Should you survey every attendee or a sample of your event audience?

You've finished a paid event, and the team wants feedback before committing to the next edition. Sending the survey to everyone feels like the obvious choice. You have the contacts, the email costs very little, and nobody wants to explain why some attendees weren't asked. But you're also planning follow-up interviews, checking a paid upgrade, and trying to avoid another fortnight of chasing replies.

For a short, routine event feedback survey, inviting every eligible attendee is often the practical starting point. Choose a selected sample when contacting or interviewing everyone would use more time and money than the decision justifies, or when you need a properly designed study. Either way, decide who belongs in the audience, how you'll reach them, and what you'll do about missing replies before sending anything.

The important distinction is between who you invite and who you hear from. An invitation to everybody gives everybody on that list an opportunity to answer. It doesn't promise that your results will describe every attendee. Getting this right at the planning stage helps you spend your research budget sensibly and explain the findings without overstating them.

What a census and a sample actually mean

The Australian Bureau of Statistics defines a census and a sample by how much of the population you study. A census covers every member of the defined population. A sample covers a selected part. In event research, your population might be everyone who attended a particular edition, everyone who bought a premium ticket, or every purchasing organisation.

Those are different populations, and they answer different questions. A purchasing organisation can explain the approval process for a team booking. An individual attendee can describe using the cloakroom. If you mix those units together, inviting more people won't make the answers easier to interpret.

There's another distinction worth making early. Sending invitations to the whole population is an attempt at a census. A complete census would require the relevant information from every eligible member. Most voluntary event surveys won't achieve that. In your report, describe the invitation method and the responses you obtained rather than calling incomplete feedback a complete account of the audience.

Even complete collection can contain mistakes. Attendees can misunderstand a question, forget an experience, or answer something they didn't use. Your team can classify a ticket incorrectly or combine answers badly. Covering everyone removes uncertainty caused by selecting only part of the population, but it doesn't remove those other problems.

Decide what you need to learn before choosing invitees

Start with the decision somebody will make. "Understand attendee satisfaction" sounds reasonable until you have to choose which people belong on the invitation list. Are you deciding whether to renew the venue, whether to keep a paid workshop, or whether to change the way group bookings work? Each decision points towards a different set of people.

Imagine a fictional paid business conference with 2,400 attendees. Of those, 300 bought a workshop upgrade. The organiser has to decide whether to sell that upgrade again and whether the general arrival process needs changing. These are illustrative figures for planning, not results from a real conference or recommended survey targets.

The arrival review needs experiences from across the conference audience, including people who reached different entrances or arrived at different times. The upgrade review needs people who bought the workshop, together with a way to distinguish buyers who actually used it. Asking all attendees to rate the workshop would introduce answers from people who couldn't judge it.

The organiser could invite the whole audience to a brief survey and route workshop questions only to eligible people. It could then invite selected workshop participants to longer interviews. That combination can be sensible because the broad survey and the interviews serve different purposes. You don't have to force every research question into the same collection method.

Write the decision and eligible audience before drafting the questions. Our guide to post-event survey questions helps you turn those decisions into a form. Here, you're deciding who should receive it and how that choice affects what you'll be able to say afterwards.

Check whether your contact list contains your audience

A ticket database can look like an attendee list while containing something quite different. One person may have bought admission for colleagues. A guest may have used a transferred ticket. Some buyers never arrived. Some people entered with complimentary admission, and some records may contain the same person more than once.

For the fictional conference, suppose the organiser has 1,700 purchasing email addresses for 2,400 attendees. Sending to every purchasing address would cover that purchasing list. It would not establish that every attendee received an invitation. A buyer who stayed home also cannot report a colleague's experience of the arrival queue simply because they paid for the ticket.

Before choosing between everybody and a sample, inspect how the records relate to people. Identify duplicate contacts, group orders, transfers, and eligibility questions you cannot settle from existing records. Don't delete an uncertain person merely to produce a tidy list. Decide whether a short eligibility question can establish whether they attended and which experience they can discuss.

AAPOR's best practices for survey research explains the sampling frame, meaning the information that lets you contact potential respondents. For your event, that might be individual attendee contacts or a purchasing list with known gaps. Write down those gaps. A random selection from an incomplete list still excludes people who were never on it.

If individual contact details are missing, decide whether you can obtain or reach the relevant attendees appropriately before collection begins. A group purchaser might forward an invitation, for example, but you won't have the same control over delivery or duplicate participation. Describe that route honestly. Don't present forwarded links as individually verified invitations.

When inviting every eligible attendee makes sense

Invite everyone when the audience is manageable, the form is short, and there is little practical benefit in excluding people. For a routine post-event review, you may value the opportunity for every attendee to mention a problem that the team hasn't anticipated. That opportunity has a purpose even when you aren't trying to estimate a population percentage.

