I grabbed a big editing collection (400 jobs), and the 1st pass is just abysmal.


The 1st pass translator didn't take into account a very basic indication from the customer about an important key term, resulting in (if I do a basic text search of the incorrect term they used) +100 corrections I have to make just on that.


But hey, we all make mistakes, right? Who didn't forgot to check the customer's indications at least once? But at some point, I realized this person was just using automatic translation without as much as giving it a glance, since I'm finding blatantly nonsensical translations, stuff like mistaking the windows of a house description for the operating system (so, leaving "Windows" in English and capitalized), that kind of thing any human would immediately notice. Out of curiosity, I ran bits of this source text through some translation engine just to compare, and they didn't even bother to use a good one at that.


This customer has been ordering several jobs like this one this days, so I'm led to suspect this person knew from previous experience that this would undergo a 2nd pass and that they could allow themselves to get sloppy, knowing well that someone else would have to clean their mess (and that Gengo reviews only go for the final pass of collections, if I'm not mistaken!).


This not only unethical and unprofessional, but also disrespectful to whomever gets to do the editing job (not to mention the customer and Gengo), since you can get in the situation of having to just retranslate entire segments for half the pay of regular jobs.


I have to admit that even with that, this specific collection is still good money, so this time I'll just finish it anyways. If it was some small collection, I'd just decline it, and maybe consider leaving a comment about the original translation's state for fellow editors.


What do you people think? Where do you draw the line?


Best regards.


  • 2


    I agree with you.

    Nevertheless, concerning your "Gengo reviews only go for the final pass of collections, if I'm not mistaken", Gengo clearly states :

    "Who gets the score if an Edit job is reviewed by a Language Specialist?

    We'll be able to review both the initial translation or the edited version, and each translator will get their individual quality score."

    You can check more details about bad initial translation here :


    Hope it helps!


  • 3

    If you sure it's an edit job*, and by your description it sounds like it, I would probably flag clearly machine translated jobs (not sure if you can do that after having edited the translation, though) and perhaps even reach out to support directly if it's really bad. Not sure what flagging actually does, i.e. if this has any consequences for the first pass translator.
    If there were several jobs like this, especially at the same time, it's possible they picked up more than the one you got by doing what you described, taking them away from translators who are working diligently and causing an even bigger problem.

    *Sometimes there are pre-filled machine translation post editing jobs, but I think those don't go under the "edit" category these days. To be sure, you can check in the right hand column after clicking on the lightbulb. MT is machine translation, TM is translation memory. Probably clear, but just in case.



  • 2

    Thanks for the info, Fudesaki! I didn't know that.


    As for Chris, thanks for your suggestions, and yeah, this is indeed an edit job, I can see on the history how another translator took it and submitted it before me and there are no.



  • 2
    carla m.

    You should report that translator to the team.

  • 4

    It happened again (same customer, same type of job, same scenario) and I followed Carla's suggestion and reported it this time.

  • 1
    Katrina Paterson

    Thanks to everyone that's contributed - and yes, as Chris and carla m. say, we'd always recommend flagging poorly-translated first-pass translations and contacting Support, providing as many details as possible, so that the situation can be looked into. The help article that Fudesaki linked to has more information about this too. As a general rule, it is worth reporting any kind of unusual translator activity because this creates a record which helps us to build a better overall picture of how translators are behaving, and it also means that we are better placed to take action where needed. It's also worth reporting technical issues, if and when you see these, for the same reason. I hope that helps.

Please sign in to leave a comment.