Data Product Owners: How do you foster a sense of ownership over your data?

Why does a data product need a clear owner?

Data plays a key role in virtually every organisation. However, raw data alone does not provide any useful insights. Value is created when data is processed into information that people can actually use to drive decision-making.

Think, for example, of KPIs such as turnover, EBITDA or churn, but also of datasets and reports used to manage processes.

A data product is a dataset, KPI, report or other data solution that is developed and managed to support a specific user need. Such a product is not usually created in one place. Data is sourced from various sources, transformed and combined with business definitions before the final result becomes available.

This makes the path from data source to insight broad and dynamic. Technical data sources change, but business processes and definitions also evolve over time.

So who maintains an overview of the whole picture?

Without clear ownership, there is a risk that individual teams will only take responsibility for their own part. For example, a technical team ensures that a pipeline works, whilst the business determines how a KPI is defined. For a reliable data product, these worlds must come together.

What happens without data product ownership?

Imagine: the data team has implemented a new calculation for turnover in the data warehouse. The new calculation takes better account of the different definitions of business units and therefore provides a more up-to-date picture of reality.

However, elsewhere in the organisation, an old definition is still being used. Those figures then end up in a strategic meeting.

The result: two different turnover figures.

Not because the underlying data is inherently incorrect, but because different parts of the organisation are not working with the same definition or version. This then leads to precisely the discussion you want to avoid: which figure is correct?

This has wider implications than a mere debate about a dashboard. Unclear ownership can lead to:

• conflicting KPIs and reports;
• duplication of effort between data and business teams;
• longer lead times for changes;
• additional remedial work when problems are only discovered at a late stage;
• diminishing trust in dashboards and other data products;
• decisions based on outdated or misinterpreted information.

The problem is not just technical. There is no one who is responsible for the data product as a whole: someone who knows where the data comes from, which definitions are used, who uses the product and what happens when something changes anywhere along the data journey.

What is a Data Product Owner and what do they do?

A Data Product Owner (DPO) is responsible for the value, quality and usability of a data product throughout its lifecycle. The Data Product Owner thus acts as the link between the business and the technical teams that build and maintain the data product.

The DPO does not need to build every data source or transformation themselves. Their responsibility lies, rather, in overseeing the whole.

Key questions in this regard include, for example:

• What definition do we use for this KPI?
• Which data sources does the data product consist of?
• Who uses the data product and for what purpose?
• What quality requirements must it meet?
• What happens when a data source, transformation or definition changes?
• Who needs to be involved in such a change?
• Does the data product still deliver the desired value to the organisation?

By designating a single, clear point of contact, you prevent responsibility from being passed back and forth between different teams. This makes the Data Product Owner a key component of data governance: the framework through which you organise, amongst other things, responsibilities, definitions and quality standards relating to data.

What is a Data Product Owner not responsible for?

Ownership does not mean that one person has to carry out all the work relating to a data product.

For example, a Data Product Owner does not have to build every data pipeline themselves, resolve every technical incident or own all the systems in which source data is generated. Nor does the DPO have to make all decisions independently.

The role is about coordination.

Technical specialists remain responsible for building and maintaining the technical solution. Business stakeholders contribute their knowledge of processes and definitions. The Data Product Owner ensures that these elements come together and that it is clear who bears which responsibility.

Responsibility thus creates not only value, but also trust.

How do you organise data product ownership within your organisation?

Introducing data product ownership is not just a technical change. It is primarily an organisational issue.

The technology must provide sufficient insight into the journey that data takes through the organisation. However, a complete overview of data sources and transformations is of little value if it remains unclear who is authorised to make decisions.

Therefore, do not start by appointing a Data Product Owner for every conceivable data product. Start with the products that your organisation actually needs to be able to rely on.

How to take control of your data products

Exactly how you organise data product ownership varies from organisation to organisation.

A company with just a few business-critical KPIs requires a different approach to an enterprise organisation with dozens of data teams and hundreds of data products. It is therefore not necessary to set up a comprehensive data governance framework from day one.

The core principle remains the same: make it clear which data products are important, how they are created, and who is responsible for their quality and value.

In this regard, technology and organisation reinforce each other. A scalable data infrastructure enables insight and control. Clear definitions ensure that everyone speaks the same language. And clear ownership ensures that someone takes responsibility when something changes.

In this way, you’re not just building data products. You’re building trust in the information on which your organisation bases its decisions.