A translation workflow design that reduces clicks and speeds job turnover.
Gengo makes it easy for anyone to publish across multiple languages. Gengo’s massive community of translators makes this possible, delivering German tweets in Polish and Japanese product descriptions in Indonesian, often in a few hours.
Job turnaround time is key to translators’ earning power and customer satisfaction, which in turn drives Gengo’s growth. Over the course of several intense weeks, we worked with Gengo’s product team to answer the question, “how can we make translation fun and efficient, whether you have a few minutes or a few hours?”
Working with AQ, we were able to test a bunch of ideas quickly, and in the end, make our translators happy and our word count go up. – Robert Laing, Gengo CEO
For many translators, Gengo is a source of extra income that can be earned in their spare time. We hypothesized that translators could easily complete bite-sized translations from their smartphone, in a few minutes at the end of a lunch break or on the train home from school.
We sketched and prototyped a mobile translation interface that focuses translators on a few lines of text at a time, with source text stacked vertically on top of the translation input. In usability tests, the vertical layout proved very efficient even on a small screen.
Writers can return to sketches days or years later to reflect upon and extend them into longer stories.
Next we took our learnings from smartphone to the desktop, where most Gengo translation happens.
Every translation interface we looked at uses a two-pane, side-by-side layout, for short translations and long. In our testing, these UIs caused too much eye-scanning and mouse movement across the screen, slowing down translation.
When working on short texts, translators kept moving between keyboard and mouse. To translate 25 short jobs required at least 75 clicks!
For longer texts, to our surprise, translators transferred text out of the interface and into a text editor like Word, then pasted the translation back at the end. We also noticed that translators worked line-by-line down the source text, deleting the source at the end to review the completed translation.
We sketched two new interfaces that embrace and improve this line-by-line workflow, one for short jobs and one for long.
Translating is going to be much faster… with far less time spent doing non-translation tasks (clicking, waiting, etc) – Gengo Translator
Taking inspiration from our mobile research with Gengo, the new short job editor focuses the translator on one line at a time, typed into a single text field. The completed jobs cascade down the page in a stream of motivating, visual proof of their productivity and earning power. With the new interface, the clicks required for those same 25 short jobs falls from 75 clicks to 1.
The new interface is proving a hit with translators, and a business win for Gengo, thanks to average job completion speed increases.
For long text translator design, we started with the Word model that translators were already used to, and enhanced it with features inspired by actual translation workflows. Translators can click anywhere in the source text to start translating. The difference in font color helps the translator focus on the translation, while keeping the source close at hand. When the job is finished, the translator can type ctrl-T to isolate the finished translation, make edits, and finally submit the job.
After showing the long form translation UI to a few translators, we realized this was a powerful workspace for translators everywhere, and with Gengo’s permission, are making it available as a working sandbox and a GitHub repository.
Design, research and prototyping conducted with Richard Hatton at Gengo.
Our first bot for distributed teams sends just-in-time meeting notifications from Google Calendar.