Texture Is Trust

How the study worked

Three boards, sourced from Google Images on the search terms “Halloween candy” and “trick or treat candy,” one board each for chocolate bars, chewy candy, and bite-size and suckers. The prompt was the whole study: drag the green dots to the candy you like the most and the red dots to the candy you like the least.

Twenty-six people completed it in October 2025. The dot maps above are the raw result. At the report’s 75% threshold the 24 scored candies sort into three groups: eight resonant, eight resistant, eight contested. The report prints every score with the count it came from, and so does this page. See the full report.

Texture is trust

More than flavour, more than brand, texture predicts the pile. The candies with satisfying physical feedback, the snap of a KitKat, the shell-crack of M&Ms, the crunch-then-gummy of a Nerds cluster, consistently beat the candies that require negotiation. People want their treats to cooperate. The effort-to-reward ratio is the whole sort.

Line the 24 candies up by the share of their dots that were green and the room does not spread its opinion across a range. It sorts into three piles, and the piles are texture.

Figure 2 · Every candy, ranked

What won: the hall of fame

Eight candies cleared the 75% positive threshold. What unites them is not just chocolate, though there is plenty of it. It is textural satisfaction, brand recognition, and the kind of flavour reliability you can set your watch to. Every one of them delivers exactly what it promises.

The Nerds surprise

Nerds Gummy Clusters are the unexpected champion: a newer entry with a chaotic texture, crunchy shell and gummy interior, and it took 87.5% on one of the bigger crowds of the chewy board. Novelty did not hurt it because the novelty is textural. It cooperates in a new way.

What lost: candy that fights back

This is where candy goes to be actively resented. Not ignored: rejected. Eight candies crossed the 75% negative threshold, three of them without a single green dot, and they share a set of failure modes: excessive hardness, sheer stickiness, polarising flavours, and textures that require negotiation rather than reward.

The nougat problem

Milky Way, Baby Ruth and 3 Musketeers all landed here. They are chocolate bars, they are famous, and they share a nougat-forward formulation. What felt substantial to a previous generation now reads as filler. These are the chocolate bars time forgot.

The coconut catastrophe

Almond Joy at 78.95% red is not an anomaly. It is a verdict. Nineteen dots landed on it, the joint-largest crowd on the chocolate board, and fifteen were red. Shredded coconut and chocolate, once novel, now signals divisive at best. The almond on top did not help.

The twin test

The cleanest pair in the data is two hard candies. Both are literally hard. One breaks for you. One makes you do the work.

Same hardness, opposite verdict. The difference is who does the breaking. That is the texture rule in two images, and it explains the nougat bars, the taffy and the coconut without a second variable.

The ambivalent middle

Eight candies landed between 27.27% and 64.29%. They inspire neither devotion nor disgust. They are the “fine, I guess” of Halloween, the ones you eat when the good stuff runs out. And the contested zone reveals something the two poles cannot: these are candies where personal history matters more than universal appeal.

The Snickers surprise

America’s best-selling candy bar landed in the middle at 64.29%. Nine of fourteen dots were green, which means a substantial minority actively objected, possibly to the peanut-nougat combination, possibly to the density. Butterfinger sits at exactly the same score for the opposite textural reason: its famously flaky, sticky crunch-dissolve alienates as many people as it attracts.

The nostalgia paradox

Candy corn, Tootsie Rolls and Smarties spark memories without sparking cravings. They are the visual vocabulary of Halloween without being the reward. Candy corn split exactly seven to seven. It survives on iconography, not flavour. Recognition is not the same as desire, which is the same lesson we found when we asked what makes people want a Coke: the icon is not the craving.

What to do with this

The decision logic translates cleanly, for candy and for any category where the product has to be reached for. The audience is scoring the experience of consuming the thing, and the wrapper is only trusted when it promises a texture the room already wants. It is the same lesson as the orange juice study: the closer the image sits to the actual sensory reward, the more the room trusts it.

One caveat, in the report’s own words: what people click may differ from what they would actually eat if handed a bowl. But directional preference? Twenty-six people, three boards, and the same rule on every one. The data is consistent enough to trust.

Frequently asked questions