How to recruit research participants beyond existing customers without losing focus

Recruit research participants beyond existing customers whenever the decision depends on people who never adopted your product, left it or solved the problem another way. Your customers can explain the experience you built. They can't explain why everyone else chose differently.

Start with the question, then pick the groups that could answer it. A random mix of customers and strangers is no more useful than loyal users alone.

Here is a made-up example used throughout: a team building staff-scheduling software wants to know why some multi-location retailers still coordinate schedules through spreadsheets and messages. The sample design is illustrative, not a claim about retail businesses.

1. Pick the comparison your decision needs

Improving an existing scheduling screen? Current users who do that task may be the right participants. Investigating why suitable businesses don't adopt the product? Include people who handle the task some other way.

GOV.UK's guidance on user needs covers actual and likely users and how they currently reach their goals. That is a useful boundary: recruit around the work, not around whether someone has an account.

Write the missing perspective as a sentence. "We have not heard from managers who considered changing their scheduling process but kept their existing approach" gives recruitment a specific job.

Not every stranger is a potential customer. They still need a relevant task and context. A retailer with a stable schedule at one small shop may be interesting, but probably not for a study about frequent changes across several sites.

2. Map the experience groups in a coverage matrix

Use the relationship to the product as one dimension and the situation the research cares about as the other.

The last group can overlap the others. A store supervisor affected by a scheduling decision may work somewhere that uses the product without controlling the purchase. Record both the account relationship and the person's operating role instead of forcing everyone under one marketing label.

3. Ask every group about the same recent event

In the scheduling example, ask each participant about a recent change to a staff schedule: what triggered it, who acted, how the information moved and how they confirmed it was done.

A shared event makes the accounts comparable. Current users describe the task in your product. Nonusers describe a spreadsheet, another service or a manual arrangement (say, a group chat named "SCHEDULE FINAL real"). Don't make everyone discuss your interface before you understand their own process.

Add group-specific questions afterward, such as what changed over time for former users or how evaluators made their comparison.

Don't build one satisfaction ranking out of these different experiences. Nobody can meaningfully rate a workflow they have never used.

4. Recruit through more than one doorway

Current users may come through a research opt-in or an established contact route. Former users and nonusers may need professional communities, relevant organizations, introductions or targeted people search.

GOV.UK notes that recruitment methods, session times, locations and activities can affect who is included, and that using different approaches can help reduce those blind spots.

Track how each participant entered the study. If every nonuser came through one partner, their agreement may reflect a shared context. That doesn't invalidate their accounts, but it limits the comparison.

Keep participation voluntary and respect contact preferences. Never use a new channel to reach a former customer who asked not to be contacted. Where direct outreach is unsuitable, use a permitted public invitation.

State any compensation plainly, and never make it depend on whether the participant likes the product or supplies the conclusion the team wants.

5. Fill the biggest gap first, not equal quotas

Equal numbers in every group look tidy without serving the decision. Put recruiting effort where the uncertainty is largest and the existing evidence thinnest.

In the scheduling study, the team might already understand current users well but know little about businesses that considered changing and decided against it. The next round could focus there, with current-user evidence kept as context.

That doesn't make the result representative of all retailers. It makes a purposeful comparison. If you need a population estimate, design a separate study for that claim.

Write an allocation note before recruiting:

"Our main gap is [experience]. We need to compare it with [reference context] because [decision]. We are not prioritizing [group] in this round because [reason]. We will revisit the plan if [new evidence or missing coverage]."

It lets a colleague spot when convenience is quietly replacing the intended comparison.

6. Report who you reached next to what you found

Record whom you actually reached: completed sessions, people who declined, unresolved eligibility and groups you couldn't recruit. Keep the reasons when known, without guessing why someone stayed silent.

Compare findings within similar contexts before contrasting whole groups. A difference between users and nonusers may come from company structure, responsibility or how often the task happens, rather than adoption alone.

Look for counterexamples. A nonuser may describe a process that works well, and a current user may report a serious gap. Both sharpen the edges of the product's value.

End the recommendation by naming which perspectives support it and which are still missing. Recruiting wider is worth it when it changes what the team understands, rather than adding names to the study.