Since Datanova: The Data Mesh Summit and our in-person executive discussions on data products and Data Mesh, we’ve been validating the data product approach — starting with identifying value and then selecting smaller use cases — that really resonated with the audience.
If you haven’t had a chance, you may still download the Data Products workbook. The workbook gently nudges you to think about your candidate data product because we’re going to be thinking about the why. More specifically: Why should the business create that data product? Identifying key performance indicators that will help the business get the most out of creating and building data products, help you fill out the workbook, and get the most out of this series.
Up to this point, we’ve discussed the why and the what of creating and building Data Products. Now, let’s talk about the how. How do we engage with the business as well as the data consumers? How do we operate this data product? How do we grow the behavior across the business? How do we evolve data products and then how do we monetize? We tackle the answers to these questions below.
Getting an executive to understand data products
Cultivating a data product mindset is a challenge, yet it’s vital because it now brings value to a technology-led business. Keep in mind, technology alone can’t create value by itself and so they need the business leader to be involved and define value.
One thing we must get right is to sell this idea to a non-data business executive. When we talk about Data Mesh, it remains data geek speak. The question becomes, “How do I get a non-data executive to care?” A data product is something they could care about, but what does even a data product look like?
One proven method is to show them: take a part of the business that everybody cares about such as finance or HR, and create a data product. And finance and HR are oftentimes easier to start because data is well understood and within one domain. Most importantly, it’s data that every part of the business needs.
We’ll take that common denominator to demonstrate what good looks like with a marquee data product. At the same time, we’ll work out how we support other products so when we are done, we’ll also have worked out the foundation to scale. Then, when you hear something along the lines of, “I love that data product; I wish all my data was data products,” now you have the playbook to move forward.
Prove value with marquee data products
As you place those initial marquee data products in production to demonstrate the value, let’s build, maintain, and update those initial data products; while working out the standards for onboarding and supporting new product development and use. Engage with existing data teams to identify the products they may already have or need.
Support various business groups
As we scale our operations, we’re going to support users that are no longer in the same line of business, but are now, across the business. The question is: how do you support these different communities who might have different levels of both tech maturity as well as expectations? Keep that user persona in mind when thinking about user management and service desk.
Set the foundation and scale
Through the initial few products that you’ve launched, you’ll likely have your initial wins which you can use to build a foundation. Make it easier to grow by offering core frameworks that you’ll need to develop, use, and monitor products at scale. A few areas to consider include: metadata hubs, data virtualization, data fabric and knowledge graphs.These are things that make it easier for people to produce products and to consume products.
Iterate and gain indirect value
Evolve to make better products that increase value through both direct and indirect monetization — increasing productivity, improving collaboration with the business, making the business safer, more efficient, more innovative. The point is that you’ve evolved beyond a one-time project to a culture that starts with looking for a data product to use. By now, you’re regularly revisiting the different data products, comparing with user demand, and assisting different groups to evolve to the next stage to answer their questions and accelerate their insights.
For instance, you might see two products and notice they’re closely aligned. Should there be a derivative product that actually combines them together? Maybe we see people searching for a type of data products that nobody has and maybe reaching out to certain groups who are most likely to be able to create those products and fill those gaps.
Organizational Planning: Roles in the Data Mesh Framework
As you’re putting together your Data Mesh framework, you’ll need to consider the key roles in your organizational planning. We highlight the vital roles below.
Data Sources
Think about who your data sources are—who within the organization owns that data? What’s the build that they need to support to be able to support the product that you had in mind? They could be internal, a third party or a partner.
Producers
Who are the producers? They are the data product owners and will be supported by a team to create the product–data stewards, business analysts, data analysts, data engineers, etc. These should be named individuals who’re actually going to be playing some of these critical roles.
Consumers and usage
Who are your end users (i.e. business users, data scientists, analysts and IT) and what are the interfaces that need to be built so that they can best be able to use that product?
Central data team
Since there is a business incentive to move quickly from a data and AI perspective, the central data team, more than other teams, will understand how exciting this is and they should be excited because data is driving every aspect of the business, and so data will be tightly integrated everywhere. So, once you have data products, much of the work is putting all the right control mechanisms to execute, no matter where your data is, and no matter who’s using the data.
In fact, you’ll actually have better governance by having a Data Mesh, than before, when the central team might have done it all. But with Data Mesh, you offer tools and processes for others to implement the hooks needed for federated computational governance.
The future role of the Chief Data Officer with Data Mesh
And so now, with the pivot towards Data Mesh, the Chief Data Officer role is also much more important than ever as the role will evolve from day-to-day tactical data exploitation to focus on the strategic growth of the business with respect to data. This move is aligned with a revenue-focused position rather than a cost-focused position.
Why? According to an analysis by Accenture, in 2021, among executives of the world’s 2,000 largest companies (by market capitalization), those who discussed AI on their earnings calls were 40% more likely to see their firms’ share prices increase — up from 23% in 2018.
Putting it all together: strategy and governance
During this series, we covered strategy and governance. We went through the why of the business vision, identifying unique data needs and the users. Keep in mind that this is an iterative process as you work with your users to confirm their needs.
After we identified the product definition, there’s one or two data products that’s worthy of launching. Essentially, you’re taking the data product team, the data product owner, the data product architect, central data team, business sponsor, product team and bringing them together to define the vision for supporting the users and to measure the value of that data product.
All the planning leads to the deciding on the strategic execution of how to put the data product into production. How are we going to manufacture it? How are we going to govern it? How are we going to service and operate this data product? And how are we going to evolve that product?
Creating data products is just the beginning. There are other dimensions that are just as important that need to be assessed such as architecture, development, regulation and ethics, and user support.
For now, a good place to start is to scope out your data product with the ingredients that you need to create your data-driven business plan.
