Making apps,owning data: Digital sovereignty and public authorities’ arrangements to “byte” back

Abstract

This study tells the story of two public transport authorities (PTA) in Sweden who developed their own apps. Although this might seem trivial and far-detached from the critical issues discussed in organization studies, this story raises questions of great relevance for this field, namely how digital sovereignty is organized. Digital sovereignty refers to governments trying to take or regain power and control over “the digital,” which a small group of large tech companies have monopolized. Drawing on a 3 year qualitative study of app making, and using assemblage theory, this study shows that digital sovereignty is not only about controlling software development or data ownership, but also about re-configuring the organization in relation to digital artifacts such as apps. By bypassing procurement procedures and paralleling their IT-departments, the PTAs here display how digital sovereignty comes from “below,” originating from frustrated civil servants within the public authorities who literally tried to “byte back” in a digital world run by large tech companies.

Introduction

“. . .there’s an app for that!” (Apple slogan in 2009) (Chen, 2010)

Mobile software applications, or apps, are so common today that we tend to take them for granted. Rarely do we reflect on how these digital artifacts affect our lives, nor how they impact the structure and routines of organizations. Since smartphones were launched in the 2000s, the number of apps has skyrocketed, and their development and use has multiplied (Dieter et al., 2019; Miller et al., 2021). Apps are a ubiquitous form of software, used on tech devices such as mobile phones, tablets, watches and computers, aiding users in everything from social media outlets, banking services to travel planning and ticketing. Paraphrasing Thrift’s (2005:153) discussion on digital software, apps have “come to intervene in nearly all aspects of everyday life” and it is therefore not surprising to find them inexorably embedded in the operations of contemporary public authorities.

Eager to deliver welfare services to their citizens, many public authorities turn to subcontractors for the design, production and organization of software. Even though local subcontractors are occasionally involved in the delivery of software, there is an ample reliance upon a handful of multinational tech-companies and consultancy firms (Fischer et al., 2021). What sets apps apart from other forms of software is their capacity to continuously generate user data (Dieter et al., 2021). This has put many organizations in new kinds of situations. No longer do they own or control the data that their services generate, and many public authorities also realize that they are starting to lose (at least some of) the control over “the digital.” Relatedly, there is a significant unease in the European Union about “global tech-driven players. . .not always obey[ing] European rules and fundamental values” (European Parliamentary Research Service, 2020:1).

Reacting to this, public authorities across Europe are taking actions to establish or regain ownership, potentially also some degree of control, over “the digital” (Pohle and Thiel, 2020). These actions have given birth to the concept of digital sovereignty. Floridi (2020:369) argues that “[d]igital sovereignty, and the fight for it, touch[es] everyone, even those who do not have a mobile phone or have never used an online service” as it concerns corporate power. After all, the large tech companies have extensive expertise, know-how and computational capabilities, which they have been able to use aggressively to leverage power and gain nearly unlimited influence over the digital realm (Floridi, 2020). As such, the concept of digital sovereignty also intersects with contemporary debates on public values as both address issues of governance, control, and autonomy. Since public values also encompass normative principles such as justice, fairness, transparency, equity, and the common good (Bryson et al., 2014), digital sovereignty is frequently invoked with reference to these principles. While the study of digital sovereignty is in its infancy, lending itself to philosophical considerations, legal principles and technical issues (e.g. Floridi, 2020), it also brings forth organizational matters, as measures taken to control the production and results of software development ostensibly imply an organization of “the digital.” Indeed, as the concept of digital sovereignty has primarily been used to conceptualize why public authorities are beginning to take action in relation to data and digitalization (Couture and Toupin, 2019), less attention has been given to understanding how digital sovereignty is achieved by public authorities and what organizational arrangements are put in place as a result. ¹

Our aim is to do precisely this and explore how public authorities move toward digital sovereignty by making, owning, and controlling their own apps. Rather than purchasing apps off-the-shelf or procuring tailor-made solutions from established tech-companies, the public authorities in this study went through considerable organizational reconfigurations to make apps. To explore app-making and data ownership as instances of digital sovereignty, we have used concepts associated with assemblage thinking. Originally a French term (agencement), assemblage “refers to the action of matching or fitting together a set of components. . ., as well as to the result of such an action” (DeLanda, 2016:1). As such, we regard app-making as a socio-technical assemblage and this allows us to make sense of how various components (actors, digital artifacts and resources) are related and forged together in organizational reconfigurations. Our empirical field of study is public transport in Sweden. As in all Scandinavian countries, public transport in Sweden plays a significant role in how people commute and move around in metropolitan areas. However, digitalization in general and the advent of apps in particular has changed not only the way people travel but also the way public transport authorities (PTAs) plan and organize their operations. Our analysis builds on interviews with key stakeholders and significant documents such as project applications, official reports, and evaluations. Through these materials, we explore how the PTAs have re-configured themselves to develop and launch app-based ticket systems, ameliorating their respective “service offerings.”

The paper has the following structure. First, we conceptualize apps as digital artifacts. Here we draw explicitly on assemblage theory and its analytical concepts. Second, we provide details about who we interviewed and what documents we gathered, as well as how we built the formative stories pivotal to our analysis. Third, we turn to the cases, in which we narrate the app-making processes. What emerges here is a story of two organizational processes, one describing the paralleling of established ways of working with digitalization and IT-projects and the other describing how routines of public procurement are bypassed. Fourth, we weave together the concepts from assemblage theory with the formative stories to offer an analysis of how the PTAs not only re-configured and organized themselves to design and develop these apps but also moved toward digital sovereignty. Finally, in the conclusion, we return to our initial problematization, present implications of our analysis and pose some questions for future studies.

