Stop Building Data Strategies That Don't Deliver
The breakthrough in data strategy rarely comes from better technology. It comes from clarity about what you're actually trying to change.
Lizzie McKinnell leads our Data & AI practice at Egremont. Having worked with FTSE 100 and 250 organisations across retail, utilities and financial services, she has seen first-hand what separates data strategies that deliver from those that stall. Here, she shares her perspective on how to build one that works in practice.
The real issue isn't access to data
"We have so much data, we don't know what to do with it." It's a familiar position - and one Lizzie hears often.
The right data should change how you approach delivery, provide feedback on performance, highlight where to act and open up new opportunities, including new revenue streams. It should take you from information overload to clarity, but in many organisations, it does not.
When data is used well, it sharpens decisions. When it isn't, organisations default back to instinct, experience or incomplete views, even at the most senior levels. The issue isn't access to data, it's how it's used.
Data strategy is often presented as a roadmap of platforms, tools and initiatives. Lizzie is direct about this: that framing is part of the problem.
“In practice, a data strategy is a set of decisions about how your organisation uses data to deliver its strategy. It tells you how to construct something, but first you need to decide what to build. The roadmap should follow from those decisions, not the other way around.”
Most organisations already have lots of data, capable platforms and skilled teams. Yet value remains inconsistent, and this is rarely a technology constraint. It is usually a lack of clarity and alignment;: which decisions should data improve, and what does that mean for how teams operate?
Start with decisions, not data
"The focus is often 'making better decisions with data, or 'becoming data led', but an effective data strategy needs to be clear; what are the decisions data can help with? Where do we want to focus?”
That clarity is everything. If a strategy does not change how key decisions are made, it will not deliver value. Starting with tools, architecture or data assets creates activity, but not necessarily impact. Starting with decisions forces focus and prioritisation.
Which decisions matter most? Pricing, personalisation, supply chain planning, risk management? The specifics will vary, but the focus needs to be explicit. Once those decisions are clear, the rest follows: what information is needed, how frequently, and by whom? This is what aligns data strategy with business strategy in practice.
Be specific about value - then prove it
"'Becoming data-driven' is one of those phrases that sounds like a destination but doesn't tell you where you're going. What does good actually look like for your business? Start there."
Value needs to be defined in terms that matter to the business - reducing customer churn by 5%, improving forecast accuracy by 10%, cutting operational costs in a specific process. That level of specificity does two things: it focuses effort and creates a basis for prioritisation.
In less mature organisations, there is often a push for more reporting or more spreadsheets. That can help, but it is rarely enough. The bigger opportunity usually comes from redefining metrics, improving access to tools or embedding data directly into workflows.
And the value needs to be demonstrated, not assumed. Running a pilot, a new tool, a new way of working, or an embedded analytics team member, lets you show early proof points, build credibility and make it easier to scale.
Treat data like a product, not a by-product
Data is often treated as a by-product of systems, something that accumulates in the background and gets queried when needed. The problem (and business risk) is that often no one truly owns it, no one is responsible for its quality, the governance is opaque, and by the time someone needs it, it often isn't fit for purpose.
Lizzie draws a useful parallel: think about how a product team operates. They know who their users are, they care about quality, they iterate based on feedback and they make sure what they build is actually useful. Treating data the same way changes the dynamic entirely.
"One organisation I worked with had seven different versions of the same customer metric sitting in different teams. Nobody was wrong, they'd each built what they needed. But nobody could agree on a single definition, which made any kind of joined-up decision-making almost impossible. Getting clear on data ownership and agreeing a single source of truth unlocked months of stalled progress."
In practice, this means clear responsibility for quality and availability, a governance approach with controls aligned to risk, and documentation that supports real use, not just compliance. When it is done well, duplication and friction reduce, trust improves and delivery accelerates.
Invest in people, not just platforms
"I've yet to see a strategy fail because the technology wasn't good enough. I have seen plenty fail because the people expected to use it weren't brought along."
Technology investment is visible and often easier to justify. Capability is less visible, but more important. The data operating model is a critical success factor: without the right structure, skills and ways of working, even well-funded strategies stall.
This goes beyond the data team. It means thinking about how business and technology functions interact, how non-technical users are developed, how specialists progress, and how data literacy is built across the organisation. If data is going to influence decisions at scale, the people making those decisions need to trust it and know what to do with it.
This is where many strategies ultimately succeed or fail.
Final thought
"A good data strategy isn't defined by how much data you have or how advanced your tools are. It's defined by whether it changes how your organisation operates. The test is simple: are decisions better, faster or more consistent? If not, go back and ask why."
That is the breakthrough: when data stops being an initiative and starts changing how decisions are actually made. The organisations that get there stay close to outcomes, test what works and scale deliberately. That is where data starts to deliver real impact.
If you are thinking about your data strategy and would like to talk it through with Lizzie or the Egremont team, please get in touch, we would be glad to hear from you.