Quick answer — Prepare incident templates in every language in advance, because nothing gets translated well at 2am during an outage. Live updates can run in English with a translated summary if the templates cover the rest.
Vitra.ai Universe ships product and customer content in every language.
Nothing gets translated during an incident
An outage is the worst possible time to start a translation workflow. The person writing the update is also handling the incident, the wording changes every twenty minutes, and accuracy matters more than polish.
So the work happens before, not during.
Prepare the templates
Most incident communication is more formulaic than it feels while writing it.
| Template | Prepare |
|---|---|
| Investigating | Yes |
| Identified | Yes |
| Monitoring | Yes |
| Resolved | Yes |
| Scheduled maintenance | Yes |
| Post-incident summary | Partly — structure only |
Translate those once, in every language you support, and an incident becomes choosing a template and filling a short variable rather than composing under pressure.
What to do with the live detail
Two honest options, and both are better than silence. Run the detailed timeline in English and publish a translated status line — investigating, identified, resolved — so a customer in any language knows the state without waiting. Or translate updates automatically and label them as machine translated, the same convention as user-generated content. Speed matters more than polish here, and a labelled fast translation beats an accurate one that arrives after the incident closed.
Component names stay in English
Status pages list services and components, and those names should match what customers see in the product and the API.
Translating them creates a customer who cannot map the affected component to the thing they are using, which is the one job a status page has.
The post-incident report is different
It is written afterwards, read carefully, and often by procurement or security teams evaluating whether to stay.
That one deserves proper translation with review, held in memory so the terminology matches the terms and any commitments referenced in it. Run quality control before publishing — a post-incident report is a document customers keep.
FAQ
Should incident updates be translated live? Not composed live. Prepare the standard templates — investigating, identified, monitoring, resolved — in every language in advance, so an incident becomes filling a variable rather than writing under pressure.
What if the detail cannot be translated fast enough? Publish the status line in every language and run the detailed timeline in English, or translate automatically and label it as machine translated. Both beat silence during an outage.
Should status page component names be translated? No. They should match what customers see in the product and the API, or a customer cannot map the affected component to the service they are using.
How should post-incident reports be handled? Properly translated with review. They are read carefully, often by security and procurement teams deciding whether to stay, and customers keep them.
