Managing Asymmetries for Data Mobilization under Digital Transformation

Traditional resource mobilization literature

Research on resource mobilization has predominantly concentrated on the initial phases of the process, emphasizing the actions required by early-stage resource seekers to secure vital resources, particularly financial, human and social capital. As they need resources, these seekers are portrayed as orchestrators of resource mobilization, navigating the complexities of resource identification, access and transfer (Clough et al., 2019). Information about their capabilities and intentions is initially limited among potential resource holders. Therefore, studies tend to focus on the alleviation of information asymmetry between resource seekers and holders (Akerlof, 1970) in order for financial resources to be mobilized and on how seekers signal their attributes and qualities to resource holders (Moss, Neubaum and Meyskens, 2015; Shane and Cable, 2002; Vanacker et al., 2020). Here, scholars show that early-stage seekers strategically employ a variety of signalling mechanisms, such as showcasing expertise, previous successes or partnerships (Hallen and Eisenhardt, 2012). These signals are given to reduce uncertainty and increase the likelihood of attracting attention from resource holders (Vanacker et al., 2020). Given the liability of newness, early-stage seekers may rely on alliances provided by third parties (Bafera and Kleinert, 2023), such as universities (Colombo, 2021). Expanding on this point, research indicates that those holding resources often ignore the quality signals of seekers unless they are accompanied by reputable third-party affiliations (Plummer, Allison and Connelly, 2016). Indeed, the literature shows that third-party entities, such as universities, play a crucial role in resource mobilization, highlighting their efforts in forming partnerships to secure financial resources and in utilizing their intellectual capital to advance research and innovation (Bonando et al., 2011; Colombo, Meoli and Vismara, 2019). Trust in reputable and prominent universities enhances the legitimacy of resource mobilization, as shown by prior literature referencing connections to quality-signalling organizations (Baum and Oliver, 1992; Podolny, 2001).

However, there are some notable limitations in extant studies. First, signalling theory tends to prioritize the specific dyadic direction as the seeker-to-holder relationship (Clough et al., 2019), often overlooking the wider relational dynamics and social interactions that may shape resource mobilization (e.g. Shane and Cable, 2002). Some studies, however, delve into the initial signalling process of resource holders (Murray, Kotha and Fisher, 2020). For example, resource holders may signal the quality of their resource by advertising, offering warranty, or using cues such as brand name or price (Kirmani and Rao, 2000). They may also signal resource scarcity through strategic connections, regular meetings, and planned interactions during key events (Hallen and Eisenhardt, 2012). Research, here, indicates that these efforts may be vital for seekers aiming to forge connections with holders with uncertain and intangible resources (Vissa, 2011). However, further investigations into holders-seekers interactions are rare.

Despite this limitation, signalling theory remains pivotal, although it assumes an equal power distribution between seekers and holders, overlooking power imbalances that exist in many mobilization contexts (Harland, 1996; Katila et al., 2022; Pfeffer and Salancik, 1978). Research suggests that resource holders hold power in this relationship (Emerson, 1962), whereas resource seekers are often far more dependent on resource holders, directly linking power and resource mobilization. For instance, in the biotechnology industry, collaboration between universities and established drug firms (Pisano, 2006) emphasizes power imbalances that can lead to situations in which promised resources are not delivered, causing mistrust and complicating alliance-building (Katila et al., 2022). Resource dependency theory addresses power imbalances but may oversimplify resource mobilization by emphasizing the holder's power and control, potentially overshadowing the seeker's agency and strategic actions (Villanueva, Van de Ven and Sapienza, 2012). Furthermore, the theory may not sufficiently account for contextual and institutional factors such as non-market-orientation (Clough et al., 2019) and organizational change, yet it remains valuable for understanding resource mobilization dynamics (Villanueva, Van de Ven and Sapienza, 2012). Here, it is suggested that more research on resource mobilization should address different market orientations (Clough et al., 2019) and in particular non-market third-party actors who place an emphasis on appropriate action rather than on self-interest (He et al., 2021).

