Introduction
Digitalization projects are widespread within public administrations across the world, yet they frequently fall short of delivering expected outcomes (Anthopoulos et al., 2016). This is also true for the Swiss public sector: media reports on IT failures in the public sector appear with striking regularity. These incidents have prompted criticism of poorly structured governance, overlapping leadership and unclear decision-making in some digitalization projects at various federal levels (NZZ, 2025a, 2025b).
Despite the public debate, there is little research into governance of digitalization projects within public administration (Magnusson et al., 2020). So far, studies have primarily examined the effect of technology and broadly digitalization on governance. However, governance itself has only rarely been treated as a potential cause for success or failure of public digitalization projects (Veeneman et al., 2017; Wilson and Mergel, 2022). There is a lack of consensus on whether appropriate governance structures exist to meet the challenges of the digital age (Hammerschmid et al., 2023; Mergel, 2019). Why do digitalization projects in the public sector fail at such high rate? Can governance provide an explanation? To examine the influence of governance on digitalization projects, the following research question is formulated: How does governance affect the outcomes of digitalization projects? This article explores whether varying forms of governance have different effects on the outcomes of public digitalization projects. To address this research question, this study examines two digitalization projects implemented by the same digitalization lab in a Swiss canton.
The study begins by outlining the theoretical foundations of the concept of governance, IT governance and project governance. On this basis, the study conceptualizes governance of digitalization projects in the public sector and develops a theoretical argument for how governance may serve as an independent variable potentially affecting the outcomes of public sector digitalization. Next, the study presents the research design, which is a comparative case study, getting its data from internal documentation and structured interviews. Through qualitative analysis of the collected data, the study identifies how governance affects the outcomes of digitalization projects. Findings indicate that the successful project is characterized by well-designed and thoughtfully implemented governance structures, processes and relational mechanisms. In contrast, the unsuccessful project lacks adequate governance structures in several critical areas. The study finds out that not only is the type of governance structure decisive, but also the extent to which stakeholders actively engage with governance issues, selecting, implementing, and continuously evaluating governance structures appropriate to their specific context.
The theoretical contribution of this article is threefold. First, it argues that governance should be considered as a factor influencing the outcomes of public sector digitalization. This challenges the perspective assuming that digitalization inherently resolves governance issues, while excluding the possibility that governance itself can shape digitalization outcomes. Second, this article conceptualizes governance as applied to public digitalization projects by bridging the literature on public governance, IT governance and project governance. Summarizing these theoretical arguments into a conceptual framework, we analyse how governance manifests in structures, processes and relational mechanisms and unfolds on the example of two projects implemented within the same digitalization lab. Third, by investigating the context of a public digitalization lab, the study contributes to the limited empirical evidence on the effectiveness of new governing structures for digitalization. Finally, practical recommendations on how governance of future digitalization projects can better guide the projects to positive outcomes can be derived from this study.
Theoretical framework
Digitalization project
We start our theoretical framework with the definition of the unit of analysis for this paper, which is a digitalization project implemented in the public sector. We define project as a temporary and innovative organizational structure that organizations need to implement for delivering ‘strategic’ (Sjöblom, 2009: 166) and ‘transformative’ change (Clegg and Biygautane, 2025: 23), since traditional line or matrix structures often cannot deliver them (Lundin and Söderholm, 1995; Mintzberg, 1989). In essence, projects are distinguished from other organizational forms by having defined timeframes and budgets, a clear objective, a designated team, and the intended changes they aim to achieve (Lundin and Söderholm, 1995).
We focus on projects because this shift toward ‘increasingly non-permanent and informal structures’ represents one of the most significant contemporary changes in public administration (Sjöblom, 2009: 165), and because in the public sector, digital transformation is mainly implemented via projects (Mergel, 2019). As for the definition of digitalization, we refer to Mergel et al. (2019) who defines digitalization as ‘transition from analog to digital public service delivery’ (Mergel et al., 2019: 10), and to Gobble (2018) who sees it as ‘the use of digital technology, and probably digitized information, to create and harvest value in new ways’ (Gobble, 2018: 56).
Project outcomes
For answering our research question, the concept of project outcomes should also be operationalized. The literature emphasizes the ambiguity of the project outcome concept, but concludes that most often it is evaluated along three axes: time, cost and scope (Conboy, 2010). Delivering a project on budget remains one of the most important elements for project performance, particularly for IT projects, as they ‘have far greater cost risk than is commonly assumed’ (Flyvbjerg et al., 2022). Time performance is closely linked to cost performance, as the latter frequently causes the former. These two dimensions of project success broadly define success as the efficiency of using dedicated resources. Focusing on scope helps us determine whether a project has also been effective, meaning the resources were not only efficiently used, but also used in order to reach the project’s goals. This provides us with three dimensions of project success: 1) whether a project has met the project timeline; 2) whether a project has met the project budget, 3) whether a project has met the project goals.
Governance
The concept of governance has evolved from traditional forms of legislation and enforcement to the governance of complex interrelations. Governance is an institutional solution to collective action problem (Filgueiras and Almeida, 2021) and ‘a holistic steering of political and societal change’ (Damkowski and Rösener, 2004: 312 as cited in Schmid, 2019: 8). The concept of governance can be interpreted very broadly, which leads to a certain lack of clarity. For this reason, its core meaning has been summarized in the following four aspects:
Governance aims to manage actors.
