Here’s an excerpt from an email thread I got last week, with a sentence that you’ve probably seen a lot of in the last few years: “Let’s not do that, budgets are really tight for the next quarter.”
The context for this budget discussion was a question: should we pay extra for some optimization work to be done by our translation vendor? The prevailing assumption was that, since the original copy’s purpose was oriented entirely around what customers were looking for in search, we wouldn’t need to optimize it again on the translation side. At first glance, it sure seems like a great way to save money, right? That is exactly the gamble made on how much to invest in the last mile of localization.
Language and Country Choices
Since we’re talking about search, there are at least 200 considerations to make for every choice. And since we’re talking about global, there are hundreds of country, language, and local considerations to make. Translation deliverables range from highly optimized, locally reviewed pieces to the same automated replacements that make the occasionally off-kilter voice transcript scripts.
Another tricky balance is determining what you win or lose by consolidating countries and languages. While it’s somewhat logical to lump countries by a common language, but then you hit problems like:
- Parisian French is not Quebecois French
- Colombian Spanish is not like the Spanish of Europe
- British English and its aversion to z’s and excessive u’s (compared to the U.S.)
- Switzerland, happy home to a trifecta of languages
There’s a risk in every efficiency step taken to deliver mass quantities of content, because much of it is done at the cost of language nuances. Aside from the high variation in local keyword volume between one synonym and the next, expediency forces a sacrifice of the niche.
Translation, Synonyms , and Nuance
Every language has its exact way of capturing nuanced and niche descriptions that cannot be perfectly translated into any other language. When you force a prescriptive 1:1 exchange as if words were numerical database entries, you lose the right word selection from multiple meanings. It’s often only revealed through syntax. You’ve shredded translation accuracy as well as chopped the long tail.
Large-scale content delivery still must be appropriate for what cultures want, too. In some countries, business is the language of an industry; keeping English copy might be appropriate, if not preferable. Reliance on communication channels varies from location to location, too. According to the 2010 TechTarget study on global media consumption trends, the type of content preferred is different for each country surveyed – as does reception to a phone call or use of search engines versus trade publishers. A BCG study found very diverse country-level Internet economies.
Making Your Call
However, a more realistic approach in today’s financial roller coaster is to craft our global approach according to the 80/20 rule. If we have to cut our costs, let’s at least attend to the vital few. If we miss on vital communications to our vital markets – both the current revenue giants and the fast-growing titans – then we might as well fire the writers, pump out jargon-laden English, and leave both search and customer experience at the door.
Realistically, if you are blind to your content quality from the beginning, localization won’t fix the fact that your words haven’t added value to the user. But if you’re going to bet on the voice of the customer, then don’t get sloppy by playing the wrong stakes globally.
Maura will be speaking on the panel, Global Search Marketing Best Practices Roundtable, at SES SF, Aug. 17, 2010.