Audience research
Part of Audience research for a lead network that separates both sides of the marketplace
Lead network buyer personas with an evidence tag instead of a biography
Build evidence-tagged buyer persona hypotheses for an England lead network without inventing people, demand, motivations or research findings.
Lead generation networks buyer personas should remain labelled hypotheses until fieldwork supports them. A useful persona records a decision, its setting and the evidence still missing. A stock photograph, invented name and confident list of motivations merely conceal assumptions.
No buyer was interviewed or surveyed for this article. The templates below were prepared on 5 September 2026 for future research in England. They are not descriptions of market segments or demand.
What to take away
- Treat every buyer persona as a labelled hypothesis until fieldwork evidence supports it.
- Record the decision, its setting and the missing evidence instead of inventing a biography.
- Use ONS population data to specify the buyer type, not to explain how enquiries are bought.
- Recruit recent users, refusers and abandoners, and record recruitment source and missing groups.
- Version each persona with research dates, sample, exclusions and contradictory evidence, and retire it when contradicted.
Replace biography with a decision record
Use a working identifier such as "local owner deciding on a paid introduction" rather than a fictional person. Then complete these fields:
Evidence-tagged decision record fields
- Decision and who has authority
- Most recent comparable event and date
- Information checked before paying
- Operational constraintslocation, response capacity
- Concern causing refusal or dispute
- Supporting record for each statement
- Unresolved assumption and next test
- decision to be made and who has authority;
- most recent comparable event and its date;
- information checked before accepting or paying;
- operational constraints, including location and response capacity;
- concern that would cause refusal or a dispute;
- supporting record for each statement;
- unresolved assumption and next test.
Tag every line H for hypothesis, I for interview, S for survey, or O for operational record. At first, every behavioural statement is H. Change the tag only when a retrievable record supports it.
Use population data as context only
The ONS 2025 business dataset distinguishes enterprises from local units and provides regional, industry and employment-size tables. That helps a researcher specify whether the prospective buyer is a legal enterprise or an operating site. It does not reveal how that buyer purchases enquiries.
A sole trader and an employee at a corporate body may also face different communication rules. The ICO's business-to-business marketing guidance distinguishes corporate subscribers from individual subscribers, including sole traders and some partnerships. Keep this regulatory classification outside the persona's presumed attitude. It is a record-level and channel decision, not a personality trait.
Create contrast without claiming segments
Three hypothesis cards can expose different product questions:
| Working card | Decision under study | Evidence needed before validation |
|---|---|---|
| Occasional local buyer | Whether one revealed contact is worth reviewing | Recent purchase sequence, acceptance criteria and refusal reasons |
| Multi-site coordinator | Whether requests can be routed to the right location | Authority map, allocation workflow and branch-level exceptions |
| Capacity-constrained specialist | Whether timing and scope justify a response | Availability rules, request detail and missed-opportunity record |
These are research prompts, not findings. Do not assign market share, willingness to pay, preferred channels or conversion expectations. Replace a card if real decisions do not match it.
Recruit against the contrast
The GOV.UK Service Manual advises researchers to begin with the problem, assumptions and information needed for a decision in its research-round planning guide. Its guidance concerns government-service teams, but the planning discipline transfers: recruit actual or likely participants and accommodate access needs.
For each card, seek recent users, refusers and people who abandoned the process. Record recruitment source and missing groups. Founder contacts may help uncover vocabulary, but they cannot validate an England-wide segment.
Collect only what the project needs. ICO guidance defines data minimisation as keeping information adequate, relevant and limited to the stated purpose. A persona document rarely needs a participant's name, exact address or raw contact details.
Publish the evidence boundary
The current MRS Code of Conduct page points to the 2023 edition and notes its requirements on participant wellbeing and representative-sample reporting. Teams should recheck the live standard before fieldwork.
Version each persona with its research dates, sample, exclusions, contradictory evidence and owner. Do not call it representative unless the sampling design supports that statement. A card can guide a pilot when it predicts a specific decision and names its uncertainty. Retire it when it becomes a decorative biography or when later evidence contradicts its core premise.
Before you act
- Tag each behavioural line H, I, S or O.
- Change a tag only when a retrievable record supports it.
- Keep regulatory classification outside the persona's presumed attitude.
- Collect only information the project actually needs.
- Recheck the live MRS standard before fieldwork.
- Do not call a persona representative without sampling support.
Common questions
What should replace a fictional name and stock photograph in a persona?
Use a working identifier such as local owner deciding on a paid introduction. Then record the decision and who has authority, the most recent comparable event and its date, information checked before paying, operational constraints, refusal concerns, supporting records and unresolved assumptions with the next test.
How should population data and regulatory rules be used in a persona?
The ONS business dataset distinguishes enterprises from local units and gives regional, industry and employment-size tables, helping specify whether the buyer is a legal enterprise or an operating site. ICO guidance separates corporate from individual subscribers. Both are record-level and channel decisions, not personality traits.
When should a persona card be retired?
Retire a card when it becomes a decorative biography or when later evidence contradicts its core premise. Replace a card if real decisions do not match it. Version each persona with research dates, sample, exclusions, contradictory evidence and an owner, and avoid claiming representativeness without sampling support.
