Item: Phrase Localization Platform
Rating: 6
Author: Verified User

Overall Satisfaction with Phrase

Use Cases and Deployment Scope

We use Phrase to translate our site text into multiple languages so that users can select their preferred language across all our domains. Our translations are done both through Phrase's automated machine translation and by translators who then upload their work into Phrase. My company also makes all copy updates by creating keys and entering the text in Phrase, allowing us to make changes to our site text without involving development work (beyond the integration of the key). Phrase allows us to be flexible in making text and translation updates with minimal development effort.

Pros and Cons

Pros

Creating and assigning jobs for translations, updates, etc.
Automatic machine translation integration is useful
Easy to switch between locales to make adjustments

Cons

All exports from Phrase to GitHub are compiled into a single PR which means that we have to create a 'queue' when merging branches. If anyone has an open PR for Phrase changes, no one else can merge and export until that PR is merged to integration, or else the second person's changes will be rolled into the first person's PR
If you create a key and use the auto-fill machine translation feature but later make edits to the key, the auto-translated text will not be updated
We've had several cases of conflicts between our yml files and the text entered into Phrase due to things like spaces or quotation marks not appearing correctly

Most Important Features

Flexible text editing
Machine auto-translation
Integration with Jira & GitHub

Return on Investment

Positive: ability to add and update text without involving our dev team
Positive: can quickly translate text using machine translation
Positive: assigning jobs allows for translators to have a clear workflow for larger translation tasks
Negative: limitations in the GitHub export process cause bottlenecks when multiple people need to make edits
Negative: keys that were created using auto-translation don't inherit any updates made to the default text once the key has been made

Alternatives Considered

(not applicable)

Key Insights

Do you think Phrase Localization Platform delivers good value for the price?

Yes

Are you happy with Phrase Localization Platform's feature set?

Yes

Did Phrase Localization Platform live up to sales and marketing promises?

I wasn't involved with the selection/purchase process

Did implementation of Phrase Localization Platform go as expected?

I wasn't involved with the implementation phase

Would you buy Phrase Localization Platform again?

Yes

Other Software Used

GitHub, Sketch, Atlassian JIRA Align (formerly AgileCraft)

Likelihood to Recommend

Phrase has lightened our dev effort significantly for making text changes. We can now assign these tasks to our product managers and the process of creating branches and keys is very streamlined. We've also found the auto-translations to be helpful for smaller pieces of text that do not require a translator; when we do use a translator, Phrase's "Job" feature allows us to simply assign certain keys.

While the workflow within Phrase is uncomplicated, we've encountered issues with how Phrase and GitHub integrate. All changes exported from Phrase go to a single GitHub PR, meaning we have to continually check amongst our team that no one has an open PR before we merge and export our branches. If someone has time-consuming changes, this creates a significant bottleneck. Also, even if the Phrase work was specific to a single locale, files from all the locales are exported to GitHub. This means that reviewers often have to sift through unnecessary files in order to find the relevant updates.

Comments

Please log in to join the conversation

Phrase: time saved on development is then wasted on workflow management