I spent Tuesday afternoon standing near the pasta sauce at the Fred Meyer on Hawthorne, which is a strange way to bill an afternoon, and watched about forty people shop. Not talked to them. Watched them.
Here is what people actually do. They pick up two jars. They turn one of them around, read maybe four seconds of the back, and set it down without ever turning the other. They look at the price, then the shelf tag underneath, then the price again. Roughly a third pick a jar, walk two steps, come back, and swap it. Almost nobody reads the front label past the brand name and one adjective.
On my desk that same week sat a survey. Six hundred respondents, and eighty-one percent said ingredient quality was the single most important factor in their decision.
Both of those things cannot be true.
People are honest with their hands and unreliable with their mouths, and most research budgets are pointed straight at the mouth.
Not because respondents lie. Because a survey asks someone to become, for ninety seconds, a thoughtful person with reasons. Nobody wants to answer a questionnaire by admitting the jar was at eye level and had a red lid. So they reconstruct a decision that sounds like the kind of decision a reasonable adult makes, and we take that reconstruction, put it in a deck, and build a positioning platform on top of it.
The aisle tells you different things, and better ones. Four seconds is your actual window on the back of the pack. The swap at two steps means the decision was not finished at the shelf and you can still lose it after they walk away. The glance at the tag before the price means people are checking whether they are being treated fairly, not whether they can afford it, and that is a completely different message to write toward.
Salt & Straw worked this out with a sampling spoon rather than a panel. The whole shop is built around noticing what somebody reaches for a second time. Trader Joe's crew members are, functionally, full-time observers who report back on what confused a person in aisle three.
They reconstruct a decision that sounds like the kind of decision a reasonable adult makes.
None of this means kill the survey. Surveys size things. They tell you how many, and how many is genuinely useful. They just cannot tell you why, and we keep asking them why because it is cheaper than standing in a grocery store on a Tuesday feeling conspicuous.
Two hours, one notebook, no incentives paid, and it beat six hundred respondents.
Came home and watched Truffles ignore the expensive food to eat Barnaby's. If I had surveyed her, she would have said ingredient quality.