Organizing apps and apps in organizing

From hardware to software

Studies of technology go a long way back in organization studies (for overview, see e.g. Orlikowski, 2007, 2010; Ratner and Plotnikof, 2022). Although technology is a classic trope in organization studies (see Beyes et al., 2020), and ICT has breathed new life into this field of research over the past 20 years, studies of the relationship between data, apps and organizing remain wanting (see Fischer et al., 2021; Wirtz and Daiser, 2018). This is also true for much social theory, even though apps have received increasing scholarly attention. A growing literature exploring the social dynamics of apps—including studies of health-tracking apps and platforms (Dieter et al., 2021; Lupton, 2014), social media apps (Miller et al., 2021), and dating apps (Thylstrup and Veel, 2020)—has led to the formation of the field of “app studies” (see e.g. Lupton, 2014; http://appstudies.org). While not confined to media studies, the field of app studies still carries traces of its origin as a subfield within digital media studies, with its primary focus on sociological explorations of end-users and their experiences. But there are exceptions. In a special issue devoted to “apps and infrastructures,” Gerlitz et al. (2019) argue that the study of apps can be divided into three strands of literature. The first considers apps as “mobile media” and concerns issues of mobility, location and communication. The second focuses more on the business side of apps and the commodification of app based data. To this strand of research we can add a great variety of studies related to platform markets, such as Apple’s app store or Google Play for instance (see Gawer, 2014; Ghazawneh and Henfridsson, 2015). In the third strand, researchers mainly devote their attention to media infrastructure looking at network connections, data traffic flows and detecting app content or platform logics (Gerlitz et al., 2019:6). Beyond these three strands of research, Dieter et al. (2019:1) make a methodological contribution to the literature and propose that the study of “apps as software packages” is to focus on app stores, app interfaces, app packages, and app connections. Similarly, Gerlitz et al., (2019) propose seven “points of entry” to the study of apps. Notably, none of them propose organization or organizing as entry points or as a level of analysis.

Among organization-oriented researchers (see e.g. Mullan and Wajcman, 2019; Schüll, 2016), it is the smartphone, not the apps, that has attracted most attention. While the smartphone is described as a multi-purpose device with “substantial organizational force” (Whyte, 2020:430), it is the apps that make the device interactively operational and which assemble and coordinate both resources and actors across time and space. Indeed, the companies behind the most popular apps have become massive tech-giants and they operate in what seems to be a continuously growing app-market (Dieter et al., 2021). Subsequently, the app is the primary digital artifact as it serves both as a product-in-use - referring to the specific software on the hardware device—and as a product-in-the-making—a software evolving through continuous upgrades in design, services and ad-ons based on end-user experiences. Dieter et al. (2021) also point to this duality of apps: they are crafted to operate as concrete software objects, but they also undergo continuous transformations as they are exposed to, and interact with, various socio-technical environments (see also Zittrain, 2009 on software as “contingently generative”). Another organization-related distinction worth mentioning is that apps may be purchased off-the-shelf in the form of “white-label” apps, which include some generic features to which logos and brands are added, or, alternatively, they can be custom-built for a particular industry (such as education or public transport) or tailor-made to match an organization’s operations. In either case, end-users often play a significant role in the development. Unlike other types of software, the end-users of apps are usually not the employees of the organization but rather people external to it, like customers or citizen-clients.

We draw inspiration from prior research on software in organizations, such as studies on the use of enterprise resource planning (ERP) systems (Pollock and Williams, 2009), algorithms (Glaser et al., 2021), and digitalization (Ratner and Plotnikof, 2022). We also argue for organizational scholars to consider perspectives from social and digital media research (see Dieter et al., 2019; Gerlitz et al., 2019) and subsequently shift focus slightly, from hardware to software, from smartphones to apps. Despite our emphasis on software, our primary interest lies not in how software is used in or adopted by organizations, but in how apps are involved in processes of organizing. Although we agree with Whyte that “the smartphone is a small device that impacts large-scale organizing” (Whyte, 2020:433), we observe that it is the construction, distribution and use of apps which generate data, determine the phone’s functionality and its primary significance for organizing. Indeed, it is the apps that enable and hinder certain actions (Mullan and Wajcman, 2019). For example, some apps blur the boundaries between work and home (Derks et al., 2016) as “the office” moves into the commuter train and the homes of employees’, while domestic activities move into “the office” (Cecchinato et al., 2014; Lanaj et al., 2014). Our point, however, is that despite the considerable technological advances in the public sector over the last few decades, where analog and paper-based processes have transitioned into digital ones, neither apps nor the gradual erosion of digital sovereignty have received much attention thus far (Vincent, 2018:353; see also Fischer et al., 2021).

Assemblage theory

We approach app-making as an assemblage. We understand the processes of app-making as a provisional and situated composition of actors, gathered around the organization and governance of public transport and the becoming of the digital artifact: the app (Briassoulis, 2019). Assemblage theory is a relational approach that emphasizes emergence, becoming and the socio-material (Müller, 2015). While actors and artifacts are transient and exist in constant processes of making and re-making, the actors as well as the relations between them and the artifacts need to be repeatedly performed not to be dissolved (see Clegg et al., 2006:238).

