Audience research
Part of Proposal software audience research without the guesswork
Who is the proposal software target audience and how do you size it?
A UK method for identifying proposal software buyers by role, sizing seat and account pools with official data, and testing segments through buyer interviews.
What to take away
- GOV.UK's Business population estimates 2024 put the UK's private sector business count at around 5.5 million. Most of those firms are small, so define your audience by deal shape rather than by industry label.
- For proposal software, buyers are job roles, not company types. Whoever owns a quote and its approvals is the person your product has to convince.
- Occupation data sizes the seat pool and business counts size the account pool. Check both before committing to a segment.
- Interview five matched buyers before you build a pitch. Drop the segment if quotes are rare.
Who actually buys?
Start with the job that hurts when a quote takes days to leave. In most B2B firms, this is a sales leader, a bid manager, or the finance reviewer who approves a discount.
Industry label matters less: a 30-person software company and a 300-person manufacturer can qualify. Test: do several people create quotes, and does someone else approve them?
Segments that pass that test include agencies and professional services firms with repeat quoting. Segments that fail include single-operator businesses that quote once a quarter.
If you are building a list from nothing, the Audience Research Guide for England shows how to turn job titles into a segment you can size and test.
Which sources size the audience?
Two official datasets do most of the work. The ONS Annual Population Survey (APS) dataset EMP04, Employment by occupation gives employment counts by occupation across the UK. Select the relevant occupation groups and geography, then sum the employed-person counts. Map buyer job titles to occupation groups carefully: roles such as bid manager or finance approver may not map neatly, and EMP04 is a proxy for a gross seat pool, not a count of likely buyers.
Two Pools, Never Multiply
Occupation pool
- Source
- ONS occupation data
- Counts
- Sales roles
- Ceiling
- Seat pool
- Filter
- Job titles
- Risk
- Proxy for buyers
Employer pool
- Source
- Business population
- Counts
- Employers
- Ceiling
- Account pool
- Filter
- Headcount
- Risk
- Undefendable if multiplied
Business population estimates answer the other half of the question by counting employers. Use its employment-size bands — no employees, 1–4, 5–9, 10–49, 50–249 and 250 or more — to filter the account pool before filtering by industry. Exclude bands that do not fit your team-software buyer, and keep the two pools separate, because multiplying them produces a market size that nobody can defend.
Use the same release to do the arithmetic: the account pool is the sum of business counts in your chosen industry and employment-size-band cells; the seat pool is the sum of the relevant occupation counts in EMP04. The around 5.5 million UK business count is the starting ceiling, not your segment estimate. Record the filters and report the two totals separately; do not apply a guessed percentage or multiply the pools.
Filtering by headcount removes the smallest employers, but it also removes the firms most likely to try a free tier. Decide early which end of that market you want.
How do you compare audience segments?
Put every candidate segment through the same five checks. A signal that looks weak on its own often combines into a clear yes or no.
| Segment signal | What to check | Why it changes the sale |
|---|
Five Segment Checks
- Sales headcounthow many touch a quote
- Quote volumedeals quoted each month
- Approval layerswho signs off price
- Current toolingCRM, spreadsheets or nothing
- Deal valuetypical contract size
Apply these five checks to every candidate segment:
- Quote volume:Strong when buyers describe regular quoting as part of the workflow; weak when quotes are rare or occasional; unknown when you have not checked recent examples.
- Multiple people and approvals:Strong when one person prepares quotes and another reviews or approves them; weak when one person handles the whole process; unknown when the approval route is unclear.
- Complexity and version control:Strong when revisions, pricing or approval steps create delays or confusion; weak when quotes are simple and standardised; unknown when you have not seen the process.
- Budget and authority:Strong when the buyer owns the budget or can name the approval route; weak when there is no budget or route to approval; unknown when this has not been discussed.
- Reachable account pool:Strong when your chosen industry and employee-size bands produce a countable pool of suitable businesses; weak when few accounts fit; unknown until you have filtered the business data.
Mark a check strong or weak only when interviews or records support it. Use unknown when evidence is missing, and turn each unknown into a question to test.
Drop any segment that scores badly on quote volume. The other signals only matter once quotes happen often enough to be a process rather than an event.
Score each signal as strong, weak or unknown, using evidence from buyer interviews, quote records and the filtered business data. Keep unknowns visible as questions to investigate rather than treating them as assumptions.
How do you test a segment before you invest?
Interviews are the cheapest test you have. Ask five people who match the definition to walk you through the last quote they sent, and time how long each step took. The Customer Interview Method for England keeps those conversations comparable, so you spot the pattern rather than the loudest voice.
Record the same fields every time: who wrote the quote, who approved it, and which tool held the version history.
Budget is the second test. Sales pay design shapes what a manager can approve. Salesforce's Trends in Sales Compensation Report covers how pay plans and quotas are set, which helps when the licence sits on a variable cost line.
A segment that survives five interviews has earned a place in your pitch. One that does not survive is cheaper to drop now than after a quarter of unrewarded pitching.
Common questions
How many people should be in a segment before you sell to it?
Enough to justify a licence for the buyer, not for you. A team quoting thirty deals a month with three approvers is a real segment even when the firm is small.
Should you size by business count or by seat count?
Both, but keep them apart. Business counts CAP how many accounts can buy, and occupation counts CAP how many people inside them might use the tool. Multiplying the two gives a figure nobody can check.
Do Scotland, Wales and Northern Ireland need separate research?
The occupation data and business population estimates cover the whole UK, but validate public-sector buying routes separately. Check Contracts Finder and the Find a Tender service, as well as Public Contracts Scotland, Sell2Wales and eTendersNI for relevant notices. The Procurement Act 2023 applies in England, Wales and Northern Ireland; Scotland has its own regime, including the Public Contracts (Scotland) Regulations 2015.
What if your buyers sit outside the UK?
Keep the method, but replace ONS EMP04 and Business population estimates 2024 with the target country's occupational-employment and business-population datasets. For UK context, see the England market outlook; do not carry UK counts across borders. The interview script travels unchanged.
