In my previous essay, I argued that the data strategy document — the tiny pink box — is the smallest part of a data leader’s job. The real work happens in the ecosystem around it.
My friend Linn Tove read it and asked the question I’d been avoiding:
if everyone agrees with the strategy but nobody does anything differently, what do you actually have?
That uncomfortable question is where this essay begins.
Strategy Is Change
A strategy that doesn’t change behaviour isn’t a strategy. It’s a description.
Most data strategies describe a future state — the target architecture, the capabilities we’ll build, the governance we’ll implement. They paint a picture of the destination. What they rarely do is name the change required to get there.
A CFO reads the document and thinks “sounds sensible” but has no idea what it means for their decisions. An engineer sees infrastructure projects, not behaviour change. A business leader nods along, then continues exactly as before.
If everyone agrees with your strategy but nobody does anything differently, what do you actually have?
And here’s the harder truth:
people rarely change behaviour unless their incentives change.
If the strategy doesn’t touch how people are measured and rewarded, they will default to old habits regardless of what the document says.
The Pain That Isn’t Felt
This points to a deeper problem. Data strategies often float above the organisation, solving problems that aren’t experienced as problems by the people who would need to change.
Is the data quality issue actually felt by the sales team, or only by the data team? Does the CFO experience the lack of integrated reporting as pain, or as a minor inconvenience they’ve learned to work around? Is the absence of governance a crisis, or just background noise?
If the pain isn’t felt, no document creates urgency.
You end up with a small group of believers trying to convince everyone else — and that’s an exhausting position to occupy.
The Boundary Tangle
There’s another question worth sitting with:
where does data strategy end and technology strategy begin?
Data touches everything. But so does technology. So does process. So does culture. When we carve out “data strategy” as a distinct domain, are we clarifying the problem or fragmenting it? If we spend our energy debating definitions, ownership, and boundaries, who is solving the underlying issues?
Perhaps the tangle that runs through the organisation cannot be neatly separated into lanes. Perhaps the attempt to do so is part of why progress feels so slow.
And when official strategy fails to address felt pain, the organisation routes around it. Spreadsheets proliferate. Shadow systems emerge. Business units solve their own problems in ways that make the data team’s job harder later — but at least they get solved.
What Would Change Look Like?
A behaviour-focused data strategy would look different. It would name specific decisions that should be made differently, and by whom. It would acknowledge that most people experience data through systems and tools — not through governance frameworks — and meet them there. It would be honest about which problems are actually felt and which are only visible to the data team.
Most importantly, it would answer the question that most strategies avoid: if this succeeds, what will people do tomorrow that they don’t do today?
The measure of a data strategy isn’t the elegance of the framework. It’s whether anyone does anything differently because of it.
What behaviour is your data strategy actually trying to change?
Read the first essay:
