In our guide to managing freelance qualitative recruiters, the recruiter scorecard started with one measure: show rate. It is the number every field manager watches on the night, and the one that most often turns a well-run job into an awkward call with the client.
This guide takes show rate as its subject. Not as a recruiter metric, but as an operational system: how to set over-recruitment from evidence rather than habit, which reminders actually change behaviour, what UK rules apply to the messages you send, and why, in online qual, the harder question is no longer whether someone turns up but whether the person who turns up is the person you recruited.
It is written for operations directors, field managers and agency owners running consumer and B2B qualitative fieldwork in the UK, in person and online.
"An empty seat costs you once. A seat filled by the wrong person costs you the incentive, the client’s trust and the data."
The true cost of a no-show
Most agencies count no-shows as a nuisance: an unpaid incentive and a phone call to the client. The real cost sits elsewhere. The venue, moderator, viewing facility, client travel and catering are fixed the moment the group is booked. A no-show removes value from all of them at once, and an underfilled group puts the agency on the back foot for the rest of the project.
Take the illustrative project from our participant incentives guide: six consumer groups, eight seated per group, a £70 incentive, and the standard practice of recruiting two extra people per group. That guide showed how the over-recruits quietly cost £750 when the quote did not allow for them. The question here is the opposite one: was two extra the right number in the first place?
For most agencies, the honest answer is that nobody knows. “Recruit ten for eight” is a rule of thumb inherited from an era of landlines and printed confirmation letters. It may be right for your audiences. It may be costing you money on easy jobs and failing you on hard ones. The only way to tell is to do the maths with your own show rates.
What show rates actually look like
Published benchmarks for qualitative show rates are thin, and most come from UX research platforms rather than market research fieldwork. They are still a useful anchor:
- Around 5% to 10% no-shows for moderated sessions. MeasuringU’s 2025 analysis reports a 9% no-show rate across roughly 14,000 studies on the User Interviews platform and 5% across its own studies, and recommends planning for 10%.
- Around 20% over-recruitment in practice. The same analysis measured actual over-recruitment of 19% across 30 studies and 1,183 participants, with almost no difference between in-person (18%) and remote (19%) work.
- A wider 10% to 20% range is often quoted for moderated research generally, with individual studies sometimes losing half their participants.
Treat these as a floor, not a target. UX platform figures are dominated by one-to-one remote sessions with participants who chose a time slot themselves. Evening consumer groups in a viewing facility, hard-to-reach B2B audiences, and health or vulnerable audiences behave differently. Weather, transport strikes, school holidays and a big football match all move show rates in ways that no global benchmark captures. Your own history, broken down by audience, method and recruiter, is the only benchmark that matters.
A quick check: pull your last 20 qualitative jobs. For each one, divide the number who attended by the number confirmed. If you cannot produce that figure in under an hour, you are setting over-recruitment by instinct.
The maths of “recruit ten for eight”
Over-recruitment is a probability problem. If each confirmed participant turns up independently with the same probability, the chance of filling every seat follows a binomial distribution. You do not need to do the calculation by hand, but you do need to see what it says, because the results are less intuitive than most field teams expect.
The table shows the probability that at least eight people attend a group, for different show rates and different numbers recruited.
| Show rate | Recruit 9 | Recruit 10 | Recruit 11 | Recruit 12 |
|---|---|---|---|---|
| 95% | 93% | 99% | Over 99% | Over 99% |
| 90% | 78% | 93% | 98% | Over 99% |
| 85% | 60% | 82% | 93% | 98% |
| 80% | 44% | 68% | 84% | 93% |
Three things stand out.
1. “Ten for eight” is only reliable at high show rates
At a 90% show rate, recruiting ten fills a group 93% of the time. At 80%, the same rule leaves you short roughly one group in three. If your audience is harder to get through the door, the inherited rule is not cautious. It is under-recruiting.
2. The odds compound across a project
Clients do not judge a project by one group. At an 85% show rate, recruiting ten gives an 82% chance of filling any single group, but only about a 30% chance of filling all six groups in our example project. Recruit eleven and that rises to about 65%. The difference between those two numbers is the difference between a smooth project and one where the client remembers the empty chairs.
3. Every extra recruit has a predictable cost
The cost of over-recruiting is the people who turn up and are sent home, who should be paid in full. At an 85% show rate, moving from ten to eleven recruits raises the expected number sent home per group from about 0.7 to about 1.4. At a £70 incentive, that is roughly £50 more per group in incentives, plus the recruiter fee for the extra recruit. Set against a 93% rather than 82% chance of a full group, that is usually a good trade on a high-value job, and a poor one on a job where the client is relaxed about seven in the room.