Governance is based on institutionalized rules that actors are expected to follow.
Governance defines patterns of interaction that emerge during steering processes.
The course of these processes, organizational boundaries or even the boundaries between state and society can be crossed (Möltgen-Sicking and Winter, 2019).
According to Markus and Meuche (2022), governance has a task of monitoring the achievement of objectives. For a long time, conditions in public administration were stable. As a result, control was a suitable governance instrument. Today the environment is characterized by uncertainty, ambiguity, and constant change. The ability to anticipate potential changes is becoming increasingly important (Markus and Meuche, 2022), making it necessary to find new approaches to governance (Jahn et al., 2019; Mergel, 2019). The digital age demands that governance supports change processes, fosters relationships with and among stakeholders, focuses on promoting transparency, more direct and efficient forms of communication, and introduces new roles and governance mechanisms that span all organizational levels (Ritz and Thom, 2019). Furthermore, all stakeholders must be aware of governing rules—in other words, governance must be transparent (Filgueiras and Almeida, 2021; Schwertsik et al., 2010).
Given all these requirements, it is not easy to find an appropriate governance form, nor is it clear which factors must determine this decision (Kuhlmann et al., 2019). At the same time, continuous improvement must be possible without constantly introducing fundamental changes (Rentrop, 2024). Further question arises as to where the governance of digitalization is actually enacted: is it situated at the organizational level, within the organization’s IT department, or at the level of individual projects? In the subsequent paragraphs, we systematically examine each of these levels in turn and their implications for governance of public sector digitalization.
When applied to the context of digitalization, governance has been commonly conceptualized as IT governance, considering IT as the organization’s resources which should be governed through a combination of dedicated governance structures, processes, and relational mechanisms (Denford et al., 2015; Rentrop, 2024; Weill and Ross, 2004). Structures refer to the presence of leaders and committees within the organization who have decision-making authority. Processes enable participation in decision-making and oversight, defining how decisions are made, controlled, monitored, and evaluated, and requiring ongoing adaptation to the changing environment (van Grembergen et al., 2007). Relational mechanisms include strategies that promote good working relationships, reciprocal communication, effective collaboration, and the leveraging of synergies between the business and IT functions of an organization (Tonelli et al., 2017). The aspects of structures, processes, and relational mechanisms are implemented differently in each organization—depending on which governance decisions are made as the manifestation of contextual factors.
Focusing precisely on structures, processes, and relational mechanisms, appropriate governance can be defined as follows. Appropriate governance structures mean that they are defined and well known, with the ultimate goal to avoid informal and parallel governance structures (Rentrop, 2024). Processes are appropriate when they involve specialist teams in decision-making, discuss goals with stakeholders and integrate their perspectives, ensure compliance and risk management (Rentrop, 2024; Schwertsik et al., 2010), while remaining continuously evaluated and adjusted (van Grembergen et al., 2007). For relational mechanisms, clear communication of goals and tasks, good working relationships, effective collaboration, and the use of synergies between business and IT functions within an organization are of highest importance (Tonelli et al., 2017).
The current reality though clearly demonstrates that digitalization is not implemented via IT units of organizations anymore: digitalization projects happen quite routinely in every department of an organization, and even more, they frequently span across the silos and organizational borders, leading to the creation of cross-departmental governing units aimed at digitalization (Mergel, 2019). Separate organizational reform units are commonly created for the purpose of exploration (Fuchs et al., 2019). Unlike the core organization, such units can operate with greater flexibility and in an experimenting nature (Fuchs et al., 2019). For instance, innovation labs are seen as islands within a public organization where experiments take place (Tõnurist et al., 2017). With the increasing push for digitalization in the public sector, dedicated governance units focusing on digital transformation have emerged. Mergel (2019) considers digital service teams as a ‘new IT governance structure’, a ‘new form of organizing and implementing digital transformation’ and as ‘independent units focusing on high profile projects’ (Mergel, 2019: 3). Clarke (2020) calls them digital government units and argues that they attempt to operate ‘free of the legacies of hierarchy, silos and traditional bureaucratic processes’ (Clarke, 2020: 364). Despite the different naming, all of these units have something in common: they are separate organizational structures, designed in a way which is supposed to be favourable for reaching positive outcomes of digitalization, and they main mode of delivery is projects. However, such units are not novel. They were also implemented for other type of reforms, such as New Public Management, where trial regulations and separate reform units were implemented to allow public organization to go beyond their mandate (Ritz, 2003). Research so far has been concerned with comparing such units across the countries or regions, and not focusing so much on trying to explain why even within such units their projects reach different degrees of success. Given that previous research has demonstrated that specialized governing units created for digitalization frequently do not have an option to fail, despite the presence of the failure tolerant culture – due to temporality of such units, they need to ‘demonstrate success early in order to justify investment in their resources and powers’ (Clarke, 2020: 368) – the question of varied success levels of projects within such units is even more worth investigating. From this development the need for investigating governance structures also at the project level arose (Magnusson et al., 2020).
