There’s a quiet assumption growing in the sector – that faster, more automated data flows mean we need fewer people to manage them.
With the Office for Students bringing forward in-year student data returns, the timing of this assumption matters. We’re moving toward a world where data must be cleaner, sooner – and policymakers expect to act on it in near real time.
But here’s the problem: systems may be getting faster, but people are leaving.
Since the rollout of Data Futures, I’ve noticed something I can’t ignore. Skilled, experienced data professionals, the ones who could spot errors before they became headlines, the ones who knew the quirks of a course code or the edge cases of a fee regime, are quietly stepping away. Some are burned out. Some are disillusioned. Some simply saw the writing on the wall and decided their energy would be better spent elsewhere.
And what we’re left with, in too many places, is a beautifully constructed pipeline – but fewer people with the expertise to interpret what’s flowing through it.
Why understanding still matters
Automation can pull data from source systems. It can even flag when a field is missing or a validation rule is breached. But automation can’t ask the right questions. It can’t spot that a student was coded incorrectly because the underlying policy changed mid-cycle. It can’t explain to senior leadership why a trend line looks suspicious, or why that dashboard won’t help them make the decision they think it will.
That kind of understanding – contextual, relational, often institutional – is the thing we’re in danger of losing.
And it matters now more than ever. If OfS is planning to use in-year data to identify risks, monitor compliance, or inform funding decisions, then accuracy and nuance aren’t luxuries. They’re essential. Poor-quality data interpreted in haste is more dangerous than a slow return.
What’s at stake
This isn’t about resisting change. I’ve been part of this shift toward more timely, automated, joined-up data. It’s the right direction.
But we need to talk about what’s been lost along the way.
We need to recognise that data strategy is more than dashboards and data flows – it’s a people business. And people need time, training, space to think, and a sense that their role is valued.
If we keep designing systems that assume human oversight is optional, we’ll find ourselves solving the wrong problems with the wrong data – and no one left who feels confident enough to say so.
What we need now
If we want to make in-year returns work – really work – we need three things:
- Retention: keep the people who understand how the data gets made
- Recognition: treat data interpretation as a strategic skill, not just back-office admin
- Resourcing: ensure that automation supports humans, not replaces them
Because automation doesn’t replace understanding. It only makes it more visible when it’s missing.