehicle Fleet for Austin, Texas, market. Transportation Research Record: Journal of the Transportation Research Board, (2536), pp. 98-106.

Featherman, M. S., & Pavlou, P. A. (2003). Predicting e-services adoption: a perceived risk facets perspective. International journal of human-computer studies, 59(4), pp. 451-474.

Feeney, M. (2015). Is Ridesharing Safe? Cato Policy Analysis, 767, 2.

Finnan, D. (2016, April 14). Cash is king for Uber’s expansion into Kenya. RFI. Retrieved from http://en.rfi.fr/africa/20160414-cash-uber-taxi-kenya-nairobi- africa

Fisher, J. (2013, September 20). The science that drives Uber. Atlassian Blogs [blog]. Retrieved from http://blogs.atlassian.com/2013/09/the-science-that-drives- uber/?utm_source=twitter&utm_medium=blogs&utm_campaign=uber

Flyvbjerg, B. (2006). Five misunderstandings about case-study research. Qualitative inquiry, 12(2), pp. 219-245.

Fowler, S. J. (2017, February 19). Reflecting on one very, very strange year at Uber. Retrieved from https://www.susanjfowler.com/blog/2017/2/19/reflecting-on- one-very-strange-year-at-uber

Gallagher, B. (2012, September 12). CEO Travis Kalanick Says UberX Numbers Are “Probably At Or Above” Lyft. TechCrunch. Retrieved from https://techcrunch.com/2012/09/12/ceo-travis-kalanick-says-uberx-numbers- are-probably-at-or-above-lyft/

Gao, J., Li, D., & Havlin, S. (2014). From a single network to a network of networks. National Science Review, 1(3), pp. 346-356.

Garcia Soriano, J. (2016, April 13). What are the pros and cons of being a driver for Uber or Lyft in Mexico City? Quora. Retrieved from https://www.quora.com/What-are-the-pros-and-cons-of-being-a-driver-for- Uber-or-Lyft-in-Mexico-City

Gawer, A. (2014). Bridging differing perspectives on technological platforms: Toward an integrative framework. Research Policy, 43(7), pp. 1239-1249.

Gawer, A., & Cusumano, M. A. (2008). How companies become platform leaders. MIT Sloan management review, 49(2), 28.

Ge, Y., Knittel, C. R., MacKenzie, D., & Zoepf, S. (2016). Racial and gender discrimination in transportation network companies. (No. w22776). National Bureau of Economic Research.

129

Geidt, A. (2011, March 2). Do Uber drivers actually rate their passengers? Quora. Retrieved from https://www.quora.com/Do-Uber-drivers-actually-rate-their- passengers

Geradin, D. (2016, April 5). Online Intermediation Platforms and Free Trade Principles – Some Reflections on the Uber Preliminary Ruling Case . Retrieved from https://ssrn.com/abstract=2759379

Gerring, J. (2004). What is a case study and what is it good for? American political science review, 98(02), pp. 341-354.

Ghazawneh, A., & Henfridsson, O. (2013). Balancing platform control and external contribution in third‐party development: the boundary resources mode. Information Systems Journal, 23(2), pp. 173-192.

Goode, L. (2012, Janaury 2). Uber CEO Responds to New Year's Eve Complaints, Plans More Surge Pricing. AllThingsD. Retrieved from http://allthingsd.com/20120101/uber-ceo-responds-to-new-years-eve- complaints-considers-dynamic-pricing-for-weekends/

Grandoni, D. (2015, August 6). Among the States, Self-Driving Cars Have Ignited a Gold Rush. The New York Times. Retrieved from https://www.nytimes.com/2015/08/07/automobiles/self-driving-cars-ignite- gold-rush-among-states.html?_r=0

Graves, R. (2010a, May 10). Benefits to UberCab. [UberCab Blog]. Retrieved from http://ubercab-blog.tumblr.com/

Graves, R. (2010b, September 13). Uber means business. UberCab [Blog]. Retrieved from http://ubercab-blog.tumblr.com/

Graves, R. (2010c, December 22). 1 + 1 = 3. [Uber Newsroom]. Retrieved from https://newsroom.uber.com/1-1-3/

Greenwood, B. N., & Wattal, S. (2015). Show Me the Way to Go Home: An Empirical Investigation of Ride Sharing and Alcohol Related Motor Vehicle Homicide. Fox School of Business Research Paper No. 15-054.

Griswold, A. (2014, October 27). In Search of Uber’s Unicorn. Slate. Retrieved from http://www.slate.com/articles/business/moneybox/2014/10/uber_driver_salary _the_ride_sharing_company_says_its_drivers_make_great.html

Griswold, A. (2017, January 19). Dry January: Uber has quietly abandoned its sweeping January price cuts. Quartz. Retrieved from https://qz.com/885455/uber-quietly-ends-january-price-cuts/

Gurley, B. (2014a, March 11). A Deeper Look at Uber’s Dynamic Pricing Model. Above the Crowd [Blog]. Retrieved from

130

http://abovethecrowd.com/2014/03/11/a-deeper-look-at-ubers-dynamic- pricing-model/

Gurley, B. (2014b, July 11). How to Miss By a Mile: An Alternative Look at Uber’s Potential Market Size. Above the Crowd [Blog]. Retrieved from http://abovethecrowd.com/2014/07/11/how-to-miss-by-a-mile-an-alternative- look-at-ubers-potential-market-size/

Ha, A. (2010, October 26). Uber(Cab) CEO: No plans to shut down despite legal spat. Venture Beat. Retrieved from http://venturebeat.com/2010/10/26/ubercab- ryan-graves-cease-and-desit/

Hagiu, A. (2014). Strategic decisions for multisided platforms. MIT Sloan Management Review, 55(2), 71.

Hagiu, A., & Biederman, R. (2015). Companies Need an Option Between Contractor and Employee. Harvard Business Review.

Hagiu, A., & Rothman, S. (2016). Network effects aren’t enough. Harvard business review, 94(4), 17.

Hall, J. V., & Krueger, A. B. (2016). An analysis of the labor market for Uber’s driver-partners in the United States. (No. w22843). National Bureau of Economic Researc.

Hanseth, O., & Lyytinen, K. (2010). Design theory for dynamic complexity in information infrastructures: the case of building internet. Journal of Information Technology, 25(1), pp. 1-19.

Harding, S., Kandlikar, M., & Gulati, S. (2016). Taxi apps, regulation, and the market for taxi journeys. Transportation Research Part A: Policy and Practice, 88, pp. 15-25.

Harris, S. D., & Krueger, A. B. (2015). A Proposal for Modernizing Labor Laws for Twenty-First-Century Work: The “Independent Worker”. The Hamilton Project, Discussion paper, 10.