To meet the challenges of the information age, different governance forms have been developed to lead digitalization projects (Thesing et al., 2021). Classical project management is characterized by clear planning with the set deadlines, implementation, and distinct communication at the very beginning of the project (Boots et al., 2024; Thesing et al., 2021). Such approach promises ‘standardized project workflows, a common understanding among all participants, improved internal and external collaboration, interchangeability of project participants, comparability of projects and their work progress, and improved management of portfolios and programs’ (Hilmer and Krieg, 2014: 48–49). The disadvantages, on the other hand, are requirements that are too abstract or misinterpreted, which lead to unforeseen delays or problems during implementation, because correcting errors is too expensive when everything has been planned (Thesing et al., 2021). Furthermore, customers are often overwhelmed with having to formulate all requirements at the beginning of a project, despite the uncertainty. Ultimately, this can mean that a lot of time is invested in planning something that later cannot be implemented (Thesing et al., 2021).
Due to these disadvantages, other project governance approaches have been developed, characterized by less standardization, higher flexibility and greater user centricity (Clarke, 2020; Thesing et al., 2021). The most well-known one is agile governance. Agility offers flexible, adaptable, and responsive organizational structures (Fischer et al., 2020). The twelve principles of agility can be divided into three core concerns: ‘promoting collaboration in interdisciplinary, cross-functional, and largely self-organized teams, cultivating customer-centred innovation, and systematically involving customers, establishing a mode of experimentation and learning’ (Stucki-Sabeti et al., 2022: 116). Whether agile project governance is more promising than classical one depends on a project. Furthermore, a combination of both methods can be pursued (Thesing et al., 2021).
Despite the variety of existing governance forms, authors agree that suitable governance for public sector digitalization is still lacking (Liu and Yuan, 2015). Governance must fit both the organization and the specific project (Luciano et al., 2017). In many cases, a corresponding governance has neither been defined nor established (Markus and Meuche, 2022). When an organization talks about, defines, and communicates the corresponding governance, a good basis for a digitalization project is already laid (Markus and Meuche, 2022). Summarizing these theoretical arguments into a conceptual framework, in our comparative case study we analyse how governance which is manifested in structures, processes and relational mechanisms unfolds on the example of two projects implemented within the same digitalization lab. That being said, aligned with the literature we focus on three aspects of governance – structures, processes and relational mechanisms, and analyse them at two levels – at the level of individual projects and at the level of a digitalization lab within which these projects are located. The figure below presents the interplay between different levels at which governance of public sector digitalization can potentially be analysed (see Figure 1). The Figure 1 presents in light blue the two levels that will be analysed in this paper. It also indicates which aspects of governance we will be focusing on within these two levels. Further studies can analyse governance at all three levels, at a combination of two (especially in cases when a specialized digitalization unit is not present), or at one level. Further studies might also adopt another operationalization of governance, and, therefore, focus on different aspects of it by slightly adjusting this figure for individual purposes.
Figure 1
Cascading governance exploration (source: authors).
Methodology
To answer the research question of how governance affects the outcomes of digitalization projects, we follow a qualitative approach. We design this study as a comparative case study (Yin, 2009). The unit of analysis is individual digitalization projects implemented under the steering of the cantonal digitalization lab. This allows the two projects to be explored within the similar organizational context using the embedded research design (Cresswell, 2007). The two selected projects were of the first projects to be implemented by the lab, so learning from these cases might be of particular importance for the future of the lab, thus, promising practical benefits of the study. The organizational conditions in the studied canton are extraordinary due to the presence of this lab, making the context of significant interest for other cantons, federal countries and other jurisdictions considering the creation of the specialized organizational units for digitalization.
The theory shows that the governance structure of both the organization and the project can have a significant impact on project success. Our design allows to capture these influences and therefore go ‘beyond an individual project level’ (Mergel, 2019: 2). To avoid selecting only successful, or only failed cases, we followed a polar case sampling technique, and chose two projects with varied outcomes: one successful project, and the other which has experienced some difficulties in reaching the expected outcomes. This allows us to control for the context variables, assuming they are the same to a larger extent.
Data collection
The data collection spanned from July to December 2024. In the preliminary interview with the program manager responsible for digitalization in the canton, we received the overview of all digitalization projects completed within two years from which we selected two polar cases. In the next step, the project managers were contacted for the information on organizational structures and project documentation. For both projects, we received 23 document items, including contracts, organigrams, status reports, and communication concepts, to name a few. The thorough analysis of this documents allowed us to enhance the interview guidelines with more detailed knowledge about the project, and to get the overview of the governing stakeholders, so that we could select the interview partners purposefully, as per their governing role in the project, rather than snowballing (see Appendices for the general interview guideline). To obtain different perspectives, participants in various roles were interviewed. Seven interviews were conducted per project (see Table 1). The format of structured online interviews was selected. The average length of an interview was 57 minutes. All interviews were recorded for the further analysis.
