
Audience research
How do video marketing agencies carry out audience research?
Audience research for video marketing agencies: the data to gather, methods that work in the UK, how to brief a supplier, and rules on targeting and data brokers.
What to take away
- The ONS publishes household characteristics and household composition data covering England, Wales, Scotland and Northern Ireland, which is free and can inform audience segments before any paid work begins.
- Audience research should cover four thingswho the audience is, what they already watch, what job the video does, and how success will be measured.
- The ICO sets out what organisations must do when buying audience data from brokers, so brief your agency on lawful data sources from the start.
- UK digital inclusion evidence shows digital skill levels vary widely across the country, so production choices such as captions and pacing should follow the audience, not the other way round.
- The ASA and CAP publish research into advertising effectiveness and audience issues, which is useful when you need to justify a research budget.
Buyers often ask what a video marketing agency actually does before a single frame is shot. The answer starts with audience research, because a video aimed at everyone tends to land with no one. This guide covers the data you can gather yourself, the questions to put to a supplier, and how audience research feeds into targeting and reporting.
What does audience research for video marketing agencies involve?
Audience research is the work of finding out who will watch a video, what they already watch, and what would make them act. It happens before scripting and before production. It is not the same as testing a finished cut, though the two connect.
Four research outputs
- Audience definition as segments
- Existing viewing habitsplatform, device
- What the video must achieve
- A measurement plan
A video production company will often ask for this material before quoting. If you cannot supply it, a good supplier will offer to gather it, but that work has a cost. Deciding who does it is one of the first commercial choices you make.
Which data sources should you use first?
Start with what you already hold. Website analytics, enquiry records, sales notes, support tickets and existing customer lists all describe real people who chose you. This is the cheapest research available and the least likely to mislead.
Data sources in order
- Start: your own analytics and records
- Next: ONS household characteristics
- Then: LinkedIn audience targeting
- Use public data to size, not stereotype
Public data comes next. Household characteristics data from the Office for National Statistics household characteristics collection can tell you how households are composed in a target area, which shapes messaging about family, location or spending. Use it to size a segment, not to stereotype it.
If your audience is professional, platform data is useful. LinkedIn audience targeting lets you see how many people match job titles, industries and locations, which is a fast way to test whether a B2B segment is big enough to justify a video budget.
How do you build audience segments you can actually use?
A segment is useful only if it changes a production decision. If two segments would get the same script, the same length and the same channel, they are one segment.
Is it one segment?
Would both get the same script, length and channel?
treat them as one segment
keep them separate
Build segments around behaviour rather than demographics alone. What someone watches and when they watch tells you more about the video you should make than their age band does. Demographics then help you size and buy media.
For a structured approach to segment definition, the guide to video marketing agencies buyer personas in England walks through the fields worth capturing and the ones that add noise. Keep each segment to a page.
What research methods suit different budgets?
Interviews with five to eight customers give you language you cannot guess. They are cheap to run and slow to analyse, so allow time. Record with permission and store the recordings lawfully.
Methods by budget
Interviews
- Cost
- Cheap
- Speed
- Slow to analyse
- Output
- Real language
- Sample
- 5-8 customers
Surveys
- Cost
- Scales cheaply
- Speed
- Fast to field
- Output
- Message testing
- Sample
- Customer list
Desk research
- Cost
- Lowest
- Speed
- Fast
- Output
- Category gap
- Sample
- Competitor output
Surveys scale what interviews reveal. A short survey sent to a customer list can confirm which of three messages lands best. The piece on video marketing agencies survey tools in England compares the practical options for building and distributing one.
Desk research covers competitor output. Watching what rivals publish tells you what the category already says, which is where you find the gap. A method for doing this systematically sits in the guide to video marketing agencies competitor research.
Where does digital literacy change the plan?
Assumptions about how easily an audience follows a video are often wrong. GOV.UK's digital strategy sets out that digital skills and inclusion vary across the UK population, and that a substantial share of adults lack the full set of essential digital skills.
Low digital confidence format
- Keep on-screen text large and brief
- Add captions
- Avoid app navigation instructions
- Test rough cut with real audience
Where you can, check the GOV.UK digital skills and inclusion evidence before agreeing a format, particularly for public-facing campaigns in England, Scotland, Wales or Northern Ireland.
What does the law require when you target an audience?
Audience research often involves buying or enriching data. The ICO is clear that organisations using the marketing services of data brokers remain responsible for how that data is used. Consent, transparency and lawful basis sit with you, not with the broker.
Before passing any list
- Check ICO data broker guidance
- Confirm where the data came from
- Confirm people expect to hear from you
- Record your lawful basis
Read the ICO guidance on organisations using marketing services of data brokers before you pass any list to an agency or platform. Check where the data came from and whether the people on it expect to hear from you.
