It’s easy to forget that the dashboard is only the end of the journey. If you want meaningful insight, you have to start much earlier – with the raw data.
Refining data doesn’t mean building the perfect model or learning a new tool. Often, it’s about getting a solid grasp of what your data actually represents, making sure it’s structured clearly, and doing the legwork before it gets near a visualisation.
Raw Doesn’t Mean Useless – But It Does Need Work
I recently worked with the OfS individualised datasets, which contain a wealth of information but aren’t immediately friendly. The values are often coded, calculations are left to the user, and the structure doesn’t lend itself to easy exploration.
But with a bit of groundwork – using Power Query in my case – I was able to:
- Do the required calculations once, consistently
- Convert numeric fields to readable text
- Create a dataset that made sense to others, not just to me
The result? A complex dashboard that was still genuinely usable. Colleagues could explore access and participation metrics with confidence, because the background work was already done.
Fixing Problems Early Saves Time Later
This is the key point. If you put the effort in up front, you only have to do it once. If you try to fix problems later, in each dashboard, each analysis, each time someone asks for a new view, you’ll find yourself duplicating work and creating inconsistencies.
It’s a simple principle: define, clean and structure your data early, and everything else gets easier.
Some Practical Reminders
You don’t need to be a data engineer to make big improvements. Start with these kinds of questions:
- Are your columns clearly named and consistently formatted?
- Do your codes and categories mean something to the people using them?
- Are calculations baked into your source data, or repeated in every report?
- Can someone else pick up your dataset and understand what it shows?
The tools you use matter less than the habits you build. Whether it’s Excel, Power BI, or a data warehouse, the goal is the same, make the data meaningful, and make it make sense early on.
Fit for Purpose Means Fit From the Start
Data doesn’t become fit for purpose by accident. It takes decisions. It takes care. But it doesn’t have to take forever – and it pays off quickly.
In the next post, we’ll look at how to measure progress. What does good look like? How do you know if you’re improving data fitness? And how can you prove it to others?