How to Recruit and Pay Research Participants (2026 Guide)

You need 25 interviewees by Friday. Your customer list has been emailed four times this quarter, and the reply rate on the last invite was half of what it was in the spring. This guide gives you an eight-step method to recruit and pay research participants with nothing more than a spreadsheet, a form, a calendar link, and email. It also shows where BuildBetter takes over the parts that break once you run studies every month: finding the right customers among thousands of calls and tickets, running AI-moderated interviews, and paying people on time. The steps are written for product managers, UX researchers, CS leads, and founders at B2B companies who have no research ops team behind them.

What Recruiting and Paying Research Participants Actually Means

Research participant recruitment is the process of finding people who match a study's criteria, confirming they qualify through a screener, scheduling them or sending them the study, and compensating them for their time. Participants come from two sources: your own customers (first-party) or strangers supplied by a research panel (third-party). Most B2B teams default to the first source and only add the second when the customer list runs dry or the question requires outside voices.

Paying research participants means offering an incentive, usually a digital gift card or cash equivalent, sized to the time you ask for and how scarce the person's profile is. You state the amount up front and deliver it promptly after the study is complete. A 15-minute survey for a broad consumer audience and a 60-minute interview with a VP of Finance are priced very differently, because the second person's time is scarcer and more expensive.

The method below covers three study formats:

• Moderated interviews: a live session with a researcher, usually 30 to 60 minutes.
• Unmoderated or AI-moderated interviews: participants answer on their own schedule, with an AI interviewer asking follow-up questions.
• Surveys: structured and open-ended questions sent by email, link, or in-app.

The recruiting mechanics are the same across all three. What changes is incentive size, the over-recruit buffer, and fraud controls. Async formats remove scheduling no-shows but increase the need for answer-quality checks, because nobody is watching the participant in real time.

Why Recruiting Breaks: The Concrete Failure Modes

User research recruiting usually fails for predictable, fixable reasons, and most of them come down to who gets invited and how the process is tracked. These are the patterns that show up again and again in B2B teams without dedicated research ops:

• The overfished list. The same 40 friendly customers answer every study. You keep hearing the same opinions, and eventually those customers stop replying.
• Selection bias toward happy users. CSM-sourced invites skew toward accounts with good relationships. Churn risks, detractors, and non-users get left out, which is exactly the group you most need to hear from.
• Wrong people pass the screener. A leading question such as “Do you manage a product team?” lets anyone qualify. Professional respondents on panels learn to answer whatever gets them in.
• No-shows wreck the timeline. Book exactly the number you need and every absence pushes the study past its deadline.
• Payment friction. A CSM buying gift cards on a personal card, or a reimbursement spreadsheet nobody owns, makes payouts slow. Slow payouts damage trust and future participation.
• Customer-company rules. Some employers prohibit staff from accepting gifts. Public-sector and regulated buyers often do too. Offering cash to them can create an awkward conversation or a compliance issue.
• Consent and data gaps. Recordings with no documented consent, PII sitting in shared drives, and panel data kept longer than its stated purpose requires. Under the GDPR, that last one breaks the storage limitation principle.

Recruiting is also a funnel with four leaky stages, not just a show-rate problem: candidates × invite acceptance (or screener pass) rate × show rate × quality pass rate = usable sessions. Teams usually plan for show rate and forget the other three. A 25% acceptance rate multiplied by an 80% show rate means you need about five invites for every completed customer interview.

The Method: 8 Steps to Recruit and Pay Participants This Week

Everything in this method works with a spreadsheet, a calendar link, an email account, and a free form tool. No paid software is required.

Step 1. Decide: own customers, panel, or both

Pick the source based on the question you need answered, not on what is easiest to reach.

Mixed studies are often the right answer. They keep you out of an echo chamber.

Step 2. Write the participant profile, then hard criteria

Describe the ideal participant in one paragraph. Then convert it into 3 to 6 hard criteria (role, company size, a specific behavior in the last 90 days, tool used) and 2 to 3 nice-to-haves. Write down who you explicitly exclude: competitors' employees, anyone interviewed in the last 90 days, and internal staff.