Hawkins, A. J. (2016a, March 23). Uber accuses Indian rival Ola of booking 400,000 fake rides. The Verge. Retrieved from http://www.theverge.com/2016/3/23/11292310/uber-ola-lawsuit-india-fake- rides-lyft

Hawkins, A. J. (2016b, April 26). Google, Ford, and Uber just created a giant lobbying group for self-driving cars. The Verge. Retrieved from http://www.theverge.com/2016/4/26/11510076/self-driving-coalition-ford- google-uber-lyft-volvo-nhtsa

131

Hawkins, A. J. (2016c, August 31). You can now use pretax dollars to pay for UberPool, which is huge. The Verge. Retrieved from http://www.theverge.com/2016/8/31/12727908/uber-pool-wage-works-pretax- dollars-transit-benefits

Hawkins, A. J. (2016d, October 27). Surge pricing is coming to UberEats in select cities. The Verge. Retrieved from http://www.theverge.com/2016/10/27/13436334/uber-eats-surge-pricing- phoenix-houston-dallas-dc-miami-atlanta

Hawkins, A. J. (2016e, December 16). Uber is stubbornly refusing to apply for a $150 permit for its self-driving cars. The Verge. Retrieved from http://www.theverge.com/2016/12/16/13990318/uber-refusing-permit-self- driving-car-california-dmv

Hawkins, A. J. (2017a, January 30). Lyft surpasses Uber in app downloads for the first time ever. The Verge. Retrieved from http://www.theverge.com/2017/1/30/14443560/lyft-surpass-uber-app- downloads-deleteube

Hawkins, A. J. (2017b, February 6). Uber sues Seattle over law allowing drivers to unionize. The Verge. Retrieved from http://www.theverge.com/2017/2/6/14524792/uber-lawsuit-seattle-law- drivers-unionize

Hawkins, A. J. (2017c, March 6). Can Uber be saved from itself? Verge. Retrieved from http://www.theverge.com/2017/3/6/14791080/uber-sexism-scandal- strike-waymo-lawsuit-travis-kalanick

Hawkins, A. J. (2017d, March 10). Alphabet’s Waymo asks judge to block Uber from using self-driving car secrets. The Verge. Retrieved from http://www.theverge.com/2017/3/10/14884666/alphabet-waymo-google-uberlawsuit-testimony-block

Hawkins, A. J. (2017e, March 29). Lyft Shuttle mimics mass transit with fixed routes and fares. The Verge. Retrieved from http://www.theverge.com/2017/3/29/15111492/lyft-shuttle-fixed-route-fare-sf- chicago

Helderman, R. S. (2014, December 13). Uber pressures regulators by mobilizing riders and hiring vast lobbying network. Washington Post. Retrieved from https://www.washingtonpost.com/politics/uber-pressures-regulators-by- mobilizing-riders-and-hiring-vast-lobbying-network/2014/12/13/3f4395c6- 7f2a-11e4-9f38-95a187e4c1f7_story.html

132

Hempel, J. (2015, September 12). The UberEATS standalone app has nothing to do with rides. Wired. Retrieved from https://www.wired.com/2015/12/ubereats- is-ubers-first-app-thats-not-about-rides/

Henfridsson, O., & Bygstad, B. (2013). The Generative Mechanisms of Digital Infrastructure Evolution. Mis Quarterly, 37(3), pp. 907-931.

Hern, A. (2014, November 19). Uber investigates top executive after journalist's privacy was breached. Guardian. Retrieved from https://www.theguardian.com/technology/2014/nov/19/uber-investigates-top- executive-after-journalists-privacy-was-breached

Hill, K. (2014, October 3). 'God View': Uber Allegedly Stalked Users For Party- Goers' Viewing Pleasure. Forbes. Retrieved from https://www.forbes.com/sites/kashmirhill/2014/10/03/god-view-uber- allegedly-stalked-users-for-party-goers-viewing-pleasure/2/#27c

Hill, S. (2015). Raw Deal: How the "Uber Economy" and Runaway Capitalism Are Screwing American Workers. Macmillan.

Hofstede, G. H., & Hofstede, G. (2001). Culture's consequences: Comparing values, behaviors, institutions and organizations across nations. Sage.

Holland, J. H. (2006). Studying complex adaptive systems. Journal of Systems Science and Complexity, 19(1), pp. 1-8.

Holloway, C. (2015). Uber Unsettled: How Existing Taxicab Regulations Fail to address Transportation Network Companies and Why Local Regulators Should Embrace Uber, Lyft, and Comparable Innovators. Wake Forest J. Bus. & Intell. Prop. L., 16, 20.

Holmes, D. (2012, August 11). Stop Hailing A Taxi And Come Ride With Uber CEO Travis Kalanick. Fast Company. Retrieved from https://www.fastcompany.com/1839520/stop-hailing-taxi-and-come-ride-uber- ceo-travis-kalanick

Holson, L. M. (2014, November 21). To Delete or Not to Delete: That’s the Uber Question. The New York Times. Retrieved from https://www.nytimes.com/2014/11/23/fashion/uber-delete-emil-michael- scandal.html?_r=0

Howard, A. (2014, December 14). The sharing economy: Will self-regulation by startups suffice to protect consumers?. TechRepublic. Retrieved from http://www.techrepublic.com/article/the-sharing-economy-will-self- regulation-by-startups-suffice-to-protect-consumers/

Huet, E. (2014, December 17). Uber Says It's Doing 1 Million Rides Per Day, 140 Million In Last Year. Forbes. Retrieved from

133

http://www.forbes.com/sites/ellenhuet/2014/12/17/uber-says-its-doing-1- million-rides-per-day-140-million-in-last-year/#739800a87a68

Huet, E. (2015, September 11). Uber Raises UberX Commission To 25 Percent In Five More Markets. Forbes. Retrieved from https://www.forbes.com/sites/ellenhuet/2015/09/11/uber-raises-uberx- commission-to-25-percent-in-five-more-markets/#3e5c745e3214

Iansiti, M., & Levien, R. (2004). The keystone advantage: what the new dynamics of business ecosystems mean for strategy, innovation, and sustainability. Boston, Mass.: Harvard Business School Press.

Isaac, E. (2014). Disruptive Innovation: Risk-Shifting and Precarity in the Age of Uber. Berkeley Roundtable on the International Economy, BRIE Working Paper 2014-7.

Isaac, M. (2014, November 19). Uber’s Privacy Practices Questioned by Senator Franken. The New York Times [Bits blog]. Retrieved from https://bits.blogs.nytimes.com/2014/11/19/senator-questions-uber-on-privacy- practices

Isaac, M. (2017, February 2). Uber C.E.O. to Leave Trump Advisory Council After Criticism. The New York Times. Retrieved from https://www.nytimes.com/2017/02/02/technology/uber-ceo-travis-kalanick- trump-advisory-council.htm

Jain, A. (2016, December 11). Take the Pledge to #SwitchtoPOOL. [Uber Newsroom]. Retrieved from https://newsroom.uber.com/india/switch2pool/

Jain, A. (2017, March 3). Driving with Uber: Setting the record straight. [Uber Newsroom]. Retrieved from https://newsroom.uber.com/india/drivingwithuber/

Jari, A. (2016, July). Life of Uber taxi drivers in Nigeria. Naij. Retrieved from https://www.naij.com/871747-life-uber-taxi-driver-nigeria.htm

Joyce, J. (2012, January 1). A Walk Through Surge Pricing, 2010-2012. [Uber Newsroom]. Retrieved from https://newsroom.uber.com/take-a-walk-through- surge-pricing/

Jøsang, A., Ismail, R., & Boyd, C. (2007). A survey of trust and reputation systems for online service provision. Decision support systems, 43(2), pp. 618-644. Retrieved from pp. 618-644

Kalanick, T. (2010, December 22). Uber’s Founding. [Uber Newsroom]. Retrieved from https://newsroom.uber.com/ubers-founding/