That last point matters. The right over-recruitment depends on the minimum acceptable group size agreed with the client, the cost of the incentive and the consequences of falling short. It should be a decision made per job, from data, and written into the quote, as we set out in the over-recruitment section of our fieldwork costing guide.
A caveat on the maths: the binomial model assumes no-shows are independent. In reality they cluster. A rail strike, a snowstorm or one recruiter’s weak confirmation work can take out several people in the same group. Where you see clustering in your data, add a margin on top of the table, or spread each group’s recruits across more than one recruiter.
Why people do not turn up
Before reaching for reminders, it helps to know what you are fixing. No-shows have a small number of causes, and each has a different remedy.
| Cause | What it looks like | Remedy |
|---|---|---|
| Long lead time | Recruited three weeks out, forgotten or double-booked by the night | Recruit closer to the date where incidence allows, and add a mid-point reconfirmation |
| Weak commitment | Agreed on the phone to be polite, never put it in a diary | A clear verbal commitment at recruitment and a calendar invite sent the same day |
| Logistics | Could not find the venue, no parking, the link did not work | Specific directions, entry instructions and an online tech check |
| Incentive doubt | Unsure how or when they will be paid, or whether it is worth it | State the amount, method and timing in writing, and pay promptly every time |
| Impersonal process | Felt like one of a crowd, assumed nobody would notice | A named contact and a personal call before the session |
| Over-booked respondents | Participants juggling several studies, or not who they claimed to be | Frequency checks and cross-identity matching in your database |
The last row links show rates to data quality. Participants who sign up for everything are more likely to drop out of anything, and a database that tracks outcomes per person will show you who they are. We covered how to record attendance, no-shows and sent-home outcomes against each participant in our guide to managing a research participant database.
A confirmation sequence that works
The strongest evidence on reminders comes from healthcare rather than research, but the behaviour is the same: people forget, and a timely nudge helps. A Cochrane review of eight randomised trials found attendance of 67.8% with no reminder, 78.6% with text message reminders and 80.3% with phone call reminders. Texts and calls performed about the same, but two studies found the cost per attendance of text reminders was 55% to 65% lower. The evidence is rated low to moderate certainty and comes from medical appointments, so treat the size of the effect as indicative.
The practical lesson is to automate the routine reminders and spend human time where it changes behaviour: the personal call shortly before the session. This is the sequence we would start from and then tune against your own data:
- At recruitment. A confirmation email or text within the hour, with date, time, location or link, incentive amount and method, and a named contact. Attach a calendar invite.
- One week before. A short reconfirmation asking the participant to reply to confirm. A non-reply is your first early warning and triggers a call.
- 48 hours before. A personal call from the recruiter or field team. This is the step that most improves show rates and the one most often skipped when the team is busy.
- On the day. A morning text with directions, parking or the joining link, and a number to call if anything goes wrong.
- Online sessions. A tech check at least a day before, and an invitation to join ten minutes early. Many online no-shows are participants who could not get the platform working.
Two refinements make the sequence far more useful. First, record every response, so a participant who has not confirmed by the 48-hour point is visible to the field manager and can be replaced while there is still time. Second, record which step each no-show failed at. Over a few months, that tells you whether your problem is commitment, logistics or a particular recruiter.
What PECR allows in a reminder
Texts and emails to participants are electronic communications, so the Privacy and Electronic Communications Regulations apply alongside UK GDPR. The key distinction is between service messages and direct marketing. The ICO describes service messages as messages sent “for purely administrative or customer service purposes”, and its guidance on regulatory communications makes clear that a routine message with marketing elements becomes direct marketing.
A neutral reminder about a session someone has agreed to attend is a service message. Add “join our panel for more paid opportunities” or a promotion for another study, and the message is likely to become marketing, which needs the participant’s consent for texts and emails. Keep reminders strictly about the session. Invitations to future research belong in a separate, consented channel, which is the recontact question we covered in the database guide.
Incentives, standbys and sent-home participants
Incentive policy shapes attendance more than most agencies realise. Participants talk to each other, and a reputation for slow or disputed payments shows up in show rates months later. Three policies help:
- Say exactly what will be paid, how and when. “£70 by bank transfer within three working days of the session” is a commitment. “An incentive will be provided” is a reason to stay at home.
- Pay standbys and sent-home participants in full. They did what you asked. Paying them is good practice, protects your reputation and makes people willing to be standbys next time. The commercial question is whether the client pays, which should be settled in the quote.
- Record no-shows against the participant. Not paying someone who did not attend is normal. Recording it is what matters, so repeated no-shows can be taken into account the next time a recruiter wants to use the same person.
