Varieties of Gig Work: Germany’s Unique Development in the Platform-based Food Delivery Sector

1 Introduction

Cyclists with colorful backpacks delivering food from restaurants to consumers have become a global symbol of the gig economy. When the discourse around the globally expanding gig economy began, it was assumed that digital platforms would establish a global presence with a single disruptive business model. A particularly disruptive aspect of the expected new business model is the fragmentation of required work into short-term assignments and project-based tasks that can also be performed by independent contractors (Kirchner 2019; Schmidt 2017). This form of work organization contrasts with the traditional full-time employment model and challenges established norms of stable, long-term employment. Gig work platforms allow workers to connect directly with customers or clients, potentially eliminating the need for permanent employment with employers or staffing agencies.

However, it has become abundantly clear that the initial assumption of a global spread of a single business model was fundamentally flawed (Azzellini et al. 2022). International platforms learned that business models cannot be easily and effortlessly rolled out globally. Startups began to experiment with different forms of work organization in their respective home markets (Muszyński et al. 2022). Within the gig economy, diverse business and work models are emerging based on work demands, task complexity, and spatial constraints (Vallas and Schor 2020). In multifaceted fields of power, digital platforms, as new market actors, have to adapt their strategies to specific circumstances (Pernicka & Schüßler 2022).

In this context, the European landscape shows a remarkable diversity, characterized by different forms of employment, including contract, employment, and subcontracting arrangements. These variations depend on regulatory frameworks and regional contexts (Ametowobla & Kirchner 2024). On a global scale, the differences between countries are even more pronounced (Davis & Sinha 2021).

This article aims to provide insights into the diversity of work models and the divergent development of the platform-based food delivery sector^[1] in different countries by examining an exceptional case. The country under study is Germany. Given the initial expectation that platform-based food delivery would be carried out by independent couriers paid for individual gigs, Germany serves as an extreme case in comparison. In Germany, all major food delivery platforms currently classify their workers as employees rather than independent contractors. As a result, these workers are entitled to the same rights and benefits as other employees, including minimum wage, paid vacation, and sick leave. The development of the German food delivery market has been different from that of other countries.

For experts in comparative institutional research, it is not surprising that the German case emerges as an extreme within the spectrum of possibilities. Germany is seen as an exemplary case of a coordinated market economy (Hall & Soskice 2001) and an extensive social security-based welfare state that social forces are keen to protect against potential erosion (Ebbinghaus et al. 2022). German labor laws are strict, and there is a relatively robust trade union presence compared to the international context (Deakin & Sarkar 2008).

Nevertheless, this situation still requires explanation, as one might have assumed that food delivery, often considered low-skilled work in a segment of the dual labor market, would not face the same pressures to conform to the pattern of regular employment (Eichhorst & Marx 2011). Even in Germany, not all platform-mediated services are provided by employed workers. For example, the platform business model of „Helpling“ specifically revolves around the use of self-employed cleaning staff (Koutsimpogiorgos et al. 2023: 14–16), and the ride-hailing sector in Germany also differs in this respect from platform-based food delivery.

The German case is also not as straightforward as the Spanish one, where an employment relationship was legally mandated for delivery services (Eurofound 2021). The question therefore arises as to what factors or mechanisms are responsible for the fact that the German case in the field of food delivery has lived up to its reputation, even though the initial expectation of digital disruption did not anticipate this outcome.

To answer this question, the paper conducts a historical-sociological case study of the German case, examining the developments in the platform-based food delivery sector. In doing so, it assesses which of the various heuristics for divergent paths and unique developments established in the literature have explanatory power for the German case.

The article is organized as follows. It begins with a review of current research on food delivery platforms. Next, various heuristics from comparative historical research and path analysis are introduced to support the exploration of different developmental trajectories. This is followed by a description of the data and analytical methods used in the case study. Subsequent sections present the evolution of the sector and identify explanatory mechanisms, followed by a discussion of the results of the analysis and a concluding summary.

2 State of Research on Food Delivery Platforms

The social process of digitalization has profound implications for work and working conditions. Accordingly, a thriving and dynamic field of research has emerged in recent years that focuses on debates surrounding the digitalization of work and its impact on working conditions. This research examines, among other things, how digitalization is reshaping traditional work processes, changing employment relationships, and influencing the social dynamics of work environments (Pfeiffer et al. 2024). Within the narrower field of food delivery platforms, a considerable amount of research has been conducted to describe the phenomenon and examine legal and regulatory aspects, especially in the early days of the discourse (Schmidt 2017, Stewart & Stanford 2017).

Since then, the topic of work organization and control has received the most attention in the research literature. Research shows that work organization in food delivery relies heavily on digital platforms that connect customers with delivery drivers (Woodcock & Graham 2019). Workers register with a platform and can then accept delivery orders as they become available. Research has found that it is the platforms that essentially control the assignment of orders, tracking of deliveries, and processing of payments (Ivanova et al 2018). The platform also sets the payment rate, which is typically based on a per-delivery or per-hour rate. The asymmetric distribution of control is sometimes interpreted as algorithmic domination (Muldoon & Raekstad 2022) and algorithmic surveillance (Newlands 2021).

The business activities of platforms are time-critical and location-dependent. Research has shown that platforms implement control regimes that rely not only on algorithmic management (Duggan et al. 2020), but also on complementary control of work schedules. Platforms organize intra-platform markets in which workers compete for shifts based on performance (Heiland 2022). Platforms exert significant control over workers, as they can vary wage rates, deactivate workers’ accounts, and change working conditions. Workers must adapt to platforms’ carrot-and-stick strategies (Huang 2022), and sometimes endure being denied work (or pay) for shifts they have signed up for (Griesbach et al. 2019). Conversely, research has also shown that food delivery workers are not isolated individuals, but collective actors who learn from each other. They develop their own strategies for managing and, in some cases, subverting the system (Sun 2019).

In addition, the working conditions of gig workers have been a topic of interest for many researchers (Goods et al. 2019). Food delivery drivers face high levels of precariousness (MacDonald & Giazitzoglu 2019) and health and safety risks in their work (Christie & Ward 2019). However, some drivers also value the flexibility of their role (Barratt et al. 2020) and the experience of delivering food (Wang et al. 2021). Previous studies have explored how drivers cope with challenges such as unpaid waiting time (Franke & Pugliano 2023) and how they balance work and leisure activities (Veluchamy et al. 2021). Consumer research has shown how much customers care about drivers' working conditions when ordering food (Healy et al. 2020, Smith et al. 2021).

Another important area of research is worker activism. Despite the high levels of precarity and atomization that characterize this segment of the gig economy, worker activism and unionization have emerged as important responses to prevailing working conditions in the platform-coordinated food delivery sector (Chesta et al. 2019, Stuart et al. 2023). Research shows that gig workers are increasingly mobilizing to advocate for their rights and improved working conditions. They are responding to the partly varying work organization models of the platforms with micro-political activities (Heiland & Brinkmann 2020). Their activism also takes various, often unconventional forms, such as unofficial strikes, protests, and social media campaigns to pressure platform companies to address their concerns (Tassinari & Maccarrone 2020). In some regions, unionization efforts have become more prominent, with the goal of gaining collective bargaining rights and formalizing worker representation. In rare cases, these efforts have led to the establishment of collective agreements (Selberg 2023).