134

Kalanick, T. (2011a, February 2). Untitled. (Kalanick speaks in a company event). Youtube [Video file]. Retrieved from https://www.youtube.com/watch?v=1l_4Hxs88c4

Kalanick, T. (2011b, May 3). Uber NYC has launched. [Uber Newsroom]. Retrieved from https://newsroom.uber.com/uber-nyc-launches-service/

Kalanick, T. (2011c, August 16). Travis Kalanick of Uber - TWiST #180. Youtube [Video file]. Retrieved from https://www.youtube.com/watch?v=550X5OZVk7Y&index=1&list=PL24nOp PUQlbYSiRgcsABbkbOlvbP6BSq7

Kalanick, T. (2011d, September 22). Halloween Surge Pricing: get an Uber at the witching hour. [Uber Newsroom]. Retrieved from https://newsroom.uber.com/halloween-surge-pricing-get-an-uber-at-the- witching-hour/

Kalanick, T. (2011e, November 3). FailCon 2011 - Uber Case Study. Youtube [Video file]. Retrieved from https://www.youtube.com/watch?v=2QrX5jsiico

Kalanick, T. (2011f, December 7). We’re Going Global with Big Funding. [Uber Newsroom]. Retrieved from https://newsroom.uber.com/france/were-going- global-with-big-funding

Kalanick, T. (2013, August 23). Bits & Atoms. [Uber Nwesroom]. Retrieved from https://newsroom.uber.com/bits-atoms-2/

Kalanick, T. (2014, January 9). uberX: changing cities. [Uber Newsroom]. Retrieved from https://newsroom.uber.com/situation/

Kalanick, T. (2015, June 2). 5 Year Anniversary Remarks from Uber CEO Travis Kalanick | Uber. Youtube [Video file]. Retrieved from https://www.youtube.com/watch?v=idjrouG_8vY

Kalanick, T. (2016a, February 2). Celebrating Cities: A New Look and Feel for Uber. [Uber Newsroom. Retrieved from https://newsroom.uber.com/celebrating- cities-a-new-look-and-feel-for-uber/

Kalanick, T. (2016b, April 21). Growing and growing up. [Uber Newsroom]. Retrieved from https://newsroom.uber.com/growing-and-growing-up/

Kalanick, T. (2016c, August 1). Uber China Merges with Didi Chuxing. [Uber Newsroom]. Retrieved from https://newsroom.uber.com/uber-china-didi/

Kalleberg, A. L. (2013). Globalization and Precarious Work. Contemporary Sociology, 42(5), pp. 700-706.

135

Kaminska, I. (2016, December 1). The taxi unicorn’s new clothes. Financial Times. Retrieved from https://ftalphaville.ft.com/2016/12/01/2180647/the-taxi- unicorns-new-clothes/

Kane, K. (2015, March 19). The big hidden problem with Uber? Insincere 5-star ratings. Wired. Retrieved from https://www.wired.com/2015/03/bogus-uber- reviews

Karjalainen, J., & Gang, X. (2016 ). Social manufacturing and business model innovation. In Service Operations and Logistics, and Informatics (SOLI), 2016 IEEE International Conference on (pp. 18-23). IEEE. .

Katz, L. F., & Krueger, A. B. (2016). The Rise and Nature of Alternative Work Arrangements in the United States, 1995-2015. (No. w22667). National Bureau of Economic Research.

Katz, V. (2015). Regulating the sharing economy. Berkeley Tech. LJ, 30, 1067.

Kelestyn, B., Henfridsson, O., & Nandhakumar, J. (2017). Scaling the User Base of Digital Ventures Through Generative Pattern Replication: The Case of Ridesharing. Proceedings of the 50th Hawaii International Conference on System Sciences.

Kenney, M., & Zysman, J. (2016). The Rise of the Platform Economy. Issues in Science and Technology, 32(3), pp. 61-69.

Khrennikov, I. (2015, August 13). Yandex Triples Russian Online-Taxi Revenue to Stay Ahead of Uber. Bloomberg. Retrieved from https://www.bloomberg.com/news/articles/2015-08-13/yandex-triples-russian- online-taxi-revenue-to-stay-ahead-of-uber

Khrennikov, I. (2017, January 13). Uber Russian Drivers Quit After Putin’s Tax on U.S. Tech Giants. Bloomberg. Retrieved from https://www.bloomberg.com/news/articles/2017-01-13/uber-russian-drivers- quit-after-putin-s-tax-on-u-s-tech-giants

King, D. A., Peters, J. R., & Daus, M. W. (2012). Taxicabs for Improved Urban Mobility: Are We Missing an Opportunity? In Transportation Research Board 91st Annual Meeting (No. 12-2097).

Kirkpatrick, D. (2011). The Facebook effect: The inside story of the company that is connecting the world. Simon and Schuster.

Kjeldstadli, K. (1999). Fortida er ikke hva den en gang var: en innføring i historiefaget. Oslo: Universitetsforlaget.

136

Knight, S. (2016, April 27). How Uber conquered London. Guardian. Retrieved from https://www.theguardian.com/technology/2016/apr/27/how-uber-conquered- london

Kolodny, L. (2010, October 24). UberCab Ordered to Cease And Desist. TechCrunch. Retrieved from https://techcrunch.com/2010/10/24/ubercab-ordered-to-cease- and-desist/

Koopman, C., Mitchell, M. D., & Thierer, A. D. (2015). The Sharing Economy and Consumer Protection Regulation: The Case for Policy Change. The Journal of Business, Entrepreneurship & the Law, Vol. 8 Iss 2, 2015.

Kosoff, M. (2015, June 1). Everything you need to know about 'The Fountainhead,' a book that inspires billionaire Uber CEO Travis Kalanick. Business Insider. Retrieved from http://www.businessinsider.com/how-ayn-rand-inspired-uber- ceo-travis-kalanick-2015-6/?r=US&IR=T&IR=T/#avis-kalanick-is-a-fan-of- rands--but-especially-the-fountainhead-1

Kosoff, M., & McAlone, N. (2015, June 4). Where are they now: Meet 12 of Uber's first employees — three of them are now billionaires. Business Insider. Retrieved from http://www.businessinsider.com/ubers-first-employees-2015- 6?r=US&IR=T&IR=T

Kovach, S. (2014, Janaury 1). Uber Did Its Best To Warn You About New Year's Eve Surge Pricing, But Everyone Complained Anyway. Business Insider. Retrieved from http://www.businessinsider.com/uber-new-years-eve-surge- pricing-2014-1?r=US&IR=T&IR=T

Kuo, l. (2016, August 2). Drivers in Kenya are protesting against being “Uber slaves”. Quarts. Retrieved from https://qz.com/748149/drivers-in-kenya-are- protesting-against-being-uber-slaves/

LA Weekly. (2014, October 27). Confessions of an Uber Driver. Retrieved from http://www.laweekly.com/news/confessions-of-an-uber-driver-5173883

Lacy, S. (2011, June 15). Uber Out-Maths Google on NYC ETAs. TechCrunch. Retrieved from https://techcrunch.com/2011/06/15/uber-out-maths-google-on- nyc-etas/

Lacy, S. (2014, October 22). The horrific trickle down of Asshole culture: Why I've just deleted Uber from my phone. Pando. Retrieved from https://pando.com/2014/10/22/the-horrific-trickle-down-of-asshole-culture-at- a-company-like-uber

Lagorio-Chafkin, C. (2013). Resistance Is Futile. Inc. Magazine, July/August. Retrieved from http://www.inc.com/magazine/201307/christine-lagorio/uber- the-car-service-explosive-growth.html

137

Laizos, C. (2010, April 9). Entrepreneur-Turned-Angel Travis Kalanick to Founders: Get Some Fangs. Pe Hub Network. Retrieved from https://www.pehub.com/2010/04/entrepreneur-turned-angel-travis-kalanick-to- founders-get-some-fangs/#

Lashinsky, A. (2015, June 3). Uber: An oral history. Fortune . Retrieved from http://fortune.com/2015/06/03/uber-an-oral-history/

Laurell, C., & Sandström, C. (2016). ANALYSING UBER IN SOCIAL MEDIA — DISRUPTIVE TECHNOLOGY OR INSTITUTIONAL DISRUPTION? International Journal of Innovation Management, vol. 20, no. 5, (19 pages).

