Screener Design and Logic
Overview
Screeners define reusable target audiences for testing with key strategic groups. The goal is not hyper-granular customization, it's creating stable, repeatable audiences that produce comparable business learnings over time.
Effective screeners balance:
- Strategic targeting
- Respondent experience
- Feasibility of recruitment
- Consistency across studies
Best Practice Guardrails
| Guideline | Recommendation |
|---|---|
| Screener length | ≤ 1 minute |
| Total length (screener + survey) | 7–11 minutes |
| Number of screener questions | 5 or fewer |
| Answer options per question | ~10 max |
| Minimum incidence rate (IR) | 20% |
| Ideal incidence rate (IR) | 50%+ |
Keeping screeners short and IR healthy reduces survey fatigue, drop-off, and recruitment cost and improves respondent quality.
Question Types
Screening Questions
Determine whether a respondent qualifies for the survey overall.
- Define the target audience and directly impact IR
- Use qualifying, filler/dummy, and disqualifying logic
- Can terminate respondents
- Should stay broad enough to maintain healthy IR
- Use strategically and sparingly (they're the main lever on recruitment feasibility)
Common criteria: category usage, purchase behavior, brand familiarity
Filter Questions
Organize or segment respondents after they've already qualified — they don't affect eligibility.
- Do not terminate respondents
- Used for reporting, analysis, or segmentation
- Let teams compare responses across subgroups without restricting recruitment
Common criteria: income, product ownership, frequency of usage, brands, retailers
Supported Question Formats
- Single Choice — respondent selects one option
- Multiple Choice — respondent may select multiple options
Answer Option Classifications
Every answer option must be assigned one of these:
| Type | Behavior |
|---|---|
| Qualifying | Selecting at least one qualifying answer allows the respondent to continue. |
| Filler/Dummy | Neutral options respondents can select; they don't qualify on their own but pass through when paired with a qualifying answer. Used to reduce bias, mask targeting intent, and improve respondent experience should stay realistic, relevant, and easy to understand. |
| Disqualifying | Selecting this ends the survey immediately. This overrides all other logic, including any qualifying selections. |
Answer Display Settings
- Fixed — anchors an option in place while other choices are randomized
- Randomized — shuffles order to reduce order bias
- Exclusive — deselects all other options when chosen (e.g., "None of the Above")
Current Logic Limitations
The screener experience prioritizes simplicity, consistency, and respondent ease over advanced survey programming. Not currently supported:
- Survey piping — responses can't dynamically change later questions (e.g., "If Option A selected in Q1, show Q4")
- Advanced conditional routing — questions can't be shown/hidden based on prior answers
- Min/max selection rules — multi-select questions can't enforce an exact, minimum, or maximum number of selections; respondents only need one qualifying answer
- Grid/matrix questions
- Open-ended responses — including "Other, please specify" style options
- Images, video, or other media within questions or answer options
FAQ
Why are advanced logic features intentionally limited? The screener experience is optimized for simplicity, scalability, faster fielding, better respondent experience, and stable longitudinal audiences. Advanced logic tends to increase respondent confusion, reduce IR, and create inconsistent audiences across studies.
Why can't screeners support niche audiences? Surveys are distributed broadly through external suppliers and panel partners, who monitor performance closely. Screeners that terminate too many respondents may get paused or deprioritized, making recruitment difficult or impossible to complete.
Why aren't open-ended questions supported? Screeners are meant to stay easy to complete and operationally scalable. Open text increases respondent burden and can reduce completion quality.