How to Recruit Participants for User Research

On this page

• Why Recruitment Determines Research Quality
• Step 1 — Define your participant profile before you write a single question
• Step 2 — Write a screener survey that filters without telegraphing
• Step 3 — Choose the right sourcing channel for your study
• Step 4 — Build and maintain a research participant panel
• Step 5 — Avoid the five most common sampling biases
• Turning recruited participants into research insight
• Frequently asked questions

Why Recruitment Determines Research Quality

Poor recruitment is the most common way a research project fails — and the least discussed. You can write a brilliant discussion guide, run sessions with precision, and synthesise findings rigorously, yet still walk away with conclusions that do not reflect your actual users. The reason is almost always who was in the room.

Recruitment and research method are separate decisions. Recruitment answers who you talk to; method answers how you talk to them. This guide focuses entirely on the first question, because it is the prerequisite to everything else. Before you think about conducting user interviews or running a round of guerrilla usability testing, you need confidence that the people you are recruiting actually represent the behaviour or context you are trying to understand.

Sampling bias — systematically over- or under-representing certain types of participant — is the specific risk this guide helps you manage. It rarely announces itself. The findings simply look cleaner and more consistent than reality warrants.

---

Step 1 — Define your participant profile before you write a single question

Start with your research objectives, not a demographic wish list. The question to answer is: what behaviour, context, or decision are you trying to understand? That answer determines who belongs in your study.

Once you have your objective, separate your criteria into two tiers.

Primary criteria are must-haves that directly affect the validity of your findings. Common examples include: job role or function, frequency of using a relevant product or workflow, domain experience, and — for B2B studies — whether the participant has purchase or procurement authority. If a person does not meet your primary criteria, their data will not answer your research question, regardless of how well the session goes.

Secondary criteria are useful but not essential. They might include company size, geographic location, or tenure in a role. These help you build a well-rounded participant set, but missing one does not invalidate a session.

The practical risk at this stage is over-specification. Stacking too many primary criteria — requiring a specific seniority level, a precise usage frequency, a particular industry vertical, and a narrow company-size band all at once — slows recruitment and increases the chance that the few participants who qualify are unrepresentative outliers. If you find yourself with seven or more primary criteria, challenge each one: does this criterion genuinely affect whether this person can answer my research question, or is it a preference?

The output of this step should be a single-page participant profile your whole team can review before the screener is drafted. One page forces prioritisation and gives recruiters, panel managers, and stakeholders a shared reference point so that compromises on criteria are made deliberately rather than quietly.

---

Step 2 — Write a screener survey that filters without telegraphing

A screener survey has one job: qualify the right participants and disqualify the wrong ones, in under five minutes of the respondent’s time. It is not a research instrument — it is a filter.

Structure matters. Open with behaviour-based questions before you reveal the study topic. Asking someone how often they perform a relevant task before mentioning what the study is about reduces social-desirability bias — the tendency for respondents to shape their answers toward what they think you want to hear.

Use multiple-choice answers with deliberate distractors rather than binary yes/no options. If you are screening for people who use a particular type of software daily, include plausible-but-disqualifying alternatives in your frequency scale. Binary questions invite straight-lining; a well-constructed set of options requires a respondent to actually consider their answer.

Include at least one hard disqualifier. Standard examples: works at a direct competitor, has participated in paid research in the last six months, or works in market research or UX themselves. These questions protect against professional respondents — people who have learned to game screeners — and remove obvious conflicts of interest.

Keep the screener to 8–12 questions. Completion rates fall sharply beyond that. Every question you add should be justified by a specific disqualification or qualification decision it enables. If a question is merely interesting, cut it.

End with availability and consent to contact. This step is frequently skipped in the rush to get recruitment moving. Skipping it creates legal and practical problems. Confirm preferred contact method and a rough availability window before you close the screener.

Sample screener question formats to consider:

• Behavioural frequency scale: “How often do you [relevant task]?” — Never / Less than once a month / 1–3 times a month / Weekly or more
• Open task description: “In a few sentences, describe the last time you [relevant scenario].” — Flags low-effort respondents and provides useful context.
• Role taxonomy: A list of job functions with a plausible mix of qualifying and non-qualifying options.

---

Step 3 — Choose the right sourcing channel for your study

There is no single best channel for recruiting research participants. The right choice depends on your study type, timeline, and budget.

Your own customer database is the fastest and cheapest channel, and participants arrive with context you already understand. The risks are over-researching the same cohort — which fatigues your best advocates and skews your findings toward their perspective — and confirmation bias, since existing customers have already accepted your product’s current direction. Reserve this channel for evaluative studies on existing features, and rotate participants carefully.

Third-party recruitment panels are the most reliable route to hard-to-reach B2B personas or consumer segments you do not yet serve. A specialist panel provider can reach senior procurement professionals, niche technical roles, or demographic groups your own database cannot. The trade-offs are cost and lead time: expect to pay more per participant and allow more time for fulfilment, particularly for senior or specialist profiles.

Social and community channels — LinkedIn, professional Slack communities, industry forums, relevant subreddits — work well for niche professional segments. Posts that make the research sound prestigious or the topic exciting will self-select for what practitioners sometimes call “research enthusiasts”: people who are unusually engaged with the subject matter and may not reflect typical users. Keep outreach neutral and specific about who you are looking for.

Intercept and guerrilla recruiting is appropriate for early generative work or when timelines are extremely tight. See our guide to guerrilla usability testing for fast intercept recruiting for execution detail.

Study type	Timeline	Budget	Recommended channel
Evaluative (existing product)	Flexible	Low	Own database
Generative (new segment)	Flexible	Medium–high	Panel provider
Niche B2B persona	Extended	Medium	LinkedIn / community
Early generative	Tight	Low	Intercept / guerrilla

One important caution: mixing channels within a single study can introduce comparability problems. Participants sourced from a warm customer database may have systematically different attitudes to your product than cold-recruited panel participants. If you mix, track the source and treat it as a variable during analysis.

