Category Convention

What it is

The two results behind it are Karen B. Schloss, Rosa M. Poggesi and Stephen E. Palmer's affiliation study and the ecological valence theory it tests: if a colour preference is the average of one's affective responses to the things that colour is attached to, and if belonging to an institution raises liking for its colours, then preference inside a category is downstream of the category's existing palette, which is what Berkeley students preferring Berkeley's colours and Stanford students preferring Stanford's, with the gap growing by self-rated school spirit, demonstrates in people who had not chosen a university on colour grounds.

The absence is the field test, because nobody appears to have tested blue against a matched alternative on a real commercial outcome and published the result, which was searched for and not found; the nearest peer-reviewed field experiment with real sales tests red against yellow in an Italian store chain and does not test blue, and the most repeated example of a blue field test has no published methodology, no peer review and no primary source, so neither its design nor its reported figure appears here. Put those together and the convention explanation is the one left standing, since blue is widely liked, it is the least emotionally distinctive colour in the largest cross-national dataset on record, and it is what every incumbent in several categories already uses.

In effect

It changes what a positive result means, because if a blue mock-up tests better than a green one for a bank then the finding is consistent with blue doing something and equally consistent with blue matching what a bank is supposed to look like, and the design that separates those is a cross-category test which has not been run. It also explains convergence without attributing a motive, since no source establishes any firm's actual reasoning for choosing blue and this site does not assert one, so convergence is a fact about the market while the motive is not evidenced. For anybody working in a category the usable version is that the convention is the baseline rather than the variable, because departing from it is the thing that carries risk and information while conforming to it buys legibility rather than persuasion.

What it does not say

It does not say colour does nothing, because Domicele Jonauskaité and Christine Mohr find systematic colour and emotion correspondences across a hundred and thirty-two articles and 42,266 participants, and Karen B. Schloss, Christoph Witzel and Lai find that redness and anger genuinely does track hue, so the honest claim is narrower: the effects are real at the level of association, small where measured behaviourally, driven more by lightness and saturation than by hue, and almost never tested in the form the trade uses them. It is not itself a tested hypothesis, since nobody has manipulated a category's convention and measured what happens to colour preference inside it, which makes this an interpretation that fits the available results rather than a finding with its own evidence. It does not establish how much of a measured colour effect is convention either, because the affiliation study shows the mechanism exists without apportioning it, and no study in this record decomposes a brand-colour result into convention and colour. And it cannot be settled by the trade literature, because the one document that literature cites most is a narrative review with no data whose headline figure carries no citation anywhere in it.

---