Step 3. Source your own customers by signals, not friendliness

Pull candidates from what customers actually said or did: support tickets on the topic, call notes mentioning the problem, usage of the relevant feature, and NPS comments. Log each candidate in a spreadsheet with these columns:

• Name, account, and evidence (why they qualify)
• Last contacted date
• Invite sent, response, scheduled, attended, paid

For small numbers, invite by personal email. For larger numbers, send a link to a signup form plus a calendar booking page. Every invite should include the specific reason you are asking this person, the time required, the incentive, the date range, and a one-click booking link. Send it from a named person, never a no-reply address.

Step 4. For strangers, use a panel that sends participants to a link

The main research panel options:

• Prolific: fast, with broad consumer and some professional audiences.
• Respondent and User Interviews: strong for B2B professionals and scheduled interviews.
• CloudResearch Connect: academic-grade online samples.
• Dynata: large-scale survey sample.

The mechanic is the same across all of them. You define criteria, paste your study or screener link, set the incentive, and the panel routes qualified people to you. Panel choice affects data quality: a 2023 head-to-head comparison in PLOS ONE (Douglas, Ewell & Brauer) found Prolific and CloudResearch produced higher attention, comprehension, and honesty scores than several other online sample sources. Check each panel's current fee schedule before budgeting.

Step 5. Write a screener survey that can't be gamed

Good screener questions ask about behavior, not identity. Instead of “Do you approve purchases?”, ask “In the last 30 days, which of the following did you do? (select all)” and mix purchase approval among 6 to 8 plausible activities. For tool usage, include distractor options, including one fake tool name. Add one open-text question you read before approving anyone. Keep the screener to 5 to 8 questions, and never reveal which answer qualifies.

Step 6. Set fair incentives and pay by gift card

Size research participant incentives by time and by profile scarcity. The values below are starting assumptions to adjust, not market data.

A 30-minute interview with a VP of Finance is not half of a 60-minute consumer interview. The binding constraint is the opportunity cost of that person's time. Ethics guidance also separates fair compensation from undue influence: an incentive so large that people lie to qualify drives screener fraud.

Rules for participant compensation:

• State the amount up front, in the invite.
• Pay within 1 to 3 business days.
• Offer a charity-donation alternative for people who can't accept gifts.
• Decide your policy for no-shows and late reschedules before launch.

Gift cards count as cash equivalents for US tax purposes. Total payments to one person above the IRS reporting threshold can trigger a 1099-MISC. That threshold was historically $600 and rises to $2,000 for payments made from 2026 under the 2025 tax law. Confirm with a tax advisor; this mostly matters for repeat participants.

Step 7. Over-recruit and track show rates

Sessions to book = target completes ÷ expected show rate, always rounded up. For 10 completes at an 80% show rate, book 13. Nielsen Norman Group once reported an average no-show rate of about 11% for usability participants, but treat that as a historical benchmark. Practitioner estimates for scheduled B2B interviews today often run 10 to 20%, and own-customer no-shows are frequently worse than paid-panel no-shows because customers have less money at stake. Start from an assumption, then replace it with your measured rate after the first study.

Send reminders 24 hours and 1 hour before each session. Track show rate separately for customers and panel participants.

Step 8. Protect consent and data

Get written informed consent that covers the recording, how it will be used, who sees it, and how long it is kept. Let participants withdraw at any time. Consent to process data and consent to record are separate: several US states, including California, Florida, Illinois, Washington, Pennsylvania, and Maryland, require all parties to agree to a recording. Getting explicit recording consent from everyone removes the jurisdiction question.

Under the GDPR, research data needs a lawful basis (Article 6) and must respect purpose limitation, data minimisation, and storage limitation (Article 5). Store PII separately from findings, delete recordings on a set schedule, and keep a last-contacted date so no customer is asked more than once per quarter.

Worked Example: A Mixed Study of 10 Customers and 15 Panel Participants

This example shows the full method applied to a real-world constraint: 25 thirty-minute interviews, starting Monday, due Friday. A B2B PM needs to understand how teams handle approval workflows.

The decision. 10 existing customers explain current behavior inside the product. 15 panel participants who use other tools or homegrown processes keep the study from becoming an echo chamber.

Customer side. The PM pulls 60 candidates from tickets and call notes mentioning approvals. Eight were contacted within 90 days and are excluded, leaving 52 invited by personal email from the PM. At an assumed 25% acceptance rate, 13 book. At an assumed 80% show rate, about 10 complete. Incentive: a $50 gift card, with a charity option.

