Expert Interview: Become the Master of Your Data
Becoming truly data-driven requires much more than just tools or platforms. It requires a clear strategy, reliable data, strong leadership, and a culture in which data is an integral part of daily decision-making.
In this interview, Erik Jan Hengstmengel, author of the book “Mastering Data” (NL: Data de Baas, D: Datenkompetenz Für die Unternehmensführung), shares his vision on what it truly means to take control of your data, from leadership and organizational change to governance, adoption, and long-term competitiveness. Together with data engineer Stan Wintraecken, who explains how these challenges manifest in practice, the conversation bridges the gap between strategic vision and practical implementation. The result is a practical and honest perspective on why so many data initiatives fail, what successful organizations do differently, and how companies can lay a scalable foundation for the future.
Q: Everyone is talking about “data-driven” work. What does that actually mean in practice for an organization?
Erik Jan Hengstmengel: Being data-driven means using data to actively support decisions, processes, and daily operations across the organization. Employees should be able to ask “Where and how can we use data to perform better?” and have access to reliable and understandable insights. This goes beyond dashboards or AI tools. It requires trust in data, clear ownership, and embedding insights into everyday workflows. Ultimately, it means making data a core part of how the organization creates value.
Q: In your experience as a data engineer, what does an organization’s data environment typically look like before transformation or optimization efforts begin?
Stan Wintraecken: I’ve worked with organizations across a wide range of industries, and when it comes to data, I see the same pattern time and time again. To be honest, it’s almost never clean. Most organizations have built their data landscape over many years, so what you usually see is a mix of legacy systems, cloud applications, APIs, and reporting solutions that were never really designed to work together. A lot of the time, there’s no single source of truth. KPI definitions differ between reports; ownership is unclear, and documentation is missing. In some cases, organizations already have a modern analytics platform in place, but the underlying foundation such as integration, governance, monitoring, still isn’t there. So, in practice, data teams often spend more time fixing issues and tracing data than actually creating new insights.
Q: Technology is rarely the issue, so what actually causes data initiatives to fail?
Erik Jan Hengstmengel: Most data initiatives fail because organizations focus too much on fragmented technology and too little on people, processes, and adoption. The real challenges are usually the lack of strategic alignment, poor data quality, unclear ownership, resistance to change, and of course governance requirements. Many organizations also start with isolated pilots that are never connected to long-term business value or embedded in an overall technology infrastructure. In reality, becoming data-driven is not primarily a technology transformation, it’s an organizational transformation.
Q: Trust in data is often a major barrier. What typically prevents organizations from trusting their data?
Stan Wintraecken: The biggest issue is inconsistent metric definitions. When the same metric means different things across reports, it’s easy to lose trust. Undocumented transformations, spreadsheets, layered logic, and unclear ownership make it difficult to verify numbers, causing teams to create their own versions of the truth. The solution is a governed source of truth with documented definitions, lineage, transparency, and shared ownership between business stakeholders and engineers. The real barriers are usually a lack of alignment, poor data quality, unclear ownership, and of course a fragmented technology platform.
Q: Even with the right foundations in place, adoption can still fail. What needs to change to truly embed data into daily decision-making?
Erik Jan Hengstmengel: I think adoption only succeeds when data becomes part of how people think and work, instead of feeling like an extra reporting layer. That means organizations need to focus just as much on culture as on technology. Employees need to trust the data, understand its value, and have easy access to relevant insights in their daily work. Leadership is also extremely important. In my experience, when leaders actively use data in decision-making, the rest of the organization follows much more naturally.
Q: As a data engineer, can you explain how a strong data foundation looks like in practice?
Stan Wintraecken: For me, a strong foundation is mainly about structure and consistency. You want a setup where raw data flows into a curated layer with standardized definitions, and from there into analytics-ready datasets and dashboards. At the same time, good engineering practices like version control, CI/CD, monitoring, and documentation should simply be standard. If that foundation is in place, organizations can deliver insights much faster and with a lot more confidence in the data.
Q: What does it truly mean to „become the master of your data“ at an organizational level?
Erik Jan Hengstmengel: In my opinion, becoming the master of your data means treating data as a strategic asset instead of just a byproduct of systems. It’s about taking control over how data is managed, shared, and used across the organization. That requires a clear strategy, governance, trusted data, leadership commitment, and a culture where people actively use data in decision-making. And most importantly, it’s not a one-time project. It’s an organizational capability that continues to evolve over time.
Q: If a company wants to take control of its data, what are the first technical steps they should take?
Stan Wintraecken: From my perspective, the first step is actually to define the architecture before building anything. A lot of organizations jump straight into dashboards or pipelines without thinking about how everything should fit together long term. After that, I’d recommend starting small but structured. Set up the platform properly, implement standards like Git and CI/CD early, and bring one important data source through the full process end-to-end. And honestly, involving the business early is just as important as the technology itself. Otherwise, you still end up with different interpretations of the same data.
Q: The gap between ambition and reality often becomes most visible when organizations try to scale. Why do so many organizations struggle to move from pilots to scalable, production-ready solutions?
Erik Jan Hengstmengel: I think many organizations underestimate how different a pilot is from a scalable solution. A pilot is built to prove potential quickly, but scaling requires a solid infrastructure, governance, ownership, alignment, and long-term commitment. What often happens is that pilots remain isolated experiments instead of becoming part of a broader organizational and technology strategy.
Stan Wintraecken: Pilots are usually optimized for speed, not for scalability. Once you move toward production, suddenly things like security, monitoring, integration, and performance become critical. From my experience, the organizations that scale successfully are the ones that build a reusable foundation early and treat data products as long-term products, not one-off projects.
Q: Becoming data-driven is not just about today, but about staying competitive tomorrow. What will happen to organizations that fail to take control of their data in the next 3–5 years?
Erik Jan Hengstmengel: In my opinion, organizations that fail to build a strong data and technology foundation will increasingly struggle to adapt and compete. Decision-making will stay slower and less transparent, while competitors move faster with AI and data-driven operations. The gap between organizations that can scale AI and those that cannot will grow very quickly.
Stan Wintraecken: From a technical perspective, I think the cost of fragmentation will only keep increasing. Organizations will remain dependent on vendors, struggle with inefficiencies, and stay stuck in pilots without real operational impact. A strong data foundation is no longer just about AI ambitions. It’s what enables organizations to move faster, make better decisions, and stay competitive for the long term.