Lawler, R. (2012a, August 14). Uber Hit With Cease And Desist In Boston, Promises To Continue “Full Speed Ahead”. TechCrunch. Retrieved from https://techcrunch.com/2012/08/14/uber-c-and-d-boston/

Lawler, R. (2012b, September 13). Look Out, Lyft: Uber CEO Travis Kalanick Says It Will Do Ride Sharing, Too. TechCrunch. Retrieved from http://techcrunch.com/2012/09/12/uber-ride-sharing/

Lawler, R. (2012c, November 14). California Regulator Issues Citations And Fines Against Transportation Services Uber, Lyft, And SideCar. TechCrunch. Retrieved from https://techcrunch.com/2012/11/14/cpuc-uber-lyft-sidecar- citations/

Lawler, R. (2013a, January 31). It’s On. Uber Will Soon Go Up Against Lyft And SideCar With A Ride-Sharing Service Of Its Own. TechCrunch. Retrieved from https://techcrunch.com/2013/01/31/uber-ride-share/

Lawler, R. (2013b, April 12). Uber Moves Deeper Into Ride Sharing, Promises To Roll Out Services Where Regulators Have Given ‘Tacit Approval’. TechCrunch. Retrieved from https://techcrunch.com/2013/04/12/uber-rideshare-almost-everywhere/

Lawler, R. (2014a, April 7). Uber Experiments With A Courier Service In New York City. TechCrunch. Retrieved from https://techcrunch.com/2014/04/07/uberrush/

Lawler, R. (2014c, August 20). Uber Opens Its API With 11 Launch Partners, Including OpenTable, TripAdvisor, and United Airlines. TechCrunch. Retrieved from https://techcrunch.com/2014/08/20/uber-api-part-deux/

Lazo, L. (2016, August 16). Arlington studying a plan that would pay for your Uber to Metro. The Washington Post. Retrieved from https://www.washingtonpost.com/news/dr-gridlock/wp/2016/08/16/arlington- planning-to-subsidize-rideshare-trips-to-the-metro/?utm_term=.f47eedfcf3a0

138

Lazzaro, S. (2016, August 23). We Tried Juno, the Company Trying to Overthrow Uber by Giving 50 Percent Equity to Drivers. Observer. Retrieved from http://observer.com/2016/08/we-tried-juno-the-company-trying-to-overthrow- uber-by-giving-50-equity-to-drive

Lee, H. (2014, September 17). Uber Is 'Ethically Challenged'; Peter Thiel. CNBC . Retrieved from http://www.cnbc.com/id/102008782 (visited Jan 03, 2017).

Levin, S. (2016, April 13). Uber lawsuits timeline: company ordered to pay out $161.9m since 2009. Guardian. Retrieved from https://www.theguardian.com/technology/2016/apr/13/uber-lawsuits-619- million-ride-hailing-app

Levinson, P. (2015, October 16). 3 reasons why Uber will win in the end. Business Insider. Retrieved from http://www.businessinsider.com/why-uber-will-win- in-the-end-2015-10?r=US&IR=T&IR=T

Lomas, N. (2014, December 14). France Bans UberPop Starting January 1. TechCrunch. Retrieved from https://techcrunch.com/2014/12/15/uberpop-non/

Lu, A. (2015, September 10). 8 Travis Kalanick Quotes That The Uber CEO Definitely Regrets. Bustle. Retrieved from https://www.bustle.com/articles/109720-8-travis-kalanick-quotes-that-the- uber-ceo-definitely-regrets

Lu, K. (2016, August 1). Launch Everywhere—How Uber Won 450+ Markets. OneSky [blog]. Retrieved from http://www.oneskyapp.com/blog/uber-global- expansion/

Lunden, I., & Buhr, S. (2015, September 3). Lyft, Didi, Ola And GrabTaxi Partner In Global Tech, Service Alliance To Rival Uber. TechCrunch. Retrieved from https://techcrunch.com/2015/12/03/lyft-didi-ola-and-grabtaxi-partner-in- global-tech-service-alliance-to-rival-uber/

Lynley, M. (2016, October 19). Travis Kalanick says Uber has 40 million monthly active riders. TechCrunch. Retrieved from https://techcrunch.com/2016/10/19/travis-kalanick-says-uber-has-40-million- monthly-active-riders/

Malin, B. J., & Chandler, C. (2016). Free to Work Anxiously: Splintering Precarity Among Drivers for Uber and Lyft. Communication, Culture & Critique.

Mangalindan, J. P. (2014, May 28). In price wars, some Uber and Lyft drivers feel the crunch. Fortune. Retrieved from http://fortune.com/2014/05/28/in-price-wars- some-uber-and-lyft-drivers-feel-the-crunch/

139

Mangalindan, J. P. (2015, March 4). Uber acquires mapping startup deCarta. Mashable. Retrieved from http://mashable.com/2015/03/03/uber-acquires- mapping-decarta/#ozznxt2d7aqH

McAlone, N. (2016, February 10). This is how Uber used to look when it first started out — and how it's changed over time. Business Insider. Retrieved from from http://www.businessinsider.com/ubers-design-history-2010-2016-2016- 2?r=US&IR=T&IR=T/#2010-uber-launches-in-san-francisco-as-ubercab-this- was-its-initial-logo-1

McCarthy, C. (2011, January 21). Can an SF start-up help hail a cab in NY? Cnet. Retrieved from https://www.cnet.com/news/can-an-sf-start-up-help-hail-a-cab- in-ny/

McChrystal, S., McChrystal, S. A., Collins, T., Fussell, C., & Silverman, D. (2015). Team of teams: New rules of engagement for a complex world. Penguin.

McGregor, J. (2014, August 26). Uber Trials Fast Food Delivery Service 'UberFRESH'. Forbes. Retrieved from https://www.forbes.com/sites/jaymcgregor/2014/08/26/uber-trials-fast-food- delivery-service-uberfresh/#15f308965c17

McMullen, B. (2012, December 5). Uber 2.0 – #TheRideAhead. [Uber Newsroom]. Retrieved from https://newsroom.uber.com/uber-2-0-therideahead

McMullen, B. (2013a, May 15). Ride with Uber, Pay with Google. [Uber Newsroom]. Retrieved from https://newsroom.uber.com/ride-with-uber-pay-with-google/

McMullen, B. (2013b, July 15). Fare Splitting = Fair Splitting. [Uber Newsroom]. Retrieved from https://newsroom.uber.com/faresplit/

McNeal, S. (2016, Janaury 1). People Woke Up And Realized They Spent Hundreds Of Dollars On Uber For New Year’s Eve. BuzzFeed. Retrieved from https://www.buzzfeed.com/stephaniemcneal/uber- hangover?utm_term=.usR90RLxAm#.umLpaNA3M9

Means, B., & Seiner, J. A. (2015). Navigating the Uber Economy. UCDL Rev., 49, 1511.