Assemblage theory is built on Deleuzian and Guattarian concepts (DeLanda, 2016) and has been used to explore organizational processes in various contexts. We build on this stream of research and follow Savage, who suggests that an assemblage is a relational construct, composed of “heterogeneous and emergent component parts that are arranged together toward certain strategic ends” (Savage, 2020:319)—in our case, the production of novel traveling and ticketing apps. Assemblages are, as such, not a “random assortment of things”(Savage, 2020:325), but an arrangement of strategic relations between various components (groups of people, digital artifacts, policy documents, laws etc) that share certain capacities when arranged in a particular way. With an assemblage approach it is possible, we argue, to focus on these relations and the dynamics of power that enable certain arrangements to emerge, while constraining others from doing so. The specific manner in which components are assembled shapes the properties of the arrangement. If the components were to be organized differently, or if new components were added or removed, it would result in different properties and outcomes. As an assemblage is constantly in the making, it evolves in many different places at the same time (e.g. when the tech-provider updates the code for an app, when the PTA changes fare prices or when the traveler uses the app to find a particular destination). It is the relations between the different parts that produce the overall organizing properties of an assemblage, and according to Ureta (2015:12) there are always tendencies where some components work to stabilize the assemblage while others work to transform it.

In organizational terms, the properties of an assemblage are enacted, meaning that they may involve very different and seemingly unrelated practices, which together make up the assemblage, aiming to fulfill a strategic end (see Köhne, 2014:8). In these enactments of assemblage - or “the on-going labor of bringing disparate elements together and forging connections between them” (Murray Li, 2007:263—apps, or digital artifacts, play a significant role. Through an assemblage theory perspective, digital artifacts are regarded to be inscribed and embodied with ideas and interests. They “become holders and dynamic vehicles of human agency, therefore replacing humans in doing things and performing functions in complex networks of human and non-human agents (Francesco and Morner, 2020:73; Gehman et al., 2022).

We draw on assemblage theory to probe how app-making ostensibly orbits around the organizing of digital sovereignty. That said, assemblage theory should be regarded more as a set of loosely connected concepts than a coherent conceptual model. In addition to the overarching concept of assemblage, key concepts for our analysis are distributed agency, heterogeneous elements and forged connections—concepts that allow us to pose a set of questions to the studied phenomena and our empirical material. Distributed agency refers to how agency is distributed among actors, or components, in an assemblage and more precisely how human subjects inscribe agency onto artifacts, which then take over action programs and capabilities (Francesco and Morner, 2020). Heterogeneous elements are related to the multiplicity of components and how certain arrangements hold together without each component losing its distinctive character (Allen, 2011). An assemblage therefore has no definable essence beyond the relations between its components, as Savage (2020) argues. The concept of forged connections refers to the process that characterizes the relations between components and heterogeneous elements (Murray Li, 2007) and how these have been strategically arranged to form “an apparatus for governing” (Savage, 2020:325). So, when agency is distributed to (or delegated to, assigned to or absorbed by) digital artifacts, they become active components in relation to other components (subjects, policy documents organizations etc.) constituting a hybrid collective. If these relations are not repeatedly performed, or forged by multiple actors, the assemblage starts to dissolve and enter a process of disorganization.

Reflections on method and materials

Sweden in general and public transport in particular, is a rich site to study app-making. Despite being a rather small country (approx. 10 million inhabitants), the number of public transport users is relatively high, where some cities experience roughly 80% of all motorized travel via public transport. Compared to other similar countries, Sweden is also a forerunner when it comes to public transport app-making as several PTAs have moved away from smart cards to app-based ticketing systems (UITP, 2020).

Our analysis is based on a research project in which we followed the development of two public transport apps. The apps were developed by two PTAs. The first case is an app developed by Västtrafik, the PTA responsible for a daily ridership of 950,000 people in western Sweden. The second case is an app developed by Skånetrafiken, the PTA responsible for a daily ridership of 460, 000 people in the region of Scania. As with other Swedish PTAs, their income derives from 50% ticket sales (the so-called farebox recovery) and 50% taxes. In terms of organization, they are both part of the regional public administrations and as such are politically run organizations.

We have conducted a series of interviews with key actors, lasting about 1 hour each (approx 600 minutes in total) (see list of interviewees in appendix). We also read all documents related to app-making in both cases (approx 10 formal documents, or more than200 pages in total) and did participatory observations at bi-weekly online app development meetings (approx 10 hours in total). In addition to these specific cases we have long experience of researching public transport authorities, stretching back 7 years or more, which has given us a solid foundation for interpreting the material.

Following assemblage theory, we traced the development of the apps through their life-cycles, from an idea of how to reorganize sales to fully functioning ticketing systems. Using the key concepts, we traced how agency was inscribed onto the artifacts, how heterogeneous elements emerged as different groups of people and artifacts were brought together to develop the apps. Within each collective, we traced the relations and how the connections between different actors and artifacts were forged. This analytical approach has allowed us to construct what Rutzou and Elder-Vass (2019) call formation stories—stories that account for how “things come to be the way they are” (p. 411). In doing so, we traced how the apps developed and took the shape they did, as well as how the organizations meanwhile transformed.