In addition to the aforementioned aspects of work organization, working conditions and representation, which make up a significant part of the research on platform-based food delivery, there are individual studies looking at the service quality of these services (Belanche et al. 2021), the sustainability of the business model (Lord et al. 2022), the generation of economically exploitable ‚hidden data‘ (van Doorn & Badger 2020) and gender-related aspects of platform work (Milkman et al. 2021).

The German food delivery market has been characterized by intense competition and rapid growth over recent years. According to the Digital Market Insights report from Statista (2024), the revenue generated by the food delivery market in Germany increased exponentially, more than tripling between 2017 and 2024, reaching approximately 7.67 billion US dollars.

In Germany, gig work within the domain of platform-coordinated food delivery was also carried out by independent drivers in the past, depending on the platform. Consequently, the displacement of self-employment represents a relatively recent phenomenon that warrants in-depth investigation. Although this particular aspect of Germany’s evolving landscape has been sporadically recognized in the scholarly literature (Ewen 2021, Koutsimpogiorgos et al. 2020: 529). However, besides Legantke (2024), it has not yet been the subject of systematic exploration. While this development may be perceived as a „race to the top“ phenomenon (Saikawa 2013), given Germany’s reputation for being relatively favorable to workers in the international gig work context, it is important to highlight that previous comparisons in Germany between platforms utilizing either self-employed or employed drivers did not reveal significant disparities in working conditions, except those directly stemming from contractual arrangements (Ivanova et al. 2018, Heiland 2021).

3 Research Design and Heuristics for Unique Developments

This paper aims to contribute to the multifaceted body of research on platform-based food delivery. Unlike previous studies, it addresses a specific research puzzle related to the divergent development of this sector across countries. Addressing this puzzle seeks to shed light on the factors and mechanisms that drive these different development trajectories. With a stronger focus on historical developments, the perspective shifts from describing and analyzing work organization to placing greater emphasis on the competitive dynamics that influence the implementation of work organization models. In addition, this paper departs methodologically from previous research by using a historical-analytical approach to unravel the research puzzle.

Given the relatively short history of platform-based food delivery and the industry’s ongoing competitive dynamics, the historical-analytical perspective may seem unconventional. Nevertheless, the historical perspective promises to shed light on developments that appear remarkable relative to initial expectations. Unique developments often follow an evolving logic. Comparative historical research has already gained considerable recognition for deciphering the processes that lead to the development of diversity. Historical sociology and historical institutionalism focus on how time, processes, and path dependencies influence institutions and social, political, and economic behavior and change. As a result, this research provides a comprehensive set of heuristic and analytical tools that are particularly valuable (Mahoney 2000; Beyer 2025).

The variety of heuristics that have been developed in comparative historical research to explain unique developments and institutional diversity can be described by the following five explanatory perspectives: Increasing Returns, Reactive Sequences, Critical Junctures, Path Creation and Incremental Institutional Change. Each of these perspectives is based on different assumptions about social change, the course of social processes, the relevance of events and the potential of actors to consciously influence the course (see Table 1).

The first heuristic to be considered, increasing returns, refers to the concept of path dependence in its original version. Path dependence emphasizes how initial choices influence long-term developments, often leading to unique outcomes and institutional diversity. This idea, formalized by economist W. Brian Arthur (1983), highlights that increasing returns occur when positive feedback loops enhance the appeal of an initial choice, encouraging its continued adoption. Arthur demonstrated that in markets characterized by increasing returns, early and seemingly small events can result in substantial advantages over time. Such dynamics mean that as more entities adopt a particular option, it becomes increasingly attractive for others to follow suit. Social science research has also embraced this explanatory framework, acknowledging that early events can significantly shape future developments. Once a „lock-in“ situation occurs, changing the course becomes challenging, as reversing the chosen path is fundamentally difficult. This heuristic has been utilized in comparative historical research to explain phenomena like unique employment regulation developments (Pedriana 2005) and the diversity of welfare state regimes (Pierson 2000). In the context of food delivery platforms, this heuristic would suggest a trajectory in which early events in the sector set in motion a self-reinforcing process based on increasing returns that may have made the employee courier business model in Germany particularly resilient. As a result, platforms employing couriers may have remained committed to this model despite the availability of potentially better alternatives or have become so entrenched in it that other options are no longer considered viable.

With the integration of path dependence theory from economics into the social sciences, new perspectives have emerged that attribute path dependencies to entirely different mechanisms. Among these is the concept of reactive sequences, introduced notably by Mahoney (2000). Reactive sequences are temporally and causally linked chains of events in which each event responds to the previous one and influences the next. This sequence follows a logical progression: event A triggers event B, which leads to event C, and so on, culminating in a final outcome. The occurrence of the final event depends on the initial event, emphasizing the importance of early events. Unlike the increasing returns heuristic, where patterns repeat, reactive sequences set off a series of interconnected reactions and counter-reactions that can be influenced only slightly or moderately. The events are closely linked by what Griffin and Ragin (1994) call „inherent sequentiality“. That is, they represent a cascade of events that follow the initial action (Falleti & Mahoney 2015). An example of a unique development explained using this perspective is the Industrial Revolution in England (Mahoney 2000), while the example of the emergence of the multi-technology lighting industry has been explained using reactive sequences to account for diversity (Onufrey 2017). When looking at the German food delivery sector, it is necessary to identify critical early events to apply this approach. Unlike the increasing returns model, this perspective requires mapping a coherent sequence of events leading to the adoption of employee-based delivery models. Each step – from event A to B, then to C and D – should follow a clear, identifiable logic.

The third heuristic considered relevant in historical-analytical research to explain diversity and uniqueness comes from comparative social research. Collier and Collier (1991) introduced the concept of critical junctures to explain the varying influence of labor movements on regime dynamics in Latin America. This concept has since been widely applied to exceptional events, such as the Charlie Hebdo attack (Della Porta et al. 2020). Critical junctures contribute significantly to understanding divergent social change by illustrating how paths branch at crucial points or during periods of significant change. They are defined as relatively short periods when the likelihood that agents' decisions will affect outcomes is significantly increased (Capoccia & Kelemen 2007). Their importance lies in their ability to initiate divergent paths, with decisions made at these junctures leading to alternative trajectories. A variety of event-driven development processes, including organizational decision making, can be interpreted using critical junctures. Research focuses on identifying moments or periods when decisions altered developmental trajectories. In examining Germany’s unique development of platform-based food delivery through the lens of critical junctures, it would be essential to identify the moments or periods that were crucial in charting an alternative path.