Miles, I. (2016). Twenty years of service innovation research. In Service Innovation (pp. 3-34). Springer Japan, 3-34.

Milian, M. (2014a, January 22). Uber Steps on the Gas in Asia Expansion. Bloomberg. Retrieved from https://www.bloomberg.com/news/2014-01- 21/uber-steps-on-the-gas-in-asia-expansion.html

Milian, M. (2014b, November 20). Uber's International 'Launch Playbook' Includes Some Tough Lessons. Bloomberg. Retrieved from

140

https://www.bloomberg.com/news/articles/2014-11-20/ubers-international- launch-playbook-includes-some-tough-lessons

Moazed, A., & Johnson, N. L. (2016, February 27). Why Clayton Christensen Is Wrong About Uber And Disruptive Innovation. TechCrunch. Retrieved from https://techcrunch.com/2016/02/27/why-clayton-christensen-is-wrong-about- uber-and-disruptive-innovation/

Morales, N. J. (2015, May 13). Philippines issues rules on ride-sharing services, as taxis threaten court. Reuters . Retrieved from http://www.reuters.com/article/us-philippines-uber- idUSKBN0NY15Y20150513 (Visited December 15, 2016)

Morris, C. R., & Ferguson, C. H. (1992). How architecture wins technology wars. Harvard Business Review, 71(2), pp. 86-96.

Mosquera, L. J. (2016, July 15). Uber Drivers on Strike in Mexico Demand End to UberPool. Panam Post. Retrieved from https://panampost.com/luis- mosquera/2016/07/15/uber-drivers-strike-mexico-demand-end-uberpool

Mulligan, B. (2012, Janaury 1). Redesigning the Uber Surge Pricing Screen. Retrieved from http://startingup.me/post/15141134089/redesigning-the-uber- surge-pricing-screen

Mungai, R. (2017, March 6). Uber seeks to resolve price war with drivers. The Star. Retrieved from http://www.the-star.co.ke/news/2017/03/06/uber-seeks-to- resolve-price-war-with-drivers_c1518810

Narducci, D. A. (2012, January 2). What happened with Uber on New Year's Eve (Dec. 2011)?. Quora. Retrieved from https://www.quora.com/What-happened- with-Uber-on-New-Years-Eve-Dec-2011

Newcomer, E. (2016a, January 4). GM Invests $500 Million in Lyft. Bloomberg. Retrieved from https://www.bloomberg.com/news/articles/2016-01-04/gm- invests-500-million-in-lyft-to-bolster-alliance-against-uber

Newcomer, E. (2016b, May 24). Toyota to make strategic investment in Uber and team up on auto leasing program. Bloomberg. Retrieved from https://www.bloomberg.com/news/articles/2016-05-24/toyota-to-invest-in- uber-and-team-up-on-auto-leasing-program

Newcomer, E. (2016c, October 13). Can Uber Conquer Latin America?. Bloomberg. Retrieved from https://www.bloomberg.com/news/articles/2016-10-13/can- uber-conquer-latin-americ

Newcomer, E. (2017, February 28). In Video, Uber CEO Argues With Driver Over Falling Fares. Bloomberg. Retrieved from

141

https://www.bloomberg.com/news/articles/2017-02-28/in-video-uber-ceo- argues-with-driver-over-falling-fares

Newton, C. (2014, August 26). This is Uber's playbook for sabotaging Lyft. The Verge. Retrieved from http://www.theverge.com/2014/8/26/6067663/this-is- ubers-playbook-for-sabotaging-lyf

Nguyen, T. (2015, November 3). ETA Phone Home: How Uber Engineers an Efficient Route. [Uber Engineering]. Retrieved from https://eng.uber.com/engineering-an-efficient-route/

Nuzzi, O. (2014, March 28). Uber’s Biggest Problem Isn’t Surge Pricing. What If It’s Sexual Harassment by Drivers?. The Daily Beast. Retrieved from http://www.thedailybeast.com/articles/2014/03/28/uber-s-biggest-problem-isn- t-surge-pricing-what-if-it-s-sex

NYC T&LC. (n.d.). Vehicles Approved as Taxicabs. Retrieved from http://www.nyc.gov/html/tlc/html/industry/taxicab_vehicles_in_use.shtml

O'Donovan, C., & Singer-Vine, J. (2016, June 23). How Much Uber Drivers Actually Make Per Hour. BuzzFeed. Retrieved from https://www.buzzfeed.com/carolineodonovan/internal-uber-driver-pay- numbers?utm_term=.lsv34gmKQr#.ojjjaVP32

Olanoff, D. (2012, September 17). Lyft’s Focus On Community And The Story Behind The Pink Mustache. TechCrunch. Retrieved from https://techcrunch.com/2012/09/17/lyfts-focus-on-community-and-the-story- behind-the-pink-mustache/

Onu, E., & Alake, T. (2016, October 19). Uber Partners First Bank to Offer Loans to Nigerian Drivers. Bloomberg. Retrieved from https://www.bloomberg.com/news/articles/2016-10-19/uber-partners-with- first-bank-to-offer-loans-to-nigerian-drivers

Pagliery, J. (2014, November 25). Uber removes racy blog posts on prostitution, one- night stands. CNN. Retrieved from http://money.cnn.com/2014/11/25/technology/uber-prostitutes/

Panzarino, M. (2013, March 14). Founders 2013: Uber’s Travis Kalanick on surge pricing, corruption and ‘being baller’ in SF. TheNextWeb. Retrieved from https://thenextweb.com/insider/2013/03/14/f-ounders-2013-ubers-travis- kalanick-on-surge-pricing-corruption-and-being-baller-in- sf/#.tnw_82gKKnGy

Park, J., Kim, J., & Lee, B. (2016). Are Uber really to blame for sexual assault?: evidence from New York city. In Proceedings of the 18th Annual

142

International Conference on Electronic Commerce: e-Commerce in Smart connected World (p. 12). ACM.

Parker, G. G., & Van Alstyne, M. W. (2005). Two-sided network effects: A theory of information product design. Management science, 51(10), pp. 1494-1504.

Paul, K. (2016, July 1). Is Uber Doing Enough to Protect Women from Drivers Who Stalk Them?. Broadly. Retrieved from https://broadly.vice.com/en_us/article/is-uber-doing-enough-to-protect- women-from-drivers-who-stalk-them

Perez, S. (2016, March 15). UberEATS’ standalone food delivery app launches in its first U.S. cities. TechCrunch. Retrieved from https://techcrunch.com/2016/03/15/ubereats-standalone-food-delivery-app- launches-in-its-first-u-s-cities

Piketty, T. (2014). Capital in the 21st Century. Cambridge: Harvard University Press.

Porter, E. (2016, December 13). Where Were Trump’s Votes? Where the Jobs Weren’t. The New York Times. Retrieved from https://www.nytimes.com/2016/12/13/business/economy/jobs-economy- voters.html?_r=0

Posen, H. A. (2015). Ridesharing in the Sharing Economy: Should Regulators Impose Uber Regulations on Uber. Iowa L. Rev., 101, 405.