From paper to data, from maps to apps: Public transport ticketing

Before narrating the development of apps, we would like to historicize ticketing technologies and contextualize our study. In Sweden, public transport ticketing has followed the same trajectory as in most other countries in Europe. For roughly 100 years, public transport ticketing systems did not experience any fundamental changes. Pre-purchased punch cards or paper tickets purchased on-board the bus or tram worked as proof of payment. Conductors, ticket inspectors, bus drivers or automatic turnstiles have all verified tickets in various ways throughout the years, relying on different technologies. At the beginning of the 1990s, many cities and PTAs underwent quite a dramatic transformation, as they moved from paper-based tickets to prepaid magnetic cards as payment for trips, with the Oyster Card in London being the prime example.

Following the introduction of smartphones in the mid 2000s, PTAs around Europe started to use apps for timetabling, allowing commuters to search for trips and connections between different modes of transport. After a couple of years, some of these apps also allowed commuters to purchase tickets (UITP, 2020). However, in many cases this took time. Long-term contracts with tech companies providing IT architecture have resulted in public transport ticketing systems being locked into specific solutions for periods extending up to 10 years or more, which often include prepaid, chip-based card systems. Consequently, the efforts of PTAs to digitize their services carry significant organizational implications. Bearing this context in mind, we transition to explore the formative narratives of app development in two PTAs.

Launching a new app in Västtrafik—a move toward digital sovereignty

“The new Västtrafikkortet is so smart that it is possible to put it in the hands of a child. [. . .] ” (Västtrafik, 2009)

In 2009, Västtrafik introduced its own prepaid smartcard using check-in and check-out (CiCo) readers, placed on-board vehicles and on train platforms. Even though this system was already in use in many cities around the globe and widely promoted by transport engineers, not least for its provision of detailed level of trip-data, it was soon criticized, both by commuters and politicians.

“We had a mixed pricing system with zone-structure and kilometer pricing. So, you had to be some kind of genius to be able to calculate how much a trip would cost.” (Susan, Head of Sales at Västtrafik)

In the region, there were over 70 zones, some of which overlapped, so pricing was not solely based on zones, but also on the distance traveled. With more than 200 different combinations of tickets, it was a very complex system (Västtrafik, 2013). This picture also emerged from surveys amongst travelers and on online forums (e.g. Uddevallabloggen, 2009a, 2009b).

These observations prompted Västtrafik to involve the users in their development projects. What the customers needed and how they viewed Västtrafik became crucial for how Västtrafik planned its operations and organized its work. But this insight is hardly new. According to Pinch (2016:331) “[c]ompanies deal with ‘user representations’ all the time and increasingly test their products or prototypes. . .” But for Västtrafiik, this meant adopting a branding-orientation and to forge closer connections to its users to understand how they perceived Västtrafik and its services. Based on this, Västtrafik soon realized that it needed to simplify its ticketing system.

Politicians in the region echoed the criticism voiced by the commuters. By inscribing agential properties to the CiCo card-based system, it was the technological system that failed, not the users (e.g. Francesco and Morner, 2020). The ticketing system therefore needed a major overhaul, and this coalesced in internal investigations. While these investigations looked to transform the geographical zones and pricing structures, the idea of a “digital sales channel” unearthed in 2014. “Well, it did not have to be an app, but we wanted to simplify payment, that was the important thing,” said Alice, the Manager of Innovation at Västtrafik. Early on, though, the idea of an app took hold, and through that the digital artifact was inscribed with agential force:

. . .for our part, it became very much that, okay, how do we do this: innovation in terms of organization, working methods, to really throw ourselves into something that we have never done before, to make sure we get a development organization, and then I really mean IT-developers, programmers, get agile principles put to work, with the roles that often agile projects entail with a Scrum Master and product owner.” (Alice, Manager of Innovation at Västtrafik).

Developing an app was new to Västtrafik. Even though the ambition was to develop a “digital sales channel,” while also simplifying the pricing and zone structures, this translated into an organizational reform, brought about by ideas of how digital innovations emerge. Hybrid elements started to take form, with new connections being forged between emerging, established and new groups of people around the idea of an app (e.g. Murray Li, 2007). Scrum Master and programmers were recruited, who came with new ideas and who wanted to explore how the app could be used. We were told by James, one of the newly recruited product-owners at Västtrafik, that both the CEO and the heads of sales supported the app development, but then you had “10 people who thought it was a playhouse and superficial, or whatever they said”.

Since app development was a new task for Västtrafik, new forms of working and routines were perceived as necessary. Susan, the Head of Sales, said that the most crucial thing in this process was “the non-involvement of the IT-department.” Avoiding their involvement, she argued, meant to intentionally bypass a department that mainly works with the administration of existing (IT)systems, not developing new ones. However, more changes were needed to the established routines. Usually, PTAs in Scandinavia use public procurement to purchase IT-services/products. The first app used by Västtrafik was developed by a German company, whose whole business model revolved around selling a generic app in the form of a search engine for timetabled trips available to PTAs across Europe. But rather than continue with this app, procure a new one or involve the IT department, Västtrafik recruited their own product-developers and IT-consultants with “the right'” know-how and expertise. By doing so, they also managed to avoid long-term contracts, which would have tied them to one or many contractors, who would not only own the app, but also charge extra for upgrading the app’s functionalities in the future.

