How to tell a stranger they’ve got food in their teeth
- Assumes the respondent has a resolution — an assumption that often turns out to be wrong
- Respondents without a resolution have nowhere to go
- Yes/No screener first — only people with a resolution answer Q2
- Uncovers and corrects the bad assumption before it contaminates the data
- One of the most common fixes: switching from a presumptive question to a two-step format
Don’t make assumptions about what shoppers know.
That might sound obvious. But the nature of assumptions is that you don’t notice them.
In one Roast, I saw the question, “What’s your New Year’s resolution?” — which assumes people had one.
The fix wasn’t to rewrite the question. It was to ask first: “Do you have a New Year's resolution?”
That two-step format — yes/no, then follow-up — is one of the most common survey design moves I recommend.
It acts as a filter. You only ask the real question to people who actually do the thing you’re studying.
But more importantly, it exposes bad assumptions.
A good metaphor is a friend telling you that you’ve got lettuce stuck in your teeth.
With the right signal, it’s easy to fix. Without it, you just walk around unaware.
Want to make sure you're not making assumptions in your surveys?
If it feels like your survey has a bit of lettuce in its teeth, book a Roast.
I’ll send you a 15-minute Loom with clear, copy-and-paste edits to tighten things up.
Two-step questions are just one of many small changes that make surveys easier to answer — and much more useful.
Cheers,
Sam