“Fit for purpose” is one of those phrases that gets thrown around a lot in data conversations. It sounds sensible. It feels right. But when you stop and ask, fit for whose purpose, and how do we know it’s fit?, the answers can get surprisingly vague.
In higher education, we’re often good at making data valid. We tick the boxes. We pass the HESA quality rules. We submit on time. But being valid is not the same as being useful, and that’s where fitness comes in.
Regulation Isn’t the Whole Picture
A lot of data work in HE is shaped by external demands: funding bodies, regulators, league tables. These purposes dominate how we think about data quality. But when you start to use the same data for internal planning, student support, AI tools or dashboards for senior staff, you quickly realise that ticking a box doesn’t mean the data is fit to answer a real question.
Take the example of course delivery information. It might be coded correctly for returns, but if academic staff aren’t aware that this data also feeds into timetable planning or student communications, the details may not be reliable enough for those uses. That disconnect between collection and use is where fitness breaks down.
Data as an Asset
Part of the problem is cultural. We often treat data as admin, not as an asset. Something to get through, rather than something to invest in. But data is a long-term asset, especially in a sector where strategic decisions increasingly rely on historic trends, projections, and analysis.
When we see data as an asset, we start to ask better questions:
- Can someone trust this?
- Will it make sense out of context?
- Is it structured in a way that works beyond its original use?
- Could we automate anything without introducing risk?
Fitness isn’t just about what data looks like today. It’s about preparing it for reuse, reshaping it over time, and making sure it keeps its value.
A Working Definition
So what does “fit for purpose” mean in practice?
It means that data is usable and trustworthy in the context of its intended use. It doesn’t have to be perfect. It just has to be good enough for what you’re trying to do – and that means different things in different contexts.
It also means future-proofing. Whether you’re looking to implement predictive analytics or roll out AI-powered tools, your data needs to be structured, consistent and clear enough to support those ambitions. If you don’t lay the groundwork now, you’ll hit a wall later.
What to Do Next
Start by asking: what are the key uses of your data beyond regulation? What decisions rely on it? What would go wrong if it were wrong?
Then talk to colleagues. Do the people entering data know how it’s used? Are the people analysing it aware of its limits? Often, small gaps in understanding create the biggest risks.
In the next post, we’ll look at the practical side – how raw data becomes usable, and what you can do to improve it without major investment.