Your data strategy isn't failing in delivery. It's failing in ambition.
By Gabbi Stubbs
Most organisations can tell you what they have spent on data. Far fewer can tell you what the business does differently because of it.
That gap is where data strategies quietly die. The pattern is familiar. A strategy lands on the executive team's desk, and it is well made: a maturity model, a target architecture, a roadmap in three horizons, a governance framework and a business case with a satisfyingly large number at the bottom. It is approved. A programme is stood up. Platforms are chosen, and consultants and hires arrive.
Two years later, the same organisation is commissioning a refresh. The platform is built, but nobody is quite sure what it is for. Dashboards multiply while decisions stay stubbornly the same. The data team is busy. The business is unimpressed.
The post-mortem blames execution: the migration ran late, adoption was patchy, data quality was worse than expected, change management was underfunded. All of that may be true. But it skips the question that matters most. Did the organisation ever agree what it wanted its investment in data to make possible?
Usually, it didn't. The ambition was to build a capability, not to change the business. And a capability, however well engineered, creates nothing on its own.
The conversation leaders avoid
The question is deceptively simple: what are we trying to achieve as a business, and how will data help us do it differently, better or faster?
Not "what is our data vision?" or "how do we become data-driven?" Those questions invite slogans. The useful version is concrete. Which outcomes matter most over the next eighteen months? Where could data and intelligence change the equation? And what are we prepared to stop, start or reallocate to make it happen?
This is where it gets uncomfortable. The conversation exposes competing priorities. It reveals that two executives hold different views of what the business is trying to achieve this year. It challenges the comforting assumption that more information leads to better decisions. Sometimes the problem isn't a lack of data at all. It's that the organisation is unwilling to act on what it already knows.
So the conversation gets skipped. Teams move straight to the things that feel like progress: choosing a platform, defining an operating model, hiring a Chief Data Officer. These are tangible. They can be budgeted, planned and measured. The hard conversation doesn't come with a Gantt chart.
Better decisions are only one answer
Most data strategy conversations narrow quickly to a single idea: giving people better information so they make better decisions. It is a good idea, and it is far too small. Data creates value in at least five ways:
A strategy that only asks "what do we want to know?" will never find most of these. The business may not need better information to make a decision. It may need to stop making that decision manually, redesign the experience around it, or commercialise something it has been giving away.
AI raises the stakes. The question is no longer just what people need to see. It is what the organisation knows, what it can infer, what it can automate and where it can build something competitors cannot easily copy. A business can have excellent data infrastructure and still create nothing if it hasn't decided how intelligence will be built into its products, its operations and its decisions.
The warning signs are easy to spot
The symptoms of a strategy that has lost sight of the business are remarkably consistent.
The strategy is about data, not the business.
Count how often the document mentions customers, revenue, margins, risk, service levels or growth, and compare it with platforms, pipelines, lakes and governance. If the vocabulary is overwhelmingly technical, it has probably been written for the data team, by the data team, to justify the data team.
Everything is a priority.
Without an agreed view of where value will be created, every stakeholder gets a slice of the roadmap. The programme spreads across a dozen use cases, and none has a clear owner or a compelling case.
Success is measured in activity.
Tables migrated, datasets catalogued, users onboarded, dashboards launched. These are outputs. They tell you the machine is running, not whether it is producing anything the business can use.
The business isn't accountable for the outcome.
Ask who owns the value of a data initiative and you will usually hear the name of someone in the data function. The people who would have to change how they work are "stakeholders", kept informed.
Governance arrives before purpose.
Policies, councils and stewardship roles are set up to manage assets before anyone has agreed which ones matter or what they are meant to deliver. The result is bureaucracy with no clear beneficiary, and the business learns to route around it.
Why smart organisations still get it wrong
This isn't a failure of intelligence. It is a failure of incentives, leadership and sequencing.
Platform decisions look like momentum. A difficult argument about priorities and trade-offs does not. Data teams are often positioned as service providers, so they take the brief rather than challenge it, and asking senior leaders what they actually want can feel like overstepping, particularly once the organisation has declared that it will be "data-driven".
There is also a persistent belief that capability comes first and value follows. Build the foundations and the use cases will emerge. Sometimes they do. More often the organisation ends up with beautifully engineered foundations and no compelling reason for the business to use them.
None of this is an argument against foundational investment. Security, regulatory compliance, data quality, architecture and resilience matter, and the business cannot always name the opportunities it hasn't yet had the capability to discover. But foundations need to be justified like any other investment: against the strategy, the risks being managed and the value being enabled. Otherwise the organisation is paying for capability with no view of the return.
Start with the business, not the roadmap
A better strategy begins with a different set of questions, asked by the leadership team before any money is committed.
What outcomes matter most?
Not the corporate vision. The two or three outcomes that will make the biggest difference over the next twelve to eighteen months: growth, margin, retention, cost to serve, risk, a new line of business. Be specific enough that someone could tell whether it had been achieved.
Where can data and intelligence change the equation?
Find the decisions, processes, experiences or capabilities that drive those outcomes. Where would better information help? Where would automation remove friction? Where could intelligence create something customers would pay for? Don't assume every opportunity needs a new platform, and don't assume every one is a better report.
What would actually change?
For each priority, say what the organisation would do differently. Would someone make a different decision, a process be redesigned, a customer be served better, a risk be caught earlier? If the answer is unclear, the value proposition isn't ready. If the honest answer is "probably nothing", better to learn that before building anything.
Who owns the outcome?
Every priority needs a named executive outside the data function who is accountable for the result and has committed to changing how their team works. If you can't find one, that is a signal about how much the priority really matters.
What are we choosing not to do?
A strategy is a set of choices. Back the opportunities with the greatest value, the clearest ownership and a credible path to delivery, and explicitly defer the rest. The list of things you are not doing protects the programme from being pulled in every direction.
Only then should architecture, tooling, governance, skills and operating model be designed. They become answers to specific questions, rather than a catalogue of generic good practice.
The real return on a data strategy
The cost of getting this wrong is rarely a dramatic failure. It is a multi-year programme with no line of sight to revenue, margin or risk, which becomes the first thing questioned when budgets tighten and has no good answer when it is.
When organisations get it right, three things change.
The strategy gets shorter. Anchored in a few outcomes that matter, it stops trying to cover every conceivable capability and becomes a plan rather than a catalogue.
The investment gets easier to defend. A business case built on clear outcomes, named owners and measurable value is far more credible than one built on "improved insight" and industry benchmarks.
And the data function earns a different relationship with the business. It stops being the group that delivers platforms and reports and becomes a partner in creating value. That is a much more durable position, and a far more satisfying job.
Before you approve the roadmap
If you are about to start, or refresh, a data strategy, try a simple test. Ask five senior leaders, separately, to name the three outcomes that matter most to them, the decisions or capabilities that will drive them, and what they would do differently with better data or intelligence.
Then compare the answers. If they are vague, contradictory or disconnected from the business strategy, you have found your starting point. It isn't a technology choice.
The organisations that create the most value from data won't necessarily be the ones with the most sophisticated platforms or the most comprehensive strategies. They will be the ones that know what they want to achieve, where intelligence can make a difference, and what they are prepared to change to get there.
The question isn't whether your organisation has a data strategy.
It's whether your business has a strategy for creating value with data.
Want to talk strategy?
If this resonates, let's have a direct conversation about your business.