Audience research
Reputation management audience research: separate the populations
Plan reputation management audience research in England: separate the populations, build a sampling frame, handle consent, then judge review software.
What to take away
- Begin with one written decision and the operating unit that owns it, with England as the stated scope.
- Separate the populations before recruiting, so one group's experience cannot stand for another's.
- Keep a sampling frame with owner, extraction date, eligibility rule and missing groups.
- Budget illustrative targets of 8 to 12 interviews per population and 60 to 120 survey responses.
- Judge review platforms and reputation-management software against the workflow problem the research finds.
What this guide covers and leaves out
This is a method guide for planning reputation management audience research in England ahead of a reputation software decision. It sets out the populations, the sampling frame and the consent record against which any shortlist of review or reputation-management platforms has to be judged.
The worked case is fictional: a home-services marketplace in England deciding whether to replace a spreadsheet and shared inbox that record authorised reviews, verify status, route responses and keep corrections. No real company or participant was studied.
Where reputation-management software fits
The research exists to support a decision about tooling. The fictional marketplace is weighing its spreadsheet-and-inbox setup against software: review platforms such as Trustpilot, Google Business Profile, Reviews.io and Feefo for collecting and answering authorised reviews, and reputation-management platforms such as Yext that extend routing, verification and reporting across several review sites.
Fieldwork runs on familiar tools as well: Smart Survey or Qualtrics for questionnaires, SurveyMonkey for short forms, UserTesting for task observation and Prolific for panel recruitment. The plan names them so that the frame, the consent record and the field dates can be judged together.
Judge any shortlist on the workflow the research identifies: coverage of the review sites that matter, a named owner for each response route, an auditable correction trail, verification status, reporting against stated denominators, retention and data residency. No platform wins on all of them.
Separation of populations before recruiting
The word audience hides groups with different knowledge, authority and risk. Mixing them produces testimony that cannot be compared.
Define an eligibility rule for each group before building any contact list. An operator might need recent, authorised experience of routing a review case. A customer might need to have used the correction route inside a stated period.
Do not infer England residence from a postcode fragment or writing style. Record how scope is set, usually the operating unit that handled the case, and label UK-wide findings as UK-wide.
Audience populations
Group
- Affected people
- Used review or correction route recently
- Decision-makers
- Can approve budget, risk, workflow
- Operational users
- Recent authorised England case routing
- Assurance reviewers
- Examine privacy, consumer, security, accessibility
- Data subjects
- Identifiable in notes, recordings, screenshots
- External observers
- Saw public response, no direct case experience
Eligibility rule
- Affected people
- Decision-makers
- Operational users
- Assurance reviewers
- Data subjects
- External observers
Separation of populations
| Population | Eligibility rule | Testimony it yields |
|---|---|---|
| Affected customer | Used the route in the last six months | How the route felt and where it stopped |
| Decision-maker | Owns the workflow budget | Constraints and acceptance criteria |
| Operational user | Routed a case in the last three months | Steps, tools and workarounds |
| Assurance reviewer | Audits responses or corrections | Rules followed, evidence kept |
| Data subject | Named in a case record | Consent and withdrawal experience |
| Platform observer | Handles reviews on Trustpilot or Google Business Profile | Platform rules and response timing |
Example: a sampling frame record for one England decision
A sampling frame is the list or rule from which invitations are drawn. Here it combines authorised case references, staff role records and an approved route to prospective customers.
The record holds:
- Frame owner and stated purpose
- Extraction date and eligible period
- England scope rule
- Inclusion, exclusion and duplicate logic
- Unreachable contacts and missing groups
Sampling frame record
- Owner
- Extraction date
- Eligible period
- England rule
- Inclusion and exclusion logic
- Duplicates
- Inaccessible contacts
- Missing groups
Sampling frame record
| Method | Illustrative target | Claim it supports |
|---|---|---|
| Interviews | 8 to 12 per speaking population | Reconstruct one recent case |
| Survey | 60 to 120 eligible responses | Describe that respondent set only |
| Observation | 5 to 8 task runs | Whether the tested route worked |
These ranges are planning figures, not benchmarks. Justify any sample against the decision.
