3 barriers to successful data collaboratives

Technology entrepreneurs at a Nairobi co-working space. Credit: rvdw images.

Data collaboratives have proliferated in recent years as effective means of promoting the use of data for social good. This type of social partnership involves actors from the private, public, and not-for-profit sectors working together to leverage public or private data to enhance collective capacity to address societal and environmental challenges. The California Data Collaborative for instance, combines the data of numerous Californian water managers to enhance data-informed policy and decision making. 

But, in my years as a researcher studying more than a hundred cases of data collaboratives, I have observed widespread feelings of isolation among collaborating partners due to the absence of success-proven reference models. 

Surprisingly, what causes practitioners to struggle most is not the technical aspects of the partnership (e.g., data interoperability, storage, and analysis) but the governance ones, which involve managing collaborative logistics such as partner onboarding, trust building, negotiating agreements, securing funding, etc. As a result, most data collaboratives fail to advance beyond the pilot stage due to governance-related challenges including lack of funding, poor conflict management, limited stakeholder engagement and communication, and low adaptation and flexibility to external events.

Surprisingly, what causes practitioners to struggle most is not the technical aspects of the partnership but the governance ones.

To address this, practitioners should refocus attention to start dealing with data collaboratives not only as technological phenomena but also as collaborative ones, requiring dedicated resources and specialized knowledge for their management and governance.

Below, I provide an overview of three governance challenges faced by practitioners, as well as recommendations for addressing them. In doing so, I encourage every practitioner embarking on a data collaborative initiative to reflect on these challenges and create ad-hoc strategies to address them. 

Thanks for reading The Data Values Digest! Subscribe for free to support our work.

1. Overly relying on grant funding limits a collaborative’s options.

Data Collaboratives are typically conceived as not-for-profit projects, relying solely on grant funding from the founding partners. This is the case, for example, with TD1_Index, a global collaboration that seeks to gather data on Type 1 diabetes, raise awareness, and advance research on the topic. Although grant funding schemas work in some cases (like in that of T1D_Index), relying solely on grant funding makes a data collaborative heavily dependent on the willingness of one or more partners to sustain its activities and hinders its ability to achieve operational and decisional autonomy.

Operational and decisional autonomy indeed appears to be a beneficial condition for a collaborative to develop trust, involve other partners, and continuously adapt its activities and structure to external events—characteristics required for operating in a highly innovative sector.

Hybrid business models that combine grant funding with revenue-generating activities indicate a promising evolutionary path. The simplest way to do this is to monetize data analysis and data stewardship services. The ActNow Coalition, a U.S.-based not-for-profit organization, combines donations with client-funded initiatives in which the team provides data collection, analysis, and visualization services. Offering these types of services generates revenues for the collaborative and gaining access to them is among the most compelling incentives for partners to join the collaboration.

In studying data collaboratives around the world, two models emerge as most effective: (1) pay-per-use models, in which collaboration partners can access data-related services on demand (see Civity NL and their project Sniffer Bike) and (2) membership models, in which participation in the collaborative entitles partners to access certain services under predefined conditions (see the California Data Collaborative).

2. Demonstrating impact is key to a collaborative’s survival. 

As partners’ participation in data collaboratives is primarily motivated by a shared social purpose, the collaborative’s ability to demonstrate its efficacy in achieving its purpose means being able to defend its raison d’être. Demonstrating impact enables collaboratives to retain existing partners, renew commitments, and recruit new partners.

However, measuring the impact of data collaboratives is neither simple nor fast; it requires specific skills, resources, and time. Moreover, data collaboratives frequently operate in the area of raising awareness or improving decision-making; their outputs are therefore distant from the final beneficiaries, making it even more difficult to quantify the social change generated.

Possible solutions include adopting an impact management perspective from the outset of the partnership via the collaborative’s theory of change design and agreeing on output indicators to be used as proxies for outcomes and impacts. Clarifying the partnership’s output and impact objectives from the very beginning may also aid in aligning the individual partners’ expectations with the collective value proposition. NeedsMap, for instance, relies heavily on impact measurement and impact results’ communication to engage and motivate corporate partners to join the project.

3. Finding a balance between collaborative approaches and effective decision-making and flexibility is like walking a tightrope.

Data collaboratives are multi-stakeholder initiatives, the existence of which depends on the participation of multiple partners (data owners, data providers, tech providers, beneficiaries/citizens etc.). Involving stakeholders in the decision-making process is a way to keep them engaged and prevent tensions that could force a partnership to suspend its operations. However, stakeholder engagement requires resources and time and may slow down decision-making processes. On the other hand, having the ability to rely on an agile decision-making process allows the partnership to be adaptable in the face of internal and external threats and opportunities.

To balance the two is comparable to walking on a tightrope. Nonetheless, specific governance settings may facilitate achieving this balance.

This is the case for collaboratives that rely on a third-party organization to serve as an intermediary between the stakeholders involved (see PopGrid). Typically, these intermediaries are legally constituted as distinct entities from the partnership’s founders, with whom they may be connected through equity participation, voting rights, or ad-hoc agreements. This type of arrangement facilitates equal and direct relationships among all the partners and involves them in strategic decisions; at the same time, it relieves partners from day-to-day decision-making, allowing the collaborative to make fast and agile decisions. 

Calling on practitioners to address governance challenges to unlock the potential of data collaboratives 

Data collaboratives hold the key to addressing many of today’s biggest social and environmental challenges, yet their success is too often stunted by the lack of shared knowledge and experience on nearly universal challenges. Significant recent knowledge creation efforts have in this area have produced a wealth of resources to build successful partnerships (see the GovLab’s datacollaboratives.org, the European Commission’s report on Business-to-Government Data Sharing, and the Global Partnership for Sustainable Development Data’s work on public-private data sharing, for example). But more work is required to provide practitioners with the appropriate governance reference models and tools for governing their initiatives.  

The governance challenges mentioned here are only a subset of those that practitioners around the world are facing. Being able to overcome these obstacles will help practitioners form stable collaboratives and so to develop their impact-generating activities over a longer period. This will allow demonstration of the effectiveness of data-enabled social intervention in generating shared social benefits and therefore scale this practice and create positive systemic impact through data.

Thanks for reading The Data Values Digest! Subscribe for free to receive new posts and support this work.

Want to get in touch with this author? Email federico.bartolomucci@polimi.it.

Related Articles


Your email address will not be published. Required fields are marked *