A further limitation is that the literature predominantly emphasizes acquiring financial resources, overlooking other resources that may demand more distinct mobilization strategies (Clough et al., 2019), including legitimacy, narratives, and intellectual property (Delmar and Shane, 2004; Hsu and Ziedonis, 2013; Martens, Jennings and Jennings, 2007). For instance, seekers engage in processes to augment other resources, enhancing human capital through a founding team, early employees and informal advisors as well as securing financial capital via external investments (Aldrich and Kim, 2007; Vissa, 2012). There is also a growing consideration of other intangibles and uncertain resource types, such as public goods (Mittermaier, Shepherd and Patzelt, 2021), where these resources are challenging to amass and retain for future utilization (Farny et al., 2019). However, how non-financial resources are mobilized remains an under-explored area (Clough et al., 2019). This lack may be due to the popularity of quantitative data on mobilizing financial capital and the relatively easier access to it (Alrich and Kim, 2007).

Resource mobilization via platforms

The advent of resource mobilization via platforms, notably crowdfunding, introduces a contemporary dimension to the traditional literature. Crowdfunding has significantly changed the resource mobilization landscape by democratizing financing and innovation (Cumming, Johan and Zhang, 2019; Mollick and Robb, 2016), allowing non-traditional investors to finance and enabling early-stage seekers to secure financial resources (Murray, Kotha and Fisher, 2020). Crowdfunding via platforms departs from conventional methods of resource mobilization in several ways. First, it allows seekers to attract resources from a dispersed pool of holders in exchange for equity, interest, rewards or even no direct return (Belleflamme, Omrani and Peitz, 2015; Mollick, 2014). Second, crowdfunding-platform owners take an active role in preselecting new ventures. Here, the prior literature shows that crowd serves as a stamp of approval, as ventures funded through crowdfunding campaigns are more likely to attract follow-up investments (Roma, Vasi and Kolympiris, 2021). These platforms have the resources to access more information by following a thorough, multistep due diligence process on behalf of the potential investors before the campaign (Löher, 2016), making the pre-campaign phase a critical step for success (Kleinert et al., 2021). Thus, it is argued that the due diligence conducted by crowdfunding-platform owners can mitigate the information asymmetries between the holders and seekers (Löher, 2016; Vismara, 2018), with larger platforms more likely to conduct due diligence of seekers (Cumming, Johan and Zhang, 2019). Third, with only a fraction of early-stage seekers being accepted by crowdfunding platforms among the total applications (Kleinert et al., 2021), emphasis is placed on employing various signals that seekers can use to counteract the impacts of information asymmetry (Vismara, 2018). Research indicates that utilizing platforms enable seekers with no financial track record to access funding from potential investors globally (Ahlers et al., 2015). The difference is that while the quality signals from seekers might be the same for both resource holders and crowdfunding platform owners, the latter may evaluate quality signals differently (Kleinert et al., 2021). Fourth, investigations into resource mobilization via platforms for other non-financial resources (Ray, Menon and Mookerjee, 2020) are limited, although some studies show that platforms such as MTurk or Upwork mediate the mobilization of non-financial resources such as skills, matching workers and employers (Chandna, 2022).

Finally, while recent studies have provided insights into the dynamics of resource mobilization through crowdfunding platforms, limited studies address resource mobilization through other types of platforms. There remains a gap in addressing the strategies and processes available to all firms to mobilize resources on other types of platforms. Our goal, in particular, is to explore multi-sided platforms that enable digital transformation and entrepreneurship by facilitating interactions among diverse user groups (Nambisan, Wright and Feldman, 2019; Parker, Van Alstyne and Jiang, 2017). In particular, we aim to explore how all stakeholders have agency to impact resource mobilization, specifically their ability to mobilize data and subsequently other resources via platforms.

Data sources

Much academic work on inductive theory building allows exploration through rich qualitative data collection (Eisenhardt, 1989) from several sources. This study used primary and secondary qualitative data comprising 157 quotations from project materials, industrial reports and web pages. Additionally, one of the authors, an embedded researcher at AnimalCo, observed more than 200 hours of meetings, which were recorded in field journals. This researcher became deeply immersed in the resource mobilization phenomenon by personally contributing to some of the activities, watching processes unfold over time, and bringing the relevant literature to bear on the study (Eisenhardt and Graebner, 2007).