Method choice and the claim it supports
Interviews reconstruct a recent event, not prevalence. Questionnaires describe a defined respondent set when wording, field dates and denominators survive. Observation tests one task in one configuration. Desk research verifies published notices, not configured behaviour.
Recruitment choices create bias, and access needs need deliberate attention, as GOV.UK guidance on finding research participants explains.
Pilot questionnaire wording and order with a small number of survey routes before fielding, then report invitations, eligible responses, partial responses and exclusions separately.
Consent, special category data and incentives
Write a short information record before collection: purpose, fields, recordings, recipients. Add storage, retention, withdrawal and contact points.
Reputation cases can expose health details, political views, ethnicity, religion or trade-union membership. Those are special category data, needing an Article 6 lawful basis and an Article 9 condition. Decide early whether the decision can be answered without them.
Do not pull every address from a reputation database because it is there. The ICO's data minimisation guidance says personal data must be adequate, relevant and limited to what is necessary.
Incentives should pay for time, not opinion. Concealed incentives and misleading review information sit inside the Consumer Protection from Unfair Trading Regulations 2008, which prohibit misleading commercial practices. Research answers must never become testimonials.
Analysis and the bounded decision
Build an evidence matrix with rows for observed problems and columns for populations. A customer may report that the correction route was hidden. An operator may show it exists under different wording. Keep both.
Code against a versioned scheme. Hand labels such as abuse, fake or urgent to the named privacy and assurance roles, because those words carry legal weight.
The decision pack holds the research question, populations, frame and denominators. It also carries missing groups, the evidence matrix, the privacy record and limitations. Findings cite evidence IDs, never "customers want".
Proceed to a reversible synthetic-data trial only when research shows a reproducible workflow problem, a named owner and acceptance evidence. Pause if participant authority is unclear or withdrawal cannot be honoured.
Final checks before you act
- Write the one-sentence decision and its England owner.
- Confirm an eligibility rule per population.
- Record the frame owner and extraction date.
- Fix the consent version before the first invitation.
- List missing groups in the decision pack.
- Judge the shortlist against the workflow the research identified.
Common questions
How many people do I need?
For one bounded decision, 8 to 12 interviews per speaking population and 60 to 120 eligible survey responses are illustrative starting targets. Justify them against the decision and report non-response.
Is this a reputation software review?
No. It covers the research that comes before a shortlist: define the populations, gather the evidence, then assess review platforms such as Trustpilot and Google Business Profile, or reputation-management platforms such as Yext, against the workflow problem you found.
Which England rules matter most?
The Data Protection Act 2018 and UK GDPR govern participant data, while the Consumer Protection from Unfair Trading Regulations 2008 cover misleading review practices. The Market Research Society's Code of Conduct sets the professional standards for UK research practice, and registered care providers in England answer to the Care Quality Commission.
Does the study need special category data?
Usually not. Ask whether case type and outcome answer the decision without health, ethnicity or political detail.
In this guide
- Reputation management target audience: who to recruitDefine a reputation management target audience by recent task, authority and evidence, with UK GDPR rules, research standards and review-platform sources.
- Reputation management buyer personas: research guideA guide to reputation management buyer personas for review software, with role prompts, workflow evidence, source gaps and privacy safeguards.
- Reputation management customer interviews: five approachesFive evidence-led approaches for reputation management customer interviews on Google Business Profile, Trustpilot, TripAdvisor and Feefo.
- Survey routes for reputation research, tested on the method before the platformCompare survey routes for reputation research in England using one evidence grid, a manual baseline and an explicit empty product shortlist.
- Researching reputation competitors without inventing their customers or trafficResearch reputation-management competitors through dated public evidence, narrow claims and explicit unknowns without inventing customers, traffic or market share.
