Run one study in 20 languages

Run a multilingual survey in Versive: AI translates one study into up to 20 languages, with editable text and results that stay in sync.

Versive can translate one study into up to 20 languages: Arabic, Chinese, Dutch, English, Finnish, French, German, Hindi, Indonesian, Italian, Japanese, Korean, Polish, Portuguese, Russian, Spanish, Swedish, Turkish, Ukrainian, and Vietnamese. Translation covers every question plus the welcome, thank-you, and screen-out messages. Participants can take the study in their preferred language while you maintain a single source study.

This guide explains language settings, the participant experience, translation review, and multilingual results.

Turn on the languages your study needs

Language settings live under Settings → Languages, and you can edit them while building the study. Configure three settings:

Once you add a supported language, Versive translates the study in the background and shows the translation status for each language.

For example, write the study in English and set it as the default language, add Spanish, French, and German as supported languages, and leave the starting language unset so everyone lands on English unless you route them otherwise. Use the per-link override below to send each cohort to its language.

What participants see

A participant's starting language is whatever you set (or the default language, if you didn't). You can also override the starting language per link by appending ?lang= and a language code to the study URL:

https://getversive.com/s/your-study-id?lang=de

Use this parameter when recruiting separate cohorts by language and sending each group a link that opens in its language. See How to recruit research participants (4 ways) for recruiting options. Within the study, participants can switch between any supported language.

The live preview and AI simulation render in the language selected in the editor. Use them to review each translated participant experience before distributing the study.

Review and edit the AI's translations

AI translation produces the first draft. You or a native-speaking teammate can then switch languages in the study editor and revise any translated question, welcome message, thank-you message, or screen-out message. The editor shows the text that participants in that language will see.

Manual edits are tracked and preserved. If you later change the source study, re-translation updates only the fields you have not edited by hand. Review idioms, register, and industry terms in particular, since a literal translation may be accurate but sound unnatural. You can revise individual translations after launch without having those edits overwritten by a later re-translation.

Questions can also use a different image or media asset per language. This is useful when the stimulus contains text, such as a screenshot of a localized product screen. A language without its own asset falls back to the default, so add variants only where the assets differ.

Study structure stays the same

Translation changes text and media while preserving the study's structure. Branches, screening rules, and question order stay as you built them in every language. See Skip logic and branching, explained for help setting up that structure.

Keep your results together

Bulk transcript exports include translated transcripts, so responses from several languages can be reviewed in one export. An analyst who does not speak every participant language can work from a consistent translated set. When you are ready to synthesize the responses, see Build a research report stakeholders will read.

Building multilingual studies programmatically

If you're building studies through automation rather than the editor, the same capabilities are available through the API: you can trigger translation, read the AI-generated translations, and submit corrections programmatically. The study creation agent also accepts a language parameter directly, so a multilingual study can be built and translated end to end without anyone opening the app.