One big survey or many small ones for event marketing audience research

What to take away

Why one research cycle beats five rushed ones

Most agencies in England do research the wrong way round. A pitch lands, someone needs a persona by Friday, and a junior researcher interviews three people they already know. The output looks like evidence but it is really a summary of existing assumptions.

The alternative is a standing research cycle that runs quietly in the background and feeds every pitch. You keep one panel of past attendees, one panel of prospects and a short list of client-side decision makers. When a brief arrives, you already have the raw material.

Recruit from your own ticket lists first. People who have already paid you money are the easiest group to reach and the most likely to reply. The week after an event, while the memory is fresh, is usually the best window.

Then add a smaller prospect panel drawn from professional networks. Prospect panels are slower to build and harder to keep, so treat them as a long-term asset rather than a quick fix.

This pillar sits above five supporting guides. Read the event marketing target audience in England piece first if you need to define who you are actually selling to. Everything else depends on that definition being sharp.

What counts as audience research?

Audience research covers anything you do to understand the people who might attend, sponsor or speak at an event. That includes desk research, surveys, interviews, behavioural data from previous events and third-party statistics.

Primary research is data you collect yourself. Secondary research is data someone else collected, such as the ONS leisure series or a trade body report. A sensible cycle uses both, starting with the secondary work because it is cheaper and faster.

A useful distinction is between attitudinal data (what people say they want) and behavioural data (what they actually did). Both matter. Behavioural data is harder to collect but more reliable.

It is not the same as competitor research, although the two overlap. Competitor work tells you what other events charge and how they position themselves. Audience work tells you what your people want and what stops them coming. The event marketing competitor research in England guide covers the first; this article covers the second.

Which questions should you ask before you collect anything?

Write down the question you are trying to answer.

Good research questions sound like: should we move the venue closer to a train line? Would a 4pm start attract more senior delegates? Is the £450 ticket price a barrier for public sector buyers?

Bad research questions sound like: what do people think of our brand? That is a broad question, and it will produce a report nobody reads.

Aim for three to five questions per cycle. More than that and your survey becomes a burden, your response rate falls and your analysis gets shallow.

Agree those questions with the people who will act on the answers. If the client's sales lead cares about pricing and you care about venue, write both down and say plainly which one the data can settle.

Who should be in your sample?

Define your population before you define your sample. If your event targets finance directors at UK manufacturers with 50 to 500 staff, that is your population. Everything else is noise.

For most B2B events in England, a sample of 80 to 150 completed responses gives you directional findings. That is an illustrative range, not a statistical guarantee, and it comes from practical experience rather than a published standard.

Worked example: population of 1,200 past attendees, target 100 completes, expected response rate 20 per cent, so invite 500. Set a quota of 50 public sector and 50 private sector buyers. At 100 completes, the margin of error is about 9.8 percentage points at 95 per cent confidence for a 50 per cent result.

Set quotas before the survey opens. If half your audience works in the public sector, do not let a 90 per cent private sector response rate distort the picture. Check the mix each week and top up the groups that lag.

If you need statistical confidence, you need a sampling specialist. For marketing decisions, directional data is usually enough.

Do not forget the people who did not attend. Non-attendee research is often more valuable than attendee research because it explains the barrier you have not yet removed.

How do you reach a B2B audience without breaking the rules?

Paid social and professional networks give you targeting options that are hard to replicate elsewhere. LinkedIn lets you filter by job function, seniority, company size and industry through its audience targeting for professional events. That is useful when you need to reach a specific slice of a sector.

Email remains the workhorse. If you are contacting corporate subscribers, the rules are lighter than for sole traders and unincorporated partnerships, but consent and opt-out requirements still apply. Check the current guidance before you send anything.

For consumer-facing events, consent must be explicit and recorded. A tick box that is pre-ticked is not consent, and the ICO has been clear about that for years.

A modest incentive lifts response rates. For example, a £10 voucher or a free place at a paid session tends to work better than a large payment, which attracts people with no stake in the topic. Record what you offered and to whom.

Keep the invitation short and say how long the survey takes. A clear subject line and a plain statement of purpose will do more for your response rate than any design trick.

What about surveys, interviews and focus groups?

Surveys scale. Interviews give depth. Focus groups generate ideas but are poor at measuring prevalence.