Table 1
Overview of the interview partners.
| FOCUS | POSITION | DATE |
|---|---|---|
| Cantonal digitalization lab | Program manager responsible for digitalization in the canton | 04/07/2024 |
| Project A | Project sponsor | 06/12/2024 |
| Quality risk manager | 19/12/2024 | |
| Project leader | 29/11/2024 | |
| Supplier-side project leader | 04/12/2024 | |
| Sub-project leader | 04/12/2024 | |
| User representative | 27/11/2024 | |
| User representative/Member of the sub-project | 09/01/2025 | |
| Project B | Project sponsor and Member of the project steering committee | 16/12/2024 |
| Member of the project steering committee | 16/11/2024 | |
| Member of the project steering committee | 25/11/2024 | |
| Project leader | 19/12/2024 | |
| Project team member | 13/12/2024 | |
| Lawyer | 13/12/2024 | |
| User representative | 02/12/2024 |
Data analysis
The interview transcripts were coded manually by two authors. Categories and corresponding codes were derived from theory as well as created during the interview analysis (see Table 2), following the abductive approach (Miles et al., 2014). In the first step, all transcript sections were deductively assigned to the pre-defined codes. In the second step, codes were created inductively based on descriptive coding by topic. All inductive codes were then integrated into one of the deductively developed themes, namely, Governance structures, Governance processes, Relational mechanisms, and Outcomes (see Code tree in the Appendices). The quality criteria that come into play are construct validity, internal validity, external validity, and reliability (Yin, 2009). Construct validity is increased by using multiple sources of evidence. Furthermore, the analysis process is described transparently and followed accordingly. Internal validity is supported by the consistent implementation of the chosen research design. External validity can only be achieved through further studies. Since this is an exploratory study, it cannot be assumed that the findings are universally applicable to all other digitalization projects. Reliability refers to whether the study is replicable. We ensure this through the usage of standardized interview guides and transparency in dealing with the transcripts, the codes development, coding and analysis.
Table 2
Codes development.
| CATEGORY | CODE | LITERATURE SOURCES |
|---|---|---|
| Governance structures | Structure of the project | Sjöblom, 2009; Clegg and Biygautane, 2025; Rentrop, 2024 |
| Changes in structure | ||
| Legal foundations* | Emerged inductively from the data | |
| Role description | Denford et al., 2015; Rentrop, 2024; Weill and Ross, 2004 | |
| Time of joining the project | ||
| Full-time position | ||
| Temporary position | ||
| Involvement in other projects | ||
| Task definition | ||
| Unclear tasks in role* | Emerged inductively from the data | |
| Dissatisfaction with role | Denford et al., 2015; Rentrop, 2024; Weill and Ross, 2004 | |
| Resource feasibility | ||
| Project leadership | ||
| Governance processes | Governance processes | Möltgen-Sicking and Winter, 2019; van Grembergen et al., 2007; Rentrop, 2024; Schwertsik et al., 2010 |
| Participative processes* | Emerged inductively from the data | |
| Processes for determining framework conditions | Lundin and Söderholm, 1995; van Grembergen et al., 2007; Rentrop, 2024; Schwertsik et al., 2010 | |
| Flexibility in processes | ||
| Impact of leadership* | Emerged inductively from the data | |
| Relational mechanisms | Challenges in coordinating stakeholders | Denford et al., 2015; Rentrop, 2024; Weill and Ross, 2004; Tonelli et al., 2017 |
| Communication* | Emerged inductively from the data | |
| Contact with stakeholders | Denford et al., 2015; Rentrop, 2024; Weill and Ross, 2004; Tonelli et al., 2017 | |
| Collaboration in the project | ||
| Collaboration across departmental boundaries | ||
| Collaboration with external parties | ||
| Project outcomes | Meeting project budget | Conboy, 2010; Flyvbjerg et al., 2022 |
| Meeting project timeline | ||
| Achieving project goals |
[i] The asterisk (*) indicates codes that emerged inductively from the data.
Case description
The legal mandate of the cantonal digitalization lab is to represent the canton and its municipalities in the field of public sector digital transformation, promote cooperation with the federal government and other public institutions, and foster effective and efficient service delivery by the canton and municipalities in this area. The unit defines their main task as providing electronic services for citizens and businesses. This relatively new digitalization unit is well-organized, with secretariat, the audit office, cooperation committee and planning committee, composed of members from the cantonal government, cantonal administration and municipal representatives.
Both projects were of the first projects of this lab. Project A is introducing a new digital solution which usage is mandatory for the canton and its municipalities. This solution digitalizes existing processes, aiming at standardizing and simplifying them. The development of the technological solution was outsourced to a private company through the public tender procedure. The project B is introducing an application for the canton and its municipalities. It aims to make existing processes more efficient and transparent. The specialized application will replace all previous tools used, introducing a canton-wide standard. Other project goals include simplifying and ensuring uniform, fully digital processes and templates, using the same data within the system, and increasing service orientation towards the public. The involved actors are to be connected digitally and seamlessly through integrated processes.
Results
The result section presents the analysis of two levels of governance: first, at the level of the cantonal digitalization lab which steered two analysed projects, and then at the level of the individual projects, comparing them along three elements of our governance framework – structures, processes and relational mechanisms (see Table 3 below for comparison summary).