During this time, in 2015–2016, the provider of the smartcard in Västtrafik was at the brink of bankruptcy and informally announced that they could not maintain and upgrade the underlying IT-systems. Västtrafik had 2400 buses using this system and the entire public transport system was virtually “dependent on this sole provider” (James, product-owner, Västtrafik). This boosted the processes of developing the app in-house. And recruiting highly qualified staff to work on the app went smoothly, said Susan, Head of Sales. For the development team, a lot of work went into simplifying the differentiated pricing structure and the geographical zone structure. Simplifying these things became the first priority, much aligned with the newly adopted brand-orientation at Västtrafik. Susan, Head of Sales, also told us that “the ticketing system is in itself quite uninteresting, what is crucial is what value it creates for the customer.”

The UX [user experience] part was super-good. And then we had a guy, he drew sketches and then we - three, four, five people who helped each other - went out to the Central Station in Gothenburg and sat and talked to travelers, and walked around and talked to people: If Västtrafik were to make an app, and it would look something like this, what would you say about that? (Susan, Head of Sales at Västtrafik)

Many thought it would be “cool” with such an app.

So, we got very good feedback from the beginning, and then we started, like, how should the buttons be placed, what tickets to include, and then. . . I do not know how many times [we did this], but we eventually had a prototype in the mobile (Ibid).

Speaking with the words of Murray Li (2007), the users were indeed one heterogeneous element forged into the app-making assemblage, as this was the place where the new developers worked and they wanted to test their ideas and gain some first hand UX-feedback.

Eventually, the CEO at Västtrafik followed the suggestions by the sales and development team, and decided that all ticket sales henceforth would be made through the new app, which would be marketed as To Go. When the app was launched on a small-scale trial, the group of trial-users could report to the development team directly through the app what was working or not working. Many ideas and suggestions emerged. But most of them were saved for later use. Gaining UX-feedback via the app echoes Pinch’s (2016:329) reflection, that the online world allows organizations to interact with its users in new ways, as “[u]sers can be solicited for their opinions and with digital products themselves.” Although forged connections between heterogeneous users and producers could be observed in this assemblage, the much-heralded ideal of co-production with end-users did not really materialize.

The assemblage literally bypassed public procurement routines and worked in parallel with the IT-department and established new structures and sales channels within the organizations. This dual bypassing and paralleling were not without complications, but it met with resistance and took its toll on some of the app developers. James, one of the key product-owners of the To Go app, left Västtrafik as a form of protest.

“. . .I resigned in protest precisely because it has been a rather, yes, cumbersome journey and you only have the strength to do good things for a while when you are opposed. If you constantly have to fight for everything and then when you have done it, it turns out that everyone thinks it’s great. If you do not have the strength, it gets a little stressful and you become an angry and bitter person. So, I actually resigned for that reason. Otherwise I had a lot of visions and thoughts on what the next area to work-through would be. But this did not happen, not this time.” (James, Product owner, Västtrafik)

James represented a new competence in the organization, being a product-owner in the app-development team. With lots of UX-thinking and ideas about the potential of the app, he challenged the established routines in the organization. His story reveals that app-making is not only about introducing or adopting a new digital artifact, but equally about bringing in new people with new knowledge and connecting them to existing components and thus setting processes of organizing in motion. Indeed, all this led to friction and eventually to his exit from the organization. However, since the app was launched in 2016, hundreds of thousands of travelers use it everyday to search and pay for trips. On the Apple Store, it has more than 70,000 reviews and an average rating of 4.4 out of 5 (Apple Store, 2022). It is continuously updated and new features are added with some intervals, which suggests that the assemblage is transforming with new heterogeneous elements being forged together and new constellations of actors coming together with this digital artifact.

App-making by Skånetrafiken—organizing digital sovereignty

An important step toward the development of Skånetrafikens’ app was taken in early 2014 when an effort was launched to transform Skånetrafiken as an organization. This effort was led by the Head of Digitalization who can be regarded as one of the components that initiated the transformation of the existing assemblage (Ureta, 2015). He describes how he “started to prepare the organization” for change and saw the development of an app as a natural step toward a digital transformation of Skånetrafiken. John, as we call him, has a long background in the telecom sector and for him the future for public transport ticketing and mobility services was obviously to be developed in relation to mobile phones.

Backed by the Transport Director at that time (the highest ranking manager at the PTA), and with some extra funding from the Swedish Innovations Agency Vinnova, John hired and set up a team consisting of a user-experience-designer (UX), a technical manager, an IT architect and an information architect. In terms of organization, these were people hired from “outside” Skånetrafiken and as such new kinds of knowledge tied to the organization. The specific objectives for this team “was not very super defined” as John phrased it, “but looking at the next generation travel system was like the idea in general.” However, John and his team early on started to think in terms of app development specifying what such an app may entail, that is, inscribing agency into the device of an app-in-the-making (see Francesco and Morner, 2020)

For the larger parts of Skånetrafiken’s organization, this team involved new kinds of roles that brought with them new kinds of expertise, knowledge, and ideas—challenging the existing assemblage and how the PTA used to organize development work. And as this app development assemblage started to take form, the app (or the idea of a possible app) was inscribed with agential forces as “someone” that could perform (or not) specific tasks or services and “something” that other actors could, and should forge connections with (see Murray Li, 2019). As an example, one of the first tasks for the team was to develop a prototype of what an app might look like and what kind of features it could have, and this was done in close proximity to the travelers. Inspired by “design thinking,” the team regarded the travelers not only as customers (buying tickets and riding buses and trains) but also as co-designers of the app and as such an important resource or component in the materialization of the app. Very similar to how they worked in Västtrafik, the team courted travelers at the train and bus station showing them drawings and cardboard box-versions of what an app might look like and how it would function.