Another perspective on social change in comparative historical analysis focuses on the deliberate creation of path-dependent developments (Garud & Karnøe 2001). The concept of path creation is a derivative approach of path dependence research. It was originally developed in management studies but is now also common in historical institutionalism. Proponents argue that paths are not simply the result of random events or external shocks, but are the result from mindful deviations and deliberate actions taken by actors to shape the future. This perspective suggests that initial conditions are constructed, contingencies emerge as embedded contexts for ongoing action, and self-reinforcing mechanisms are strategically manipulated by actors. The potential to influence developments is considered moderate to high. This approach emphasizes the role of inventors, innovation pioneers, and institutional entrepreneurs (Garud et al. 2007). Research using the path creation heuristic includes the unique development of Finland’s knowledge economy (Schienstock 2007) and processes of company diversification (Li et al. 2024). With regard to Germany’s exceptional development in the food delivery sector, it would be essential to identify actors who have influenced this development through deliberate decisions, institution-building activities, or innovative actions. These could be individual or collective actors interested in preventing the expected disruptive development based on gig work opportunities. Documenting the decisions and actions deliberately contributing to this unique development would be crucial.

The fifth and final heuristic, developed by Streeck and Thelen (2005), is the concept of incremental institutional change, which describes forms of gradual but transformative change that influence social dynamics. Key variants include displacement, layering, drift, and conversion. Displacement introduces new practices that gradually become dominant, replacing older ones. Layering adds new rules to existing structures to modify them without a complete overhaul. Drift occurs when institutions evolve unintentionally through shifts in practice or interpretation without deliberate reform. Conversion occurs when actors interpret unchanged rules to suit their interests and reorient institutions toward new goals. This framework has been applied in historical-analytical research to explain phenomena such as the unique development of the Danish road and park sector (Lindholst et al. 2016) and the diversity of housing regimes (Bengtsson & Kohl 2020). In the platform-based food delivery sector, identifying incremental institutional change would require observing incremental change over time, with one of these types of change driving the specific trajectory of the industry.

Based on the heuristics discussed, a historical perspective must consider several factors to understand a unique development trajectory. These include the initiation of platform-coordinated food delivery, the sequence of event chains, potential critical moments and decision phases, the actors and interest groups that served as institutional entrepreneurs, and whether institutional change was intentional or occurred incrementally and unintentionally.

These five different heuristics are used because they represent common historical-analytical interpretations of the shaping of social change. They have influenced the methodological approach and are appropriate to facilitate the interpretation of the findings. However, it’s important not to misunderstand them as alternative hypotheses. The aim of the analysis is not to test hypotheses, but rather to reconstructively decode the mechanisms that have influenced the course of social change. It is conceivable, for example, that a recombinant interpretation may become necessary, combining aspects of different heuristics to solve the puzzle at hand. It is also possible that the analysis will point to aspects that have not been considered by the common heuristics.

4 Data Collection and Methodology

Methodologically, this paper is based on a historical-sociological case study of the platform-based food delivery service sector. The central focus of this study is to reconstruct the sector’s developmental trajectory and the decision-making processes that have contributed to its distinctive development in Germany. This reconstruction follows an inductive approach aimed at identifying conditions that explain the choice of food delivery platforms over the self-employed model.

The historical-analytical case study is motivated by an interest in comparative development. Specifically tailored case studies are a common strategy in comparative sociology to investigate the specificity of cases (Dogan 2002). Germany’s departure from the self-employed model places it at an extreme position within the spectrum of possibilities. The analysis of an „extreme case“ is particularly useful from a comparative research perspective because extreme cases often have unique, complicated, and multifaceted histories. Analyzing such cases can lead to comprehensive case studies that yield rich qualitative data that is invaluable for understanding the complexities of a particular phenomenon (Siggelkow 2007). Extreme cases can also shed light on causal mechanisms. This can help researchers identify the main drivers of change or stability in a given context (Gerring 2007: 101–105).

The case study that underpins this paper used a variety of data collection methods. To reconstruct the history of the food delivery platform industry in Germany, newspaper articles reporting on food delivery platforms were analyzed. For this purpose, a database of 973 press articles from the period 1997–2023 was compiled. These press articles came from financial and business newspapers (Handelsblatt, Börsenzeitung), national dailies representing a range of perspectives from conservative (Frankfurter Allgemeine Zeitung) to critical journalism (TAZ), as well as weekly newspapers and magazines (Die Zeit, Der Spiegel). An initial examination of the coverage of these companies allowed for the reconstruction of sequences of events and the media interpretation of day-to-day events from the history of food delivery platforms in Germany. The press articles were subjected to a content analysis with a special focus on information relevant for „path tracing“ (Sotarauta & Grillitsch 2022), and evidence of events was specifically collected and combined into a timeline.

For more in-depth analyses, which also addressed questions about the business models of the platforms and the relevant factors influencing the design of work models, additional data collection was conducted. The method of expert interviews was used for this purpose (Gläser & Laudel 2010). Expert interviews are suitable as a data collection method for historical-sociological analyses for several reasons. They serve to verify the chronological sequence of events and contribute to a better assessment of the significance of individual events within the timeline. Expert reports and assessments also serve as an important corrective derived from press releases. In reconstructing processes and the driving mechanisms behind them, interviews are of paramount importance, especially when they directly address the perception of decision-making situations and the background of decision-making.^[2]

The analysis is based on a total of 30 expert interviews. Fifteen expert interviews were conducted with digital platform managers and decision-makers, policymakers and advisors, trade union representatives, and social security experts, all of whom are or have been directly responsible for the German context. A further 15 interviews were conducted with international decision-makers, stakeholders and policymakers from other country contexts and the European decision-making level in order to bring external perspectives on Germany and comparative views into the study.

All interviews were selected using a process of reactive sequencing (A leads to B and B leads to C …), meaning that the information obtained in the interviews was taken into account in the selection of further interviewees and the interview questions were tailored to the specific interview situation with the respective expert. This approach resembles the stepwise approach in grounded theory research (Glaser & Strauss 2017) but differs in that it did not prioritize maximum contrasts for case selection. Instead, it was driven by an interest in continuing the search for clues while incorporating newly acquired information. The interviews took place from December 2021 to January 2023 and, like the newspaper articles before them, were subjected to path-tracing analysis. The focus of the analysis was to identify explanations for the choice of work models and to identify relevant contextual conditions.

5 The Development of the Food Delivery Industry in Germany

The following section describes the development of Internet-based food delivery services in Germany. According to the historical-analytical heuristics mentioned above, it is advisable to reconstruct developments from their origins, because early influences on development trajectories must be taken into account, whether due to path-dependent increasing returns or reactive sequences.