Advertising claims and audience treatment also fall under the CAP Code. The ASA publishes its own research into advertising effectiveness and audience issues, which is worth reading if you need to justify a research spend to a finance team.
For the wider set of rules covering video campaigns, including consent for filming and accessibility, see the video marketing agencies UK rules and compliance guide.
What should you ask a video marketing agency to do?
Ask for the research method in writing before you approve a budget. You want to know how many people will be spoken to, how they will be recruited, and what you receive at the end.
Ask how the research will change the video. A supplier that cannot explain what decision the research informs is selling process, not insight. Ask for the raw material too: interview notes, survey questions, segment definitions.
Ask who owns the data and how long it is kept. If the agency holds your customer list, that arrangement needs a written basis. Ask about deletion at the end of the project.
A useful test is to request one page describing the audience before the creative treatment. If the two documents agree, the research has done its job.
How do you scope and price audience research?
Research is usually charged as a separate line, either as a fixed phase or as days within a retainer. For example, a team paying £400 a month for ongoing video production might add a one-off research phase priced as several days of work.
Scope by decision, not by deliverable. If the decision is which of two messages to lead with, you need enough interviews or survey responses to distinguish them. That is a smaller job than a full segmentation study.
Be cautious with very cheap research. If a quote covers interviews, analysis and a written report for a token sum, it is likely desk research wearing a different label. Nothing wrong with desk research, provided it is described honestly.
Which approach fits your situation?
Choose
- First video, no existing audience data
- Five to eight customer interviews plus platform sizing
- Established customer list, message unclear
- Short survey to the list, then two rough cuts
- B2B campaign with a named job title
- Platform audience sizing and industry desk research
- Public-facing campaign with mixed digital confidence
- Accessibility-led format testing with real viewers
- Buying third-party audience data
- Documented lawful basis and ICO checks
- Tight budget, fast deadline
- Desk research on competitor output, clearly labelled
Avoid
- First video, no existing audience data
- A full segmentation study before you have a script
- Established customer list, message unclear
- Another round of internal opinion
- B2B campaign with a named job title
- Consumer panel research
- Public-facing campaign with mixed digital confidence
- Assuming app-native behaviour
- Buying third-party audience data
- Passing lists to a platform unchecked
- Tight budget, fast deadline
- Presenting desk research as primary research
How does audience research feed targeting and measurement?
Research outputs should map onto targeting settings. Each segment needs a matching audience definition, whether that is a list, a platform interest set or a lookalike. If a segment cannot be targeted, it is a persona document, not a media plan.
Measurement should be agreed before launch. Decide which metric shows the video worked for each segment and where that number is recorded. Views alone rarely settle the question.
A written plan for video marketing agencies target audience work helps here, because it links the research to the platforms and the reporting rather than treating them as separate exercises.
What goes wrong most often?
Research is done after the script. By then the creative decisions are made and the findings cannot change anything. Sequence it first.
Segments multiply. Teams add segments until no single video can serve any of them. Two or three segments is usually enough for one campaign.
Findings are not written down. If the insight lives in one person's memory, it disappears when they change role. Keep a short, dated document.
Compliance is treated as an afterthought. Data bought for targeting carries obligations that survive the campaign, so record your lawful basis when you gather the data, not months later.
Common questions
How long does audience research take before a video project?
A focused round of interviews and desk research typically fits into two to four weeks. Larger surveys or segmentation studies take longer. Agree the deadline alongside the production schedule so research does not delay the shoot.
Can we skip audience research if we already know our customers?
You can reduce it, but not remove it. Existing knowledge usually covers who buys, not what they watch or which message lands. A short survey or a handful of interviews is enough to test the assumptions you already hold.
Who owns the audience data gathered during the project?
That depends on your contract. Confirm in writing whether customer lists and interview recordings stay with you, what the agency may keep, and when each copy is deleted. Data protection obligations remain with your organisation.
Do we need research for every single video?
No. Research sets the audience and message for a campaign. Individual videos within that campaign can reuse the same findings unless the audience or objective changes.
In this guide
- How a video marketing agency builds buyer personas from evidenceA practical guide to building buyer personas for a video marketing agency, covering buying groups, data sources, interviews, UK GDPR and sample sizes.
- What buyer persona work means for video marketing agency briefsHow England agencies build video marketing agency buyer personas, covering data protection, ONS health data, connectivity assumptions and interview method.
- Non-leading customer interviews for video marketing agenciesA practical guide to video marketing agency customer interviews, with neutral question examples, a sample discussion guide and UK data-protection points.
- Why video marketing agencies survey tools cannot fix a weak question setSurvey tools decide how fast video marketing agencies test creative and profile audiences, but wording, quotas and consent decide whether the numbers hold up.
- Nine checks for video marketing agencies competitor researchMost teams compare showreels and stop there. These nine checks turn video marketing agencies competitor research into evidence you can defend when budgets are set.