Pradhan, N. (2016, December 27). What is the average daily income of an Uber cab driver in India?. Quora. Retrieved from https://www.quora.com/What-is-the- average-daily-income-of-an-Uber-cab-driver-in-India

Prassl, J., & Risak, M. (2015). Uber, Taskrabbit, and Co.: Platforms as Employers- Rethinking the Legal Analysis of Crowdwork. Comp. Lab. L. & Pol'y J., 37, 619.

Prigg, M. (2014, July). Uber agrees to stop 'surge' pricing during emergencies and natural disasters after charging EIGHT TIMES its normal fare during snowstorms in New York. Daily Mail. Retrieved from http://www.dailymail.co.uk/sciencetech/article-2685112/Uber-forced-stop- surge-pricing-emergencies-natural-disasters-charging-EIGHT-TIMES-normal- fare-snowstorms-New-York.html

Prigg, M., & Liberatore, S. (2016, February 5). Uber's design boss leaves following disastrous rebranding. Daily Mail. Retrieved from http://www.dailymail.co.uk/sciencetech/article-3434051/Uber-s-design-boss- leaves-following-disastrous-rebranding-says-using-recent-successes-chance- time-rest-reflect-recharge.html

143

Prithivi, S. (2013, August 29). Hello India! Secret Ubers Have Arrived In Bangalore. [Uber Newsroom]. Retrieved from https://newsroom.uber.com/india/hello- india-secret-ubers-have-arrived-in-bangalore/

Public Citizen. (2016, May 16). Disrupting Democracy: How Uber Deploys Corporate Power to Overwhelm and Undermine Local Government. Retrieved from https://www.citizen.org/documents/uber-disrupting-democracy- corporate-power-report.pdf

Radhakrishnan, M. (2014, July 15). "Designing for Flexibility" Mina Radhakrishnan, #ProductSF. Youtube [Video file]. Retrieved from https://www.youtube.com/watch?v=WuAAqYgdmaY

Rahel, S. (2016). Economics of the Taxi Industry: An Uber Shake-up. Wyoming Scholars Repository.

Rai, S. (2016, October 10). Uber aims to unseat Ola in $10 billion India ride-hailing market. Livemint. Retrieved from http://www.livemint.com/Companies/jljItP5FzCAkvtXLwJhcfM/Uber-aims- to-unseat-Ola-in-10-billion-India-ridehailing-ma.html

Ram, A., & Kazmin, A. (2017, February 13). Uber’s Indian drivers strike for fourth day over earnings squeeze. Financial Times. Retrieved from https://www.ft.com/content/9653ace2-f1d9-11e6-8758-6876151821a6

Ranchordás, S. (2015). Does sharing mean caring: Regulating innovation in the sharing economy. Minn. JL Sci. & Tech., 16, 413.

Rao, L. (2011, July 22). Ticket To Ride: Uber Testing Reduced Pricing In San Francisco. TechCrunch. Retrieved from https://techcrunch.com/2011/07/22/ticket-to-ride-uber-testing-reduced-pricing- in-san-francisco

Rao, L. (2012a, April 18). Uber Experiments With Lower-Priced Taxis In Chicago Through Newly Launched Labs Group, ‘Garage’. TechCrunch. Retrieved from https://techcrunch.com/2012/04/18/uber-experiments-with-lower-priced- taxis-in-chicago-through-newly-launched-labs-group-garage/

Rao, L. (2012b, October 5). Uber Sued By Taxi And Livery Companies In Chicago For Consumer Fraud And More. TechCrunch. Retrieved from https://techcrunch.com/2012/10/05/uber-sued-by-taxi-and-livery-companies- in-chicago-for-consumer-fraud-and-more/

Rawlins, L. K. (2016, June 2). Inside Uber SA. ITWeb. Retrieved from http://www.itweb.co.za/index.php?option=com_content&view=article&id=15 3166

144

Reinhold, E. (2016, April 20). Rewriting Uber Engineering: The Opportunities Microservices Provide. [Uber Engineering]. Retrieved from https://eng.uber.com/building-tincup/

Revesz, R. (2016, Septemeber 19). Uber defends itself against criticism it raised prices in Manhattan after New York City bombing. Independent. Retrieved from http://www.independent.co.uk/news/world/americas/uber-surge-pricing- new-york-bombing-chelsea-twitter-refund-a7317286.html

RideSharing Guy. (2016, March 29). Driving for UberEats: What it’s like delivering food for Uber. Ridesharing Guy [Blog]. Retrieved from http://www.ridesharingdriver.com/driving-for-ubereats-what-its-like- delivering-food-for-uber

Ries, B., & Ryall, J. (2014, December 14). Uber intros surge pricing during Sydney hostage siege, then backtracks after user outcry. Mashable. Retrieved from http://mashable.com/2014/12/14/uber-sydney-surge-pricing/#_aayB3A8H5qi

Rochet, J. C., & Tirole, J. (2003). Platform competition in two-sided markets. Journal of the european economic association, 1(4), pp. 990-1029.

Rogers, B. (2015). The social costs of Uber. University of Chicago Law Review Dialogue, Forthcoming.

Rogers, E. M. (2010). Diffusion of innovations. Simon and Schuster.

Rothman, S. (2016, April 20). Why Uber Won: The Startup Steroid Era and the Use of Capital as a Performance Enhancing Drug. Greylock Partners. Retrieved from https://news.greylock.com/why-uber-won-5598a2a66561#.6i19k031e

Roushan Ali, M. (2015, November 21). Uber rich drivers find taxi business is less luring. Deccan Chronicle. Retrieved from http://www.deccanchronicle.com/151121/nation-current-affairs/article/uber- rich-drivers-find-taxi-business-less-luring

Rubinstein, D. (2012, August 17). Judge declares Bloomberg's borough taxi law 'null and void'. Politico. Retrieved from http://www.politico.com/states/new- york/city-hall/story/2012/08/judge-declares-bloombergs-borough-taxi-law- null-and-void-00000

Rusli, E. (2010, November 3). Uber, Not UberCab, Launches Android App. TechCrunch. Retrieved from https://techcrunch.com/2010/11/03/uber-not- ubercab-launches-android-app/

Russell, J. (2015a, January 8). Uber’s Latest Experiment Is Uber Cargo, A Logistics Service In Hong Kong. TechCrunch. Retrieved from https://techcrunch.com/2015/01/08/uber-cargo/

145

Russell, J. (2015b, January 8). Uber Announces Price Cuts In 48 U.S. Cities, But Guarantees Its Drivers Will Earn More. TechCrunch. Retrieved from https://techcrunch.com/2015/01/08/uber-us-price-cuts/

Russell, J. (2015c, August 18). GrabTaxi Lands $350M From China’s Top Uber Rival Didi Kuaidi And Others. TechCrunch. Retrieved from https://techcrunch.com/2015/08/18/grabtaxi-lands-350m-new-backers-include- chinas-top-uber-rival-didi-kuaidi/

Russel, J. (2017a, February 6). Uber’s newest service in India lets customers book a car for multiple trips. TechCrunch. Retrieved from https://techcrunch.com/2017/02/06/uber-hire-india-pilot/

Russell, J. (2017b, March 3). Uber plans to turn its app into a ‘content marketplace’ during rides. TechCrunch. Retrieved from https://techcrunch.com/2017/03/03/uber-plans-to-turn-its-app-into-a-content- marketplace-during-rides/

Santi, P., Resta, G., Szell, M., Sobolevsky, S., Strogatz, S. H., & Ratti, C. (2014). Quantifying the benefits of vehicle pooling with shareability networks. Proceedings of the National Academy of Sciences, 111(37), pp. 13290-13294.