A practical mix for an agency is one survey per quarter, six to ten interviews per year and a single focus group before a major launch. That is an illustrative pattern, not a rule.

When you build the survey, keep it under ten minutes. Every additional question costs you completions. Put the demographic questions at the end, not the start.

Pilot with five colleagues before it goes out. A pilot always catches one ambiguous question, and you cannot rewrite a survey after 200 responses.

The event marketing survey tools: data and sources article compares the main platforms and explains how to weight results when your sample skews.

How do you turn findings into personas?

A persona is a decision aid, not a character sketch. If your persona has a name, a stock photo and a favourite podcast, it is probably decoration.

A useful B2B persona contains four or five fields:

Build one persona per buying role, not one per industry. A finance director and a marketing manager at the same company will book for different reasons.

Review each persona every time you refresh the research. One that survives three cycles unchanged is either unusually stable or simply not being tested.

The event marketing buyer personas in England guide walks through the template and shows how to keep personas current as your audience shifts.

Does the ASA and CAP research matter to event marketers?

The ASA and CAP publish research on advertising issues. That evidence base sits outside the one-survey-versus-many question, so check it only if your event creative makes outcome claims.

How do you keep research lawful and ethical?

Start with a lawful basis. For most audience research, legitimate interests or consent will be the right choice, and you must record which one you are relying on.

Write a privacy notice in plain English. Say what you collect, why, how long you keep it and who you share it with. Link it from every collection point.

Collect only what you need. Special category data, such as health information or political opinions, needs a specific condition and a good reason before you touch it.

Anonymise where you can. If you are quoting a delegate in a pitch deck, get written permission or remove anything that identifies them.

Store research data separately from your CRM. Research data has a shorter useful life and a different retention period.

What does good analysis look like?

Analysis is where most research projects fail. The data is fine, but the report sits unread.

Start with the question you wrote at the beginning and answer it directly. Then note the surprises. Then note what you still do not know.

Cross-tabulate by the variables that matter. If seniority is your key segmentation, every finding should be split by seniority.

Report the findings that contradict your plan as clearly as the ones that support it. A report that only confirms what you already believed will not be trusted next year.

Write a one-page summary. If it needs more than a page, it will not be read by the people who approve budgets.

How often should you repeat the cycle?

For a stable audience, once a year is enough. For a fast-moving sector, twice a year.

Repeat the same core questions each time so you can track movement. Add two or three new questions per cycle to explore shifts.

Keep a research log. Note the date, the method, the sample and the headline finding. This becomes your institutional memory and stops the same debate recurring every quarter.

Put the review date in the calendar at the start of the year and treat it like a client deadline. Otherwise it slips behind delivery work every time.

How do you connect research to delivery?

The findings must land in the operational plan, or the research was theatre. Ticket tiers, session formats, catering and timing should all trace back to something you learned.

Brief the delivery team face to face. A summary attached to an email is easy to ignore, while a short walkthrough makes the reasoning stick.

Build the research calendar alongside the delivery calendar so the two do not compete for the same weeks. The event marketing: operations and delivery guide for 2027 sets out how to sequence the year and where research fits.

If a finding contradicts your plan, say so in writing. That is the point of doing the work.

What does community engagement add?

Research is not only extraction. Events that build a community generate better data because people want to contribute.

Eventbrite's guide to increasing community engagement for events covers practical tactics such as pre-event prompts, follow-up loops and member-led sessions, all of which create natural research touchpoints.

Treat your panel as a community, not a database. Give them early access, share findings back and thank them properly.

Ask panellists what they want to hear about next. That single question often tells you more than a formal survey.

Glossary

Common questions

How many responses do I need for a useful B2B event survey?

For directional marketing decisions, 80 to 150 completed responses is a workable target for most UK B2B events. If you need statistical confidence, speak to a sampling specialist.

Can I survey people who have never attended my event?

Yes, and you should. Non-attendee research explains the barriers that keep people away, which attendee research cannot reveal.

Do I need consent to email corporate subscribers about research?

The rules for corporate subscribers are less strict than for individuals, but you still need a clear opt-out and a privacy notice. Check the ICO guidance for your specific case.

How long should I keep research data?

Keep it only as long as you need it for the purpose you collected it for. A common approach is 24 months for research data, reviewed annually, but your own retention schedule should reflect your lawful basis and sector.