Skip to main content

Select your location

Two individuals looking at digital screen

Think differently about data: 5 steps to enable Data-as-a-Product thinking

We've got decades of data challenges and failures behind us. Various data management approaches may have worked for a while, but so far none have proven flexible or scalable enough as data sources increase and volumes grow.

Product thinking has been applied successfully to other areas of IT like application development--and now thought leaders and technology executives are looking at this approach for data management. Is this the solution to tapping the potential that is locked in enterprise data stores?

I recently hosted a discussion with a community of CIOs across industries, in which we narrowed in on the importance of treating your data like a product - aka Data-as-a-Product thinking. Together, we addressed the all too common questions leaders across organizations are facing today when it comes to managing their data:

  • How does my team need to think differently about data?
  • How do I sell these ideas to business leadership?
  • How do I get my business partners on board?
  • How should my team integrate governance into the data-as-a-product mentality?

Here are the top key takeaways from the discussion and actionable steps CIOs and leaders, like yourself, should take to apply Data-as-a-Product thinking across your teams:

Unleash the potential energy stored within your data

Download one-pager

What does product thinking look like when it’s about data?

The goal in a Data-as-a-Product maturity model is a data function that lets users easily access and extract maximum value from data with complete trust in the system and the quality of the data they retrieve.

Think in terms of a pharmacy: When you go in to purchase medication, that’s just what you do. You don’t contact the pharmaceutical company to ask about the chemical compound of the medicine or how they produce it. It wouldn’t even occur to you to do so. You trust that the medicine and dosage you need are in the container you purchase.

This is the goal of Data-as-a-Product. Your user only focuses on what they need. They don’t question the system or how the data is managed. They simply “buy” what they need and move on with their work.

As far as first steps in the right direction, think in terms of enabling and embracing product thinking.

Step 1: Enable product thinking for data by asking the right questions

Enabling product thinking around data is, at least, 80% people and 20% technology. In some companies, it may actually be 85/15 or even 95/5. The key point is that it’s a mindset shift or a cultural shift more than something accomplished through technology.

Try this exercise with your data managers and teams: Get their answers to these questions and consider the results.

  • What are our five most valuable data assets?


  • Can we describe that data?


  • Do we know what value those assets are driving into the organization?


  • Do we know who's the best person to talk to about that data?


  • Do we trust that data?


  • Do we know where that data escapes and grows legs around their organization?


  • Who's using it, and for what reasons?


There is no technology involved in these questions. They all revolve around the business and the people in the business that are affected by it. Gathering this information should move your teams in a “Data-as-a-Product” direction.

Step 2: Embrace product thinking from the top down

Building a system is easy. Making a culture shift is a challenge. Like any culture change initiative, top management is key. When the CEO, other executives, and business managers embrace data as a product, the organization as a whole will move in that direction.

For example, it is not unusual for leaders to announce what is being done in the business, but it is unusual for them to include why a particular direction is being taken. Matching data to decisions (“these numbers show us that….” or “the data indicates…”) begins to show the importance of data in the business.

Step 3: Create motivation and momentum

If you don't actually activate participation, let people know it's there, or give people a reason to want to use that data, why would they? The best shop in the world is no use without any customers.

To create momentum and incentives use, look at what you’re currently building & plan ahead. Talk to the people on those teams and show them the pain points that data engineers are experiencing trying to clean up the data they are producing.

Motivate them by showing where their data goes and the value of it. Create motivation and momentum by involving the team in planning by hosting workshops or whiteboard sessions to work out how to create value and modernize it. Gain buy-in by finding ways to incentivize.

Data mesh and Data-as-a-Product

Turning attention to that 5 to 20 percent of the equation, technology is still how we deliver change. But what technology best suits Data-as-a-Product?

Data mesh is a scalable way of applying order to the data we have today vs data lakes where data is centralized to one place and difficult to move, scale, and to get the agility we need.

Imagine, for example, if Amazon in its early days had decided to stockpile every book into one distribution center, and then tried to distribute those books everywhere in the world from once single location. Sounds a little bit ludicrous, right? You can almost imagine the scale of that and how it would grind to a halt very quickly. It would become very difficult to manage, and very disorganized.

Companies like Amazon have always acknowledged the fact that the best way to scale is to distribute these effort through its network, hubs, and resellers. And data is a similar thing which is why data mesh is a better solution when preparing ourselves for a product thinking approach.

Data should be as easy to consume as over-the-counter medicine. When we all go into the pharmacy, we take something off the shelf, we trust the brand, we trust the instructions, and we just use it.

Karl Hampson - CTO Data & AI, Kin + Carta

The four stages of the Data-as-a-Product model

 As organizations and IT leaders start on a journey towards data maturity, they should leverage this four-stage proposition model to unlock the same set of tools that product owners use to bring teams together.

  1. Domain-oriented decentralized data ownership and architecture: The ecosystem creating and consuming data can scale out as the number of sources of data, number of use cases, and diversity of access models to the data increases.
  2. Data-as-a-Product: Data users can easily discover, understand and securely use high-quality data distributed across many domains, and it’s a delightful experience.
  3. Self-serve data infrastructure as a platform: Domain teams can create and consume data products autonomously. They do not perceive the complexity of the building, executing, and maintaining secure and interoperable data products.
  4. Federated computational governance: Data users get value from aggregation and correlation of independent data products. The mesh is behaving as an ecosystem following global interoperability standards and standards that are baked computationally into the platform.

Getting started ownership is key

The starting point of the Data-as-a-Product journey will vary from company to company. One of the basics that applies across the board is ownership of the data. “Who owns the data?” should not have “IT” as the only answer. Ownership needs to be cross-functional and self-sufficient so that data engineers and domain owners work together to control data creation and determine how it's supplied to the rest of the business.


Managing data as a product is in its very early stages, and we can expect to see more conversations about this by thought leaders as time passes. We are likely heading toward an “x-factors for data,” similar to the 12 factors for cloud-native apps, which define behaviors as well as processes.

Kin + Carta was named top Data Management service provider by Forrester

Learn more