The early phase of internet-based food delivery in Germany is significant in a comparative context because online food ordering has been available in the country since 1997 (StuttgarterZ_2012.07.22).^[3] A similarly early adoption of internet-based food ordering can only be found in the United States, where companies like „World Wide Waiter“ launched in the San Francisco Bay Area in 1995 and „SeamlessWeb“ followed in New York in 1999 as ordering platforms. In Germany, it was the website „Bringdienst.de“ that allowed users to order food from restaurants over the internet at a relatively early stage. „Bringdienst.de“ was developed as a proof of concept by the advertising agency Asco GmbH. Unlike the initially limited business models of U.S. pioneers – such as „SeamlessWeb“, which initially focused on employees from 200 law firms, brokerage houses, and investment banks – „Bringdienst.de“ aimed at the entire Federal Republic of Germany from the outset. However, „Bringdienst.de“ was not a food delivery platform at that time; it was an online portal that facilitated contact with restaurants that were capable of delivering food themselves.

The innovation of „Bringdienst.de“ was the aggregation of offers from different providers. For instance, users could click through to an offer from a specific restaurant in a city district. Ultimately, they end up on the restaurant’s homepage, which was not integrated into the online portal’s ordering interface (web.archive.org/www.bringdienst.de). Bringdienst.de charged restaurants a commission for listing on the web portal, but payment for individual orders was not processed by „Bringdienst.de“. At the turn of the millennium, electronic payment systems were not widely used in Germany, so payments were typically made in cash upon delivery (Zeit_2020.08.27). „Asco GmbH“, the operator of the „Bringdienst.de“ portal, profited not only from the commissions generated by the referral business, but also from the fact that it developed and marketed merchandise management systems for restaurants and delivery services tailored to Internet sales.

The previously broader product range of the „Bringdienst.de“ ordering platform became more focused with the integration of the „Pizza.de“ subsidiary in 2007. Since then, business has been conducted exclusively via the „Pizza.de“ Internet portal. „Pizza.de“ was founded in 2001 and was originally a search engine for pizza offers. After the merger, ordering and payment could now be done through „Pizza.de“ (web.archive.org/www.pizza.de)

Until 2014, „Pizza.de“ was the market leader in Germany for online food orders (FAZ_2014.08.14). At that time, restaurants paid a commission of 13 % of the order value for referring orders and delivered the orders themselves (HB_2014.08.15). The business model was profitable for the platform. It should be noted that „Pizza.de“ in Germany was the market leader in online delivery orders without employing its own delivery staff, either as permanent employees or as independent contractors. Instead, for more than 17 years, the market leader has helped establish a distribution system for restaurants in Germany that has motivated restaurants to deliver themselves. The founders of „Bringdienst.de/Pizza.de“ pursued a strategy of incremental business growth. A risky rapid scaling of the business model with the support of external investors was not the goal of the founders.

Figure 1:

Food Delivery Platforms, Timeline of Market Presence, 1997–2023

Note: Grey areas refer to legal predecessors and former brand names: „Foodora“ was a platform owned by „Delivery Hero“, and its predecessor was the platform „Volo“, founded in 2014 and acquired by „Delivery Hero“. „Lieferando“ has its origins in the two platforms „Lieferservice.de“ (2008–2011) and „Yourdelivery“ (2009–2011), which were merged to form the „Lieferando“ platform. „Pizza.de“ operated under the name „Bringdienst.de“ until 2007. The name change was due to a company acquisition and a focus on meal delivery.

In 2014, „Pizza.de“ was sold to the company „Delivery Hero“ for the sum of 290 million euros. Delivery Hero (founded 2011 in Berlin), which aims to become the leading delivery holding company in the world and seeks to expand its market presence with a wide range of local brands, fundamentally changed the strategy of „Pizza.de“. The new approach focuses on rapidly scaling business operations through significant advertising investments (HB_2014.01.06). „Pizza.de“ served as a platform and brand of „Delivery Hero“ in Germany until 2019 (see Figure 1).

Starting in 2008, „Pizza.de“ encountered new challengers in the German market who employed assertive market entry and swift expansion strategies to secure their foothold. The first was the Dutch food delivery service „Thuisbezorgd.nl“, which established a German subsidiary named „Lieferservice.de“. The Dutch parent company, renamed „Takeaway.com“ in 2010, quickly sought to establish itself in various European countries. Furthermore, in a rapid succession of events from 2009 onwards, several startup companies emerged in Germany to compete directly with „Pizza.de“. One of these was „YourDelivery“, founded in 2009, which swiftly transitioned from a B2B focus to targeting end consumers. Following this, in 2010, „Lieferheld“ was established. In 2012, „Delivery Hero“ acquired „Lieferheld“, making it one of its German subsidiaries while retaining the name „Delivery Hero“ for the entire group. Furthermore, in 2011, „Lieferservice.de“ and „YourDelivery“ merged to form a new German subsidiary under „Takeaway.com“ with the moniker „Lieferando“ (BöZ_2014.08.15).

This tumultuous founding phase, spanning from 2008 to 2015, was marked by the active involvement of incubator firms, business accelerators, venture capitalists, and various investment entities. Numerous venture capital firms, both domestically (e. g., Team Europe Ventures, Hasso Plattner Ventures, Rheingau Founders, DuMont Venture, Rocket Internet) and internationally (e. g., Macquarie, Moraún Investments, Mountain Super Angel, Wayra), played a pivotal role as financiers (HB_2013.01.04, HB_2015.10.30, BöZ_2014.04.15).

Competition escalated significantly during this period as the new platforms quickly engaged in cut-throat competition (Interview 12), even resorting to aggressive tactics such as undermining competitors by strategically placing fictitious orders at restaurants (tz_2021.02.20). The intensifying competition among these platforms, which included unfair practices, played a significant role in the decision to divest „Pizza.de“, despite its prominent position in the market.

In addition, this competitive landscape was profoundly influenced by the emerging concept of „winner take all“ in the digital economy (Kuchinke & Vidal 2016), which found favor with investors (Masters & Thiel 2014). The expectation that rapid growth in size was necessary to ultimately increase the chances of emerging as the winner in the „winner take all“-competition led all platforms to pursue rapid scaling of their operations, reduce commissions for restaurants and customers, and, like the original market leader „Pizza.de“, which was struggling to maintain its market position, significantly increase advertising spending. As a result of the increased competition, the commissions of the ordering platforms had fallen to as low as 8 % of the order value during this period, and the high advertising costs severely impacted the profit outlook. Although new companies entered the market, delivery remained the responsibility of individual restaurants in the vast majority of cases (BöZ_2014.04.15).

The journey to food delivery using platform gig workers began in this tumultuous founding phase, with a small startup called „Food Express“, founded in 2013. However, this company, which operated until 2016, was not a platform for customers to place their food orders, but rather a service provider for restaurants that preferred not to handle their own deliveries. „Delivery Hero“ was a co-owner of „Food Express“ from 2014 to 2015, gaining valuable insight into a possible change in business strategy, but then decided against the full acquisition of „Food Express“ that had already been negotiated, driving the company into bankruptcy (deutsche startups_2025.11.09). As a result, the role of the first food delivery platform to deliver food exclusively with gig workers fell to „Volo“, a Munich-based startup founded in October 2014. „Volo“ developed an innovative algorithm that could calculate the most efficient routes between drivers, restaurants, and customers while facilitating the acceptance of additional orders. This innovation sparked optimism within the company, as it aimed to reduce the logistical burden on restaurant owners managing their own delivery operations and allow customers to receive orders from multiple restaurants simultaneously (business insider_2015.05.27).

