Audience research
Part of How do talent agencies carry out audience research in England?
How to pick talent agencies survey tools for UK audience data
A practical comparison of talent agencies survey tools and data sources, covering panels, owned lists, employment status checks and UK privacy rules.
What to take away
- Talent agencies survey tools are the platforms and data sources an agency uses to gather structured feedback from audiences, creators and clients. Most projects mix paid panels, owned lists and desk research.
- The sample matters more than the software: a cheap panel with unverified respondents gives weaker data than a smaller list you own.
- Email and cookie-based collection must follow the ICO's Guide to Privacy and Electronic Communications Regulations, which sets consent rules for direct marketing.
- The employment status and employment rights guidance helps frame questions where surveys touch freelance creator contracts.
- Budget for the sample, not the subscription. A £50 monthly tool with a poor panel can waste more on decisions than a £300 one with verified respondents.
Panels, owned lists or desk research
Paid panels are the fastest route to a quota sample, but recruitment method stays invisible. Ask how respondents are verified and whether the vendor supplies raw field dates.
Owned lists come from your creator roster, newsletter and past campaign respondents. They cost less per response and reflect your actual audience, though the pool is smaller.
Desk research uses published statistics, platform analytics and public reports. It is not a survey, but it can validate a panel result before you commit budget.
Most agency projects blend two of the three, using panel data to size a market and an owned list to test the wording of a proposition. The blend is what keeps the cost per usable answer down.
Matching the tool to the question
A survey tool is only as good as the question it answers. Audience sizing, sentiment, pricing tolerance and creator satisfaction each need a different instrument.
Pricing work using conjoint or Van Westendorp needs trade-off support, while sentiment suits a short Likert scale on an owned list.
Creator-side samples often run under 100 people, where a spreadsheet and a structured interview guide can outperform an enterprise platform.
If you are building respondent profiles before choosing a tool, the talent agencies buyer personas in England article sets out the demographic and behavioural fields worth capturing first.
Data sources by research type
| Before: source chosen by habit | After: source matched to the decision |
|---|---|
| One panel vendor for all projects | Panel for quota samples, owned list for sentiment |
| Free online poll shared on social | Paid or owned sample, verified field dates |
| Email list bought from a broker | Opt-in list collected under PECR |
| Creator contract questions written ad hoc | Questions checked against employment tests |
| Single source per report | Two sources compared before reporting |
The after column costs more per project, but it cuts the chance of a decision resting on a sample nobody can defend. Label each source in the report, so a reader can see whether a number came from a panel, an owned list or a published statistic.
Production safety in creator research
Audience research sometimes involves filming contributors or running sessions on location. The HSE entertainment and leisure guidance covers risk assessment for shoots and events. That applies to a focus group as much as to a brand content day.
Put the safety check in the research brief, not after the venue is booked. Note the session date and venue in the same log as the questionnaire.
Budget and sample size
Survey costs scale with completes, not tools. A team paying £400 a month for a panel subscription might get 200 completes, while a £150 owned-list send yields 80 responses from a better-matched group.
Quote the incentive separately from the software fee, because response rates move with the reward, not the interface.
Agree the minimum sample before comparing vendors, so a weak result is not reported simply because it arrived first.
Keep field dates and the questionnaire in the file when research supports a client pitch. An unverifiable number damages a pitch more than a small sample does.
For the full method, including sampling frames and reporting standards, the audience research guide for England covers how these sources fit together.
Common questions
Are free survey tools good enough for agency work?
For internal testing and creator feedback, yes. For client-facing audience data, sample quality matters more than the platform, so a free tool with a verified owned list can beat a paid panel.
Do I need consent to survey my own email list?
For marketing, yes. PECR requires consent for most electronic marketing, and survey invitations that double as promotion fall under the same rules. Pure research invitations to existing customers are treated differently.
How many responses make a survey reliable?
There is no single number. It depends on the decision, the subgroup sizes and the margin of error you accept, so state the limitation in the report.
Can I ask creators about their employment status in a survey?
You can ask, but the answer is not a legal determination. Employment status depends on the working arrangement rather than self-description, so take separate advice on contracts.