Sawers, P. (2012, July 3). Taxis, black cars, hybrids, SUVs...Uber announces ‘freedom to choose’ for US and Canadian riders. TheNextWeb. Retrieved from https://thenextweb.com/apps/2012/07/03/taxis-black-cars-hybrids-suvs-uber- announces-freedom-to-choose-for-us-and-canadian-riders/#.tnw_aLFwhFQF

Scheiber, N. (2017, April 2). How Uber Uses Psychological Tricks to Push Its Drivers’ Buttons. The New York Times. Retrieved from https://www.nytimes.com/interactive/2017/04/02/technology/uber-drivers- psychological-tricks.html?_r=0

Schonfeld, E. (2011, March 7). Uber's Private Cars Are Preparing For A Secret Push Into New York City. TechCrunch. Retrieved from https://techcrunch.com/2011/03/07/uber-new-york/

Schor, J. B. (2015). The sharing economy: reports from stage one. Boston College.

Schumpeter, J. A. (1994). Capitalism, socialism and democracy. London: Routledge.

Schwietert Collazo, J. (2014, October 16). Uber: The answer to Mexico City's transportation problems?. Latin Correspondent. Retrieved from http://latincorrespondent.com/2014/10/uber-solving-mexico-citys-biggest- transportation-problems/#7Tq75tLGTOFiVatM.97

Scoble, R. (2016, January 27). Don’t worry Uber/Lyft drivers: self-driving cars won’t take your job for at least a decade. Medium. Retrieved from

146

https://medium.com/@scobleizer/don-t-worry-uber-lyft-drivers-self-driving- cars-won-t-take-your-job-for-at-least-a-decade-3b8c58a7f102

Seawright, J., & Gerring, J. (2008). Case selection techniques in case study research: A menu of qualitative and quantitative options. Political Research Quarterly, 61(2), pp. 294-308.

Shaheen, S. A. (2016). Mobility and the sharing economy. Transport Policy.

Shapiro, B. D. (2015). O'Connor v. Uber Techs., Inc., 82 F. Supp. 3d 1133. W. St. UL Rev., 43, 325.

Sharma, S., Mishra, D., & Arora, K. (2016, January 19). Ultimately, all rules have to bend towards people and progress: Travis Kalanick, Uber. The Economic Times. Retrieved from http://economictimes.indiatimes.com/opinion/interviews/ultimately-all-rules- have-to-bend-towards-people-and-progress-travis-kalanick- uber/articleshow/50637018.cms (Visited Dec. 19, 2016)

SherpaShare. (2015, July 15). Report: What Uber, Lyft drivers earn per trip. SherpaShare Blog. Retrieved from http://www.sherpashareblog.com/2015/07/what-uber-lyft-drivers-earn-per- trip/

Shontell, A. (2011, May 16). The truth about Uber: The Personal Driver App, Tried And Broken Down For You. Business Insider. Retrieved from http://www.businessinsider.com/what-is-uber-tour-new-york- city?r=US&IR=T&IR=T

Shontell, A. (2014a, January 11). All Hail The Uber Man! How Sharp-Elbowed Salesman Travis Kalanick Became Silicon Valley's Newest Star. Business Insider. Retrieved from http://www.businessinsider.com/uber-travis-kalanick- bio-2014-1?op=1&r=US&IR=T&IR=T

Shontell, A. (2014b, January 11). How A Curious VC Found Uber Before Anyone Else, Back When It Was Only Worth $4 Million. Business Insider. Retrieved from http://www.businessinsider.com/first-round-capital-ubers-first-investor- 2014-1?r=US&IR=T&IR=T

Siegler, M. (2010, October 24). UberCab Cease & Desist Means One Thing: They're Onto Something. TechCrunch. Retrieved from https://techcrunch.com/2010/10/24/ubercab/

Silverstein, S. (2014, October 16). These Animated Charts Tell You Everything About Uber Prices In 21 Cities. Business Insider. Retrieved from http://www.businessinsider.com/uber-vs-taxi-pricing-by-city-2014- 10?r=US&IR=T&IR=

147

Sims, P. (2017, March 2). The real story behind Uber CEO. CNN. Retrieved from http://edition.cnn.com/2017/03/01/opinions/real-story-behind-uber-sims- opinion/

Singer, N., & Isaac, M. (2015, May 9). An App That Helps Drivers Earn the Most From Their Trips. The New York Times. Retrieved from https://www.nytimes.com/2015/05/10/technology/a-dashboard-management- consultant.html

Singh, M. (2017, February 23). Here's why Uber drivers keep going on strike in its fastest growing market. Mashable. Retrieved from http://mashable.com/2017/02/23/uber-ola-strike-bangalore- india/#HDDtJneWgOqz

Slaughter, A. M. (2017). The Chessboard and the Web: Strategies of Connection in a Networked World. Yale University Press.

Smart, R., Rowe, B., Hawken, A., Kleiman, M., Mladenovic, N., Gehred, P., & Manning, C. (2015). Faster and Cheaper: How Ride-Sourcing Fills a Gap in Low-Income Los Angeles Neighborhoods. BOTEC Analysis Corporation.

Standing, G. (2011). The Precariat: The New Dangerous Class. A&C Black.

Stiglitz, J. E. (2012). The price of inequality: How today's divided society endangers our future. WW Norton & Company.

Stone, B. (2017). THE UPSTARTS: How Uber, Airbnb and the killer companies of the new Silicon Valley are changing the world. New York: Random House.

Stone, B., & Chen, L. Y. (2016, October 6). Uber Slayer: How China’s Didi Beat the Ride-Hailing Superpower. Bloomberg Businessweek. Retrieved from https://www.bloomberg.com/features/2016-didi-cheng-wei/

Summers, N. (2013, March 13). Uber launches on BlackBerry and Windows Phone as Android app gets latest design. TheNextWeb. Retrieved from https://thenextweb.com/apps/2013/03/25/uber-redesigns-its-android-app-and- launches-on-blackberry-and-windows-phone/#

Sundararajan, A. (2016). The sharing economy: The end of employment and the rise of crowd-based capitalism. Mit Press.

