The translation was technically accurate. It was also completely wrong for the context. The recall cost €2.3 million.
Technical translation failures make headlines when they cause spectacular problems – medical devices with mistranslated instructions, industrial equipment with confusing safety warnings, automotive recalls triggered by documentation errors. But for every headline, hundreds of smaller failures create costs that never make the news: rework, delays, customer confusion, lost credibility, audit findings.
Understanding what goes wrong – and why – is the first step toward preventing it. These aren’t failures of translator competence. They’re failures of process, communication, and oversight that competent translators couldn’t overcome.
The specification that killed a project
A UK engineering firm won a contract to supply components to a German automotive manufacturer. The specifications were clear, the engineering was sound, the quality was proven. The documentation was translated by a general business translation agency.
The German engineers rejected the first delivery. Not because of the parts – because of the documentation. The translated specifications used terminology that didn’t match the customer’s existing systems. Part numbers were formatted incorrectly for German conventions. Reference standards were cited in ways that didn’t align with German industry practice.
The parts were fine. The documentation made them useless. The customer’s procurement system couldn’t process the specifications. Their engineers couldn’t match the documentation to their existing component database. The relationship never recovered.
The root cause wasn’t translation quality in any conventional sense. Every word was accurately translated. The failure was contextual – the translators didn’t know the customer’s terminology requirements, the industry conventions, or the specific standards framework the documentation needed to fit within.
When safety instructions don’t protect
Medical device documentation follows strict requirements for good reason. Instructions for use must be clear, unambiguous, and written for the actual users of the device. When translation fails to meet these requirements, the consequences can be severe.
A home-use medical device was translated from English into several European languages. The English instructions used the phrase “ensure device is at room temperature before use.” The translation rendered this accurately in each language. But “room temperature” means different things in different contexts – and the translators weren’t told that the device required a specific temperature range to function safely.
Some users, particularly in Southern Europe, stored the device in locations that exceeded the actual safe operating range. The instructions didn’t specify the range because the English writer assumed “room temperature” was sufficient. The translators didn’t question it because they weren’t told there was a critical temperature threshold. The feedback loop that should have caught this – from translation back to source documentation – didn’t exist.
The fix required rewriting the source documentation to specify the actual temperature range, then retranslating for all markets. The cost included not just translation but regulatory resubmission in multiple jurisdictions.
The glossary that diverged
A manufacturing company maintained product documentation in twelve languages. They’d worked with the same translation provider for years. Quality was good. Consistency was good. Then they changed providers.
The new provider was competent, professional, and cheaper. They delivered accurate translations on time. But they used different terminology for key components – terminology that was equally correct but different from what the existing documentation used.
The result was documentation that contradicted itself. The new user manual used different terms than the existing maintenance manual. Training materials didn’t match operating procedures. Spare parts lists didn’t match the component names in the instructions.
Each document was individually correct. The documentation set was collectively incoherent. Fixing it required auditing every document across all twelve languages to restore consistency – a project that cost more than the savings from changing providers.
The root cause was a glossary that existed but wasn’t transferred. The original provider had built a termbase over years of working with the client. That asset stayed with the provider when the relationship ended. The new provider started from scratch, making defensible but different terminology choices.
Process failures, not people failures
These examples share a pattern. The translators weren’t incompetent. The translations weren’t wrong in any simple sense. The failures were systemic – gaps in process, communication, or context that no amount of translator skill could overcome.
The specification failure happened because the brief didn’t include the customer’s terminology requirements. The translators couldn’t apply standards they didn’t know existed.
The safety instruction failure happened because critical information wasn’t in the source text and the feedback loop to question it didn’t exist. The translators couldn’t add information they weren’t given.
The glossary failure happened because terminology assets weren’t treated as assets. The translators couldn’t maintain consistency with a termbase they never received.
In each case, the fix isn’t “better translators.” It’s better process – clearer briefs, better reference materials, functioning feedback loops, terminology management, and oversight that catches problems before they become expensive.
Red flags worth watching
Certain patterns predict translation failures before they happen.
Rushed timelines create risk. Translation quality depends on translator availability, reference material review, and quality checking time. Compressing timelines compresses quality. When a project “absolutely must” be delivered in half the normal time, something will be sacrificed – usually the careful review that catches problems.
Incomplete briefs create risk. Translators make decisions constantly – terminology choices, style choices, interpretation of ambiguous source text. Without context, they make reasonable decisions that may not be right decisions. The less information translators have, the more likely their reasonable decisions will miss the mark.
Multiple handoffs create risk. When source text passes through multiple parties before reaching translators, context gets lost. When translations pass through multiple reviewers with different priorities, consistency suffers. Simpler chains of communication produce better results.
Cost pressure creates risk. Translation pricing reflects the time and expertise required to do the work properly. Substantially lower prices typically mean less experienced translators, less review, or less time per project. The savings often cost more than they save.
Treating translation as commodity creates risk. When procurement treats translation as interchangeable words-by-the-thousand, they optimise for the wrong things. Translation quality depends on context, consistency, and expertise – none of which appear in unit pricing.
What good process looks like
Preventing translation failures isn’t mysterious. It requires treating translation as a professional discipline rather than a mechanical conversion.
Good briefs include context. Who will read this document? What decisions will they make based on it? What terminology must be used or avoided? What standards apply? What reference materials are available? Translators who understand the purpose of a document make better decisions about how to translate it.
Good terminology management preserves consistency. Termbases capture approved terminology and make it available to translators. When terminology decisions are documented and shared, consistency becomes systematic rather than accidental.
Good review processes catch problems. Technical review by someone who understands both the subject matter and the target language catches errors that linguistic review alone will miss. Review against existing documentation catches consistency issues. Review at draft stage costs less than review after publication.
Good feedback loops improve future work. When errors are found, understanding why they happened prevents recurrence. When translations work well, understanding why helps replicate success. Translation quality improves over time when learning is captured and applied.
At Bubbles, we’ve learned from both successes and failures over nearly 25 years of technical translation. We’ve seen what goes wrong when process fails, and we’ve built systems to prevent the failures that matter most. Because in technical translation, the cost of getting it wrong is always higher than the cost of getting it right.