„Volo’s“ time in Munich was short, however. It was acquired by Berlin-based tech-investment firm „Rocket Internet“, which had previously invested in Belgian startup „Take Eat Easy“, known for a similar service model. After the acquisition, „Volo“ was rebranded as „Foodora“ and quickly expanded to other German cities and several European countries. Notably, „Foodora“ chose not to rely on independent delivery drivers in Germany (taz_2016.05.07), and instead operated its delivery service with employed drivers (Ivanova et al. 2018). In 2015, „Rocket Internet’s“ second venture, „Take Eat Easy“, also attempted to implement its own delivery model in Germany using self-employed bicycle couriers, but exited the market after only a few months of operation (business insider_2015.09.28).

British industry leader „Deliveroo“ entered the German market in 2015 with a business model centered on delivery by self-employed bicycle couriers, embodying the typical gig work model.^[4] By 2019, „Deliveroo“ had built a roster of 1100 freelance delivery riders and around 100 employees (Spiegel_2019.08.21). Despite significant efforts to solidify its presence in Germany, the company only managed to capture a low single-digit percentage of the market share for all platform-mediated food deliveries.

„Delivery Hero“ also responded to the new challenge posed by „Foodora“ and „Deliveroo“ by creating subsidiaries that also focused on delivery with their own delivery drivers. The companies „Valk Fleet“ (B2B) and „Urban Taste“ (B2C) were founded in 2015, but were only active for a few months each (MM_2015.11.10).

Starting in 2015, when a variety of different platforms were shaping the market, a period of market consolidation began, which ultimately resulted in only „Lieferando“ remaining as a food delivery platform by 2019. At the end of 2015, „Rocket Internet“ decided to discontinue its former brands „Take Eat Easy“ and „Foodora“. „Foodora“ was sold to „Delivery Hero“, in which „Rocket Internet“ subsequently took a stake. In 2016, in addition to „Take Eat Easy“, „Food Express“, „Urban Taste“ and „Valk Fleet“ exited the market, where they had each operated for a relatively short time. By 2017, the competitive landscape had already narrowed down to five remaining food delivery platforms, three of which belonged to the „Delivery Hero“ group. However, the multi-brand strategy did not pay off for „Delivery Hero“ (deutsche startups_2016.02.09). The former market leader „Pizza.de“ had already fallen to third place in 2017 with a market share of 25 %, ceding its leading position to „Lieferando“, which reached a market share of 43 % in 2017. „Lieferheld“, the second platform of the „Delivery Hero“ group, also overtook „Pizza.de“ with a market share of 28 %. The two platforms „Foodora“ and „Deliveroo“, which only deliver using their own couriers, lagged significantly behind with market shares of just 2 % each. While these two companies were able to charge significantly higher commissions from restaurants due to their delivery model and were able to target restaurants that did not offer their own delivery, they struggled to establish a strong presence in the German market, which has a long tradition of restaurant-managed delivery (see Figure 2).

Figure 2:

Market Share Food Delivery Platforms in Germany, March-August 2017

Source: Sullivan 2019

In early 2016, „Lieferando“ decided to enter the active delivery business in Germany in response to competition from „Deliveroo“ and „Foodora“. They offer this service on a limited basis to restaurants that do not handle their own deliveries (Interview 7). In 2019, only 3 % of all deliveries were made by „Lieferando’s“ own couriers (HB_2019.01.11). Therefore, three of the five companies in the market in 2017 had to choose between using self-employed gig workers or employed couriers for their food deliveries. These three platforms took different approaches: „Lieferando“ and „Foodora“ employed couriers, while „Deliveroo“ opted for a gig worker-only model.

During the market shakeout phase between 2015 and 2019, there were a number of protests and strikes among self-dispatching providers. „Deliveroo“ in particular was hit hard.^[5] In 2015, there were protests over the removal of weekend bonuses for riders. In 2016, protests focused on the refusal to cover the cost of repairing bikes. In 2017, protests broke out again, simultaneously at „Foodora“, over insufficient coverage of work-related expenses (such as bicycles, repairs, and mobile phone costs) and questions about shift allocation rules (tagesspiegel_2017.07.30). Protests against „Deliveroo“ resurfaced in 2018 over surveillance apps and suspected exploitation of bogus self-employment (taz_2018.04.12). In 2019, however, „Lieferando“ and „Foodora“ were the targets of protests over violations of minimum wage laws and unilateral, unsolicited terminations of fixed-term contracts.

In 2019, „Deliveroo“ exited the German market, marking the departure of the only provider that continued to operate with independent couriers in Germany. Shortly thereafter, in 2019, „Lieferando“ acquired the entire German business of „Delivery Hero“, along with all brands present in the market, which were then gradually phased out (Interview 22). As a result, „Lieferando“ became the only active food delivery platform in Germany for a short period of time (from late 2019 to mid-2020). With „Lieferando“, an organizational model had prevailed that relied primarily on delivery by restaurant staff, with temporary workers used only in rare cases and at significantly higher fees.

„Lieferando’s“ monopoly in the German market has stimulated the ambitions of other food delivery platforms, especially the international leaders „Uber Eats“ and „DoorDash“, which were not previously present in the country. In 2020, the Finnish company „Wolt“ entered the German market. While „Wolt“ operates internationally using the independent contractor model, it chose to directly employ couriers in Germany from the beginning (Interview 1). In 2021, in the midst of the COVID-19 pandemic and the resulting surge in demand for delivery services, three more food delivery platforms entered the German market, all of which primarily operate internationally with independent self-employed couriers. However, all three decided to adopt an employee model in Germany. „DoorDash“ announced its intention to employ its couriers through partner companies, but ultimately acquired „Wolt“, which was already established in the German market, prior to its planned entry in November 2021. Nevertheless, „DoorDash“ continued to operate in Germany for a few months before deciding in June 2022 to continue exclusively with „Wolt“ under its existing model (Welt_2022.06.01). In May 2021, „Delivery Hero“ returned to the German market with its brand „Foodpanda“. Like „Wolt“, „Foodpanda“ chose to operate with its own employees. However, „Foodpanda“ exited the German market just a few months later. Also in mid-2021, „Uber“ launched its food delivery service, „Uber Eats“, in Germany, choosing to work with logistics companies that employ couriers (Interview 25).^[6]

This overview of the development of food delivery platforms in Germany shows that the independent contractor model played a relatively minor role in this industry. Between 2015 and 2019, however, several platforms made use of it. In particular, the „Deliveroo“ platform, which was dominant in the UK at the time, tried to enter the German market with this model (Interview 7). In addition, subsidiaries of the food delivery holding „Delivery Hero“ and the tech investment company „Rocket Internet“ tested this business model, but exited the market after a short time. Since 2019, no platform has operated with independent delivery couriers in the German market, which is noteworthy because the international platforms „Wolt“, „Foodpanda“, „DoorDash“, and „Uber Eats“, which have entered the German market since then, all operate with the ideal typical delivery model using independent gig workers in their home base and in other countries.