Swan, R. (2014, August 15). Uber Promises New Drivers $5,000 a Month. SF Weekly. Retrieved from http://archives.sfweekly.com/thesnitch/2014/08/15/uber- promises-new-drivers-5000-a-month

Swisher, K. (2014, November 5). Man and Uber Man. Vanity Fair. Retrieved from http://www.vanityfair.com/news/2014/12/uber-travis-kalanick-controversy

148

Tam, D. (2013, November 18). Uber picks up PayPal to drive member numbers. Cnet. Retrieved from https://www.cnet.com/news/uber-picks-up-paypal-to-drive- member-numbers/

Tassi, K. (2015, May 28). An Update on Privacy at Uber. [Uber Newsroom]. Retrieved from https://newsroom.uber.com/an-update-on-privacy-at-uber/

Taylor, A. (2012, October 1). This African Megacity May Have The Worst Traffic Jams In The World. Business Inside. Retrieved from http://www.businessinsider.com/lagos-traffic-the-worst-in-the-world-2012- 10?r=US&IR=T&IR=T

Telegraph Reporters. (2016, July 20). Uber driver earned less than minimum wage, tribunal told. The Telegraph. Retrieved from http://www.telegraph.co.uk/technology/2016/07/20/uber-driver-earned-less- than-minimum-wage-tribunal-told

Telegraph Staff. (2015, September 30). Proof that Uber costs less than black cabs. The Telegraph. Retrieved from http://www.telegraph.co.uk/technology/uber/11902613/Proof-that-Uber-costs- less-than-black-cabs.html

Tepper, F. (2016, July 18). Uber has completed 2 billion rides. TechCrunch. Retrieved from https://techcrunch.com/2016/07/18/uber-has-completed-2-billion-rides/

Tiku, N. (2012, September 6). Taxi & Limousine Commission Tells Uber It Can’t Legally Operate a Taxi App In NYC, Uber CEO Disagrees. Observer. Retrieved from http://observer.com/2012/09/taxi-and-limousine-commission- tells-uber-they-cant-legally-operate-in-new-york-city-uber-ceo-disagrees/

Tiku, N. (2014, November 18). Uber CEO Posts 13-Tweet Apology Without Answering a Single Question. ValleyWag. Retrieved from http://valleywag.gawker.com/uber-ceo-posts-13-tweet-apology-without- answering-a-sin-1660229634

Tilson, D. L., & Sørensen, C. (2010). Research commentary-digital infrastructures: the missing IS research agenda. Information systems research, 21(4), pp. 748- 759.

Tiwana, A., Konsynski, B., & Bush, A. A. (2010). Research commentary—Platform evolution: Coevolution of platform architecture, governance, and environmental dynamics. Information Systems Research, 21(4), pp. 675- 687.

Tsotsis, A. (2012a, July 1). Uber Opens Up Platform To Non-Limo Vehicles With “Uber X,” Service Will Be 35% Less Expensive. TechCrunch. Retrieved from https://techcrunch.com/2012/07/01/uber-opens-up-platform-to-non-limo- vehicles-with-uber-x-service-will-be-35-less-expensive/

149

Tsotsis, A. (2012b, September 7). Uber Wants To Offer A Low Cost Option In All Of Its Markets. TechCrunch. Retrieved from https://techcrunch.com/2012/09/07/uber-wants-to-offer-a-low-cost-option-in- all-of-its-markets/

UberCab. (2010, June 1). UberCab Demo.Youtube [video file]. Retrieved from https://www.youtube.com/watch?v=L8hJHAJARJY

Van Zyl, G. (2017, April 11). Some Uber drivers unhappy with price cuts. Fin24. Retrieved from http://www.fin24.com/Tech/Opinion/uber-profits-at-the- expense-of-drivers-20160411

Verma, S. (2015, September 30). Uber starts to lease cars to its drivers on the lines of Ola. Livemint. Retrieved from http://www.livemint.com/Companies/8MOz38G6Z8VcLcew0QcWNI/Uber- starts-to-lease-cars-to-its-drivers-on-the-lines-of-Ola.html

Visentin, L. (2017, February 27). Uber increasingly popular in Sydney as taxi use stands firm. The Sydney Morning Herald. Retrieved from http://www.smh.com.au/nsw/uber-increasingly-popular-in-sydney-as-taxi-use- stands-firm-20170227-gulydf.html

Voytek, B. (2012, March 26). Rides of Glory. #uberdata [Blog]. Retrieved from https://web.archive.org/web/20141118192805/http://blog.uber.com/ridesofglor y

Wagner, K. (2014, September 8). Uber Adding 50,000 New "Driver Jobs" a Month, Up From 20,000 in May. Recode. Retrieved from http://www.recode.net/2014/9/8/11630668/uber-adding-50000-new-driver- jobs-a-month-up-from-20000-in-may

Wagner, K. (2017, February 19). Uber’s CEO promises an ‘urgent investigation’ into a former employee’s sexual harassment claims. Recode. Retrieved from https://www.recode.net/2017/2/19/14665076/ubers-travis-kalanick-susan- fowler-sexual-harassment-investigation

Wahito, M. (2016, July 28). Kenya: Uber Slashes Taxi Charges By 35% in Nairobi. All Africa. Retrieved from http://allafrica.com/stories/201607290648.html

Wallsten, S. (2015). The competitive effects of the sharing economy: how is Uber changing taxis? Technology Policy Institute, 22.

Weise, K. (2015, June 24). This Is How Uber Takes Over a City: To conquer America’s quirkiest city, the company unleashed its biggest weapon. Bloomberg Businessweek. Retrieved from http://www.bloomberg.com/news/features/2015-06-23/this-is-how-uber-takes- over-a-city

150

Wesangula, D. (2016, September 16). Uber and Little Cab's battle to control the streets of Nairobi. Guardian. Retrieved from https://www.theguardian.com/sustainable-business/2016/sep/16/uber-little- cabs-kenya-nairobi-battle-taxi

Wilhelm, A. (2013, August 22). Google Ventures Puts $258M Into Uber, Its Largest Deal Ever. TechCrunch. Retrieved from https://techcrunch.com/2013/08/22/google-ventures-puts-258m-into-uber-its- largest-deal-ever/

Wilhelm, A., & Lawler, R. (2013, October 3). In Another Strike Against The Competition, Uber Lowers UberX Prices In San Diego, LA, And DC. TechCrunch. Retrieved from https://techcrunch.com/2013/10/03/in-another- strike-against-the-competition-uber-lowers-uberx-prices-in-san-diego-la-and- dc/

Wirtz, J., & Tang, C. (2016). Uber: Competing as Market Leader in the US versus Being a Distant Second in China. SERVICES MARKETING: People Technology Strategy, pp. 626-632.

Wohlsen, M. (2014, July 8). Uber's brilliant strategy to make itself too big to ban. Wired. Retrieved from https://www.wired.com/2014/07/ubers-brilliant- strategy-to-make-itself-too-big-to-ban/

Woo, C. P., & Bales, R. A. (2016). The Uber Million Dollar Question: Are Uber Drivers Employees or Independent Contractors? Mercer Law Review Forthcoming.

Wortham, J. (2011, May 4). With a Start-Up Company, a Ride Is Just a Tap of an App Away. The New York Times, p. B6. Retrieved from http://www.nytimes.com/2011/05/04/technology/04ride.html?_r=0

Wu, R. (2014, February 17). Uber Shanghai Has Officially Launched! Uber 正式进驻中国上海！. [Uber Newsroom]. Retrieved from https://newsroom.uber.com/china/uber-shanghai-has-officially-launched-uber- %E6%AD%A3%E5%BC%8F%E8%BF%9B%E9%A9%BB%E4%B8%AD% E5%9B%BD%E4%B8%8A%E6%

Zittrain, J. L. (2006). The generative internet. Harvard Law Review, pp. 1974-2040.

Zuboff, S. (2015). Big other: surveillance capitalism and the prospects of an information civilization. Journal of Information Technology, 30(1), pp. 75-89.

151

Appendix A

More information at https://www.crunchbase.com/organization/uber#/entity

152