Pretranslation in Pontoon: upcoming general availability and opt-in guidelines

Hello localizers!

We already wrote on Discourse about the pretranslation feature in Pontoon, but it’s time to provide you with an update. In case you missed it, we just published a blog post describing the results of the Alpha and Beta testing, and we’re quickly approaching general availability for all supported locales.

This means that locales will be able to opt in directly from Pontoon to use pretranslation in projects, and we need guidelines to evaluate these requests, similarly to how we manage requests to add locales to projects.

These are the guidelines we plan to start with. It’s important to note that these are not strict criteria: members of staff will evaluate each request to opt in individually, based on their knowledge of the project and direct experience with the locale.

Criteria for enabling pretranslation for a new locale

  • Request needs to come from translators or managers active within the last month (translating or reviewing).
  • There is an active manager for the locale (last activity within 2 months).

Criteria for enabling pretranslation for a new project

  • Less than 400 missing strings, except for projects or locales where existing pretranslation statistics provide high-confidence.
  • Average review time for pretranslations in existing projects is faster than 3 weeks.

Criteria for disabling the feature for a locale or a project

  • Approval rate drops below 40%.
  • Average review time for pretranslations is slower than 6 weeks.

Note that disabling a project would always involve a conversation with reviewers for the locale.

Here are a few additional notes on how we formulated these guidelines:

  • Only translators and managers have the permissions required to review pretranslations. This is why we want the request to originate from a translator/manager active within the last month, as that signals their availability to review translations within the expected timeframe. Note that, for similar reasons, contributors won’t be able to request opt-in.
  • We also want an active manager present, because it might be necessary to promote users to account for the increased review load and the overall different approach to localization.
  • 400 missing strings is an indicative limit. It was chosen because most new projects should remain below this threshold at the start, and therefore be eligible for pretranslation. On the other hand, pretranslating a large batch of strings will create a sudden increase in review load, and we want to avoid that unless the locale has shown sufficient review bandwidth in the past.
  • For the review time, we chose 3 weeks since most of our products ship on a 4-week release schedule. That increases the chances that content will be reviewed by an experienced localizer before it reaches general audience.

Initially, we’ll ask new locales to start with a small number of projects, to make sure that the performance is aligned with the results we had during Beta.

We’re happy to hear your thoughts and feedback about this.