Our primary qualitative data was collected through in-depth interviews between May and July 2021. We sampled multiple professionals working in animal healthcare and data science within the setting. A total of 25 interviews, targeting the most experienced stakeholders (Table 2), were conducted in the setting, and covered various facets of resource mobilization. These stakeholders were the most informative and appropriate sources of information for our study because of their deep understanding of actions taken for data mobilization, as they understood how companies leverage data for innovation and how to gain a first insight into the value of animal health data. By using open-ended questions, ensuring anonymity and employing triangulation with the secondary qualitative data, we gained a comprehensive understanding of processes for data mobilization. All interviews were audio-recorded and transcribed. Our approach with diverse data sources and strategic interviews provided convergence information and effectively minimized informant bias (Brown and Eisenhardt, 1997). In summary, by leveraging many data sources, the team could piece together a thorough chronicle of the endeavours undertaken by actors in our setting (Eisenhardt and Graebner, 2007).

Analytic approach

After collating documentary evidence, assembling observations and conducting interviews, we built a database using NVIVO 14 to analyse our data systematically. We assessed the data through the lens of our research question (e.g. Graebner, 2009). We followed an iterative process by immersion in the data, reading and re-reading the field notes, documents and interviews, and engaging in discussions between the authors, paying attention to emerging concepts and, in particular, to the actors’ activities about sources of data (e.g. sensor data) and control over those data. Doing so meant identifying relationships, patterns and themes and ensuring that these mapped to our resource mobilization phenomenon of interest. Consistent with the principles of inductive reasoning, our theoretical analysis of resource mobilization and the interactions with the actors evolved alongside our empirical analysis. We sought to develop a broad conceptual framework, following a sequence of steps to reveal how the actors could mobilize data and other resources under the conditions of digital transformation.

First, we focused on the relationships, patterns and themes addressed in the existing literature on the process of mobilizing resources (Clough et al., 2019; Vanacker et al., 2020). Consistent with the resource mobilization literature, we observed that the firms in our study referred to the identification of resource-holding locations as searching, emphasizing securing access as ensuring data-sharing agreements with holders and the transfer and deployment of resources through instituting governance mechanisms (Clough et al., 2019). The difference between our setting and the literature is that seeker firms rely heavily on data from various holders that may not be directly accessible to attract further resources such as research funding, whereas holder firms need more convincing to release their data. Thus, our context provided us with a new orientation for resource mobilization. For example, the increasing need for seeker firms to signal their quality became obvious through statements such as, ‘[we] put together pieces of data telling a story …“something is wrong with the pet”’. Thus, they use narratives and storytelling to persuade resource holders to commit resources (Garud et al., 2022). As another example shows, whereas extant studies have assumed that only holders have agency in mobilization studies (Vissa, 2012), we found in our study that other players were concerned with locating and incentivizing holders to share their data, as demonstrated in the statement, ‘the root of all barriers resides in the fact that the owners of data of interest are numerous, both internal and external to [AnimalCo], with no obvious incentive and benefit to share their data’, showing seeker agency in reaching out to the holders (Clough et al., 2019). Overall, our initial analysis showed how the perspectives on data shaped the considerations for data mobilization in the ecosystem. Through open coding, we identified these distinct and recurring topics as first-order codes (Corbin and Strauss, 1990). The team discussed the codes and resolved any discrepancies in order to reconcile the topics with the coding (see Figure 1).

Next, we used thematic analysis (Gioia, Corley and Hamilton, 2013), a method that allows researchers to consider pre-defined initial ideas by engaging with the data inductively. The analysis provided an intermediary step between our coding and more abstract conceptualizations. Identifying mechanisms adopted for this step involved comparing and contrasting the similarities and differences in the actions and interactions of stakeholders throughout various stages of resource mobilization, which formed the basis for our second-order themes (see Figure 1). For example, with uncertainty around the provenance of the data held by holders, we recognized attempts at due diligence of data holders as well as signalling as mechanisms to increase knowledge about the quality of the data for seeker firms, as shown by the statement, ‘the quality of the shared data needs to be assessed before we can engage in any data-sharing practices’. Similarly, for university partners, to signal their quality they adopted a number of mechanisms such as matchmaking to aid the mobilization process as shown by ‘matchmaking is attracting the right people to turn data into insights’. As a final step, we further reviewed and interpreted the statements and categorized the second-order themes into aggregate dimensions of enabling data mobilization (Figure 1). Through this process, we identified three distinct dimensions that influenced data mobilization at different stages of actors’ resource mobilization attempts.

