Funding statistics can help teams understand a research workload, but only when the measure, observation date and coverage are clear. This October edition explains how to read FundingLens data. It does not present the earlier illustrative figures as verified research or claim a measured total for all UK funding.
1. Check data availability before quoting a number
Start at the UK grant statistics hub and read its availability notice. If a current snapshot cannot be retrieved, there is no basis for assigning a new observation date to old values or treating an unavailable result as zero grants. A cached snapshot may still be informative, but cite its actual date and state the coverage limits.
A dataset generation timestamp records when a file or aggregate was produced. It does not establish that every source was successfully fetched, that a funder reviewed the record, or that a listed programme is still accepting applications. Those are separate questions requiring source-level evidence. The editorial review date of this article is also separate from any dataset observation date.
2. Define the unit of analysis
One scheme can have several rounds, regional delivery pages and application strands. Counting every page as an independent grant can overstate availability. A reproducible report should explain whether it counts schemes, application rounds, source records or delivery offers and how duplicates are resolved. Keep closed rounds and expected future windows separate from currently open applications.
Sector and applicant categories may overlap: an innovation competition can serve both a university and an SME, and an energy project can also be manufacturing investment. Do not add overlapping categories to obtain a supposed national total. Describe whether categories are exclusive or multi-label, and retain an unspecified category where the source does not support classification.
3. Interpret money and deadlines carefully
A programme budget is not the same as the maximum award to one applicant, and neither necessarily describes the uncommitted money remaining. Summing maximum awards across records produces a different measure from summing disclosed scheme budgets. Any monetary headline must identify the field used, treatment of missing values, currency and duplicate rounds.
Days until a deadline measure the time remaining at observation. They do not measure the interval between an announcement and closing, or the time an applicant needs to prepare. Rolling programmes and missing or ambiguous deadlines require their own treatment. Publish a median only after describing the population and exclusions; it must not become an unsupported promise that applicants usually have a particular preparation period.
4. A checklist for a publishable funding bulletin
- Record the snapshot identifier, generation time, underlying source dates and coverage.
- State inclusion, geography, open-status and deduplication rules.
- Document each metric, denominator, missing-value treatment and calculation.
- Manually check a sample against primary programme pages and explain errors or exclusions.
- Retain the evidence used so a later source change does not make the published method irreproducible.
Use the public directory for discovery and confirm individual application rules with the funder. Before preparing a data bulletin for journalists or trade bodies, establish a reproducible observation and review its wording. No national ranking, rejection rate or historic trend is inferred from this methodology article.