Table 3
Comparison of two projects.
| COMPARISON CRITERIA | PROJECT A | PROJECT B |
|---|---|---|
| Governance structures | ||
| Governance mode | Internal hierarchies, contractual collaboration, and horizontal network | Internal hierarchies, contractual collaboration, and horizontal network |
| Project sponsor | Stable | Changed |
| Project leadership | From the respective department; Stable and positively evaluated by all stakeholders | Originally from digitalization lab – without any subject expertise, then – from respective department; Not-stable |
| Governance processes | ||
| Stakeholders’ flexibility | Required and provided | High, but limited by insufficient time and human resources |
| Project management | From waterfall changed to agile | Planned agile but implemented like waterfall |
| HERMES | Yes | Yes |
| Relational mechanisms | ||
| Communication | Significant efforts dedicated; New communication channels created | Insufficient communication |
| External solution provider | Yes, very positive experience | Yes, troubled experience |
| Project outcome | ||
| Budget | 7 million CHF budget has been met | 5 million CHF budget has been met so far |
| Timeline | 6-month delay from the planned 4 years | 12–18-month delay from the planned 3 years |
| Meeting project goals | Project goals have been met; The scope of the project remained unchanged | Project goals have not been fully met; the project’s mandate has remained the same, but the project was re-scoped |
| Overall outcome | Success | Troubled: stuck in the conceptualization phase |
Cantonal digitalization lab
The organizational structure of the lab is defined by law. According to the interviewees, the cantonal digitalization lab partially possesses the necessary know-how and skills. However, decision-making powers are unclear, and the organization operates slowly—though still helping the canton to move faster: ‘I believe that since this new organization, processes have become somewhat faster, especially internally within the canton.’ (Anonymous interview) Review processes for initiatives are lengthy and complex, and often unclear or unknown to stakeholders.
The budgeting process reportedly functions well. Legal regulations allow financial instruments to relieve line employees working on lab projects: they may invoice their services to the lab, though they are not required to.
Strategic governance by the lab’s committee is seen as useful, but the current committee is reported to be heavily influenced by political interests. Members tend to focus on their specific domains, which may limit broader perspectives and pose challenges for project work due to missing expertise. The lab was frequently described as resembling a start-up still in development.
Interviewees noted ongoing tensions among departments involved in public sector digitalization at cantonal and municipal levels. The lab was seen as drawn into political dynamics, partly due to limited leadership. Its decentralized structure was identified as a factor limiting cohesion. To advance digitalization, interviewees emphasized the need for greater investment and centralization, organizational maturity, closer adherence to the HERMES project management methodology,1 improved use of synergies, stronger networks, and increased agility.
Since the lab has only existed for a few years, it has not yet been evaluated. However, such evaluations are seen as essential for developing into a learning organization.
Projects’ comparison
Governance structures
Both digitalization initiatives are organized as projects, with defined timelines, dedicated budgets, and temporary project organizations functioning as secondary structures for the involved public bodies. They have similarities in governance structures. Both have a single project sponsor, externally staffed quality and risk managers, and project committees of ten members representing different government levels and external suppliers. Differences include Project A’s leadership split among three people (two from external supplier), whereas Project B had a single leader from the cantonal digitalization lab. Project A lacks a subject committee, while Project B originally had one. Project B also has a larger Project Management Office, including roles like Data Protection Officer, Information Security, IT-Architects, Legal, and Communication. Both projects run the same number of sub-projects; however, in Project A, these are often led by external supplier, while in Project B, they are led by staff from the digitalization lab, supported by external supplier. In Project A many team members had subject-matter background, which supported project success. In Project A, the organizational chart was published online for transparency. Transparency and the right mindset meant that extensive steering became less necessary, as competent people adhered to structures independently.
In Project A, governance structures, roles and tasks were clearly defined from the start, personnel remained stable and tasks were seen feasible. Early extensive meetings helped clarify roles and build consensus, supported by continuous adjustments and open dialogue, with users remaining central to decisions and communication. Project B, by contrast, experienced frequent staff changes and only recently clarified leadership roles. The project committee, meant to advise per HERMES, in practice had decision power, leading to uncertainty about whether it could overrule the project sponsor. It faced persistent issues with unclear decision-making, fragmented responsibilities across governance levels, internal handling of problems and underdeveloped organizational structures, all prompting a current governance redesign. The previous project leader, employed by the digitalization lab, lacked subject expertise and struggled to coordinate with canton and municipal experts. To address this, leadership was moved closer to subject matter by integrating the role into the respective public office: ‘Simply for the reason that we want to reduce the risk that if the next project leader also comes from outside and then resigns again, the pile of issues just keeps getting bigger. Because with every resignation, a huge amount of specialized knowledge is lost. Not everything is always documented 100%, and of course all the connections and networking that have been built up are also lost — which is a huge setback.’ (Anonymous interview)
Other changes included removing the specialist committee, shortening communication paths, involving office representatives in governance, and giving new tasks to the core team. Interviewees differed on what defines governance: some stressed hierarchies, others networked collaboration, and others partnerships with suppliers. One interviewee emphasized the need to combine approaches: ‘When progress stalls at a level where hierarchies are basically horizontal, vertical hierarchies are also needed somewhere.’ (Anonymous interview)
Lessons from Project B include simplifying governance to ensure timely decisions, clarifying roles with dedicated resources (including relief from regular duties), working in smaller groups to promote decision-making, and creating a user representative role to clarify expectations, involve users, improve buy-in and coordination with external suppliers. In line with one observation — ‘Honestly, I feel like no one is truly being brought along; instead, a lot of time is spent just giving the impression that people are involved.’ (Anonymous interview) — a more pragmatic approach to future stakeholder involvement was also suggested: ‘We need to focus on working with the people who do want to move forward, and push those who don’t a bit into the background. Even if politically, it might not sound so nice to say “we didn’t get everyone on board,” the truth is that we got those on board who actually want to — and they’re definitely in the majority.’ (Anonymous interview)
Ultimately, Project B’s redesigned governance now closely mirrors Project A’s, suggesting institutional learning within the cantonal digitalization lab.