6 The Continuing Influence of the Restaurant Delivery Tradition

To explain why the platform-based employment model with independent gig workers has not taken hold in Germany in contrast to other countries, several factors could be considered, such as new legislation, protests and countervailing union power, or customer reservations about this type of employment arrangement. However, based on the findings from the process reconstruction of the media coverage analysis and the evaluation of expert interviews, these factors played a secondary role.

Unlike other European countries where gig work has prompted the introduction of new legal regulations, the German federal government has not enacted any legislation apart from the regulation of ride-hailing in the amended Passenger Transportation Act. Although the Ministry of Labor, led by Social Democratic ministers since 2013, critically assessed the development of the food delivery market, no specific legislation was pursued due to unclear responsibilities and differing views between ministries.^[7] Instead, the focus shifted to regulation at the European level (Bundesministerium für Arbeit und Soziales 2021).^[8]

Regarding trade union activities, it can be stated that, particularly between 2015 and 2019, there were annual protests, spontaneous strikes, and the self-organization of delivery couriers within the Free Workers‘ Union (FAU). These activities were subsequently supported by ver.di and the Food, Beverages, and Catering Union. Strikes have impacted „Deliveroo’s“ choice of work organization (Heiland & Brinkmann 2020). However, the company responded to these strikes by exclusively relying on freelance riders, no longer employing riders at all. Conversely, we found no evidence to support a proactive response favoring the employment of drivers. The current and former leaders of the delivery platforms we interviewed did not cite forms of protest as a factor influencing their choice of delivery model in this direction. From 2019 onward, trade union activity in the food delivery sector has primarily focused on supporting the establishment of works councils at „Lieferando“ and representing workers in court cases (Neumann 2023).^[9]

For the vast majority of customers, the quality of the service is far more important than the working conditions of the providers, and as customers they are very limited in their willingness to accept higher prices for better working conditions. Only a small group of socially conscious consumers attach significant importance to the working conditions of couriers, but these consumers do not represent the typical user base of the platforms (Belanche et al. 2021). Research has shown that platform users consistently have a more positive perception of the flexibility benefits associated with rider activity than non-users (Healy et al. 2020: 14).

It is of greater consequence to consider the specific market conditions and competitive environment in Germany that have developed as a result of the actions of the original market leader, „Bringdienst.de.“ Since 1997, this food delivery service has focused its business strategy on restaurants that deliver their own food. Additionally, „bringdienst.de“ has provided support to restaurants in establishing their delivery services as a distribution channel through consultancy services. The pioneering role of this company, combined with the relatively long-standing tradition of food delivery in Germany, has created a somewhat atypical market situation for food delivery in the country.

Favorable conditions for food delivery by platforms arise when many restaurants are unable or unwilling to deliver themselves and when restaurants are willing to accept delivery fees of 30 percent of the delivery value (instead of commission rates that used to be below 10 percent and are currently around 14 percent). It would also be advantageous if there were a strong desire among customers to receive food from restaurants that do not deliver themselves and that differentiate themselves through other offerings, such as higher quality for the customer. However, these favorable conditions do not exist in Germany, as a significant number of restaurants have already established their own delivery services due to the long delivery tradition or show no interest in the delivery channel.

Potential benefits of platform-based delivery could include more efficient route planning or the ability to offer different dishes from different restaurants in a single delivery. However, these potential benefits have not proven to be central. A clear indication of the unattractiveness of platform-based delivery in Germany is provided by the business model of „Lieferando“. The company, which once had a monopoly on the German market, offers platform-based delivery for a higher fee, but has no intention of expanding this distribution channel.

„On average, we lose money on every delivery we make. […] The most profitable part of the business is simply passing the orders on to the restaurants. The marketplace model. […] and we also get cross-subsidies from it in logistics“ (Interview 22).

By focusing on facilitating orders, „Lieferando“ does not rely on the relatively few suppliers to make a significant contribution to its overall business results, as measured by the total number of orders processed. Instead, suppliers serve other important functions for „Lieferando“: they act as visible advertisers, ensure service availability, and provide options for the restaurants that do not operate their own delivery service. In order to prevent the underperforming segment of in-house deliveries from growing unnecessarily, „Lieferando“ does not offer its service around the clock and reserves the right to reject orders if its driver quota is exhausted. This creates a strong incentive for restaurants to choose the self-delivery option.

Ultimately, the gig worker food delivery model with self-employed riders failed in Germany due to the higher costs associated with it compared to restaurant delivery. Maintaining a courier fleet requires a high utilization rate of couriers as a fundamental requirement for business operations. In competition with other business models, it appears that profitable utilization cannot be achieved in Germany. The cost of deploying and coordinating the courier fleet is a significant expense for the platform, even with the low wages and payouts to the couriers, which, depending on the calculation, rarely exceed the legal minimum wage and may even fall below it if breaks and waiting times are taken into account.

Platforms that relied exclusively on self-delivery with self-employed riders struggled to keep their operating and coordination costs within acceptable limits, and most of them abandoned their business models after a short time in competition with other platforms in Germany. The longest-lived of these was „Deliveroo“, but even this company eventually exited the market very abruptly. The announcement of „Deliveroo’s“ departure from the German market was made on August 12, 2019, and the company ceased operations just a few days later, on August 16. This hasty departure was not initially expected, as „Deliveroo“ had recently announced plans to focus on major cities such as Berlin, Munich, Cologne, Hamburg and Frankfurt due to intense competition in Germany. The only reasons given for the abrupt exit were strong competition and a desire to focus on other markets. One industry insider described „Deliveroo’s“ withdrawal from the German market as the „nail in the coffin“ for the self-employment model in Germany (Interview 25).

Despite the specific market conditions in Germany resulting from a long tradition of restaurant delivery, it remains puzzling why new entrants such as „Wolt“, „DoorDash“, „Foodpanda“ and „Uber Eats“ have not attempted to re-employ independent couriers. With the exception of „Foodpanda’s“ parent company, they have no prior experience in the German food delivery market and generally prefer the ideal type of gig work with self-employed couriers in their home markets and other international locations. To understand this, it is necessary to identify a second key factor that characterizes the German case: the platforms‘ fear that labor authorities will classify couriers as bogus self-employed.

7 Fear of Courier Work Being Classified as Bogus Self-employment

From the perspective of food delivery platforms, the potential classification of independent work as „bogus self-employment“ in Germany represents a significant business risk. The verification of bogus self-employment is based on the German Pension Insurance (Deutsche Rentenversicherung, DRV). The DRV is authorized to determine whether workers are classified as employees or self-employed through a so-called status determination procedure. This determination can be made retroactively at the request of either the employee or the employer. Bogus self-employment is also checked during routine, random audits by the DRV, which can take place in any company.

