Transcreation vs translation: why your China campaign copy needs rewriting, not converting
Short version: translation converts your words into Chinese; transcreation rebuilds your message so it actually lands on a Chinese reader. For a manual, a spec sheet, or a returns policy, translation is fine and transcreation is a waste of money. For a tagline, a campaign line, a brand name, or anything meant to make someone feel something, translation gives you a sentence that is grammatically correct and emotionally dead — and sometimes worse than dead. Know which job you're buying before you sign the invoice.
The classic cautionary tale is Coca-Cola. When it first reached out for a Chinese script in the 1920s, some shopkeepers put up phonetic characters that sounded like "Coca-Cola" but read as something close to "bite the wax tadpole." The version the company eventually settled on — 可口可乐, kekou kele — keeps the sound and means roughly "tasty and joyful." That is transcreation in one line: it doesn't translate anything, because "Coca-Cola" has no meaning to translate. It rebuilds the name so the sound and the sense both do the brand a favor. A literal approach couldn't have gotten there, because there was nothing literal to convert.
What the two words actually mean
People use "translation," "localization," and "transcreation" as if they were three sizes of the same thing. They're not the same job, and the difference is what you're optimizing for.
| Translation | Transcreation | |
|---|---|---|
| Optimizes for | Accuracy — same meaning, other language | Effect — same reaction, whatever words it takes |
| Starts from | Your source text | Your intent and the brief behind the text |
| Freedom to change words | Minimal — you're bound to the original | Total — the output can share zero words with the source |
| Who does it well | A skilled bilingual translator | A bilingual copywriter who thinks in the culture |
| Right for | Manuals, specs, legal, FAQs, product detail | Taglines, campaign lines, brand names, hero copy |
| Priced like | Per word | Per idea, like creative — because it is creative |
The tell is that last row. If a vendor quotes your tagline by the word, they're going to translate it, no matter what the proposal calls it. Great campaign lines aren't a word count; they're a small creative problem. You brief them, you get options, you pick.
Where translation quietly kills the work
The famous disasters — the ones that get retold as "eat your fingers off" or "brings your ancestors back" — are mostly apocryphal in their exact wording, and I wouldn't stake a claim on the phrasing. But the underlying failure is real and I've watched milder versions of it in the work more than once. A line that carries a pun, a rhyme, a cultural reference, or an emotional register at home goes through a competent translator and comes out the other side correct and inert. Nobody flags it, because nothing is wrong. It just doesn't do anything.
That's the trap: a bad translation gets caught, laughed at, and fixed. A merely flat one ships. It runs for a full campaign flight, underperforms, and gets blamed on the media plan or the "Chinese consumer being hard to reach." The copy was the problem, and the copy passed review because it was accurate. This is why translation is the more dangerous default — its failures are invisible.
Names are the extreme case
A brand name is transcreation with the difficulty turned up, because a Chinese name has to work on three axes at once: how it sounds (ideally near the original), what the characters mean, and whether it's easy to say and remember. The good ones make all three pull the same direction. BMW became 宝马, baoma — "precious horse" — which flatters the product and sounds right. Mercedes-Benz landed on 奔驰, benchi, "to dash / speed along." Coca-Cola's kekou kele again. None of these are translations. Nobody translated "Bayerische Motoren Werke" into "precious horse." They were built to a brief: sound like us, mean something good, be sticky.
If you're at the naming stage, this deserves its own process — I've written a fuller piece on choosing a Chinese brand name that goes past the examples into how to actually run the shortlist.
How to brief for transcreation (so you get the good version)
Most weak transcreation is a briefing failure, not a talent failure. You hand a copywriter your English line and say "make this work in Chinese," which is barely more than a translation order. Give them the thing behind the line instead:
- The intent, not the words. What should the reader feel or do after reading it? That's the target, not your English sentence.
- What's load-bearing vs. disposable. Is the pun sacred, or is it just how you happened to say the idea in English? Say so explicitly.
- Guardrails. Words, claims, or associations you cannot use — regulated category language, competitor territory, anything off-brand.
- Room for options. Ask for three to five routes with a back-translation and a one-line rationale each, so you can judge the effect without reading Chinese.
- A native gut-check. Before it ships, put the finalists in front of real target-market people, not just your in-house reviewer. Tone is the thing offices are worst at judging on their own.
The back-translation-plus-rationale step is the one non-Chinese-reading marketers skip, and it's the one that lets you actually own the decision instead of rubber-stamping it.
So do you need it? A quick test
You don't transcreate everything — that's expensive and pointless. Sort your copy by one question: would a flat-but-accurate version of this still do its job? If yes, translate it and move on. If a flat version would technically inform the reader while quietly losing the sale, that line earns transcreation. Most of a website is the former. The handful of lines that carry the brand are the latter, and they're usually less than five percent of the word count and most of the impact.
Bottom line
Translation and transcreation aren't better and worse versions of the same service — they're different jobs with different failure modes. Translate the copy that only needs to be understood. Transcreate the copy that needs to move someone. The costly mistake isn't picking wrong once; it's not knowing there was a choice, and letting accurate-but-lifeless campaign lines ship because nothing looked broken. Sort your copy, brief the important five percent properly, and gut-check it with real people before it runs.
If you're staring at a China campaign and can't tell which lines are load-bearing, that sorting is exactly the kind of read I do — reach out. And for the wider job that copy sits inside, see the cross-border marketing playbook for Greater China.