This team and their prototype later on consolidated into a firmer, or official, project in 2016 constituting what Savage (2020) would call a strategic “apparatus for governing” the development. At this time, Skånetrafiken also declared an urgency to find a new ticketing system, as the contract with its current provider, a global tech company, was to expire in just a few years with no option to extend due to time-limitations in the contract (Public transport Board minutes, 2016). At the same time, Skånetrafiken argued that “new modern digital solutions for payment are key in achieving the visions and objectives” of their organization (Public transport Board minutes, 2016:2). Due to the fact that the current provider owned all code and backend system architecture, Skånetrafiken could not develop the existing ticketing system on its own. This left them with an option to procure a completely new solution. This, however, was met with strong counter-arguments:

A traditional procurement of such a large and flexible system is difficult to implement with quality as the market and expectations move fast, especially in the digital sector. . .In the case where a ready-made system solution (so-called “off the shelf” product) is procured, a strong dependence arises on this particular supplier’s ownership and competence. Even with a shelf product, large parts of adaptation must be added, to suit the requirements of the organization (Public transport Board minutes, 2016:6)

Following these lines of thought, the regional politicians took the decision, in late 2016, to let Skånetrafiken develop the new ticketing system of their own and continue the work that John and his team had started. In addition, they were also to restructure the prize zone system:

The goal is to have a large-scale solution for payments, information and services, which makes Skånetrafiken independent of, above all, the current supplier of ticket systems (Public transport Board minutes, 2016:2).

This was, of course, a series of rather big decisions as Skånetrafiken went from procuring a ticketing system (including maintenance and services) from “outside” their organization to incorporate not just a new ticketing system but also the knowledge, resources and personnel needed to develop and manage such a system. The existing assemblage was challenged and connections between new actors and new kinds of knowledge had to be forged and stabilized. The impact of this organizational turn toward in-house app development also went beyond Skånetrafikens’ organization. The decisions meant that neighboring regions would be excluded from the system, at least initially as their contracts with their provider (same as Skånetrafiken) run for a few more years. Ticket-integration (between the regions) would be sacrificed in the short term, but in the long term the plan was to invite them to collaborate and use the same IT-architecture and app as Skånetrafiken. However, due to Skånetrafiken’s decision to develop its own app, the former provider stopped supporting the existing system, which, of course, had a disastrous impact on the neighboring regions, who were locked into, and dependent on, that system.

Similar to Västtrafik, Skånetrafiken left their own IT department out of the process and decided to recruit new staff and place them in a newly launched Development Lab. There, they worked on the IT-architecture, revising the price zone system and developing an app. When staffing this lab, Skånetrafiken cherry-picked what was perceived as necessary in terms of resources and skills to assemble an app development team, John told us. In hindsight, this way of organizing digitalization has been regarded as successful, and subsequently been formalized into a “model for digital development” in Skånetrafiken’s entire organization. In essence, this model is a purchase-provider model where specifications are formulated by a group of executives and the work (in terms of coding, designing and implementing) is provided by a team of procured consultants (temporarily) assembled according to what is needed. The organizational configuration for how to develop an app has, as such, influenced large parts of Skånetrafiken, way beyond the people (and the digital artifact) involved in the development of this specific app.

The app-making assemblage grew in organizational complexity rather rapidly. To start with, the app project was accompanied with a backend project. So, the app initially generated two projects—one focusing on the software application as met by the traveler using a phone or computer, and one focusing on the data/information “behind the scene,” the IT-architecture. Both projects required knowledge and resources beyond Skånetrafiken’s ordinary staff (or at least that was how they argued), so great efforts were placed to incorporate new knowledge and expertise into the app development. However, two years in, the complexity of the app grew and two more projects had to be assembled, one dealing with customer accounts and the other directing the attention to different so-called sales channels. In terms of organizing, this meant that people who earlier worked with sales and accounts in more general terms (or in relation to traditional forms of tickets) now worked with a specific focus of sales and customer accounts in relation to the app, and this also called for more consultants to be procured.

While the app-making assemblage evolved over time and could be described as a compartmentalization of Skånetrafiken, not least as the organization was divided into different projects with specific objectives and goals as well as specific budgets, personnel and timeframes, its activities were at the same time in many respects governed by the agential forces inscribed into the digital artifact—the app. So, the app spawned four different but closely interrelated projects—agency was distributed between several components, and as such the need for coordination arose and an overarching program was put in place and added on top of the four projects. New people also had to be employed to coordinate these projects, including a program manager, a coordinator, a test and release manager and an information architect. However, four different projects (and a program) also meant four (at least) different conceptions of what the app is or should be. Each project had, in some sense, a different problematization of the app or different relationship to the app and perceived it somewhat differently illustrating the heterogeneous features of such an assemblage that holds together without each component losing its distinctive character (see Allen, 2011).

As the assemblage stabilized in this organizational set up, a first official release of the app came in 2017. This was a rather simple app where users could search for trips and buy single-journey tickets. Later, a series of functions were added. During the first 3 years, the app operated alongside the card-based ticketing system at the same time as users were encouraged to transfer to the app. In the beginning of 2021, tickets were sold only via the app and users not yet transferred from the old system had to visit a customer service center to do so. This caused hourly-long cues and a wave of criticism against Skånetrafiken.

