Skip to main content

Screeners

A screener is a small set of questions every potential participant answers before they enter your study. Only people who pass start the interview, survey, or live call.

Use a screener to:

  • Confirm the participant matches a niche profile that demographic filters can't capture (e.g. "uses Notion daily for project management").
  • Catch low-quality respondents.
  • Apply attention checks.

When the screener runs​

Screeners are available on:

  • AI-Curated Interviews
  • AI-Generated Surveys
  • Live Interviews

(AI-Curated Surveys don't have a separate screener; the survey itself is conversational, so qualifying logic goes in the script.)

The screener runs between the welcome page and the study. A failed answer ends the session politely; the participant doesn't take a slot.

Building a screener​

In the Screen step of the wizard, add up to about a dozen questions. Each question has:

  • A type: Text Input, Multi-line Input, Number Input, Multiple Choice, or Checkboxes.
  • Qualifying criteria. How this depends on the type:
    • Multiple Choice / Checkboxes: mark each option Accept or Reject. Choosing a rejected option fails the question.
    • Text Input / Multi-line Input: write a "Qualify with AI" description of the answer that should pass; the AI judges the response against it.
    • Number Input: set a qualifying number threshold.
  • A position in the screener order.

Logic between questions: OR within a question (any qualifying answer passes that question), AND across questions (you have to pass every question to enter the study).

Screener pass logic is unified across the study types that support screeners since the recent matcher overhaul. Older studies might behave differently.

Example screeners​

Validating product usage

  • "Which of the following do you use weekly?" → must include Notion.
  • "Roughly how many projects do you manage in it?" → must be ≥ 3.

Catching low-quality

  • A Number Input: "On a 1-10 scale, how likely are you to recommend us?" → set a qualifying threshold.
  • An attention check (Multiple Choice): "For this question, please select 'Strongly Disagree'." → mark only that option Accept.

Niche role check

  • "What is your job title?". open-ended, manually reviewed before approval.

Best practices​

  • Keep it short. Three to five questions covers most needs. Longer screeners drop completion rates.
  • Don't restate demographics. If you've already filtered the pool by gender or industry, don't ask again.
  • Front-load the deal-breakers. Strictest filter first, so unqualified participants exit fast.
  • Test it on yourself. Click the participant link and run through your own screener; you'll catch typos, broken logic, and unclear phrasing.

For filtering before the screener runs, see Audience & recruiting.