We all get the gist of data mesh: it's about scaling up data-driven value in a cohesive way. With the rise of generative AI and other high-octane techniques that need a solid data foundation to succeed, it's no wonder everyone's buzzing about data mesh. In the first entry of this series, we focused on insights from our two-year deep dive into data mesh adoption in big, intricate enterprises.
Extracting value from data on a large scale demands transformations that diverge from how companies have traditionally managed their data. This means weaving in product-centric approaches to data, building multi-tenant and self-serve data platforms, aligning data ownership with business domains, and reimagining data governance as a catalyst rather than a constraint. While the journey to adopt data mesh demands effort, it's a worthy endeavour when aligned with tangible value.
Establishing the direct link between a data mesh model and tangible business value isn't just important; it's the linchpin for modern data strategies. While this presents a nuanced challenge without a clear-cut answer, achieving clarity here is vital to garner commitment from the business and support from stakeholders. While many discuss business value and data mesh in isolation, the real breakthrough comes from intertwining them effectively. In this post, I'll unpack approaches and tactics we’ve deployed in the real world and how we’ve seen enterprises reap the benefits.
Pragmatically, this boils down to three simple, but important questions you need to keep in mind:
Business value is an informal term; there is no universally accepted definition. What I have in mind is akin to Mark Schwartz’s definition from his book, The Art of Business Value:
“Business value is a hypothesis held by the organization’s leadership as to what will best accomplish the organization’s ultimate goals or desired outcomes.” - Mark Schwartz, The Art of Business Value.
When considering data mesh adoption, it's crucial we tap into significant value that justifies the transformative changes required for the business.
The value we aim for should be:
Effective data transformation is all about aligning with the company's core strategic priorities. Data strategies defined in isolation can end up out of touch with the actual business needs, diminishing their impact. Using the company's strategy—often articulated in publicly available strategy documents—as our foundation, we need to define:
The Scoreboard: Which performance indicators and metrics will guide us and set our priorities?
The Upgrades: What net new capabilities are we enabling?
The Testing Grounds: Where within the organisation can we best demonstrate the capabilities of a data mesh transformation?
Together, these elements ensure our approach is both focused and in tune with the company's broader objectives, giving us a strong starting point.
Let me share some hands-on tools and tactics we've tapped into to pinpoint and voice the unique value data mesh will offer in various scenarios.
At the heart of data mesh is the recognition that data isn't merely a technical asset, but a fundamental contributor to value creation.
The concept of a value chain, introduced by Michael Porter in his 1985 book "Competitive Advantage: Creating and Sustaining Superior Performance," describes value chain as the sequence of activities an organisation undergoes to deliver a product or service. In today's digital age, data is woven into nearly every aspect of these activities.
It's crucial to establish value chains that highlight how data drives value creation. This not only quantifies the value of data but also simplifies communication about intricate data concepts with the wider business. This clarity becomes especially vital when the origin of the value is distant from its realisation point, and when multiple ownership layers are involved.
For a cohesive data mesh strategy, creating a domain map is a crucial navigational tool. This map should detail the core business domains, their owned data that other domains can benefit from, and the data they require to achieve their own goals.
Rather than merely reflecting the organisational hierarchy or some abstract concept about data, the domain map serves as a practical guide. Its purpose is to clarify data ownership, pinpoint where specific data is needed, and illustrate how different business domains interact with each other in terms of data. Such insights will enable the organisation to establish effective and robust data governance ensuring seamless data interoperability and aligning data product deliverables with genuine demand.
Embracing data mesh goes beyond its underlying technology. A frequent misstep is treating it like a traditional centralised and top-down architectural task, which neglects the input of data consumers and the value delivered to them.
In reality, bringing a data mesh strategy to fruition involves the creation of numerous data products, and it's imperative that these are approached and closely aligned with the value they bring.
After all, data products are still products. Adopting product thinking allows us to harness established approaches and best practices in identifying, prioritising, and defining products that deliver genuine value. Techniques like the opportunity backlog and product canvas come into play here, providing a clear, shared, and actionable understanding of potential data products and how to bring them to life.
Data mesh democratises the data by putting it in the hand of those who need it at scale. This process involves a lot of interaction with stakeholders who do not have a deep technical background and would not benefit from diving into the technicalities of building data products.
What we need instead is a data-driven, visual and interactive way of showing the users what a data product would look and feel like. This is where data product prototyping comes in as a powerful technique to achieve that. Realising the value of this approach but also the lack of tooling options in this space, we have been working on a specialised tool tailored for this very purpose.
In today's data-centric business landscape, a strategic, value-driven approach to data is non-negotiable. Data mesh, as we've explored, is more than just a technical architecture; it's a true shift in how we perceive, handle, and derive value from data. By emphasising clear domain mapping, product-centric thinking, and tangible prototyping, we can bridge the gap between the technical world of data and the broader business landscape.
At Mesh-AI, we're not just talking about these concepts; we're actively pioneering tools and strategies that bring them to life. As part of our Data Strategy Accelerator, Mesh-AI follows a structured framework to baseline our customers’ data maturity across people, process and technology. We conduct a focused assessment, analysing a handful of applications or products to establish the current state of your data transformation. You can find more information on our accelerator or by getting in touch with us.
As we continue to evolve in this data-driven era, ensuring that our approach to data remains rooted in value and strategy will be pivotal to achieving successful data transformation.