Quality control is an integral part of software localization work. That sounds self-evident, right? Like something that has to be there, because you don’t want to end up on a list like this. But the big question is how (and of course, how much).
The common approach to quality control in localization has traditionally been translating-editing-proofreading (in industry jargon simply TEP). This process consists of two or more professionals, one who translates the text and the other who proofreads it. This often contains a lot of sending files back and forth and some discussions, maybe even arguments, about terminology or grammar. As with all human processes, the end results depend on the individuals involved, but basically traditional TEP offers a good chance of getting good quality results.
But does this great process happen in real translation life? Sure, but all too often you might not want to allocate enough budget to have two professionals on board and you skip the proofreading step. Or you think you don’t have enough time and skip the proofreading step. Or you buy the whole process from someone who says “all translations are proofread by a second professional”, but who still skips the proofreading step. So you end up with a translation that your translation subcontractor (who you found on the Internet without being able to check the references) says is OK, but you have no chance of checking if it really is. Unless you want to pay more and send it to a second vendor for review. Which not only costs money, but also takes time. And if you don’t have a nice budget in your hands, you are probably bound to skip localization testing too. Which means the quality of that single professional is what your end-users will see, be it good or bad.
It is easy to see that localization quality has risks attached to it. We haven’t even discussed style issues yet: good grammar and fabulous Shakespearian wordings might still not cut it, if the software doesn’t speak the kind of language expected by the buyers.
The end-users are of course the ones who should be happy with the localized texts. What if they would be the one who helped you in quality assurance? Heck, if you have a nice app with loyal users, there’s a fair chance they’d love your software even more having been allowed to participate in the process.