Measurement
Part of Measuring startup funding across five layers, from cash need to outcomes
Startup funding measurement mistakes in England: checks for published figures
Catch startup funding measurement mistakes before you cite a figure in England: births, EIS and SEIS, regional totals, denominators and revisions.
A published funding figure is only as good as its population, date and method. Funding reports can look precise while mixing populations, dates and stages. Before a founder, adviser or programme team cites one, they should check what it counts, who is in it and whether it has been revised.
What to take away
- Business births are not funding demand, and EIS or SEIS statistics are not all equity.
- State which location a regional source uses before comparing places.
- Add deal count and distribution to any total, and a denominator to any rate.
- Keep proposals, signed documents and cleared cash apart when a provider reports results.
- Trace one headline number back to its source before you cite it.
Business births are not funding demand
Business births measure new active enterprises, not people looking for money. The ONS business demography release calculates the birth rate as the number of births as a proportion of the active businesses. It does not count funding applications.
For demand, look at evidence-based demand signals and treat births as background on the size of the business base. Funding that a company actually obtained is a different measure again, so keep the two apart in any chart.
Label the scheme and the place
A figure needs its population and its location basis stated beside it. The table shows what to ask of each source named in the funding data.
| Source | Population to label | Question to ask |
|---|---|---|
| ONS business demography | Active-enterprise births under its methodology | Does it count funding? It does not |
| HMRC scheme statistics | Defined tax schemes, EIS and SEIS | Which scheme, and are recent periods revised? |
| British Business Bank Equity Tracker | Regional deal totals | Which location does it use, and how concentrated is the total? |
| Any regional series | Registered office, operating site, founder residence or investor address | Which one places the deal? |
HMRC's scheme statistics cover defined tax schemes and can revise recent periods. EIS and SEIS are tax schemes, so do not treat their figures as all equity. Registered office, operating site, founder residence and investor address can place a deal differently, so state which one the source uses. For more context see the market size sources.
Totals need counts and distribution
A few large rounds can move a regional total. Add deal count, stage and distribution where available. The British Business Bank Equity Tracker announcement makes this caution explicit in its regional discussion.
An illustrative example: a region reports £120m across 30 deals. One £50m round means the other 29 average about £2.4m. A reader who sees only the total would assume a typical deal many times that size.
Rates need a denominator
Invited, eligible, submitted, assessed and completed cases are different populations. Put the denominator beside every rate, and do not change it between reports without saying so. Comparisons only hold across comparable groups, as funding benchmarks show.
Provider results: gross, net and cash
Provider-reported results often skip three distinctions. Gross is not net, a term sheet is not cash and a mixed count hides the route behind it. Ask which of the three the number reflects.
Show deductions, fees and timing. A larger headline round may provide less immediately usable money. Illustrative example: a £5m headline round with £400,000 fees and £1m staged over 12 months leaves £3.6m available at close.
Track proposal, signed document, conditions and cleared receipt separately. Separate conversations, proposals, signed contracts and received cash. Keep uncertain money outside reconciled cash.
Term Sheet to Cleared Cash
- Proposal
- Signed document
- Conditions met
- Cleared receipt
- Uncertain money kept outside reconciled cash
Grants, debt and equity have different applications, economics and obligations. Do not add them into one conversion rate without a justified common definition. Illustrative before and after: report 30 grants, 20 debt deals and 10 equity rounds separately; a combined 60 funded count hides the mix.
Survivors and causes
Reports that keep only successful founders create survivorship bias. Include withdrawals, rejections, partial funding and deliberate decisions not to raise, or the published success rate describes only those who stayed in the sample.
A funded company that grew did not necessarily grow because of the funding. Compare a credible unfunded case and other influences. The Regulatory Policy Committee's guidance on monitoring and evaluation addresses the need to consider external factors that might impact the evaluation. Methods are covered in attributing outcomes to funding.
Trace one number before you cite it
Run a trace test on one headline number. The test takes four steps and catches most of the mistakes above.
- Start at the published chart and follow its formula to the source records.
- Confirm the denominator, the date and the latest revision.
- Ask a reviewer to describe what the measure does and does not prove.
- If their reading differs from the intended claim, repair the label or the analysis.
Archive previous definitions and values. Explain source updates, corrections and method changes rather than smoothing the series. Keep a correction table with the old value, new value, reason and affected decisions. A metric dictionary with purpose, population, formula, source, period, geography, owner and limitation makes the trace faster.
Do not delete an inconvenient failed target or unsuccessful company after the reporting period closes. Review incentives as well: a team rewarded for applications submitted may push weak cases, so pair volume with fit, quality and later outcome measures.
Before you cite a figure
- Name the population the source counts.
- State which regional method the source uses.
- Add deal count and distribution beside any total.
- Put the denominator beside every rate.
- Separate proposal, signed document, conditions and cleared receipt.
- Run a trace test on the headline number.
Common questions
Do business births show how much funding founders want?
No. Business births measure new active enterprises under the ONS methodology. They do not count funding applications.
Why is a regional deal total not enough?
A few large rounds can move a regional total. Add deal count, stage and distribution where available.
How should revisions in funding data be handled?
Archive previous definitions and values, explain source updates and method changes, recalculate affected series, and keep a correction table.