For the conference organiser, inviting all eligible attendees to a brief arrival review is a reasonable choice if the contact list supports it. It avoids spending staff time constructing a sample merely to reduce a low-cost email send. It also allows people with unusual experiences to respond, provided the form gives them a relevant route.

The workshop population is smaller, so inviting every eligible workshop buyer may also be practical. The organiser shouldn't remove most of that group just because it has selected a smaller group from the main audience. A small population can make an invite-all approach easier to run, especially when the organiser needs to understand a specific paid benefit.

Don't promise that inviting everyone guarantees every kind of experience will appear. People can decline, miss the message, or find the survey difficult to complete. The advantage is the breadth of the invitation opportunity. You still need to make participation accessible and check which parts of the audience you actually heard from.

An invitation can also communicate that feedback is welcome. Keep that separate from the research claim. Giving all attendees a chance to comment may be good customer practice, but it doesn't make the eventual respondents a statistically representative group. You can value broad access without claiming more than the collection method supports.

When a selected sample is a better use of your budget

The cost changes when research needs a person on the other end. Interviewing every attendee, listening to recordings, and reviewing detailed accounts can use substantial staff time. A selected sample lets you concentrate that effort where it will help the decision. The right selection method depends on whether you need numerical estimates or explanations of particular experiences.

For the conference's workshop decision, detailed interviews might explore whether attendees understood the promise, used the materials, and found the format worth paying for. The organiser could deliberately invite people with different workshop experiences. That would be a useful way to investigate why an upgrade disappointed some buyers, provided the report describes those interviews as selected accounts.

A different brief might require an estimate of how the entire eligible audience experienced the workshop. That calls for a defensible selection and analysis plan, with enough time to recruit the selected people and deal with nonresponse. You shouldn't use a few convenient interviews to support a percentage about every buyer.

Sampling can also reduce the burden on people who attend many editions of your events. If every event triggers a long questionnaire and repeated reminders, the total request becomes considerable. Plan which editions and audiences need which questions. A short feedback opportunity for everyone and occasional deeper research may serve the business better than repeatedly asking everyone everything.

Consider staff capacity after collection as well. There's little value in requesting long comments from the entire audience if nobody has time to read them. Budget for reviewing, investigating, and acting on the answers. Reducing the scope of the questionnaire may be more useful than reducing the invitation list, so compare both options before choosing.

A selected sample can be random or deliberately chosen

People often use "random" to mean they didn't think very hard about whom they asked. In survey research, random selection means using a chance process. Statistics Canada's guide to probability sampling explains that members of the survey population have known selection probabilities. They need not all have the same chance in every design.

A simple random sample could select individual attendee records from an eligible list through a documented random process. Choosing the first people whose names you recognise, or emailing the contacts at the top of an export, isn't the same method. Neither is asking a colleague to pick people who seem typical.

A probability design matters when you want to estimate something about the defined population using a planned statistical method. Selection is only one part of that work. The list needs suitable coverage, questions need to work, and missing replies need attention. Random invitations don't guarantee representative completed answers after some selected people decline.

A deliberately chosen sample can be appropriate for a different purpose. You might want to understand the experience of people who bought an upgrade and left before it ended, or compare how first-time and returning attendees interpret an instruction. Choose people who can discuss those experiences and explain why you selected them.

Statistics Canada's guide to non-probability sampling describes methods including convenience and quota sampling. These can help you recruit particular people, but filling a quota doesn't turn volunteers into a random sample. Use the findings to explore the people and experiences studied, and take care with wider claims.

Plan important groups before you draw the sample

A sample that looks large overall may leave you with very little information about a commercially important group. At the fictional conference, the workshop population is only part of total attendance. If workshop decisions matter, leaving its inclusion entirely to a small general sample would make that review unnecessarily uncertain.

You can plan separate invitation groups. In a probability study, a researcher may divide the population into relevant groups and select randomly within each. This is called stratified sampling. The groups need definitions you can apply consistently before selection, such as ticket type or event day, rather than assumptions based on people's names.

The organiser might invite all workshop buyers while selecting a sample of other attendees for the general review. That's a possible design, but the invitation chances now differ. If you later combine answers to estimate the whole audience's views, you need analysis that accounts for how people were selected. Don't simply treat every respondent as equally representative because they completed the same form.

You can avoid unnecessary complexity by reporting the workshop review separately when that matches the decision. The arrival review and the upgrade review don't have to produce one combined score. Ask what the team will actually compare before requesting a design that requires specialist analysis nobody has budgeted for.

Choose only the group distinctions that matter. Splitting by every ticket category, purchase period, day, workshop, and attendance history can leave many tiny groups and a confusing recruitment plan. If you need to understand first-time arrivals, prioritise that distinction. If you can't identify first-time attendance reliably, acknowledge the gap and consider asking respondents directly.

