If you launch with a mediocre product, you’d expect low sales and high support costs, right? Well, when you enter new markets with mediocre language quality in your user support or marketing materials, you get the same results — even when your product is impeccable.
Language quality is how your target customers discover and experience your solution. Poor quality at best leaves your customers confused — at worst, it irreparably damages your brand — and in all cases, it demands expensive rework to publish corrected content later.
Some would like to believe the translation-edit-proof (TEP) process generates exactly the right quality control with no further effort required. That’s like thinking your development team could take a specification document and magically produce the exact software you’ve envisioned with no supervision, feedback, clarification, adjustments, or continual improvement required.
In reality, localization QA requires a number of tasks that occur before, during and after the TEP process. An earlier blog post details these language quality services.
Without exception, it’s better for your entire localization program to incorporate quality improvement from the onset of your global market strategy. Yet as global companies gain localization maturity, many come to quality management reactively and urgently in the face of costly quality issues.
If you suspect you’re not doing all you can to optimize linguistic quality proactively, or that quality isn’t as high as you’d prefer, here are five symptoms that can help you diagnose and escalate quality management in your localization program.
1. Negative feedback from your in-country reviewers
Many global companies use in-country employees and partners to check translated materials before they go live. When translations come back with extensive edits and comments, you need a deeper quality check and remediation before any further talk of publication.
That said, you shouldn’t infer perfection from a lack of linguistic feedback: most client-supplied reviewers have day jobs, and don’t have time to complete a thorough review. Neither are they trained reviewers, but rather bilinguals with an interest in the product, whose feedback is valuable when they have time to devote. (See here for more on how to get the most from your in-country reviewers.)
2. Lower user acceptance than expected
Let’s assume you did extensive market research and got to know your target demographic intimately. You adapted your brand for new markets, built specialized marketing materials, and translated your web content. If sales forecasts exceed the actual growth in-country, it may indicate the localization effort was sub-optimal, containing errors that alienated your potential new user base. In-country focus groups or user testing might have helped you identify the problems prior to rollout.
3. Too many linguistic bugs logged in localization testing
After the UI or website build stage, quality assurance testing may reveal linguistic or display errors caused by the engineering and localization process. If this testing reveals a lot of linguistic errors, then there may be a quality issue in the source files, in the build process, or in the translations themselves. The amount of time spent after testing for each individual language – finding, logging, confirming, escalating, fixing, reviewing and re-confirming – is enormous compared to the amount of time required to pinpoint the root cause and fix it at its source.
4. Disproportionately high support demands in one or more markets
Let’s say your phone lines and customer support chats have gone crazy in Poland and Mexico, but everywhere else they are steady or even declining. Standout problems in specific markets suggest that your marketing materials, user help materials, or FAQs might not be high enough quality to get buyers oriented and comfortable using your product.
5. High site traffic bounce rates
Bounce rates indicate how quickly web site users leave a page after arriving. Do your site visitors find your site and leave within a few minutes or even seconds? That could be evidence of a quality problem, right there, on your landing page — the one you’ve optimized to capture as much traffic as possible to draw people in.
The big picture
Nobody likes to learn content went live with poor linguistic quality. But look at the bright side: diagnosing the problem from any of these symptoms is far more pleasant than becoming the Internet’s next big laughingstock when high-profile issues go viral, or worse, when you never get feedback and your offerings fall flat. When you look at the sources of quality breaks you will then be able to put the right holistic quality program in place.
Have you ever dealt with quality issues reactively? How did you diagnose the problem, and what did it take to make things right?