The potential consequences and penalties for bogus self-employment are significant for business owners. They can include back payments of social security contributions for up to four years, fines and even imprisonment. When asked about the status determination process, a company representative explained:

„It’s about determining: are you self-employed or not? I mean really, or is it just bogus self-employment? And if you are bogus self-employed, then as an employer or client who then becomes the employer, you have a real problem. And there has been […] a lot going on over the years, where […] it became increasingly clear that the risk of reclassification exists, that this procedure will be carried out and investigated, and then it will be determined, ‚Hey, the thousands of drivers you’ve hired in recent years are not self-employed at all; they are all employees.’ That means you have to pay back social insurance contributions, you have to pay fines for incorrect registrations. The executives are personally liable“ (Interview 14).

The testimony of the company representative highlights that the company’s perception of the risk of bogus self-employment has increased over time. However, this perception of risk can only be partially correlated with enforcement practices in Germany, as there is no evidence that food delivery platforms have been specifically targeted by the DRV. Moreover, regular audits of platform companies have not revealed any specific irregularities, in particular no increased classification of courier drivers as bogus self-employed (according to DRV, Interview 24). On the contrary, political observers consider the control of bogus self-employment in the area of low-skilled jobs, such as those on food delivery platforms, to be inadequate. One interviewee described the situation in Germany as somewhat schizophrenic (Interview 21). There is insufficient monitoring of low-skilled work, while there is much more monitoring of high-skilled people, who should be less vulnerable to the risks associated with bogus self-employment.

This intense scrutiny of highly skilled freelancers by the German Pension Insurance (DRV) offers a possible explanation for the heightened perception of bogus self-employment risk among food delivery platforms. Indeed, one occupational group relevant to the platforms – independent IT experts – has increasingly faced reclassification (Verband der Gründer und Selbstständigen 2019). It is therefore likely that, consciously or unconsciously, platform-employed IT experts assume that pension insurance is strict in its enforcement of bogus self-employment. This strong perception of risk, as an unintended consequence of the rigorous monitoring of certain professional groups, has likely been projected onto couriers, who could plausibly be subject to significant scrutiny by the DRV.

A specific regulation on potential repayment obligations for misclassified employment status creates a cost risk for platforms that they are unwilling to bear under the current circumstances. In contrast to other countries where social security contributions are significantly less linked to labor contracts, the repayment risk is particularly high (Funke and Picot 2021: 358). The potential obligation to repay social security contributions increases the financial risk for continuing employers, and significant amounts can accumulate. These amounts play a role in the platforms’ decision-making processes:

„It’s not really a matter of going in and saying, ‚Okay, let’s see where this all ends up, and if it’s under a certain amount, we’ll do it, and if it’s over that amount, we won’t do it. It’s not like that. […] But just to underline the point, there was also this calculation, although it wasn’t really necessary in this case, because it was very clear that it’s not allowed, so it doesn’t make sense.“ (Interview 14).

From statements such as these, it is clear that the regulations on bogus self-employment in Germany contribute to food delivery platforms perceiving the risk of bogus self-employment as particularly high. Significant for the risk perception of food delivery platforms is the fact that they consider a court decision to be very clear, even though it does not directly relate to food delivery. In the interviews, a decision of the German Federal Labor Court (BAG) from December 2020 was repeatedly mentioned (Bundesarbeitsgericht 2020). In this ruling, the court found that an employment relationship exists between a micro-task platform and a worker, emphasizing the platform’s structuring of the work, which explicitly incentivizes continuous activity. However, the ruling is very cautiously worded, so it cannot necessarily be concluded that all platform workers are classified as employees (Stoffels 2020). Thus, it remains unclear what this ruling means for food delivery platforms, but it has significantly increased the risk perception of companies. This is particularly evident in an interview with a participant describing „Foodpanda’s“ plans to enter the German market in 2021:

„They were also looking at whether freelancers could be used at Foodpanda. There was, I really think it was at the end of 2020, just before Foodpanda launched, there was a decision in a court – I don’t know if it was a district court or what. It wasn’t about food delivery, but it was in some other industry where the court said that if […] the app serves as a means of coordinating work assignments, then […] it’s quite clearly an employee, and at that point it’s no longer a freelancer. So, okay, that was a relatively clear ruling“ (Interview 14).

Another company representative also refers to the BAG ruling when explaining that the legal situation in Germany is now such that work platforms cannot engage independent contractors. He points out that a court has decided that „if you work through an app – in a simplified way – then you are an employee“ (Interview 17). As a result of this ruling, delivery companies have come to the conclusion that they cannot employ independent contractors because the risk of platform workers being reclassified in court has become too high,^[10] according to the interviewee.

8 Discussion

In terms of the historical-analytical heuristics of social change, it is important to highlight two key points in time regarding the unusual trajectory that has led to the marginalization of the gig-worker model involving self-employed couriers. First, the starting point of the development in 1997 is noteworthy, particularly because it is relatively early by international standards, and because for a considerable period of time efforts were focused exclusively on equipping restaurants to deliver online orders. The concept of direct delivery by the platforms themselves was not recognized as a viable option until 2014. Second, an important factor in the German context was a court decision in 2020, which the platforms interpreted as a decision that greatly increased the risk of self-employed couriers being later classified as bogus self-employed. Since then, no food delivery platform that has entered the market has been willing to take on the potential risk of being classified as a company that relies on bogus self-employment.

The significant influence of an early market player that helped restaurants adapt to online orders without taking responsibility for deliveries may suggest a self-reinforcing path dependence, as this heuristic places strong emphasis on the early phases of developments. However, the increasing returns logic of this heuristic does not seem entirely applicable to the case at hand. Following the acquisition of „Pizza.de“ by „Delivery Hero“ and the entry of other food delivery platforms into the market, the profits from the original business model quickly turned into massive losses, which were accepted because the competing platforms believed they were operating in a winner-take-all market. Thus, it was not increasing returns but rather comparatively smaller losses that enabled business models that did not rely on self-employed couriers to sustain themselves longer in the competition. Moreover, all business models have been modified over time. For instance, „Lieferando“ has introduced a small-scale option for deliveries made by the platform itself. Other platforms now combine food delivery with the distribution of goods from warehouses or supermarkets. In principle, the entire sector is still searching for sustainable long-term business models today.

In light of the acknowledged significance of the early developmental phase, the heuristic of a reactive sequence represents an additional explanatory model that merits empirical investigation. In considering a sequence, it is noteworthy that the evolution of the online food delivery market from 1997 to 2023 can be divided into phases, during which the organizational forms, primary competitive goals, competitive logic, success criteria, and contractual arrangements with food couriers have changed over time (see Figure 3).

While this market development is not outside the realm of typical population ecology developments (Hannan & Freeman 1977), one might still question whether there is the „inherent sequentiality“ (Griffin & Ragin 1994) that would support the reactive sequence heuristic described by Mahoney (2000). The inherent logic leading from event A to B and C and so on is undermined by the fact that in each phase new actors emerged who influenced the market development. In the start-up phase, there were new start-ups and venture capitalists; in the market shakeout phase, trade unions were active and the courts played a role; and in the new entry phase, leading international competitors of „Lieferando“ entered the scene. The displacement of platform delivery services by independent couriers cannot be attributed as a logical consequence of the turbulent and fluctuating market development or to the specific sequence of phases, suggesting that this explanatory heuristic seems to be applicable only in certain aspects (such as the importance of the early phase).

