How Do You Write a Screener That Keeps the Wrong Respondents Out?

The screener decides who your results represent. A weak one lets in people who guessed the right answer, and a bloated one wastes questions and respondents before the survey even starts. Two questions come up again and again when teams draft screeners: how do we stop people from simply saying yes, and which profile questions do we actually need?

Why do yes or no screener questions cause problems?

Ask "Have you consumed Brand X in the past year?" and the right answer is obvious. Some respondents will say yes to qualify, others will say yes because the brand sounds familiar. Either way, unqualified people slip through and dilute your results.

The fix is to hide the target answer in a list, which removes the cue. Ask which of several brands, variants or products the respondent has used, including your target among realistic alternatives. People who genuinely use the product find it easily, while guessers have no clue which answer qualifies.

What does a strong screener sequence look like?

1. Category use first. Ask which product types the respondent has used in a defined period, as a multiple-select list with an exclusive "None of these" option.
2. Brand or variant use next. Show this only to category users, with your brand among several real alternatives and specific variant names where useful.
3. Decision role. Ask who decides or shares the decision for this purchase, and qualify only sole or joint decision-makers if that is your audience.
4. Recency or frequency. Add a time condition only if the study needs it, such as current versus lapsed users.
5. Terminate cleanly. Exit non-qualifiers politely and keep their category answers if they are useful for profiling.

Which questions can you remove from the screener?

If you field with a profiled consumer panel, many standard questions are already known. Age, gender, region and ethnicity are usually captured when people join, so targeting can be set directly instead of asked again. Repeating them adds length and gives people another chance to answer inconsistently.

If income is part of who qualifies, it does need to be in the screener, and you should check the answer ranges match the target in your brief exactly. Gaps between bands, such as a range that jumps past the threshold you care about, are a common error. Use bands in local currency, keep them mutually exclusive, and include a "Prefer not to say" option so people are not forced to guess.

What other details protect screener quality?

• Make "None of these" exclusive, so it cannot be ticked alongside other options.
• Avoid offering both "Other" and "None of these" where their meaning overlaps.
• Randomise brand lists so your brand is not always in the same position.
• Include one or two plausible brands that are not your target, so qualifying is not obvious.
• Use the brand name respondents recognise, including the parent brand if that is how people know it.
• Check that each terminating answer matches the brief, and test the logic before launch.

It is worth running a short soft launch before full fieldwork. Look at the share of people terminating at each screener question. If one question removes far more people than expected, the wording may be confusing, or the audience may be smaller than you assumed. Fixing that early is much cheaper than discovering it halfway through fieldwork.

What does this look like in a real study?

Imagine a breakfast cereal brand studying why former buyers stopped choosing one of its flavours. The first draft asks, "Have you ever bought this product? Yes, No, I can’t remember." Almost everyone would qualify. The revised screener first asks which breakfast products the respondent has bought, then shows category buyers a list of specific flavours from several brands. Only those who pick the target flavour continue.

Age and region are set through panel targeting instead of being asked. The result is a shorter screener and a cleaner sample of genuine past users, which matters because the whole study hinges on their reasons for stopping. For more on that type of study, see our guide on surveying lapsed users.

A tighter screener can also shrink your eligible audience. Before you lock it, check feasibility using our guide on checking niche audience feasibility in Malaysia.

Where does Vase.ai fit?

We build Vase.ai with screening logic, exclusive options, randomisation and panel targeting built in, so screeners stay short and qualified respondents from a verified panel of Southeast Asian consumers reach the main survey quickly.

If you are drafting a screener now, send us your research question, the audience you need and the date you need results. We will suggest a screener sequence and sample plan so you can check feasibility before fieldwork.

Frequently asked questions