Our participant incentives guide covers payment methods, reconciliation and rebilling in detail, including why incentives should be released centrally once attendance is confirmed rather than paid in cash by recruiters.
Online qual: when the wrong person turns up
Moving qualitative work online improved show rates for many agencies. No travel, no parking and a session that fits around work have all helped. It also opened a door that in-person research kept closed: the participant you recruited and the person on the screen may not be the same person, or may not be a genuine member of your target audience at all.
This is no longer a fringe concern. At an MRS Operations Network event in July 2025, a working group reported that 67% of qualitative researchers had experienced participant fraud in their projects over the previous year. Academic researchers have documented the same pattern. A widely cited 2023 letter in Health Expectations described suspected imposters across five online interview studies, and more recent papers in Qualitative Health Research and the International Journal of Qualitative Methods document the problem continuing, including interviews where participants were suspected of using AI to help answer questions.
For a fieldwork agency, an imposter is worse than a no-show. The seat looks full, the incentive is paid, and the client receives data that may quietly mislead them.
Signals worth acting on
No single signal proves fraud. The MRS discussion was careful to separate fraudulent participants from professional participants and from people who simply prefer to keep their camera off. The point of a signal is to trigger a proportionate check, not an accusation.
| Stage | Signals | Proportionate check |
|---|---|---|
| Sign-up | Many responses within minutes of an advert, similar email formats or providers, wording copied from the advert, a referral source you did not use | Cross-identity matching against your database, and a pause on any channel producing clusters |
| Screening | Answers that fit the specification too neatly, unusually fast or slow completion, inconsistent answers to related questions | Blinded screeners with mutually exclusive options and knowledge questions a genuine category user would answer instantly |
| Verification | Reluctance to give a phone number, a work email for B2B, or proof of product ownership | Proof of qualification requested promptly after recruitment, such as a receipt, product photo or professional profile |
| Session | Camera off throughout, delayed or oddly polished answers, vague detail on lived experience, a different voice or face from the screening call | A brief camera-on check-in at the start, re-screening questions, and a structured moderator flag |
| Payment | Pressure for gift cards or a particular wallet, payment details that match another participant | Pay only to verified details, and hold payment until any open flag is resolved |
The QRCA’s guidance on preventing participant fraud in qualitative research is a good practical source for screener design, and MRS members can access its Safeguarding Qualitative Research guidance through the Global Data Quality initiative.
Be careful with biometric verification
Some verification tools compare a participant’s face on camera with a photo ID. That is automated biometric recognition, and the ICO’s biometric data guidance treats biometric data used to uniquely identify someone as special category data. In practice that usually means explicit consent, a genuine alternative for people who decline, and a data protection impact assessment before you start. A moderator looking at a participant on camera is not biometric processing. Software matching their face to a document is. If you are considering it, take advice, and check the guidance is current: the ICO notes it is under review following the Data (Use and Access) Act 2025.
"The goal is not to catch every fraudster in the session. It is to make your process more trouble than it is worth, so they go elsewhere."
The show-rate dashboard
Show rate on its own hides too much. A job can hit a 90% show rate and still disappoint the client if two of the attendees were removed for quality. These are the measures we would track, per job and over time:
| Measure | How to calculate it | What it tells you |
|---|---|---|
| Show rate | Attended divided by confirmed, by audience, method and recruiter | The input to every over-recruitment decision |
| Confirmation completion | Participants who confirmed at each step of the sequence divided by those recruited | Where commitment drops away, and how early you can see it |
| Lead time | Days between recruitment and session | Whether recruiting too early is driving no-shows |
| Replacement rate | Late replacements divided by seats sold | Hidden recruiter hours that rarely appear in the quote |
| Sent-home rate | Participants paid but not seated divided by seats sold | The cost side of over-recruitment |
| Removal rate | Attendees removed or excluded from analysis for quality or identity reasons | Integrity risk, especially in online work |
| Effective fill rate | Valid participants seated divided by seats sold | The number the client actually experiences |
Effective fill rate is the one to put in front of the leadership team. It combines attendance, quality and identity into one figure and connects directly to the cost per attended participant measure in our recruiter scorecard.
Want to know your real show rate?
We will take a sample of recent jobs and calculate show rates, sent-home costs and effective fill rate by audience and recruiter, using the records you already have. You will see whether your over-recruitment rules are protecting margin or quietly eroding it, whether or not the answer is software.