Panel side. Criteria: ops or finance managers at companies with 50 to 500 employees who approve purchases weekly. The 6-question screener includes a distractor tool list and one open-text question. At an assumed 85% show rate, the PM books 18 to land 15. Incentive: $100 for 30 minutes, an assumption that reflects a professional B2B profile.

No-show policy: $0 for no-shows, full incentive for reschedules requested 24 or more hours ahead.

Timeline. Monday: customer emails sent and panel study launched. Tuesday to Thursday: sessions. Payouts go out within 48 hours of each session. Friday: synthesis.

What changed mid-study. Two panel participants gave vague, copy-paste answers to the open-text question and were declined. The PM backfilled two more from the panel instead of lowering the bar. That cost a few hours, not the integrity of the data.

Lesson. Record actual acceptance and show rates in the spreadsheet. The next study's math should use real numbers, not the assumptions above.

Common Recruiting and Incentive Mistakes

Most recruiting problems are self-inflicted, and each one has a specific fix.

• Recruiting only through CSMs. This over-samples happy accounts. Source by signals and evidence, and deliberately include detractors and churned users.
• Underpaying professionals, then blaming the panel. Slow fills, late cancellations, and low-effort answers often mean the incentive is below market. Before switching panels, raise the incentive 25 to 50% on a small batch and compare fill speed.
• No over-recruit buffer. Booking exactly what you need guarantees a missed deadline.
• Screeners that telegraph the right answer. Professional respondents read the question and pick whatever qualifies.
• Paying late or inconsistently. One slow payout undoes the goodwill of the interview.
• Ignoring contact fatigue. Without a last-contacted field, the same customers get every request.
• No charity alternative. Offering gift cards to people whose employers prohibit them puts them in an awkward spot.
• Changing questions mid-study. Early and late participants answer different things, and the data becomes incomparable.

No tool fixes a weak research question, an under-sized incentive, a missing contact-fatigue policy, or a customer's gift-acceptance rules. That includes BuildBetter. Panel fraud screening ultimately depends on your screener and the panel's own quality controls.

When You Need Tooling (and When You Don't)

For 3 to 5 friendly customers, a personal email from you beats any tool. Don't buy software for that.

The spreadsheet becomes the bottleneck once you run more than one or two studies a month, mix customers and panels in one study, pay out more than a couple dozen gift cards, or need to find the right customers among thousands of calls and tickets. Contact dates drift, payouts lag, and the evidence for why a person was invited gets lost.

The second break point is analysis. Past a few dozen interviews or hundreds of open-ended answers, manual synthesis slows the decision you ran the study to make. If that's where you are, see this guide to tools to analyze thousands of open-ended survey responses.

Tools for Recruiting, Running, and Paying Participants

The right stack depends on volume: one platform that targets, runs, and pays for studies once you're doing this monthly, or a free stack when you're not.

1. BuildBetter — Best for Signal-Led Recruiting and AI-Moderated Studies

BuildBetter runs the full method from this guide in one place: AI Surveys, AI-moderated interviews, and Synthetic Personas, with gift-card rewards built in. You can recruit your own customers by email or link, or bring participants from any panel that sends people to a link, including Prolific, Respondent, User Interviews, CloudResearch Connect, and Dynata. Gift-card rewards are paid through Tremendous and set per study.

Signal-led targeting

Describe who you want in plain language, for example “admins who complained about approval delays in the last 60 days.” BuildBetter turns that into editable filters over what customers said on calls, tickets, and Slack, through Signals. It shows each matched person with the evidence that matched them, and you approve the list before anything is sent. This automates the spreadsheet in Step 3 and gets you past the friendly-40 problem.

Interviews and surveys

AI-moderated interviews run by video or browser audio. You set the questions, time limit, language, and probing depth. Surveys go out by branded email, public link, or a verified in-app widget. AI follow-ups ask one clarifying question only when an answer is thin. Questions are versioned and immutable once sent, which prevents the mid-study drift mistake. Every completed response becomes signals with citations back to the exact answer.

Cost transparency

Credits are priced at 1 credit = $0.001. An invitation costs 4 credits, a completed response 8, an AI follow-up 1, and each signal 4. A completed email response with two follow-ups and three signals is about 26 credits, or roughly $0.026. Activation shows the estimated maximum credit exposure before sending, and previews and test responses are never billed. Seats are unlimited, and the platform is SOC 2 Type II, HIPAA, and GDPR compliant.