There are also doubts about the applicability of the path creation heuristic. A comprehensive entrepreneurial master plan of „mindful deviation“ does not seem to explain the unique development in Germany. Early pioneers were either acquired or forced out of the market, and the market domination plans of incubators and tech investors often failed, leading to frequent shifts to other markets (regions and sectors). Moreover, it would be difficult to attribute any intention for path creation to the courts that had to decide on the potential bogus self-employment of platform workers.

Figure 3:

Phases in the Development of the Online Food Delivery Market in Germany, 1997–2023

Note: Own illustration; please note that the phases have different durations

The heuristic of incremental institutional change is also rather inappropriate in light of the turbulent developments in the food delivery market. Among the various forms of incremental change, the judicial decision regarding the employment status of platform workers could, with some goodwill, be partially categorized as a phenomenon of „drift“ (Streeck & Thelen 2005). This categorization arises from the judges’ interpretation of existing labor law based on case law, without any vested interest in expanding it. As a result, they modified its application to assess the new context of platform work without necessitating legislative change.

In contrast to the heuristics discussed so far, which seem plausible only in certain aspects but not comprehensively, the critical juncture heuristic is highly explanatory for the German case. The court decision of 2020 has significantly influenced the decision of internationally leading platform companies not to rely on platform coordination of self-employed couriers in Germany. This singular event, a court decision that did not even directly concern the food delivery sector, can be identified in the analysis as so significant that it seems justified to classify it as „critical juncture“ (Capoccia & Kelemen 2007).

For the period up to 2020, the early shaping of the market environment appears to have been significant, making it more difficult for platforms that relied on the delivery of food with independent couriers as a corporate strategy to gain a foothold in the market and remain there in the longer term. This cannot be described as a phenomenon of path dependence favoring a single business model, or as a reactive sequence guided by an inherent logic from a starting point to the current state. However, the early shaping was so powerful that what was initially considered the standard model of gig work no longer appears as a viable option for the German market. To conceptualize the less pronounced impact of the early market phase compared to the heuristics of increasing returns and reactive sequence (the exclusion of an option rather than the determination of a possibility), it is useful to refer to a „shadow of the past“ (Tolkien 1999)^[11] that darkened the prospects for Deliveroo and other platforms that had experimented with self-employed couriers. In the academic literature, the metaphor „shadow of the past“ is used inconsistently (Henry 1948, Reiter 1994, Poppo, Zhou & Ryu 2008). In this context, the term is used to describe the impact of an early innovator and the role of the duration of early stages in the formation of competition in a market.

9 Conclusion

Based on the analysis of the platform-based food delivery sector in Germany, it has become clear that the development of this sector has taken a route that has moved away from the typical gig economy model common in several other countries. This divergence is characterized by the fact that delivery workers are not self-employed gig workers, but employees, which guarantees them rights and benefits such as minimum wage, paid vacation and sick days. The historical-analytical case study approach used in the analysis has examined the trajectory and decision-making processes that have influenced the development of the sector, using various heuristics of the comparative historical approach to explain this divergence.

The early shaping of the market by a pioneering online portal had a decisive impact on the market environment. This pioneering company focused on enabling restaurants to deliver online orders without taking responsibility for the delivery itself. This strategy, combined with over 17 years of uninterrupted market dominance by the market pioneer, contributed to the establishment of a robust restaurant-based delivery system in Germany, making it challenging for platforms that wanted to establish a business model with independently operating couriers to gain a foothold. For platforms that focus on the brokerage business, it is easier to handle the remaining business with employed suppliers.

Another key event for the specific course of development in Germany was a decision by the Federal Labor Court in 2020, which increased the risk perception of bogus self-employment among platforms. This decision had a significant impact on the strategic decisions of food delivery platforms and meant that new companies entering the market no longer considered the gig worker model with self-employed couriers. Although not directly related to food delivery, the ruling was interpreted by platforms as a clear indication that self-employed couriers could be classified as employees, with significant financial and legal consequences.

The study finds that the critical juncture heuristic, combined with a shadow of the past effect of market constitution, best explains the unique development in Germany. The early market constitution created a context that made the gig worker model less viable, and the 2020 court decision acted as a critical juncture that solidified the shift to employee-based models. This combination of factors led to the current situation, where all major food delivery platforms in Germany classify their workers as employees.

With regard to the common heuristics of comparative historical research, the analysis indicates that the aspect of temporal duration is underspecified as a potential influencing factor. This is why we had to use the metaphor of the ‚shadow of the past‘ to better describe the role of the early phase of development. In the future, it may be appropriate to consider the duration of time spans in the design of heuristics, in addition to events, sequences of events, and process dynamics such as increasing returns.

The analysis also highlights the importance of historical context and specific market conditions in shaping the development of the gig economy. While the typical gig economy model of independent couriers has been successful in many countries, the German case shows that local factors such as market traditions, specific regulatory frameworks (in this case, the self-employment specification in the German Pension Insurance Regulation) and legal interpretations can significantly influence the adoption and evolution of business models.

In conclusion, the study underscores the need for a nuanced understanding of the divergent paths of the gig economy in different countries. The German case serves as an extreme case in the spectrum of possibilities, illustrating how historical market shaping and critical legal decisions can shape the development of the platform-based food delivery sector. This research contributes to the broader literature on the gig economy and platform capitalism by providing insights into the factors and mechanisms that drive divergent development trajectories.

Like any analysis, this one has limitations that should be noted. It cannot be completely ruled out that the selection of the experts interviewed may have influenced the results, and that a broader inclusion of other expert opinions could change them. In particular, we note that the role of strikes and protests has not yet been sufficiently clarified. It may be that the managers interviewed did not want to be seen as affected by such activities and therefore downplayed their actual impact. Based on the chronological sequence of events alone, a greater significance of the impact of protests may be possible. Despite the existing research on protests and union activities in the food delivery sector, we see a need for further research on the importance of these aspects for business decisions.

Another limitation is that in a case study, there can be no comparison of the findings with developments in other countries and institutional contexts. The data we have collected for other country contexts are still being analyzed. It appears that the identifiable diversity of organizational models (Amentowobla & Kirchner 2024, De Groen et al. 2021) is often due to country-specific developments. Therefore, there may still be significant research needs for historical comparative studies, such as the one we conducted in this study. One advantage of such studies is that they can more thoroughly examine the factors and mechanisms that lead to different business models, varying coordination methods of work organization, and contractual arrangements for service provision. As we hope to have demonstrated, the German case has lived up to its reputation in comparative country research, but for much more specific reasons than those used, for example, in the Varieties of Capitalism research (Hall & Soskice 2001) to characterize the institutional context.