Book a free discovery call →What a well-run attendance process looks like
Whether you configure an existing field system or build around your own operation, these are the capabilities that move show rates and protect data quality:
- Over-recruitment set from history. Suggested recruit numbers per group based on your actual show rates for that audience and method, with the expected sent-home cost shown before the quote is issued.
- An automated confirmation sequence. Confirmations, reconfirmations and day-of messages sent on schedule, written as service messages, with replies captured against the participant.
- A live risk view. Participants who have not confirmed by each checkpoint flagged to the field manager, group by group, while there is still time to replace them.
- Outcomes recorded once. Attended, no-show, sent home and removed, recorded on the night and flowing straight into incentives, recruiter pay and the participant’s history.
- Identity checks built into the job. Cross-identity matching at sign-up, proof-of-qualification requests, moderator flags in the session and payment held until flags are cleared.
- Reporting that reaches the quote. Show rate, sent-home cost and effective fill rate by audience feeding back into the next costing.
That last point is where the value compounds. Every job you run produces the show-rate data you need to quote the next one accurately, which is the approach we described in quoting from your own job history. Generic professional services platforms do not model participants, confirmations or sent-home costs at all, which is why we argued in our Accelo vs Scoro comparison that the fieldwork layer is where custom software earns its place.
Where to start
In order of effort:
- Calculate your show rate. For the last six months, by audience type, method and recruiter. Even a rough figure is better than the rule of thumb.
- Check your over-recruitment against the table. Find the jobs where “ten for eight” was not enough, and the ones where it was more than you needed.
- Write down your confirmation sequence. One page covering each step, who owns it and what triggers a call. Make the 48-hour personal call non-negotiable.
- Tidy your reminder wording. Strip any promotional content from session reminders so they stay service messages.
- Agree an online fraud protocol. Screener design, proof of qualification, a camera-on check-in and a moderator flag, applied proportionately and explained to clients up front.
- Connect attendance to the rest of the job. When volume justifies it, record outcomes once and let them drive incentives, recruiter pay, participant history and the next quote.
If attendance tracking is one of several processes still held together by spreadsheets, our practical workflow for replacing spreadsheets in a research agency explains how to sequence the change without disrupting live jobs.
We build these systems in Laravel for research, fieldwork and professional services firms, including resource planning for insight consultancies, and we are currently building an operations platform with a UK fieldwork agency covering quoting, recruiters, quotas, incentives and job profitability. Arc Digital is based in Leeds and works with agencies across the UK.
Frequently asked questions
What is a normal no-show rate for qualitative research?
Published figures for moderated research typically fall between 5% and 20%, with large UX research datasets reporting around 9%. Consumer focus groups, hard-to-reach B2B audiences and evening sessions can all run higher. Your own history, broken down by audience and method, is a more reliable guide than any published benchmark.
How many people should you over-recruit for a focus group?
It depends on your show rate and how important a full group is. At a 90% show rate, recruiting ten for eight seats fills the group about 93% of the time. At an 80% show rate, the same rule fills it only about 68% of the time, and recruiting twelve is closer to the mark. Set the number per job from your own data, and agree the cost of over-recruitment in the quote.
Do text message reminders reduce no-shows?
The best evidence comes from healthcare, where a Cochrane review found attendance rose from 67.8% with no reminder to 78.6% with text reminders, similar to phone call reminders but at a lower cost per attendance. In research fieldwork, automated texts work best combined with a personal call around 48 hours before the session.
Are participant reminder texts direct marketing under PECR?
A neutral reminder about a session the participant has agreed to attend is a service message, not direct marketing. If the message also promotes other studies or panel membership, it is likely to become direct marketing, which needs consent for texts and emails. Keep reminders strictly about the session.
Should participants who are sent home be paid?
Yes. Participants who arrive as asked but are not needed should be paid in full. It is fair, protects your reputation and keeps people willing to act as standbys. Agree with the client before fieldwork whether they will cover over-recruitment.
How can you spot imposter participants in online interviews?
Look for clusters of sign-ups with similar details, screener answers that fit the specification too neatly, reluctance to verify a phone number, work email or product ownership, and in the session, cameras off throughout or polished but vague answers. No single signal proves fraud, so use signals to trigger proportionate checks rather than accusations.
Can we use facial recognition to verify participants?
You can, but it is biometric processing. The ICO treats biometric data used to uniquely identify someone as special category data, which usually means explicit consent, an alternative for people who decline and a data protection impact assessment. A moderator simply seeing a participant on camera is not biometric processing.
If you run a research or fieldwork agency and your show rates live in recruiters’ heads and the field manager’s inbox, we would be glad to take a look. We will work out what no-shows and over-recruitment really cost you and tell you plainly what to fix first, whether or not the answer is software.