Why the same name performs differently across languages

A brand tests well in three markets and badly in a fourth. The product, packaging and price are identical. In the fourth market the name shares its first syllable with an unrelated everyday word, and nobody in the naming process spoke the language well enough to hear it.

What everyone sees

International naming problems are imagined as translation accidents — the memorable cases where a name means something rude somewhere. Those are rare and easy to screen for. The common version is quieter: the same phonemes carry different associations about size, softness, speed and expense, and the shift happens without any word being invoked at all.

What is actually happening

Yorkston and Menon found that the phonetic properties of an invented brand name influenced inferences about the product — vowel quality shifting judgements of attributes such as creaminess — with the effect operating automatically and being discounted only when respondents were made aware of it. Zhang and Schmitt studied bilingual Chinese and English speakers and found that whether a name was processed for sound or for meaning depended on the writing system and language mode, so the same brand name was evaluated through different routes by the same person under different conditions. A name therefore has no single performance. It has one per language community, and the variation is not reducible to translation errors.

Why it stays hidden

The variation hides because the string looks identical in every market. Legal registration, packaging artwork and brand guidelines all treat the name as one asset, and internal reporting compares markets on everything except the thing that differs. The failing market gets explained by distribution, competition or channel mix, all of which are also true.