Yes, Of Course You Need a Data Strategy
When one of my junior colleagues is struggling to prioritise their workload with a client, I usually ask them what the strategy is. I generally get one of two answers:
- "What strategy?"
- "Well, I think I’ve worked it out, but I’m not allowed to see the official one."
Hiding a strategy from your team is like hiding the architectural plans from your builders because you’re worried they might steal the design: nonsense. Strategies are often born of the best intentions, only to be abandoned the moment execution starts. But a strategy is just a plan. It keeps everyone on the same page so you aren't all rowing in different directions. So why are they so often AWOL?
The longest data strategy I’ve ever seen was nearly 100 slides. This thing raged in words but didn’t range in depth. It was a cocktail of broad business fluff followed by 15 slides of suffocating micromanagement... about 6 times over.
I dread to think how many billable hours were incinerated on something that was only read end-to-end by three people. We fall into this trap because strategy feels vague, and we think the only cure for ambiguity is more words. It isn’t.
A data strategy can fit on one page. It only needs to answer three questions:
- What are your priorities?
- What are your constraints?
- What are your key issues? (Usually the mess that happens when the first two hit reality).
Plain, honest answers to these questions allow people across the organisation to actually make decisions without checking every single thing with you first.
What are your priorities?
If everything is a priority, nothing is. Making a choice feels risky—what if you’re wrong? What if the market shifts?
Shit happens. You might have to adjust later, and that’s fine. But your priorities should be steady enough to survive a bit of economic wind. Stop using buzzwords and say what you actually care about:
- Sharing stuff: Making sure teams can actually use each other's data without a six-week lead time.
- Tight grip: Ensuring the pipeline is secure and observable, where only specific people can pull the heavy levers.
- Getting our money’s worth: Making the most of the expensive tech stack you’ve already bought.
- Independence: Avoiding being locked into a single vendor's ecosystem for the next decade.
- Experimental: Giving teams the freedom to break things and iterate without a committee meeting.
What are your constraints?
Most organisations hate saying "we’re tight on cash," so they say "we’re being efficient." Efficiency is a priority; a constraint is a hard wall you can’t jump over.
Constraints are the choices or realities that move slowly. It might be a regulatory requirement like SOX compliance, but more often, it’s about your actual capacity.
- Brainpower: The time it takes for staff to learn a new tool.
- Security: Some data strictly cannot be shared, no matter how "agile" you want to be.
- The "As-Is": The limitations of your current software or the specific skills of the people you’ve already hired.
- Market scarcity: The fact that there are only ten people in the country who know how to fix your legacy system, and they all want £1,500 a day.
- Budget, of time, or money: no one has infinite amounts of either.
What are your key issues?
Your key issues are the friction points. They are the obstacles preventing your priorities from becoming reality, and often they are a result of your priorities colliding with your constraints.
For the folks working in infrastructure or data, identifying these is the only way to move the needle. Examples include:
- Data Literacy: You want an experimental "skunkworks" vibe, but half the department is terrified of Excel. You can't fix that with a new tool; you fix it with training.
- Culture: The organisation is allergic to change. This requires communication and stakeholder management, not better code.
- The Locked Box: You’re stuck with a vendor that’s putting up their rates or making decisions you're not a fan of, and getting out requires a massive investment in new tech and retraining. This may be particularly prevalent for European organisations focusing on sovereignty at the moment.
Answering these three questions gives you a strategy your organisation can actually live by, and it will be useful! The jumping off point for a good strategy is enabling people to make decisions, which is definitely easier to do with a blunt, one-page truth than a hundred-page, bush-beating exercise gathering dust in a SharePoint folder.
