Foundations
Part of Women founder funding foundations: reading the evidence
Women founder funding market size in England: a method
Estimate the women founder funding market size in England in five steps: a stated proxy, named data sources, separate money units and a double-count check.
No single current number exists for the women founder funding market in England, so an estimate has to be built from parts. Each part needs its own source, its own proxy and its own unit. The build runs in a fixed order, from the first definition to the final test.
What to take away
- Build four separate measures, population, demand, supply and unmet need, and never merge them into one market.
- Pick a proxy for woman founder before opening any source, and state it.
- No named source identifies women founders by itself, so every count rests on a proxy.
- Keep loans, equity and grants in separate columns.
- Publish a range and an evidence map rather than a fabricated total.
Choose the proxy first
Every source classifies leadership differently. A dataset may use directors, ownership, a senior decision maker, applicant self-identification or founding team composition, and it may use a binary gender field. The proxy you choose decides who is counted and who is left out.
Definitions change who counts
Directors
- Who is counted
- Board members
- Risk of exclusion
- Non-director founders
- Data source
- Company filings
Ownership
- Who is counted
- Shareholders
- Risk of exclusion
- Minority owners
- Data source
- Cap table
Self-identification
- Who is counted
- Applicant states
- Risk of exclusion
- Non-binary founders
- Data source
- Application form
Write the rule down in one sentence and keep it fixed across all four measures. The definition of the founder group is the first thing a reader will test.
Match each measure to a source
The sources that report below UK level are few, and none of them identifies women founders by itself. Each needs a stated proxy before it can be counted.
Four measures, never one total
- Population: addressable women-led businesses
- Demand: applications and amounts requested
- Supply: approvals, debt, equity, grants
- Unmet need: rejections, gaps, discouragement
| Source | Input it provides | What it cannot do |
|---|---|---|
| ONS, UK business: activity, size and location | Business counts by size and location for the population measure | Identify women founders |
| Companies House, company records | Registered address, current and resigned officers for a denominator | Show where the business actually operates |
| British Business Bank, small business finance data | Finance data with regional breakdowns for the supply measure | Identify women founders |
| UK Business Angels Association | Angel activity for the supply measure | Identify women founders |
| Beauhurst | Named equity deals by region | Identify women founders |
The UK business activity data from ONS provide regional context, but they do not identify women founders by themselves.
UK figures that are not an England market size
UK headline figures cannot stand in for an England total. The government's small business access to finance call for evidence puts women-led businesses at 15% of the UK business population, and women-owned and women-led businesses together contribute £221 billion in UK gross value added.
These are national contextual figures from cited underlying sources. They are not an England funding-market valuation, and should not be multiplied by an average funding amount. HM Treasury says the Investing in Women Code has grown from 12 founding signatories in 2019 to more than 300, but it covers signatories, not all funding in England.
Build the estimate in five steps
- Populationapply your proxy to a current business count from the ONS dataset, or to Companies House officers for a director-based rule.
- Demandrecord businesses considering finance, applications and amounts requested, from provider data or your own survey.
- Supplyrecord approvals, invested value, debt advanced and grants awarded by route, from British Business Bank data, angel records and Beauhurst deals.
- Unmet needmeasure discouragement, rejection, partial offers and funding gaps through surveys and interviews.
- Join the results by English region where sample size permits, and show UK-only rows plainly.
Applications show only expressed demand. So signals of funding demand that include discouragement belong in the build from the start, not as an afterthought to step four.
Measure unmet need with care
Unmet need is the hardest measure, because nobody files a record of a loan never sought. Discouraged founders appear only in surveys and interviews, so plan audience research to reach them. Report rejections and partial offers from provider data separately, as they describe a different moment in the journey.
Label the result as an estimate of a gap, not a count. A gap built from a survey carries the limits of its sample, and the write-up should say so beside the number. Never subtract supply from demand and call the remainder the market.
Use Companies House for the denominator
Companies House shows current and resigned officers and a company's registered address, so it can anchor a director-based proxy. It does not say where the work is done.
London registrations can obscure activity elsewhere in England. Compare registered-office geography with operating location where lawful data permit, and note which one each regional row uses.
Keep the money units apart
Debt is repaid, shares transfer ownership, and grants fund defined activity. Adding loan value to equity investment gives a number that is not one comparable market, so keep a column for each.
State whether each value records commitments, approvals, drawdowns or completed investments, and whether it is gross or net of fees. A reader cannot compare two totals that count different stages.
Remove double counts
Check for double counting across repeat applications, co-founders, subsidiaries and several providers reporting the same syndicate round. Write the deduplication rule beside each count.
Preserve zeroes separately from missing demographic fields. A blank record does not mean the business is men-led, so show missing gender data plainly instead of dropping those rows.
Test the estimate before release
Recalculate the whole build under a second plausible proxy. If the conclusion changes materially, show both results and explain the choice rather than selecting the larger market.
Then set the figures beside benchmarks with matching definitions and report them through a reporting dashboard.
Where this build fits
The result is an evidence map, not a valuation. For the wider route picture and the reading of the evidence, use the foundations guide, and keep the build as the method behind any figure it quotes.
Before you act
- Write the proxy for woman founder in one sentence.
- Assign each measure to a named source.
- Keep loans, equity and grants in separate columns.
- State which stage each amount records.
- Write a deduplication rule beside each count.
- Recalculate under a second proxy.
Common questions
Which source identifies women founders in England?
None of the named sources identifies women founders by itself. Each needs a stated proxy definition before it can be counted, such as directors, ownership or applicant self-identification.
Why keep loans and equity apart?
Debt is repaid, shares transfer ownership, and grants fund defined activity. Adding loan value to equity investment and calling it one market compares unlike things.
What should be published while no consistent England dataset exists?
A range and an evidence map. Report population, applications, approval rate, amount requested, amount supplied and median transaction separately, and include missing gender data.
