Rethinking the Inquiry: Reflections and Lessons from the Data Workers’ Inquiry Project

5 min read

The Data Workers’ Inquiry (DWI) is a participatory action research project that centres on data workers as co-researchers in documenting their lived experiences and challenging exploitative labour practices in the AI industry. Unlike traditional research, which often positions participants as passive subjects, DWI empowers workers to lead the inquiry, define research topics and choose creative formats for their outputs. This article reflects on the methodological principles guiding the project and the challenges of applying them in practice.

The methodology of a collaborative inquiry

This methodology builds on Karl Marx’s 1880 Workers’ Inquiry, which recognised workers as the primary authorities on their exploitation and as essential agents of social change. Our project embodies an attempt to build on this insight, along with the participatory action research tradition, to understand the digital and dispersed workforce for data work. 

Marx’s 100-question survey was designed to document labour conditions in 19th century industrial France. To make it relevant for investigating data work, we updated it by removing items about oil lamps and furnaces and introducing questions on contemporary issues such as mental health and algorithmic management. The resulting set of 20 open-ended questions covering the material and structural conditions of labour relations was intended to serve as a flexible prompt for reflection. However, it quickly became evident that the questions were not helpful for our co-researchers. One co-researcher pointed out that asking colleagues about their salaries felt redundant since everyone earned roughly the same amount. In follow-up projects, we emphasised that the 20 questions were intended as a guide or starting point for reflection, encouraging co-researchers to develop their own questions that best aligned with their lived realities.

Our methodological approach centres on workers’ epistemic authority, recognising that their daily experiences and deep understanding of industry practices position them as experts in their own right. Their direct involvement in data work reveals dynamics, challenges and nuances that we as academics, studying from a distance, cannot fully grasp through traditional research methods alone. Additionally, we break away from traditional academic research practices, which often extract information without benefiting the participants. Traditional academic research often prioritises the researchers’ interpretation, relegating participants to passive roles. We sought to invert this dynamic by inviting workers to define their own goals and emphasising tangible impacts. At the community researchers’ request, we prioritised raising awareness with a speaker series, participation in conferences, events with policymakers and this magazine series, rather than publishing academic articles. We seek to measure our success by the factors our collaborators care about, rather than by conventional academic key performance indicators (KPIs). This project is not merely about understanding; it is about creating space for the data workers to articulate their realities in their own voices.

The power and challenges of collaboration

Collaboration has been both the greatest strength and the most significant challenge of this project. Working together has been rewarding, creating meaningful relationships and pooling our diverse skills to create powerful outputs and see the impact of this project. However, the data work industry is structurally opaque. Big tech companies not only outsource data work, but also enforce strict confidentiality throughout the production chain. Speaking out about working conditions often entails risks, including retaliation, which led nearly a third of our community researchers to remain anonymous.

In this context, establishing trust was paramount. Without it, the research could not progress. The project involved building relationships, aligning goals, negotiating roles and working through the creative process together, which required a sustained effort. While it is no surprise that collaboration can be difficult, many of the challenges we faced were rooted in the inherent complexities of working across such a fragmented and sensitive landscape.

Along with trust, establishing relationships naturally also brings emotional closeness. We have a valuable opportunity to get to know the trajectories and struggles of the community researchers in depth. Unfortunately, they all share experiences of precarious and traumatising work environments. Processing and dealing with such trauma can be deeply challenging, especially in a project that places collaboration at the centre. One particularly poignant example involved Sakine, a philosopher and co-researcher based in Berlin. Sakine gave us profound insights into the intersections of data work, precarity and trauma, particularly as a migrant in Germany. Her analyses were breathtaking, but also deeply painful. Working with someone so candid about their mental health struggles was a privilege, but also placed an emotional burden on us.

Beyond the difficulties that naturally emerge in the creative process, our intended shift in the academic research roles brought its own challenges. As researchers based in Berlin, we were acutely aware of our privileged positions within prestigious academic institutions. In contrast, many of the workers faced precarious financial situations and lived in vulnerable conditions. Our positionality as researchers often positioned us as “experts”, so we made a continual effort to create open and horizontal dialogue to decentralise authority. At the same time, some community researchers also expected some direction or validation and were frustrated with our insistence that they be the ones to decide. These moments require balancing the goals of empowering workers’ perspectives with the specific needs of each individual. Being aware and mindful of the existing power imbalances is no easy task. In fact, we were continually questioning and renegotiating our roles, expectations and methods, depending on the situation. The frustration that some of our first community researchers expressed led us to be a bit more proactive in subsequent projects, for example by offering concrete suggestions for interview questions once the topic had been decided, or guiding the writing process more closely.

Reflexivity also extended to the outcomes of our work. While the project fostered empowerment and collective sense-making, it could not directly alter the structural inequalities faced by many participants. This limitation was a sobering reminder of the challenges inherent in socially engaged research. However, it did not diminish the value of the work. There were moments of intense difficulty – sitting with someone as they shared their pain, knowing that our project could not fundamentally change their material conditions. Yet these moments also underscored the power of reflexivity and collaboration. They reminded us that while we might not solve systemic inequalities, creating space for these stories and relationships is itself meaningful.

The Data Workers’ Inquiry has been a continuous learning experience in the difficult task of bridging the gap between academic inquiry and real-world impact. A single project cannot solve the systemic issues facing data workers and a reflexive methodology cannot erase the power imbalances between academics and community researchers. Nevertheless, we see this project as a testament to the power of collaboration, reflexivity and trust. And hope that it inspires others to see research as a tool for social impact, provided it centres on the voices and agency of those most affected.