Using translated text generated with Machine Translation technology can be harmful to the health of your web content! Twice in the last month we have had clients suffer negative consequences from using translated text from Google Translation as placeholder text in live websites.
Both clients were well aware that Google translation is really not ready for release on client-facing pages without a thorough review by a qualified linguist. So this certainly wasn’t a client education issue.
The conditions facing our clients are typical… in the heat of web development of a large and complex site, developers will come across sections of the site that require quick translation and rather than hold up the process, Google Translation or a similar Machine Translation technology is used. The developer figures he will just go back to that section and make sure the text is appropriate during review.
Of course, Murphy’s Law takes over, many more pages are created, other issues are resolved and the location of placeholders are forgotten! Next the site goes live and you have chunks of ridiculous text dispersed through the site.
In the case of our clients, errant text was discovered because of negative comments they received from live visitors to their website. Very embarrassing…
Our suggestion for properly handling the chunks of missing English text is for the developer to create daily logs of untranslated text and to set up a workflow that allows for the quick translation and review of the key bits of text holding up development. Daily exchange of these logs by the development team and the translation team allows for the rapid completion of the missing text while still using the translation memory, terminology database and the original team that handled the translation of the bulk of your site.
Now, I do have to say, there are valid reasons for using Machine Translation technology. I will cover that in a subsequent post.