Synthetic Personas

Use Synthetic Personas for a first read or a pressure test before you recruit, such as sharpening screener and interview questions. They are built from your real signals, with citations. They are not a replacement for real participants. For context on AI-moderated research at scale, read how ustwo tested AI-moderated interviews and synthetic personas and how Anthropic ran 1,250 AI-moderated interviews.

“I've never used a tool that actually connects the dots across all my platforms. BuildBetter does that, and it makes me so happy.” — Agnes Abel, Typeform

2–4. Panels, Payouts, and the No-Budget Stack

2. Research panels

Prolific, Respondent, User Interviews, CloudResearch Connect, and Dynata are participant sources. Choose by audience type: Prolific for consumer and general professional audiences, Respondent and User Interviews for B2B professionals, CloudResearch Connect for academic-grade samples, and Dynata for large survey samples.

3. Tremendous

If you pay participants outside a research platform, Tremendous sends digital gift cards and choice-of-reward links by email, with a charity option. That removes personal-card purchases and reimbursement spreadsheets.

4. The no-budget stack

A spreadsheet, a free form tool for screeners, a calendar booking link, and email. That is enough to run every step in the method above.

FAQ: Recruiting and Paying Research Participants

How many people should I recruit for user interviews?

Decide how many completed sessions you need, then divide by your expected show rate. For 10 completed interviews at an 80% show rate, book 13. In a single, well-defined segment, themes usually start repeating between about 9 and 17 interviews (Hennink & Kaiser, 2022); an earlier study by Guest, Bunce & Johnson (2006) found saturation within 12. If you need more, split them across distinct segments. Nielsen's “test with 5 users” rule applies to iterative usability testing of one group, not discovery interviews.

How much should I pay research participants?

Pay in proportion to the time you ask for and how hard the person's profile is to find. A short survey for a broad consumer audience earns a small incentive, roughly in line with panel minimum hourly rates. A 60-minute interview with a senior B2B specialist earns the most. Treat any table as a starting assumption, state the amount up front, and pay within 1 to 3 business days.

Should I recruit my own customers or use a research panel?

Use your own customers to study how people use your product, why they churn, or how a change affects existing workflows. Use a panel to hear from non-customers, prospects, or competitor users, or to test pricing perception without relationship bias. Combine both when you need to avoid an echo chamber.

What is the best way to pay research participants?

Send digital gift cards or a choice-of-reward link by email within 1 to 3 business days of completion. Offer a charity-donation alternative for people whose employers don't allow gifts. Avoid manual payouts on personal cards: they are slow, hard to audit, and create tax-tracking gaps.

Do I have to pay customers to take part in research?

No, but unpaid studies fill more slowly, skew toward your most enthusiastic users, and have higher no-show rates. If you can't pay, offer something of real value: early access, a findings summary, or direct influence on the roadmap.

How do I stop unqualified people from passing my screener?

Ask about past behavior instead of identity, include distractor options, add one open-text question you review manually, and never reveal which answer qualifies.

Can synthetic personas replace real research participants?

No. Synthetic personas are useful for a first read or a pressure test of questions and concepts. Decisions should rest on real customers and real participants.

Sources and Further Reading

• Informed consent (Wikipedia)
• General Data Protection Regulation (Wikipedia)
• Prolific (official site)
• Tremendous (official site)
• BuildBetter AI Surveys
• Guest, G., Bunce, A., & Johnson, L. (2006). How Many Interviews Are Enough? Field Methods, 18(1).
• Hennink, M., & Kaiser, B. (2022). Sample sizes for saturation in qualitative research: A systematic review of empirical tests. Social Science & Medicine, 292.
• Douglas, B., Ewell, P., & Brauer, M. (2023). Data quality in online human-subjects research. PLOS ONE.
• Nielsen, J. (2000). Why You Only Need to Test with 5 Users. Nielsen Norman Group.

Make churn optional.

Recruiting the right participants starts with knowing what your customers already told you. BuildBetter finds them by signal, runs AI-moderated interviews and surveys, pays them through gift cards, and turns every answer into cited signals your team can act on. Stop emailing the same 40 customers.

Make churn optional. Book a demo