---

Step 4 — Build and maintain a research participant panel

A panel is a pre-screened, opted-in pool of people you can re-contact across multiple studies. Over time, it significantly reduces per-study recruitment cost and turnaround.

Seeding a panel is most effective at moments of natural engagement: post-purchase surveys, in-app prompts after a key action, or follow-ups after a support interaction. These touchpoints reach people while their experience of your product is fresh and tend to attract participants who are genuinely motivated to give feedback.

Panel hygiene is non-negotiable. Track participation history for every contact. Enforce cooling-off periods — typically eight to twelve weeks between studies for any individual participant — to prevent a small set of highly engaged people from dominating your research. Refresh the panel regularly by retiring inactive contacts and adding new ones. A panel that is not actively maintained drifts toward a pool of research enthusiasts, reintroducing the very biases you were trying to avoid.

Segment your panel along your key participant dimensions — role, product usage tier, customer tenure, or whatever is most relevant to your research programme. Pre-segmented lists mean you can pull a targeted group for a new study without re-screening the entire panel from scratch.

Consent and data handling require careful attention. Under UK GDPR, you need a lawful basis for holding contact data and must honour opt-out requests promptly. For a thorough treatment, see our research operations hub.

Tooling: a well-maintained spreadsheet is sufficient for panels under roughly 200 participants. Beyond that, the complexity of tracking participation history, consent status, and segmentation typically justifies a dedicated tool — either a purpose-built research ops platform or a CRM with a consistent tagging system.

---

Step 5 — Avoid the five most common sampling biases

Sampling bias does not require carelessness to occur. Most of these patterns emerge from reasonable-sounding decisions that compound over time.

Convenience bias is recruiting whoever is easiest to reach: colleagues, internal employees, or a handful of vocal existing customers. These participants are accessible precisely because they are already embedded in your world, which means their mental models, vocabulary, and expectations are already calibrated to your product. They will not surface the friction a genuine new user would.

Mitigation: Explicitly exclude internal employees from most studies. Set a target proportion of participants who have had no prior contact with your team.

Survivorship bias is only recruiting people who are still using your product. Churned users and non-adopters often hold the most diagnostic signal about where a product fails to serve its intended audience. They are harder to reach but worth the effort for any research that touches acquisition or retention.

Mitigation: Include a churned-user segment in any study concerned with why people leave or do not convert.

Screener coaching and professional respondents occur when participants learn, through over-specific screeners, exactly which answers get them into a study. Signs include implausibly uniform screener responses, unusually polished recall of specific product features, or a participant who seems overly familiar with research conventions.

Mitigation: Use indirect behavioural questions rather than direct qualification statements. Vary distractor options across recruitment waves.

Homogeneity bias is unintentionally filtering out edge cases. Tight criteria produce participants who are similar to one another, and similar participants produce tidy findings. The messiest, most unusual participants often reveal the most actionable insights — they have found workarounds, encountered genuine failures, or arrived at the product through paths you did not anticipate.

Mitigation: Deliberately recruit one or two participants who meet primary criteria but represent less common contexts or usage patterns.

Recency bias in panels is returning to the same small group of engaged, responsive participants across every study because they are reliable. This is a form of convenience bias specific to established panels.

Mitigation: Set a hard limit on the number of studies any participant can join per quarter, and actively recruit new panel members on a rolling basis.

---

Turning recruited participants into research insight

Recruitment ends when a participant shows up. Research begins.

For moderated qualitative sessions, conducting user interviews is the natural next step — covering discussion guides, facilitation, and note-taking practice.

One practical note: the screener responses you collected during recruitment are themselves useful analytical context. Save them alongside your session notes. Knowing that a participant described their usage frequency one way in the screener and another way in the session is a data point, not an inconvenience.

---

Frequently asked questions

How many participants do I need for user research?

For moderated qualitative studies — interviews and usability tests — 5–8 participants per distinct user segment is the range most commonly associated with thematic saturation. Beyond that point, each additional session tends to surface fewer genuinely new insights. For unmoderated or survey-based studies, larger samples are required; 30 or more responses is a reasonable floor for directional quantitative signal, though this depends on the expected effect size and how you intend to segment the data. The right sample size depends on study type, the number of distinct segments you are investigating, and the confidence level your decisions require.

What should I include in a screener survey for user research?

A well-constructed screener should include: behavioural frequency questions tied directly to your research topic, role or context qualifiers that map to your primary criteria, at least one deliberate disqualifier (competitor employment, recent research participation, market research profession), an open-text task-description question to flag low-effort respondents, and a consent-to-contact and availability check at the end. Keep the total to 8–12 questions. Anything beyond that reduces completion rates without meaningfully improving qualification quality.

How do I recruit hard-to-reach B2B participants?

For niche B2B personas, the most reliable approaches are: LinkedIn outreach with a personalised message that clearly states the time commitment and what the participant will gain; targeted posts in professional Slack communities or industry forums relevant to the role; referrals through existing participants; and specialist B2B panel providers for senior or highly specific profiles. Whichever channel you use, the value proposition to the participant needs to be concrete. Senior professionals respond to relevance and respect for their time.

How do I avoid recruiting the same people repeatedly?

Track participation history rigorously — whether in a spreadsheet or a dedicated tool — and enforce a cooling-off period of at least eight to twelve weeks between studies for any individual. Segment your outreach so that each study draws from a different portion of your panel. Introduce new participants continuously rather than waiting until the panel feels stale. If you are using a third-party panel provider, specify that you require participants who have not taken part in a study on a related topic within a defined period.