For decades, the HESA student return was an annual ritual: a once-a-year, look-back exercise, assembled over weeks from the student record system and a dozen local sources, reformatted by hand, and submitted with a collective exhale. That world is gone. Under HESA Data Futures, the sector has moved to in-year data collection across multiple reference periods - and at VE3 we'd argue this is the single most underestimated change in university data of the last decade.

Here's why. Most institutions treated the move as a reporting-format change: same process, more often. It isn't. In-year HESA reporting is a data-foundation change wearing a compliance costume and it has quietly exposed which universities have a genuine data platform underneath them and which have been holding the annual return together with spreadsheets and a few heroic individuals. The ones in the second category are now discovering that what you can just about survive once a year becomes untenable when you have to do it continuously, to a tighter standard.

And there's an upside hiding in the pain - because the very foundation that makes in-year reporting survivable is the one that finally makes credible forecasting possible. That matters more than it has in years.

What actually changed

The shift is more fundamental than "more frequent returns." The annual single look-back has been replaced by in-year collection against a new, expanded data model and platform, with data expected at multiple points across the year. Just as importantly, the centre of gravity for data quality has moved: assurance is being pushed earlier, "to source," rather than relied upon as a post-collection clean-up after the fact. (HESA is now part of Jisc, and the direction of travel - quality at source, in-year cadence - is only sharpening.)

That last point is the one institution feel most acutely. When you returned data once a year, you could afford a frantic quality-assurance scramble at the end. When you're returning it in-year, that scramble has nowhere to live. Quality has to be built in upstream, or you simply can't keep up.

Why the old way can't keep up

We've seen the legacy pattern across the sector, and it's remarkably consistent: data pulled from the student record and a scatter of local systems, reshaped in spreadsheets or bespoke scripts, reconciled by a small number of people who understand the dark corners, and assembled into a return shortly before the deadline. At one large UK university, a single year's return ran to millions of lines of data across tens of thousands of students - and published sector case studies describe exactly the manual, reformatting-heavy effort that involves.

Do that once a year and it's painful but possible. Do it several times a year, to a standard that demands quality at source, and three cracks open at once: the manual effort no longer fits in the calendar; the key-person dependency becomes an operational risk rather than an inconvenience; and there's no time to fix data quality retrospectively, because the next reference period is already arriving. This is not a process you optimise your way out of. It's a foundation you have to replace.

In-year reporting is really a data-quality-at-source problem

Strip away the HESA-specific detail and the underlying requirement is simple to state and hard to deliver: a single, trustworthy, conformed view of student data that is correct continuously, not correct once a year after a lot of effort. That means conformance and validation moving upstream into the data platform - cleansing, standardising and quality-checking data as it lands, so that any reference period can be assembled from data that is already trustworthy rather than data that needs rescuing. In a modern medallion architecture, that's exactly what the curated layers are for. The return stops being a project and becomes a by-product of a well-run platform.

The upside: from looking back to looking ahead

Now the part strategic planning teams should care about most. The same conformed, current, trusted foundation that fixes in-year reporting is precisely what credible forecasting has always needed - and the timing could hardly be sharper.

The recruitment market is volatile. Overall student numbers have softened, international postgraduate-taught numbers have fallen markedly, study-visa applications have been down year on year, and a prospective international fee levy adds further uncertainty to income planning. In that environment, forecasting student numbers, income and retention off an annual, backward-looking dataset is planning with the rear-view mirror. What planning teams need is the ability to model on current data - to see this cycle's intake taking shape, to flag retention risk while there's still time to act, and to run scenarios as the external picture shifts.

You cannot build that on a legacy warehouse that refreshes once and conforms nothing. You can build it on the foundation in-year reporting forces you to create anyway. That's the opportunity: the compliance driver pays for the platform; the platform unlocks the forecasting.

What the foundation actually needs

In our experience, the institutions getting this right have put five things in place:

• One conformed source of truth. Student-record and local-system data integrated and conformed once, so "enrolled", "FTE" and "active" mean the same thing everywhere - no more reformatting per return.

• Data quality at source. Validation and cleansing applied as data lands, so every reference period draws on data that's already trustworthy.

• In-year, near-real-time refresh. A platform that reflects the current state, not last year's snapshot.

• Governed self-service for planning. Certified metrics and semantic models that let strategic planning and faculties answer their own questions, without five versions of the same headcount.

• A forecasting capability on top. Predictive models for student numbers, income and retention, built on the trusted layer - turning the same data from a compliance artefact into a planning asset.

The bottom line

In-year reporting is not a heavier version of the annual return. It's the sector's forcing function for a proper data foundation - and the institutions that recognise that are the ones turning a compliance burden into a strategic capability. Get the foundation right and two things happen at once: the return stops being an annual crisis and becomes routine, and your planning team stops looking backwards and starts looking ahead, on data they can actually trust.

The question we'd put to any university leader is this: when the next reference period lands, are you assembling it from a foundation that's already trustworthy - or rescuing it, again, by hand? Your answer is also your answer on whether you're ready to forecast.

Facing in-year reporting on a legacy foundation? Talk to VE3 about modernising university data so returns become routine and forecasting becomes possible.