Discussing the organizing of digital sovereignty

These two formative stories are similar in several respects, both theoretically and empirically. They both include large public organizations undergoing strategically designed digital and organizational transformations, and in both cases the development of an app was closely related to, and formative for, these strategies. Both cases illustrate a movement toward digital sovereignty. In Skånetrafiken the development of an app was an explicit part of a larger program to digitally transform the organization, while in Västtrafik it became a response to an allegedly much-needed reform of the ticketing system and pricing zones. Both cases brought forth problems related to the procurement of ready-made off-the-shelf software from established tech companies, and saw in-house app-making as a solution. This is in line with the discourse on digital sovereignty as they move toward a path, where they design, develop, and own their own software, which enables them to gain control and ownership of the data generated through the apps.

Based on these interrelated formative stories, we observed two specific features that illustrate the constitutive organizational effects of app-making as well as the subsequent movement toward digital sovereignty. Both PTAs circumvented their traditional ways of “doing” technology and/or service development in two interrelated ways; first, by bypassing public procurement and second, by parelleing their IT-departments. These organizational processes constitute what we call assemblages of digital sovereignty.

Organizational bypassing

“If you bypass someone or something that you would normally have to get involved with, you ignore them, often because you want to achieve something more quickly” (Collins English Dictionary, 2023).

Assemblages are generally guided by some form of strategic end or purpose, and in our cases we observed the bypassing of procurement as such a strategic end. Murray Li (2007:264) suggests that the purpose of an assemblage is “the will to improve: the attempt to direct conduct and intervene in social processes to produce desired outcomes and avert undesired ones.” This is evident in our cases. The PTAs wanted to simplify, modernize and possibly digitize their “service offers,” as well as avoiding technological and contractual lock-ins. In both cases, the organizations were explicit in their objectives to avoid long-term contracts with tech companies, where software updates would be expensive or even outright impossible due to technical specifications jotted down in a contract years ago. In both cases avoiding procurement was a strong argument for in-house app-making and moving the technology and data closer to the organization.

Organizational bypassing, we suggest, is a way to unsettle combinations of heterogeneous elements into an assemblage. According to Ureta (2015:12), there are often components that work to destabilize the assemblage, while others work to stabilize it. Even though both PTAs felt a need to update their ticketing systems, apps were not the only option, yet app-based ticket systems were selected nonetheless. Each app carried promissory capabilities and were consequently inscribed with agency. As artifacts become inscribed with agency, they take onboard action programs and provide new conditions for transformations (Francesco and Morner, 2020). Since the apps embodied different possibilities and constraints, the PTAs could use the digital artifacts and their proclaimed features to move beyond public procurement. Indeed, to some extent, it was the promissory capabilities of the apps that made this possible, encouraging each of the organizations to step outside their routines, or perhaps even compelling them to do so.

While components are forged together into an assemblage, each component keeps some original elements, whereby an assemblage may continue to transform “without actually ceasing to be heterogeneous” (Allen, 2011:154). Indeed, when connections were forged between the new app and the PTAs, new components, such as staff with new competencies became necessary for the PTAs. By relying on public procurement, the PTAs had previously been able to tap into cutting edge expertise offered by leading tech-companies. Buying white-labeled apps off-the-shelf would be cheaper in the short term and offer a stable design, they told us, but when updates or new features had to be added in the future, the contracted tech companies would add substantive charges. In short, as the development of the apps equipped the organizations with a new category of employees and ideas about UX and digitalization, they bypassed the usual ways of doing things and the typical channels of interaction within the confines of each organization.

Organizational paralleling

“If something has a parallel, it is similar to something else, but exists or happens in a different place or at a different time.” (Collins English Dictionary, 2023)

As Ureta (2015) argues, assemblages always include processes where some components work to destabilize or transform existing arrangements. In both our cases, the assemblages of public transport were challenged by the introduction of new components—in terms of digital artifacts through which new kinds of actors and knowledge were introduced that also encouraged new relations to take form and new connections to be forged. Through the introduction of new components, or what Deleuze and Guattari (1987) call processes of deterritorialization, the distribution of agency shifted or was negotiated. In both cases we saw that even the idea of a possible app caused the PTAs to reorganize their work and with that came a complexity of agential forces—UX expertise argued for the inclusion of the travelers, the sales departments argued for the necessity of a brand-orientation and for the app to simplify ticket sales whereas others stressed the importance of keeping up with the digital development of society.

Rather than involving their IT-departments, both PTAs established parallel processes to design and develop their apps. New staff with new kinds of knowledge were recruited. Even though these different agents had different interests and backgrounds, they were all arranged as a hybrid collective toward the strategic end of creating ticketing and traveling apps (see Savage, 2020:319). Agency—in terms of desirable actions, technological possibilities and restrictions—were placed upon, or inscribed into, the digital artifacts while in-the-making. Indeed, the app was inscribed with meanings and attributes of what it could and could not do (even long before any app materialized), and what expertise-knowledge and resources that were desired or needed. These inscriptions made some work possible within the PTAs while other work tasks became more difficult. After all, without the app and its perceived relations to sales, to digitalization and its inscriptions to various forms of potentiality, there would have been neither a development team nor specific recruitments or, for that matter, a need to engage in organizational paralleling.

