Funding rules monitored: 31 official sources · awaiting first verified checkFunding Intelligence →
Skip to the Academy content

Topic

The data quality routine that prevents most problems

A small number of checks, run every month without exception, prevent the majority of funding, performance and audit problems in further education.

Funding year2026–27Awaiting verificationReviewed18 September 2026See the sourcesReport an issue

Enough detail to do the job and hold a sensible conversation with MIS or Finance.

How to read this page
  • Official requirement: Stated by the official funding document named in the sources for this page.
  • FEFunding explanation: Our plain-English explanation. Helpful, but the official document is the rule.
  • Worked example: An illustration of how the rule applies in one situation. Not a universal rule.
  • Recommended good practice: Operational advice from FEFunding. Not itself a funding requirement.

The monthly set

Five checks, every month, in the same order.

The value is in doing the same five every month and tracking the count for each. A count that rises is a control failure, not a workload problem, and it tells you where to look.

Each of these checks covers something validation cannot see, which is why they matter more than the validation report.

FEFunding explanation

Passing validation means the data is internally consistent and structurally acceptable. It does not mean the data is true, that the learner is eligible, or that the funding claimed is correct.

Build each check as a saved report from your own data rather than relying on what is returned to you. You want to find these before anyone else does.

Record the count per period on a single trend sheet. Five numbers, twelve months, is a more informative data-quality report than most colleges produce in fifty pages.

The termly set

Three checks each term, which need more time than a month allows.

These are the checks that find the audit findings before the auditor does. They are sampling exercises rather than exhaustive ones, and a failure in a sample is a process finding rather than a record finding.

Report them as a percentage with a named cause, not as a list of learners to fix.

Official requirement

Adult Skills Fund eligibility depends on the learner's age, residency and circumstances, and the provider must hold evidence of eligibility before funding is claimed.

Official requirement

Learning support funding requires a documented chain: an assessed need, the support actually provided, the cost of providing it, and the claim made. A claim without the chain is not defensible.

Sample sizes should be proportionate but consistent so the trend means something. Test retrievability, not existence: evidence that cannot be found within a few minutes will not be found at audit either.

The annual set

Three things to do once a year, at the right point.

Each of these has a deadline built into it. Done late, none of them is useful: the specification review after the first return, the hours reconciliation after December, or the results reconciliation after the close all arrive too late to change anything.

Diary them against the return calendar rather than the calendar year, and give each one a named owner with the time allowed to do it properly.

Worked examples

Worked example

What the trend sheet shows

A college tracks its five monthly counts across a year.

  1. Months 1 to 3. In-learning-past-planned-end rises from 40 to 180. Nobody notices because each month looks small.
  2. Month 4. The trend is reviewed. The rise localises to two curriculum areas.
  3. The cause. Both areas lost the staff member who submitted leaver notifications.
  4. The fix. A process change, not a data clean-up. The count falls back within two months.

Five numbers found a process failure that a list of 180 learners would only have described.

Illustrative figures.

Common pitfalls

Measuring activity instead of control

What goes wrong: Reporting counts errors cleared rather than causes resolved.

Consequence: The team looks busy and the same errors recur every period.

Prevention: Track the count of open exceptions per check, not the number fixed.

What this means for your role

Head or director of MIS

A one-page trend sheet is a better data-quality report than a long narrative.

  • Publish the five counts monthly, by curriculum area, and take the rises to the owners.

See it play out

Knowledge check

Knowledge check

The data quality routine

2 questions. Nothing is recorded unless you are signed in, and there is no time limit.

1. Which measure tells you most about whether data quality is under control?
2. The same records appear on a monitoring report every period. What does that mean?
0 of 2 answered

Sources

These are the official documents this page rests on. Where a figure, a deadline or an exact rule matters, the document is the authority and this page is the explanation.

Funding allocations and data returns information (opens in a new tab)

GOV.UKOfficial funding documentAwaiting first verification

ILR: sources of data (opens in a new tab)

GOV.UKOfficial funding documentAwaiting first verification

Adult Skills Fund funding and performance management rules 2026 to 2027 (opens in a new tab)

GOV.UKOfficial funding document2026–27Awaiting first verification

Advice: funding regulations for post-16 provision (opens in a new tab)

GOV.UKOfficial funding documentAwaiting first verification

Apprenticeship funding rules and assessment plan guidance 2026 to 2027 (opens in a new tab)

GOV.UKOfficial funding document2026–27Awaiting first verification

ILR validation rules 2026 to 2027 (opens in a new tab)

Department for EducationOfficial technical document2026–27Awaiting first verification

Content reused from GOV.UK is Crown copyright, used under the Open Government Licence. FEFunding is not endorsed by the Department for Education.