Keep open feedback separate from selected research

An open link on social media or an exit sign can be useful for hearing what people want to tell you. It isn't a controlled invitation to a known set of attendees. Some eligible people won't see it, some people may share it, and you may not know how many people had a genuine opportunity to respond.

The conference organiser could keep an open feedback route while running selected workshop interviews. Explain those routes separately. Someone volunteering a detailed complaint through the open form may be worth following up with, but adding that person's answers to the selected study changes the collection method. Preserve where each account came from.

An on-site intercept survey, where staff approach attendees at a location, needs similar care. Asking people near the workshop exit will miss buyers who left early or never entered. Asking at the premium lounge overrepresents people using that space. A location that is convenient for staff can quietly exclude the experience you most need to understand.

If you use intercepts, plan where and when staff recruit, who is eligible, and how they choose whom to approach. Train staff to follow the method rather than approach only friendly faces. Record refusals and unavailable groups in a proportionate way. A researcher can help design this properly if the findings will support consequential population claims.

You don't need to ban informal feedback to protect a study. Keep its role clear. Open comments can identify an issue for investigation, while a planned survey addresses a defined research question. Both can inform the same operational decision without being combined into a percentage that hides their different origins.

Count the full cost of asking

The email sending charge is only one part of the cost. Somebody prepares and checks the list, tests the form, answers queries, handles participation problems, reviews comments, and communicates the findings. Longer interviews add scheduling, interviewer time, and analysis. These costs may make a smaller, carefully recruited study worthwhile.

Compare realistic versions of each option. An invite-all survey with a short core form may require less work than a complicated sample with many recruitment groups. Conversely, a selected interview programme may be manageable where an invitation to every attendee would create more demand than your team could handle. The number of contacts alone doesn't settle the comparison.

Include the attendee's effort in the decision. Test the form with people similar to your intended respondents and time the actual route they see. A workshop participant may receive more questions than someone who skipped it. Don't advertise a completion time based only on the shortest possible version.

Think about what you already know. If attendance records can answer how many people entered a session, you may not need to ask everyone to recall it. A survey can then explore why they chose it or what they thought. Check whether the records measure the thing you need before replacing a question with them.

For questions about preferences before an event, our guide to pre-event survey questions covers useful decisions to make before arrival. The same invitation principle applies. Ask the people whose answers could change the decision, and use existing information where it answers the question accurately.

Decide how you'll handle missing replies

Whichever invitation approach you choose, assume some people won't answer. Plan your reminder policy, alternative participation route, and closing date before seeing the ratings. This helps the team focus follow-up on participation problems rather than deciding to keep collecting because an early result is uncomfortable.

For a selected sample, don't casually replace someone who doesn't respond with a friendlier attendee. The substitute may have different reasons for participating. If you expect to need additional recruitment, agree the method in advance and preserve the selection record. A survey researcher should determine how replacements or additional invitation groups affect a formal study.

For an invite-all survey, reminders can reach eligible nonrespondents where your setup permits that distinction. Avoid repeatedly asking people who already completed it. If answers are anonymous and you cannot identify completions, explain that clearly when planning reminders. Don't promise a level of anonymity that your actual collection method cannot provide.

The fictional conference should check whether workshop buyers received the right route and whether group-booking guests could access the invitation. A weak response from those groups might reflect delivery or eligibility problems. It doesn't automatically reveal dissatisfaction, indifference, or anything else about their experience. Investigate the contact process before assigning motives to silence.

Statistics Canada's questions to ask about surveys covers nonresponse and other issues that affect quality. Missing replies matter because respondents may differ from people you didn't hear from. Increasing the total doesn't necessarily solve that problem if the additional people come from the same easily reached group.

Prepare the invitation and fieldwork plan together

Your invitation should explain which event the request concerns, why you're asking, who should answer, and when the survey closes. Keep the request neutral. Asking delighted attendees to tell you what they loved recruits a different response from inviting people to describe their experience, including anything that needs improving.

Give people an accurate expectation of the effort involved and explain any incentive. Don't make a reward depend on a positive answer. If the incentive is a discount on another edition, consider whether it will appeal more to people who already intend to return. Record what you offered so the eventual report can describe the recruitment conditions.

Our guide to event email marketing strategy can help you prepare clear messages and organise communication. Your survey invitation still needs a research purpose of its own. Avoid adding a sales pitch that makes the feedback request feel like an excuse to promote the next event.

Before launch, assign responsibility for list checks, form testing, delivery issues, and the final review. Set a deadline for fixing a broken link or routing mistake. Test the workshop path with an eligible participant and the general path with someone outside that group. Check the experience on the devices attendees are likely to use.

