THE TREATING DATA AS A PRODUCT AND DATA PRODUCT LANDSCAPE SURVEY 2026.
We surveyed people across the industry to understand the landscape. Here's what they told us about how Businesses Treat Data as a Product and how Data Product Management is practised today, how value is measured, and how AI is changing the practice.
1. Who responded
What this tells us
These findings come straight from the people doing the work day to day not from vendors or platforms. That makes this a practitioner's view of the discipline.
What this tells us
The survey gives us a worldwide account. Whilst skewed to the UK the themes you'll read came up consistency across the world.
2. Who they represent
Industry
Organisation size
What this tells us
Treating data as a product isn't confined to one sector or one size of company. The same questions come up whether you're a 30-person fintech or a global insurer. The responses came from a wide mix of industries and business size.
3. How mature is Data Product Management?
What this tells us
Whilst the discipline catapulted 5 years ago with the launch of the Data Mesh book, only a relatively small amount adopted it early. But what we are seeing now is a huge shift 5 years on with. the majority of companies going on the journey. With the majority developing or emerging.
4. Where the work comes from
What this tells us
For many teams, work arrives as requests rather than through proactive discovery or a roadmap. That means teams are reacting to demand rather than leading. This concept has been referred to as a Data Factory.
5. Strategic or delivery?
What this tells us
Data teams should be leading the work through proactive discovery, but over half are stuck between delivery and strategy. If we are to truly Treat Data as a Product we need data teams to lead the work.
6. How value is measured
What this tells us
Measuring outputs is easy dashboards shipped, pipelines delivered, but its a vanity metric that we cannot prove what we have data teams have done to drive the business forward. Measuring outcomes is harder, but without it you can't tell whether a data product is actually creating value for anyone. By measuring outcomes it also puts the team in control of working out the solution and reporting back on progress to something measurable and meaningfull.
7. What happens at the end of a product's life
What this tells us
You can do one of 3 things to a Data Product: Improve, Maintain or Retire. Without a deliberate process for retiring data products, old dashboards and pipelines pile up, quietly consuming cost and eroding trust. This is particularly problematic in the Data Factory example where teams are building based on demand, building up a vast estate and not retiring unwanted or unused Data Products. For Data Product Management to work as designed the estate must be manageable by the team.
8. AI is changing the work
What this tells us
For many years teams have worked in a Kanban 2 week sprint style. Affording 1 week for discovery and 1 week for delivery. AI now means that the build cycles have reduced, teams can build faster. This leaves more time for the all important discovery. Actually understanding what our users want. And with over half saying AI is changing how users consume the actual data, discovery needs more time than ever.\n
9. Discovery
What this tells us
This means most people are being given product work and then doing discovery rather than doing proactive discovery that drives the work. It's the difference between validating something you've already been handed, and letting user needs decide what gets built. This is a big call out the the industry. Data teams have all the expertise to understand their users problems so they can find the solutions. Data Teams need to build time in for Proactive Discovery to enable this shift.
10. Where AI helps today
What this tells us
AI is already useful for the core craft of product work. Used well, it frees up time for the strategic side and more proactive discovery. That means Data Teams can spend more time with their users, interviewing and understanding their needs, and less time writing up.
About this survey
The survey was conducted in September 2026 and received a total of 20 responses. Respondents included independent consultants and people employed by a range of businesses worldwide.