You can design a flawless study and still learn nothing, because the wrong eight people sat in front of you. Recruiting research participants is the least glamorous part of the craft and one of the most consequential: who you talk to sets a hard ceiling on what you can find out.
It is also where studies quietly go wrong, weeks before the first session. Recruiting takes longer than anyone budgets, screeners wave through people they should have filtered, and no-shows eat the sample. Here is how to recruit research participants deliberately, so the people in your sessions are the people your study actually needs.
Recruiting research participants: the four main sources
- Your own users. Recruit from your CRM, in-product prompts, or an opted-in research panel of customers. Cheapest and most relevant, but beware the fan effect: people who accept invitations skew towards people who like you. You also need consent and contact-permission hygiene sorted first.
- Panels and agencies. Fast, broad reach, and the only realistic route to some audiences. The trade-off is distance from real usage and a higher risk of professional participants, more on them below.
- Intercepts. Catch people mid-behaviour, with a prompt inside the product or a polite approach in the field. Wonderful for context, biased towards whoever happens to be there.
- Communities and networks. Specialist forums, professional groups, and social channels reach niche audiences that money struggles to find. Watch for snowball bias: friends of friends tend to resemble each other.
Most teams end up blending sources, and that is healthy. Each carries a different bias, and the biases partially cancel.
Screeners that actually screen
A screener is a short survey whose job is to let the right people in and keep the wrong people out. Many screeners do neither, because they announce the right answers.
- Screen on behaviour, not self-description. "Which of these have you done in the past month?" beats "Are you an experienced online shopper?" People are generous with themselves.
- Never telegraph the study. If the screener says you are looking for people who meal-plan weekly, you will meet a lot of sudden meal-planners. Describe the topic loosely and let the criteria work quietly.
- Include disqualifiers and a red herring. Offer plausible-sounding options that do not exist, a made-up brand or tool works well. Anyone who claims it is ticking every box.
- Ask one open question. A sentence or two of free text filters out low-effort respondents and gives you an early feel for how the person communicates.
- Keep it short. A screener is still a survey, so the fundamentals of survey design apply: neutral wording, one thing per question, no double-barrels.
The professional participant problem
On every panel there are people who do studies for a living. They are articulate, punctual, agreeable, and unrepresentative, and they have learned exactly what screeners want to hear. The warning signs: a suspiciously perfect match to every criterion, polished but generic answers, and enthusiasm untethered to any specific experience.
Defend with the behavioural screeners above, the red herring, and a direct question about how many studies they have taken part in recently. In the session itself, ask people to show rather than tell: real screens, real receipts, the actual app on their own phone. Keep an exclusion list so the same faces do not cycle back through your studies.
Incentives, no-shows, and the over-recruiting maths
Pay people fairly for their time. Specialist audiences, clinicians, engineers, procurement leads, genuinely cost more, and lowballing them selects for the unrepresentative few with time to spare. Be careful at the other extreme too: an incentive far above someone's normal hourly rate invites misrepresentation at the screener. Vouchers, cash, or a charity donation all work; what matters is that the amount respects the effort.
Then plan for entropy. Even with confirmations, some participants will not show, so over-recruit by roughly one in five, and consider a floater: a backup participant paid to be on standby on lab days. Confirm twice, once at booking and once the day before, and make rescheduling easy, because a rescheduled session is data and a no-show is a hole. How many participants you need in the first place is its own question, covered in our guide to qualitative sample sizes.
The ethics floor
None of this works without basic respect for the people giving you their time.
- Get informed consent: what the research involves, what will be recorded, how the data will be used and stored
- Make clear they can stop at any point, without losing the incentive
- Keep personal details separate from research data, and delete them on a schedule
- Pay people even when a session fails for reasons that are not their fault
These are not bureaucratic niceties. Participants who feel safe and respected talk more honestly, and your future recruitment depends on your past behaviour.
Good recruitment is invisible in the final report, which is exactly why it gets underrated. Nobody presents a slide titled "the screener worked". But every insight downstream stands on it, so treat recruiting as research infrastructure rather than admin: budget real time for it, write screeners that filter on behaviour, expect no-shows, and pay people properly. The right participants make an average researcher look insightful. The wrong ones make a brilliant one look lost.
People of Research is a community and newsletter for UX researchers, market researchers, and anyone whose next study depends on finding the right people to talk to. Join us and get the good stuff in your inbox.