Governance processes
In both projects, the classic HERMES methodology for project management was followed, dividing governance across steering, leadership, and implementing levels, with multiple sub-projects. Both had to balance standardized and strictly defined processes with flexible role adaptation and manage tensions between classic and agile approaches. The project A shifted to an agile approach due to stakeholder pressure:
‘Originally the plan was to handle the entire concept phase strictly according to HERMES: so, to document everything in writing — ‘we will do it like this and like that,’ lay down the process, define it formally, and so on — all without actually implementing anything in the software itself. We, as the provider, then submitted a change request to adopt a different, more intensive approach — specifically, an agile approach. That way, we wouldn’t end up with an unpleasant surprise — where we had written a lot of concepts, only to discover that half of it wouldn’t actually work.’ (Anonymous interview)
In comparison, Project B, although planned as agile, proceeded mostly in a classic way, starting with detailed concept documents. Absence of a user representative complicated requirement definition: ‘That led to tensions, because on our side there was an expectation for a perfect, fully defined concept, while the supplier said, “But it’s supposed to be agile — we don’t need to define everything at this stage”. We also weren’t able to develop many of the concepts to the extent that some people demanded, because we ourselves didn’t really know what we wanted.’ (Anonymous interview)
Relational mechanisms
In both projects, governance extended across cantonal and municipal levels, aiming to balance interests. Both faced challenges from complex structures, numerous stakeholders, and limited resources. Leadership, communication, and proactive stakeholder involvement were critical, but approached differently. Project A invested in communication, and direct stakeholder engagement, supported by external supplier. Special communication channels — an information page with schedules, checklists, and materials — became a widely used tool. Early, proactive communication through newsletters, mailing lists, in-app channels set clear expectations. Communication had dedicated project structures — a communication subproject. Second, leadership actively mediated conflicts and remained stable. In addition to fulfilling a dual role in both project and line work, the leader managed to build trust. Furthermore, Diverse competencies within shared leadership allowed better decisions.
In contrast, Project B struggled with staff turnover and unresolved stakeholder conflicts. Leadership roles were initially unclear and part-time. The resignation of the sponsor, project leader, and subproject leaders resulted in the governance vacuum and the following significant changes in the governance structures, processes and relational mechanisms, which has been argued to stall the project. Although communication resources were comparable, efforts lacked continuity and no new communication channels were introduced. Collaboration with the external supplier was difficult from the beginning — ‘What was delivered was not what had been requested or agreed upon. Deadlines were often only poorly met, and in some cases, the deliverables were simply unusable.’ (Anonymous interview) — but the project sponsor and steering committee remained unaware: ‘The first project leader — that was a bit of a problem — kept the conflicts too long at the project team or specialist committee level, so that the same issues were repeatedly discussed without anyone really realizing what kind of conflict was developing.’ (Anonymous interview) Ultimately, this led to the project leader quitting. Earlier escalation up to the steering level has been recommended for the future, supported by the introduction of improved escalation processes and increased communication frequency. In addition, the supplier appointed a new project leader and test manager.
At the implementing level in both projects, staff worked alongside full-time roles. Project A valued flexibility, which supported success but sometimes caused dissatisfaction. Commitment was high. In Project B, flexibility was also praised, especially by municipal staff, though openness to change remained low. The project struggled to build a shared identity or ‘we-feeling.’
Project A predefined escalation paths. Conflicts mostly involved the digitalization lab, perceived as inflexible. Minor issues with the supplier were resolved through dialogue. Users, who had only an advisory vote, were reported to be dissatisfied. Adherence to budget and timeline had built trust and allowed flexibility in redefining requirements and adding features. The software was rolled out gradually. In Project B, collaboration was inconsistently described—some praised transparency, others criticized limited involvement: ‘At the project level, there was little agility or collaboration […] there was little focus on team collaboration and little focus on the customer, but more on the departments themselves.’ (Anonymous interview) Multiple legally prescribed processes and dependencies on other projects added complexity. Standardizing existing processes in an environment used to highly participative processes where stakeholders traditionally determined their own processes overwhelmed the project and became an obstacle to success. Each office acted as main approval body with own processes, and every municipality worked differently. Decision-making was difficult: many members lacked authority and had to consult superiors. While decisions were formally escalated — first within the team, then to a specialist committee, then to the department — the overlapping membership across bodies meant real decision-making remained a challenge.
Project outcomes
In both projects, the original goals largely remained unchanged. However, the project B experienced re-scoping. Both projects faced timeline adjustments driven by legal, organizational, or coordination factors, and both highlighted the importance of leadership, stakeholder involvement, and resource management for project success.
