Global Marketing Translation Gets the Problem Wrong

The Nova Never Failed. That's the Point.

For decades, business professors have opened marketing lectures with the same story: Chevrolet tried to sell the Nova in Latin America and watched it flop because "no va" means "doesn't go" in Spanish. It's a clean lesson. It's also completely fabricated.

The Nova sold well in Mexico. It outperformed projections in Venezuela. Snopes traced the claim back through decades of repetition and found zero evidence of a failed campaign in any Latin American market. The Pepsi story — that "Come Alive! You're in the Pepsi Generation" was rendered into Chinese as a promise to raise ancestors from the dead — is equally unsubstantiated. Nobody can even agree on which Chinese wording supposedly caused the offense.

These myths persist because a single wrong word sinking a household brand is a satisfying lesson. The real problem with global marketing translation is less dramatic and far more expensive: campaigns that simply underperform, quietly, with no error message and no obvious villain.

Three Services Most Buyers Treat as One

The translation industry draws clear lines between translation, localization, and transcreation. Most buyers outside the industry do not, and that confusion is where budget gets wasted.

Translation changes the words. A certified translator rendering a legal contract or a technical manual from English into Spanish is doing exactly that — converting language while staying faithful to the source.

Localization adapts everything around the words. The same technical manual localized for Germany swaps US measurements for metric, reformats dates, adjusts currency, and replaces US-specific regulatory references with German ones. The text changes because the context changed.

Transcreation starts over on purpose. A transcreation brief often contains no source text at all — just a description of what the message needs to accomplish emotionally. The transcreated result might share nothing with the original wording and still be correct, because the job was never to preserve the sentence. It was to preserve the effect.

The MultiLingual piece that prompted this post lays this out clearly, quoting the American Translators Association's Judy Jenner on where translation ends and localization begins. The distinction matters because these are genuinely different scopes of work with different skill sets, different timelines, and different price points.

The One Failure That Was Real

In 2009, HSBC carried its "Assume Nothing" tagline — built around the idea that the bank made no assumptions about customers — into international markets. In several of them, the phrase translated closer to "Do Nothing." Roughly the opposite message for a financial institution.

The fix cost HSBC an estimated $10 million. They replaced the tagline with "The World's Local Bank" and moved on.

That story is the exception, not the template. Most translation-adjacent marketing failures don't produce a $10 million headline. They produce a Q3 report where the new market numbers never quite materialized, and everyone spends six months guessing why.

CSA Research's 2020 study of 8,709 online shoppers across 29 countries put hard numbers on the underlying preference: 76 percent said they preferred buying where product information was in their own language, and 40 percent said they would not buy from a foreign-language site at all. That data is five years old now, but nothing published since suggests the preference reversed.

What Word-for-Word Translation Cannot Touch

Slogans fail on literal translation first because they are built to be short, rhythmic, and emotionally loaded — the exact qualities that don't transfer by dictionary lookup. Rhyme, wordplay, and double meaning are the first casualties. A pun that works in English almost never happens to work in Japanese or Portuguese by accident.

Visual and cultural signals are the second failure point, and they don't involve words at all. White signals mourning in parts of East Asia rather than the purity it suggests in the US. Humor that reads as self-deprecating in British English can read as incompetent in German. Gesture and body language in video content carry meaning that subtitle translation never addresses.

The fix for slogans is not a better translator working more carefully on the original. It's a copywriter briefed on intent rather than wording, building something new in the target language from the ground up.

Why This Matters More in Healthcare Than in Consumer Marketing

Most transcreation discussion centers on consumer brands, and that makes sense — a failed tagline is visible and embarrassing. But the stakes are higher and the dynamics are more complex in healthcare and social services, where TXLOC does most of its work.

Chuukese and Pohnpeian, the primary languages of Micronesian communities in the US, have no direct linguistic equivalent for several concepts that American healthcare systems treat as routine. Informed consent, as a legal and procedural concept, does not map cleanly onto Chuukese communication norms around decision-making, which are heavily communal and often defer to family elders rather than the individual patient. A word-for-word translation of a consent form doesn't just sound awkward — it can produce a document the patient understands differently than the provider intended, with real clinical consequences.

Medical taboo language is the second layer. In Pohnpeian culture, certain body parts and functions carry naming conventions that differ by speaker gender and social context. A translated patient intake form that ignores those conventions can cause a patient to disengage entirely rather than answer accurately — not because they don't understand the language, but because the phrasing signals that the provider does not understand them.

Neither of these problems shows up in a machine translation quality score. MT tools for Chuukese and Pohnpeian are limited to begin with — training data for these languages is thin, and the tools that exist were not built with healthcare context in mind. What looks like a translation problem is actually a localization problem layered on top of a cultural competency problem, and it requires human linguists who carry both the language and the community knowledge.

This is also why the "good enough" instinct that drives a lot of MT adoption in healthcare is more dangerous in rare-language pairs than it appears on a budget spreadsheet.

What to Actually Do With This

If you are buying translation services for marketing or patient-facing content, the most useful question you can ask a vendor is not "can you translate this" but "what service does this content actually need."

For a product manual or a policy document, translation is probably right. For a website with localized pricing and regional references, you need localization scoped into the project from the start. For a tagline, a campaign headline, or any content where the emotional effect matters as much as the information, you need a transcreation brief — and the vendor needs to tell you that before the project starts, not after the campaign underperforms.

If your content reaches communities that speak Chuukese, Pohnpeian, or other Pacific Island languages, add one more question: does the linguist have community-level cultural knowledge, not just language fluency? For healthcare and education content especially, fluency alone is not enough.

If you are working on rare-language localization for healthcare or government content and want to understand what that scope actually involves, we are glad to talk through it.