Inspired by Beyes et al. (2020:1), we suggest that apps are not just “channels connecting one agent with another,” but also something that structures “conditions configuring the very possibility of agency.” This means that apps are not simply artifacts bestowed with agency, but also with power. As new staff come with knowledge that is in high regard, they wield influence on the organization. Thrift (2005, p 127) similarly argues that new privileged groups emerge with new forms of ICT and that these tend to take “ownership of a cultural style being retailed by the cultural circuit and more widely.” Indeed, the recruited software developers, UX designers and SCRUM-masters were not placed—as one might have thought—at the IT departments. Instead, they triggered internal reconfigurations and in both cases parallel processes were initiated.

Assembling digital sovereignty

Closing in on the core argument of this study, we now arrive at what we call the assemblage of digital sovereignty, that is, actions taken to forge connections between heterogeneous elements, including resources and actors, to gain control of “the digital” and govern its own user-generated data. After all, apps generate massive amounts of data and since this data may be used for various purposes, it has become somewhat of an organizational currency in contemporary society. Indeed, digital sovereignty lends itself to analyses of the big techcompanies’ monopoly of this increasingly valuable currency. Our study shows that digital sovereignty is not only about control and ownership of data, but also about reconfiguring the organization to fit, and be able to host and manage, a digital artifact such as an app. So, while app-making was made possible by organizational bypassing and paralleling, these configurations are interconnected with the notion of digital sovereignty in at least two ways.

First, paralleling meant bringing new knowledge and expertise into newly created units parallel to the organization and taking onboard new staff, resulting in forged connections between the organization and new “components” (at the same time as there was cognitive and social distance between “new” and “old” staff and expertise). Forging connections in this way is at the very heart of digital sovereignty as it generates a sense of control and closeness to “the digital”. Adding to this, both PTAs also gained control over their own apps and data by circumventing procurement procedures and long-term contracts with large tech companies, and avoiding the involvement of their own IT departments.

Second, it is well known that technological performance or superior design is not the main driving force in software development, but rather “non-technological forces” such as organizational acceptance and feasibility (Munir and Jones, 2004). Dominating technologies are accepted because many organizations are structured around them, by which they reify their significance (e.g. Francesco and Morner, 2020). To this we add that ownership and control of data is pivotal for understanding how “non-technological forces” shape app-based software development. An app generates data, and the owner of the app also designs how and when data is generated and potentially who could use the data (Floridi, 2020). As the two PTAs’ development teams imbued their apps with agency and bestowed them with promissory effects, related to how data could be used in the future to develop public transport services, they literally prefigured digital sovereignty (Couture and Toupin, 2019; Pohle and Thiel, 2020).

Conclusion

This is a unique moment: apps and the data they generate are underpinning nearly all economic life and thereby also processes of organizing. Our aim here has been to explore how public authorities make, own and control their own apps—or “byte” back—and by doing so move toward digital sovereignty. By analyzing app-making as an assemblage, we showed how digital artifacts bring together an array of heterogeneous elements, such as actors (commuters and citizens, IT consultants, officials), resources (funding, app-making and project management skills), and objectives (innovation, increasing market share and modal share of public transport). This meant that connections were forged between new and old elements into new arrangements (Murray Li, 2007), which, in turn, destabilized established relations and routines (Ureta, 2015). We conceptualized these new arrangements as organizational bypassing and organizational paralleling. Unlike earlier studies on technology in organization studies, which has explored how power is renegotiated with the introduction of new software or machinery (Bloomfield and Hayes, 2009; Labatut et al., 2012), or how new technology builds upon and adapts to existing practices (e.g. Munir and Phillips, 2005; Pollock and Williams, 2009, our study suggests that organizations move toward digital sovereignty by bypassing entrenched procedures (especially the public procurement of off-the-shelf software) and by paralleling existing units and departments (who otherwise work with IT, digitization and data).

While digital sovereignty has emerged as a growing concern among powerful governance actors, such as the EU and its member states, what these actors usually bemoan is the gradual loss of control over data and digital development (European Parliamentary Research Service, 2020). Underlying this call for digital sovereignty are also other concerns, for instance a perceived lack of innovation and competitiveness in Europe. Many of the most popular apps come from the US or China, not Europe. However, in our cases it was not the EU, the state or even the regional politicians who took the lead to “byte” back, but frustrated civil servants who felt locked-in by long-term contracts, and employees engaged in specific ticketing-related tasks in each of their regional public transport systems. So, unlike much of the current debates on digital sovereignty, we showed how this emerged from within each of the organizations, and not as a response to governmental policies on data and internet governance (cf. Couture and Toupin, 2019; Pohle and Thiel, 2020).

The current historical constellation we are in, marked by the prevalence of app-making and the pursuit of data ownership, presents novel challenges for both organizations and organization scholars. For instance, what are the implications for organizations when their operations become increasingly dependent on apps and the data they generate? Since apps are not just one thing, but an array of connected heterogeneous elements that take on different functionalities, such as an e-service, a platform, and a commodity (e.g. Fox, 2011; Lupton, 2014), to what extent are digital artifacts malleable to organizational practices? As these questions are preferably explored in various public sector organizations and institutional environments, our exploration of how public authorities “byte” back can hopefully serve as a starting point for such future studies at the intersection of app-making and digital sovereignty in organization studies and beoynd.