Do we have a Data Culture?
Source: https://link.springer.com/chapter/10.1007/978-3-658-27495-5_11 ↗
Kremser and Brunauer sharpen what "data culture" actually means by treating it as a subtype of organisational culture — not a technology stack or a dashboard habit but a set of shared assumptions about how decisions get made.
The paper distinguishes between organisations that claim to be data-driven and those whose structural conditions (governance, literacy, feedback loops) actually support it.
For product direction this is useful because it names the non-technical prerequisites that most data initiatives skip: you cannot build a data-informed product organisation by buying tools alone.
The framework is compact and provides a diagnostic lens for why some teams generate insight from their metrics while others drown in dashboards.
Pair with Croll and Yoskovitz's Lean Analytics for the operational complement.
Central argument
Kremser and Brunauer argue that 'data culture' is a specific subtype of organisational culture — constituted by shared assumptions about how decisions are legitimised — rather than a property of technology adoption or analytical tooling. Their central finding is that most organisations claiming to be data-driven lack the structural conditions that would make that claim meaningful: coherent data governance, widespread data literacy, and feedback loops that close the gap between metric generation and decision-making. The paper offers a diagnostic framework for distinguishing performative data culture from the real thing.
Critique
Published in 2019 and grounded in organisational culture theory, the framework risks being more taxonomic than actionable — it is better at naming what is missing than at prescribing how to build the structural conditions it identifies as necessary. There is also a tension the authors may underweight: in fast-moving product organisations, the deliberative, governance-heavy conditions they associate with genuine data culture can conflict with the speed at which decisions must be made, raising the question of whether their model implicitly favours large, stable enterprises over leaner product teams.
Why it matters for product
For a CPO, the paper's diagnostic lens is directly applicable when auditing why a product organisation drowns in dashboards without changing behaviour — the culprit is rarely the metrics themselves but the absence of shared decision norms and literacy that Kremser and Brunauer identify as prerequisites. It also reframes hiring and team design: investing in data infrastructure without first establishing governance and interpretive literacy at the product team level reproduces exactly the gap the paper describes. This is particularly sharp when scaling discovery practices, where multiple teams measuring the same funnel can reach contradictory conclusions precisely because the cultural substrate for interpreting data is absent.