Our analysis revealed that different actors used different approaches to address data searches in order to mitigate uncertainty and identify controlling resource holders' locations. Identified as data discovery, actors employed more targeted strategies beyond mere signalling, recognizing the need to incentivize resource holders to make their data available. AnimalCo sought to integrate resource holders into its operations through promoting screening to ensure the provenance and quality of the data. We noted this dimension as initiating resource mobilization.

The reluctance to share sensitive information creates barriers to collaboration and data sharing. To overcome this, AnimalCo considered a data platform allowing a wide range of data to be securely connected, valorized, shared and exploited. Their intention was to adopt a platform to make data sharing a smooth process, as stock markets do for shares. The distinct issue of trust appeared to be crucial in overcoming the challenge of data sharing, particularly regarding AnimalCo, a key player in the industry. Establishing transparent and secure data-sharing practices is vital for building trust among resource holders and seekers. We coded this dimension as enabling data sharing.

Lastly, regarding resource transfer, governance mechanisms, power and trust are crucial in data mobilization. Here, the findings highlighted that the variation of choice of governance mechanisms differed among firms: some opted for formal arrangements based on contracts and subscriptions, while others sought to choose to engage resource seekers in informal governance arrangements built on trust. As a result, greater attention was paid to organizing and institution-building at this stage. Some players preferred open systems with diverse resource holders, ensuring continued resource access. Others emphasized closed systems, building strong relationships with a select few resource holders aligned with their specific integration needs and supporting AnimalCo's mission. We coded this dimension as instituting data mobilization.

Figure 1 below provides a comprehensive overview of our data framework. We use this as a basis for our analysis to theorize how firms mobilize resources when establishing data platforms that enable resource mobilization and data sharing. By expanding on our analyses, we constructed a theoretical framework elucidating the activities employed by the actors in mobilizing data through data platforms, consolidating emergent insights into a data mobilization framework.

We employed two strategies to prevent retrofitting the data to fit our theorizing (Wodak and Busch, 2004). First, we considered multiple theoretical explanations throughout the research process, eliminating alternatives to align the findings with the clearest interpretation based on resource mobilization concepts (Dubois and Gadde, 2002). Second, we implemented triangulation between different data types, ensuring theoretical saturation (Bryman, 2008) and inter-coder consistency (O'Connor and Joffe, 2020). Through documented peer review and discussions, we addressed observer bias (Eisenhardt and Graebner, 2007), achieving external validity by reviewing the academic literature on resource mobilization. Any discrepancies were resolved through peer review and discussion (Kassarjian, 1977).

Initiating resource mobilization

However, AnimalCo encountered challenges in leveraging data. These challenges featured widely in the findings and were referred to as concerns for ‘data discovery’. For instance, seekers, including data scientists at university research institutes, express interest but struggle to find reliable data sources. Other respondents note substantial disconnected data in animal healthcare stored in isolated, unrefined silos [IP15]. Thus, the interviews revealed that seekers faced challenges initiating resource mobilization and communicating their interests directly with holders. Here, the findings confirmed that information asymmetry from a seeker's perspective was evident and that seekers needed mechanisms to signal their ‘confidence, credibility and reliability’ [IP1] in understanding the value possibilities of the data.

Second, to enable data discovery, AnimalCo explored a process in which holders’ data could be categorized as ‘core’ or ‘non-core’, with a distinct integration process required for each type. It was stated that key considerations for determining a dataset's ‘core’ status included the quality and significance of content, the potential audience for the data, and the existence of any usage restrictions. Therefore, this process implied that it was imperative to apply additional rigour to appraise the data before integrating and making it accessible. It was also stated that data with restrictions is unlikely to be considered core. Thus, AnimalCo considered initiating a process of screening data holders, which can be interpreted as a process of due diligence.

The accounts suggest that the environment for initiating resource mobilization through data discovery highlighted dual-sided information asymmetries affecting AnimalCo's orchestration efforts. Therefore, it was identified that to initiate resource mobilization, and to enable data discovery, the actors had to address the dual-sided information asymmetry between seekers and holders. Thus, seeker signalling and screening holders became crucial for initiating the mobilization of data as a resource. Seekers wanted to employ several signalling strategies to convey their credibility and seriousness to the resource holders. A screening process can be interpreted as a process of due diligence, where holders play an active role as they are interested in the outcome. It was also understood that these processes could be crucial in locating resource holders and enabling seekers to discover details about resources and signal their credentials, which ultimately needed a trusted third-party partner.