In Project A, the timeline was extended by six months to enable proper closure, not because of development delays, and the budget is expected to be met. The project kept delays under control thanks to early coordination with users and the use of a dedicated project management tool. The project leader participated directly in defining the initial plan and budget, alongside the project sponsor. Overall, success was attributed to the leadership’s diverse competencies and excellent professional and interpersonal qualities, strong commitment and high efficiency from participants, and well-set priorities — all of which helped to compensate for limited resources.
In Project B, the timeframe and budget were defined by the project sponsor and the project committee and were quite rigid, leaving the project leadership with no influence on these framework conditions: ‘And I must also say, the timeline was very tight. And the coordination efforts were somewhat underestimated.’ (Anonymous interview) Despite the rigidly defined timeframe, delays occurred. They resulted from dependencies on other projects, the absence of a legal basis for a digital only process (expected to take at least two more years to resolve), dissatisfaction in collaboration with the supplier, and the resignation of the project leader. Planning errors were admitted, in particular: the focus on analysis instead of results, the lack of iterations, and missing intermediate deliverables. Political pressure initially demanded on-time delivery, but dependencies on legal changes and external projects made this unrealistic, demanding some tolerance for delay: ‘I assume there won’t be major resistance, because everyone realizes it’s better to lose another month or two than to compromise on quality later.’ (Anonymous interview)
Discussion
This section is devoted to interpreting and connecting all unexpected findings, which are the codes that were generated inductively from the data, to the literature (see Table 2 above). Analysing two digitalization projects led by the same cantonal digitalization lab in Switzerland, this study shows that even when many contextual factors remain constant, identifying an effective governance framework remains a key challenge for achieving desired outcomes in public digitalization, especially in multi-actor and multi-organizational settings. This finding contributes new empirical evidence to existing research highlighting that even under the same lab within one canton, different governance forms can be chosen, leading to different approaches to governance structures (organization and staffing of leadership, committees and other project roles), processes (flexibility in the process of needs definition), and relational mechanisms (stakeholder management, conflict resolution), and ultimately project success (Di Giulio and Vecchi, 2021; Dolmans et al., 2023).
Starting with the unexpected findings in the governance processes, we discuss the importance of the participative processes, as the topic, emerged during the interviews. Both cases emphasized the importance of meaningful user involvement for the project outcomes. Both cases highlight that meaningful user involvement does not emerge naturally but requires both dedicated structures and processes to generate user needs, which links the aspects of governance between each other: the cases show that defining needs and translating them into requirements is a complex and resource-intensive task, requiring deliberate motivation of stakeholders and resource allocation. Carefully defined solution requirements rooted in these needs increase solution acceptance and usage, by that helping to achieve project goals. Furthermore, meaningful user involvement helps manage stakeholder expectations and structure collaboration with external suppliers by positioning the public sector as an informed customer. This demonstrates how this processual aspect of governance (meaningful user involvement through participative processes) is interconnected with relational aspects of governance (managing stakeholder expectations and collaborating). The articulated importance of meaningful user involvement aligns with observations from other contexts that even though the success of digitalization projects is linked to addressing user needs (Schmid, 2019), the reality shows little to no meaningful user participation and the lack of governance mechanisms for real involvement, and users are typically restricted to a passive role of ex-post commenting rather than co-designing and co-producing solutions (Tenney et al., 2020). Another inductive code in the governance processes is the impact of leadership, which concerns how projects approach organization and staffing of leadership positions. The studied projects approached this governance decision differently. One of the projects initially did not clarify the leadership role and appointed a project leader which lacked subject expertise and due to that struggled to coordinate with canton and municipal experts, which negatively affected the project outcomes. When these detrimental effects of governance on the project outcomes have been acknowledged, the project leadership was moved closer to subject matter by integrating this role into the respective public office. In the other project, leadership was shared, which allowed the project to benefit from diverse competencies and better decisions. Furthermore, the project leadership was staffed with the people who had a dual role in both project and line work, which helped the leaders to build trust and on this basis to actively mediate conflicts.
As for the relational mechanisms as the element of governance, the prominence of the communication topic in the interviews from both projects resulted in it becoming an inductively-informed code. Digitalization often cuts across organizational silos and government levels, especially in federal systems, making the success of a project dependent on targeted communication. As good practice, the both projects illustrate the creation of new, stakeholder-specific communication channels; dedicating substantial resources to change management; establishing full-time communication roles; and promoting digital solutions internally before rollout. These findings reinforce the literature emphasizing communication and change management in digitalization success (Bordeleau, 2019) and suggest further research on how tailored communication impacts change acceptance (Allen et al., 2007). A relevant part of communication is conflict resolution. Both cases faced conflicts typical of projects spanning departments and government levels. The interviewees highlighted that the project success depended on having governance structures for conflict resolution and the willingness to address conflicts early rather than delay.
