Umeå University's logo

With the growing influence of data-driven technologies and Artificial Intelligence (AI) across industries, questions of data governance have become a central concern. As organizations embark on digital transformation initiatives to enhance their operations, services, and strategies through AI, the effective management of data has become a key condition for success. However, most data governance frameworks focus on technical aspects, overlooking the everyday work that makes data usable, meaningful, and trustworthy. Based on qualitative research in Swedish forestry, this study traces how data are produced, interpreted, and moved across the sector. To capture this process, I introduce the concept of data journeys, the paths along which data move, transform, and sometimes get stuck as they encounter tools, people, and organizational boundaries. What enables data movement is data liquidity, the capacity of data to be reused or recombined across contexts without losing their interpretive integrity. Data liquidity, however, is not a given – it depends on a variety of socio-technical arrangements. When these fail, data friction emerges. Data friction refers to the obstacles that slow or block data movement. Yet data friction is not inherently negative. In forestry, where data must often be interpreted with care and based on ecological expertise, data friction can be productive. It draws attention to data ambiguity, prevents unwanted data sharing, and protects against context loss. This leads to the central argument of the dissertation: data governance is not just about enabling flow, it is also about negotiating the tensions between data liquidity and data friction. Effective data governance involves knowing when to enable data movement and when to slow it down.

Based on the above, the dissertation makes three contributions. First, it moves beyond traditional assumptions of data governance as a matter of formal control, data quality, or compliance. It does so by conceptualizing data governance as a practice-based, socio-technical process, enacted through the everyday efforts of making data usable across systems and contexts. Second, it develops the concepts of data journeys, data liquidity, and data friction to trace how data governance unfolds in motion. In doing so, it highlights the socio-technical frictions that shape data work in practice. These data frictions often reveal where the limitations of dataliquidity are and what work is required to restore it. Third, the dissertation challenges the assumption that data friction is inherently problematic. It shows that data friction can be protective, reflective, and even productive.