Here, the findings revealed that AnimalCo saw universities as a key third-party partner, not only for hosting data but also for playing a crucial role in facilitating the discovery of and access to relevant data sources. They saw universities as important to signal credibility: as one respondent put it, the ‘USP [Unique selling point] of University – [is the] credibility for industrial partner… as data is hosted in University –' [IP3]. But AnimalCo also recognized the multifunctional role that universities could play in enhancing screening, incentivizing, and ensuring credibility signalling from seekers. The university partners were also seen as pools of intellectual capacities in discovering and accessing relevant data that could be achieved through mechanisms such as matchmaking and hackathons.

Matchmaking was described as ‘attracting the right people to turn data into insights… attracting many data sets does not do it’. Another respondent [IP14] noted that matchmaking could provide a ‘quick provision of insights that can incentivize smaller companies lacking statistical expertise or analytical resources to seek out potential partners’. University respondents claimed that matchmaking not only builds confidence and credibility among stakeholders but also fosters collaboration. Hackathons were seen as an opportunity to create a ‘data marketplace’ for trading data between different subjects, predominantly as B2B transactions [IP5]. As such, the findings emphasized utilizing matchmaking and hackathons to highlight mutual gains and enhanced outcomes for all involved in resource mobilization. It was understood that the university partners play a crucial role in locating resource holders and enabling seekers to discover details about resources and signal their credentials, which ultimately needed a platform for creating a marketplace for data.

Enabling data sharing

However, the interviews revealed concerns about data sharing via the data platform, particularly regarding privacy, security, identifiable data, and data ownership. In particular, the data holders felt entirely visible and vulnerable to seekers, while seekers remained obscure to holders hiding behind their data capabilities, highlighting a power imbalance. The analysis of the interviews also showed that this imbalance created a sense of mistrust for the resource holders, with one stating that there is a ‘problem of inequality…: those who contribute data are not those who get value’ [IP5] . At the same time, the limited alternatives for seekers to obtain necessary data grant greater bargaining power to the resource holders. This dependence, it was claimed, would impede data mobilization, as seekers relying heavily on holders may face delays in data availability. This dependence reduces seekers' power for data access, giving the dominant power to holders, particularly those with more attractive data.

Thus, establishing data-sharing practices and protocol became a key concern in alleviating power imbalances.

Institutionalizing data mobilization

But some of the respondents wanted some form of light governance. For instance, it was claimed that this would ‘allow the creation, transformation, evolution, curation and exploitation of data sets, together with soft competencies around them’ [IP13]; that is, governance and social interactions.

The interviews and documents showed that AnimalCo relied on creating a genuine and trusted environment for resource holders through universities as trusted partners. Here it was suggested that the university partners could further reduce power imbalances by supporting data sharing because of their non-market credentials, technical know-how and expertise, which were seen as invaluable for building trust. As a result of insights, with the universities involved in data curation and governance, data sharing could be seen as both authentic and desirable, whereby holders may reciprocate in kind by offering increased access to data in return.

The interviews showed that universities prioritized non-market logic when building the platform. In particular, one respondent claimed that the leading goal of such collaborations is public good, namely increasing animal well-being, supported by data, such as early disease diagnostics and treatment. However, a broader perspective cannot exclude pure market motivations. AnimalCo's strategy included marketing and investment capabilities, which may help to find venture capital for start-ups on the platform, thus indirectly accessing data resources and exploiting the outcomes. Therefore, AnimalCo intended to be the primary beneficiary of resource mobilization, as they will sell the data commercialization outcomes back to the resource holders (e.g. as subscriptions for new apps).

However, other respondents thought that AnimalCo should pursue a closed-platform approach. They also referred to the need for data-sharing agreements, negotiations, and legal frameworks, indicating the challenges and barriers associated with accessing closed data. These respondents suggested that specific data sources, such as proprietary databases and clinical records, may not be readily available or accessible for open data purposes. They mentioned constraints in short-term contracting within university settings. This is the most significant barrier in dealing with universities as third parties during data mobilization, which suggests that closed data may require specific arrangements and permissions to be accessed.

Thus, AnimalCo sought to mobilize data resources by institutionalizing relations with its partners. Hence, AnimalCo embarked on achieving digital transformation through further phases of data due diligence, collaboration, trust-building and legitimacy building, in particular by working with their university partners to align industrial and academic objectives.