A bank spends millions undoing a slogan that told customers to do nothing. A soft-drink campaign accidentally resurrects the dead. A game’s opening line becomes an internet meme for all the wrong reasons. These are not isolated translation quirks. They are what happens when linguistic quality assurance—LQA—is treated as optional.
LQA is not the same as translation. Translation converts meaning from one language to another. LQA puts that meaning back into the actual product: the app screen, the ad placement, the game dialogue box, the button that users will tap. It checks whether the words still make sense once they sit inside a fixed layout, under cultural expectations, and under real device constraints. Without it, even accurate translations can produce confusion, offense, or broken interfaces.
Real Costs, Documented Cases
In 2009 HSBC rolled out the global campaign “Assume Nothing.” In several markets the phrase arrived as “Do Nothing.” A bank advising inaction is not a message that builds trust. The rebrand that followed cost roughly $10 million. Marketing analyses at the time pointed to the absence of contextual review: translators worked from text alone, without seeing how the line would land in ads or on local websites.
Pepsi’s older “Come Alive with the Pepsi Generation” campaign produced a similar mismatch in Chinese markets, landing as something close to “Pepsi brings your ancestors back from the grave.” In cultures that treat ancestors with deep respect, the line was not playful; it was disturbing. KFC’s “Finger Lickin’ Good” became, in one Chinese rendering, an invitation to “eat your fingers off.” These examples are old enough to feel like folklore, yet the mechanism remains current: source text leaves the spreadsheet looking clean and arrives on screen looking alien.
Games supply their own cautionary list. The European release of Zero Wing gave the world “All your base are belong to us.” Beyond the meme, the same rigid handling of text created UI problems that made parts of the game harder to play. Later titles have faced player backlash when dialogue changes or truncated strings altered character tone or blocked progression. In each case the translation itself may have been competent; the failure was the lack of in-context checking.
Industry data reinforces the pattern. Surveys of localization maturity show that weak localization can cost companies an average of 20 percent of potential revenue in new markets. Separate research finds that nearly three in ten marketers have seen a translation or cultural misstep damage brand reputation, and for a sizable share the financial hit exceeded $10,000 once lost sales, staff time, and repair work were tallied. German text routinely expands 30–40 percent beyond English; Russian and certain other languages follow close behind. Fixed-width buttons designed for “Settings” or “Continue” simply cannot absorb “Einstellungen” or its equivalents without overflow, truncation, or awkward wrapping. Mobile screens make the problem sharper: limited real estate turns a minor expansion into unreadable UI.
Why the Failures Keep Happening
Three practical pain points drive most of these incidents.
First, translators often work without visual context. A string that reads neutrally in isolation can sound sarcastic, overly formal, or simply wrong once placed next to an animation, a character portrait, or a specific button hierarchy. LQA places native speakers inside the live build so they can flag those mismatches before users do.
Second, layout assumptions built around English collapse under longer languages. Design systems that do not allow 30–50 percent expansion, dynamic wrapping, or flexible containers produce the clipped buttons and overlapping text that users notice immediately. Pseudo-localization—replacing English with deliberately longer placeholder text early in development—catches many of these issues before real translations arrive.
Third, manual regression testing across languages is slow. Teams under launch pressure sometimes skip full LQA passes, then discover problems only after reviews and social media posts. Automated tools can accelerate string extraction, length checks, and basic consistency, yet they still require human judgment for cultural tone, register, and functional fit inside the product.
A practical mobile-app LQA checklist therefore includes more than linguistic accuracy. It covers string completeness (no leftover English or missing keys), text expansion and overflow on actual devices, in-context narrative flow, cultural and regional formatting (dates, numbers, currency), and basic functional behavior (buttons remain tappable, placeholders resolve correctly). Running the longest languages first—German, Russian, French—exposes layout risks early.
What Changes When LQA Is Built In
Companies that treat LQA as a standard gate rather than an afterthought reduce the chance of public embarrassment and the downstream cost of emergency fixes. The difference is measurable: fewer support tickets about unreadable interfaces, higher retention in non-English markets, and brand messaging that does not need costly repair campaigns.
For organizations expanding across languages, the practical question is no longer whether to invest in quality assurance after translation, but how thoroughly that assurance is performed inside the finished product. The public record of slogan collapses, game memes, and truncated interfaces shows what happens when the answer is “not thoroughly enough.”
Artlangs Translation has spent more than twenty years refining exactly this layer of work. With proficiency across 230-plus languages and a network of more than 20,000 professional collaborating linguists, the company has supported clients through full translation services, video localization, short-drama subtitle localization, game localization, multilingual dubbing for short dramas and audiobooks, and multilingual data annotation and transcription. Those capabilities rest on repeated experience with the same constraints that produce the disasters described above—context gaps, text expansion, and the need for in-product verification—so that the final user experience matches the intent of the original content rather than exposing its weak points.
