Blog

  Natalia Manidis

Translator Performance Scorecards, Better Job Allocation

4 Comments
alignnone

Today we’re announcing big changes to the way we assess performance, share information and allocate jobs. Some of these are features you’ve asked for; others will help our customers as Gengo grows. We think you’ll be excited.

Performance Scorecard

Your dashboard now features a personalized scorecard showing your performance as a Gengo translator. The scorecard focuses on speed and quality, and also gives you a summary of the volume of your work at Gengo.

Quality

The main quality metric is your average job quality rating, from job reviews by Senior Translators. Job scores are based on the number of major and minor errors they find and have a range of 0 (lowest, many errors) to 10 (highest, zero errors). Not all translators have been reviewed yet. If you are new to Gengo, or you have not completed a job recently, your score may be empty. We’re working towards a goal of all active translators checked by June 2013. So, quite literally, you can watch that space for your score!

Speed

Your speed is measured by the number of units (words or characters) you have translated divided by your working hours on Gengo (the time from when you pick up a job to the time you submit it). We’ve included speed because it’s an important feature of Gengo’s service, and we also want to discourage translators from picking up jobs without starting them. We do NOT want you to sacrifice quality for speed, but we do want you to start working immediately, and get jobs back to customers as quickly as you can. We’ve set an initial sweet spot of ~250 units/hour — how does your average speed compare?

Note: These metrics are not set in stone. In the future, we’d like to differentiate between character- and word-based languages, as well as Standard and Pro. We’d also like to make improvements based on your feedback — please add it here.

Job Allocation

Recently we’ve been experimenting with better models than our “first come, first served” system for picking up jobs. Under the new system, your access to jobs will be based primarily on the things in your scorecard: the quality and speed of your translation. New translators will also be given a greater chance to access jobs, so that they can build up their score. This will give a larger group of translators access to jobs, because pickup will be determined not by the speed of your mouse, but your ability to do good work over time. We’ll continually adapt the system over time, but we’ll tell you when we make major changes.

This means you’ll be rewarded for high performance, and our customers will enjoy an improved experience. Sounds good, right?

That’s all for now! We’re confident that these changes will help Gengo become a fairer and more rewarding place for you to work. We’re keen to hear your feedback, so we’ve created a forum specifically for this topic. Please feel free to add any comments or questions here or contact us directly.

Natalia Manidis
THE AUTHOR
Natalia Manidis
Natalia is Gengo's Translator Community Manager. As a Japanese-speaking Australian with an international background (pinch of Greek, sprinkle of Dutch), she enjoys the opportunity Gengo offers to collaborate with people from all corners of the world. Natalia is passionate about building a vibrant learning community and discovering the best places to eat in San Francisco.
  • http://www.incion.com/ Kim Meredith Smith

    All of them are really awesome and all tips are unique as well..

  • Leo

    Well, i see one really important disadvantage caused by your system: imagine, that i start a bigger job and i will submit it in two days. In this case all the time will be calculated, including time for sleep, eating, walking – whatever.

    For example, the text containing 5000 words can generally be translated within 16 hours(300 words per hour). But nobody will perform the translation during 16 consecutive hours, as that is totally exhausting for a translator. So we may break our time in two parts, and work with an actual speed of 300 words, but your system will register a reasonably lower speed.

    Given that quick on-line and large off-line jobs will be counted together while calculating the translator’s speed, your system will register a significantly lower output for the translator.

    My proposal is therefore to subtract 16 hours from every 24 hours given for a job while calculating the translators performance for an off-line job.

  • Leo

    i could not send my feedback by following your link(target page is nonexistent), so please consider my previous post a feedback. Thanks!

  • Devin

    How is quality quantified exactly, and by whom?

    I have received several 10/10′s on my job performance gradings along with a short comment like “great job,” but once or twice I have gotten a lower score– once a 4.6. How is this measured? Is there an actual system that gets into tenths of points, or is the number completely arbitrary?

    It doesn’t affect me too terribly, as I receive 10′s more frequently, but I do have to wonder why I’ve gotten lower grades and who gave them. Are certain “graders” stricter, or is there an objective grading system?