Keep a copy of the invitation, eligible list definition, selection procedure, questionnaire, and planned reminders. You don't need an elaborate research department to document the basics. You do need enough information for somebody at the debrief to understand why these people were asked and how the final respondents entered the study.

Give unusual experiences a way to be heard

Some experiences matter even when you aren't trying to measure how often they occurred. An accessibility problem, a serious complaint, or an undelivered paid benefit may need a direct response. Attendees shouldn't have to be selected for a research interview to raise those issues. Keep an appropriate support route available alongside the survey and make clear when somebody should use it.

This doesn't require adding a long complaints section to every questionnaire. It means deciding who monitors replies and what happens when a respondent asks for help. If your main survey is anonymous, offer a separate way to request contact. Explain the difference so people can choose whether they want a response without misunderstanding how their survey answers are handled.

Be careful when selecting interviews from small groups. A detailed account can identify someone even after you remove their name, particularly when only a few people used a specific service. Agree how you'll use interview material before collecting it. Don't promise that a distinctive story can be published anonymously if the people involved would recognise it.

For the conference organiser, this means assigning a person to handle workshop delivery complaints while another person coordinates the broader review. A buyer can then receive a useful reply without their case becoming the sole explanation for the workshop's performance. That separation lets the business address an individual problem promptly while continuing to investigate whether the underlying issue affected other buyers. Include the person responsible for attendee support when you test these arrangements.

Keep the next edition comparable where it matters

If you plan to compare events, record how you invited people this time. A switch from inviting the whole audience to interviewing selected regulars can change the findings even when the event itself hasn't changed. The same applies when one edition reaches individual attendees and the next reaches only purchasing contacts.

You may have good reasons to change the method. Perhaps individual contacts become available, a new ticket category needs its own review, or the previous questionnaire required too much effort. Describe the change when presenting results. Don't turn a difference between two recruitment methods into a confident claim that attendees became happier or less satisfied.

For the fictional conference, the team could keep the general arrival questions and invitation approach reasonably consistent while changing the workshop research to suit its decision. That allows the operations review to continue without forcing a new paid benefit into an old survey structure. Keep a record of what changed and why, so next year's team doesn't have to reconstruct the method from an email export.

Work through the conference invitation decision

Before choosing a final method, the conference team should compare what each option requires. Start with the broad arrival review. The team wants to identify confusing instructions and understand whether particular arrival experiences deserve investigation. It can invite all eligible individual contacts, ask attendees to identify their arrival route, and keep the form brief. It should also record that some group-booking guests may not be directly reachable.

Now consider the workshop. The decision concerns a paid benefit, so workshop buyers deserve a separate invitation route or a clearly tested section within the main form. Buying and attending are different eligibility conditions. Buyers who missed the workshop can explain why they missed it. Participants can judge the experience. Separating those questions may reveal a scheduling problem that a satisfaction rating would miss.

The interviews need a further decision. If the team wants to understand why people struggled with the materials, it can recruit willing participants who describe that experience. It should also seek other relevant accounts so the discussion doesn't assume a single explanation. The interview brief should say that the work explores explanations and does not measure how common each explanation was across all buyers.

Suppose the finance lead instead needs a defensible estimate to support a substantial investment in expanding the workshop. The organiser should discuss that requirement with a researcher before selecting invitees. The required precision, available contacts, expected participation, and groups to compare all affect the study. A smaller study isn't automatically cheaper once careful recruitment and specialist analysis are included.

Finally, agree what would happen if the evidence remained incomplete. The team could improve unclear workshop instructions, investigate delivery problems, and defer expansion until it had stronger evidence. Planning that possibility now reduces the temptation to treat whichever replies arrive as permission for a decision already made. The survey can inform the investment without being expected to settle every part of it.

Choose an approach your team can explain afterwards

For the fictional conference, the practical plan could be a short invitation to every reachable, eligible attendee, with workshop questions limited to people who can answer them. Selected follow-up interviews could explore the upgrade decision in more detail. The organiser would keep the interview findings separate from numerical survey results and document the gaps in individual contact coverage.

Another event business might reasonably choose a probability sample because interviews are expensive and the decision needs a population estimate. That choice requires a suitable list, a selection plan, and analysis expertise. Get those arrangements agreed before collection. You can't recreate a planned random sample by applying a statistical label to whichever attendees volunteered.

The choice between an event census and a sample survey is therefore a planning decision about access, effort, and the claim you need to support. Invite everyone when broad participation is practical and useful. Select people when that gives you a workable study for a clearly defined purpose, and describe how you chose them.

Before sending the invitation, ask the person responsible for the next event to read the plan. They should be able to identify whose experience you'll hear about, which decision it can inform, and what would remain unknown if important groups don't respond. Resolve those questions while you can still change whom you invite.