A topic which was equally rooted in two aspects of governance – relational mechanisms and governance structures – concerned prolonged discussions not ending with any decision (coded to both inductive codes ‘Communication’ and ‘Unclear tasks in role’). The cases approach the task of ensuring decisions are made in different ways: on the one hand, they recognize the parity and collegiality characteristics of Swiss public bodies as necessary and unchangeable, though it demands more time and resources to reach consensus (thus, highlighting the importance of relational mechanisms). On the other hand, interviewees also suggested establishing rigid structures with clearly defined roles, responsibilities, and lines of accountability. The suggested strategy for avoiding ‘discussions without decisions’ was to make sure that public organizations must design governance structures that provide role clarity and entitle decision-making while maintaining flexibility needed for adaptation and innovation.
The final inductive code emerging in the governance structures concerned legal foundations. The cantonal lab provided the legal foundation for any project implemented under its roof to utilize a financial instrument to relieve line employees working on lab projects, by that potentially affecting such aspects of governance as resource feasibility, and ultimately making the achievement of project outcomes more realistic. On the other hand, for one of the projects, the absence of a legal basis to implement the designed solution has resulted in a significant project timeline prolongation and a total redesign of the governance structure. This demonstrated the necessity of treating governance as a dynamic, living structure that adapts to changing circumstances and fits project goals, which supports literature arguing that governance modes require ongoing reassessment: as project maturity advances, governance maturity evolves correspondingly (Jacobs et al., 2020; Meyerhoff Nielsen, 2019).
Conclusion
This study has set out to explore how governance affects the outcomes of public sector digitalization projects, drawing on two polar cases from the same cantonal digitalization lab. Despite identical contextual conditions, the projects developed different approaches to governance structures, processes, and relational mechanisms, that ultimately affected the project outcomes. These findings reinforce the argument introduced at the outset: governance can actively determine the success or failure of digitalization initiatives.
This research is limited by its focus on only two cases within a single canton, which may restrict broader generalization. Additionally, the way this paper operationalised governance, despite being well rooted in the literature, is not the only way to evaluate governance, and further studies might try other approaches, potentially also comparing them with the one presented in this study. This paper highlights what interviewees indicated as important elements within governance structures, processes, and relational mechanisms. For future research, we propose applying factor analysis to governance variables to a larger sample, in order to determine which of the governance constructs most strongly predict the success of digitalization projects. Furthermore, future studies could extend these insights by examining how governance evolves over time in other public digitalization contexts and how this evolution affects project success.
Data Accessibility statement
Research-generated materials, including interview guidelines, coding strategies, code trees, and anonymized interview quotes, are published alongside this study.
Appendices
Appendices
Code trees
Interview guideline
| CONCEPT | QUESTIONS |
|---|---|
| Governance structures | Can you briefly describe your role in the project? Have you been involved from the very beginning? Is this a full-time position? Is your position temporary for the duration of the project? Are you leading or working on other projects at the same time? If yes, how much of your work time do you dedicate to this project? Are the tasks you perform in this role clearly defined and feasible with the resources available? How is the project organized? Can you describe the general organizational structure? Does the project organizational structure go beyond departmental boundaries? How does the collaboration work? Do you have a project team? What roles exist in the project team, and who is part of it? Has the organizational structure changed significantly during the project? Does your IT organization have capable IT leaders and IT committees with clear decision-making authority? Can you describe the organization? Are the employees of the IT organization familiar with these structures? In your view, are there aspects of the governance structure that work particularly well or could be improved? |
| Governance processes | Are there steering processes and instruments you must follow (such as HERMES, agile elements, or IT organization processes)? If so, which ones? Many stakeholders are involved in this project. What challenges arise in coordinating these stakeholders? How are decisions made in the project? How are they communicated? What is the core basis of project leadership? Mainly internal hierarchies, partnerships with external suppliers, or horizontal collaboration, i.e., a network? Which processes are established in the IT organization for decision-making and oversight, and are these continuously evaluated? |
| Relational mechanisms | How would you describe the relational mechanisms in the IT organization? (Relationship mechanisms include working relationships, communication, and the use of synergies) Are there stakeholders who are not entirely satisfied with the role distribution in the project team? How would you describe collaboration within the project? How are conflicts managed? Do you have to cooperate with external suppliers? What has been your experience with this? How intensive is your contact with the project’s stakeholders, and to what extent do you see yourself as responsible for actively maintaining these contacts? |
| Project outcomes | Have the project goals and timeline remained the same throughout? If not, what are the main reasons for the changes? Were you able to have input on the timeline and budget, or were these predetermined for you? How would you assess the flexibility of the framework conditions in the project? |
Notes
[2] HERMES is the project management methodology developed by and for the Swiss government, and obligatory for big digitalization projects (https://www.bk.admin.ch/bk/en/home/digitale-transformation-ikt-lenkung/hermes_projektmanagementmethode.html).
Ethics and Consent
This research has undergone the ethical clearance from the Ethics Committee of the Faculty of Business, Economics and Social Sciences of the University of Bern, before the data collection.
Acknowledgements
We would like to acknowledge the interview participants, without whom this study would not have been possible, the full research team of “The role of governance in public sector digitalization (RoGoDi)” project, and the anonymous reviewers.
Competing Interests
The authors have no competing interests to declare.
Author Contributions
The foundations of this study originate from the research proposal “The role of governance in public sector digitalization (RoGoDi)” funded by the Swiss National Science Foundation. This study draws substantially on the individual research of the first author. The subsequent translation of this work into its current academic-paper form reflects equal contributions by all co-authors.