The real cost of translating an Articulate Rise library - and what changes with built-in AI translation

Translating an Articulate Rise course library is one of those L&D workflows that everyone agrees is painful and almost no-one has fixed. The mechanics haven't really changed in years: export XLIFF, send to a translation agency or vendor, wait, receive the translated file back, import it, fix the layout issues, repeat per language, repeat per update.

The cost is rarely the headline number. It's the time, the version drift, and the fact that any meaningful update to the source course resets the whole process for every language.

This piece walks through what the Rise translation workflow actually costs in time and money for a realistic library, and what changes when the same library is moved into a platform with built-in AI translation and narration.

The standard Rise translation workflow

The supported path in Rise is XLIFF export and re-import. In practice, for a team running this across a real library, the steps are:

1. Export the course as XLIFF from Rise
2. Send the file to a translation vendor or in-house translator
3. Receive the translated XLIFF back, per language
4. Import each language file back into Rise as a duplicated course
5. Review the translated course for layout breakage, truncation, image text that wasn't captured, and tone-of-voice issues
6. If audio narration is involved, separately record and upload narration in the target language, block by block
7. Re-export each language as a SCORM package
8. Deploy each language package to the LMS, track which version is current
9. When the source course is updated, repeat steps 1-8 for every language

For a single short course in two languages this is annoying. For a library of fifty courses in five languages with regular regulatory updates, it is a full-time job and a permanent source of version drift.

What it actually costs

Agency translation pricing for e-learning varies, but a reasonable midpoint for a typical Rise course - say, 4,000 to 6,000 source words - is somewhere between £400 and £900 per language for the translation alone. That excludes review, layout fixes, and narration.

For a library of fifty courses across five languages, you're looking at £100,000 to £225,000 in translation fees alone over the cycle of producing the library, before any updates. Add narration recording, project management, and the time lost to version control, and the realistic total is meaningfully higher.

The hidden cost is the one that doesn't show up on an invoice: the library effectively freezes. Updates become expensive, so updates stop happening. Courses drift out of date. Compliance teams lose confidence in what's actually live.

What changes with built-in AI translation and narration

When the same library lives in CourseAgent, translation is a one-click action on a finished course. The course is translated into any of 19 languages in minutes, not weeks. The translation is included in the subscription rather than billed per word. AI narration in the target language is generated in the same flow.

That changes three things in practice:

The unit cost collapses. A library that previously cost six figures in translation fees over its lifetime is included in a subscription that's a fraction of one of those agency invoices. The maths stops being "can we afford to translate this course?" and starts being "is there any reason not to offer this course in every language we have learners in?"

Updates stop being expensive. When the source course changes, AI Compare & Update identifies the changed sections and updates only those - then re-translates and re-narrates only what changed, in every language. The full re-translation cycle stops being a project. It becomes a re-publish.

Quality is reviewable and editable. AI translation is not perfect. The point is not that it replaces human review where the content is sensitive. It's that the baseline translation is generated in minutes, in-platform, and any edits live in the same course rather than in a XLIFF file being passed back and forth between systems.

Human review still matters for legal language, regulated content, and culturally sensitive material. What changes is that the human review starts from a usable draft generated in minutes rather than from a process that takes weeks.

"But we have all of this in Rise already"

This is the objection that usually closes the conversation. It's also why CourseAgent imports a Rise course as a native, fully editable course in under ten seconds. The translation argument is theoretical if the library can't move. Once the importer is real - and it is - the question becomes whether the translation, narration and update cost reductions justify migrating the library that's most expensive to maintain in Rise.

In most cases that's the multilingual subset of the library. That's where the savings compound fastest and where the version drift is most painful.

A practical sequence

For most teams the lowest-risk path is:

1. Identify the three to five Rise courses that currently live in more than one language, or that should but don't because of cost
2. Import them into CourseAgent - the import itself takes seconds per course
3. Translate them into the languages you actually need, generate narration, and review
4. Compare the result against the existing translated versions, on quality, time, and cost
5. Decide whether the rest of the multilingual subset moves across, then whether the wider library follows

The point isn't to make a strategic platform decision in the abstract. It's to put the work through the workflow on courses you already have, and see whether the numbers and the output actually hold up. They usually do, and they usually surprise the team that runs the test.

The library is the lock-in. Once that's no longer true, the translation maths is hard to argue with.