> For the complete documentation index, see [llms.txt](https://docs.elevateab.com/elevate-helpcenter/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.elevateab.com/elevate-helpcenter/page-testing/shopify-translate-and-adapt.md).

# Shopify Translate & Adapt

***

If your store uses Shopify's Translate and Adapt app, you need to translate your experiment variations yourself. Translations are not carried over automatically.

This applies to any experiment that creates a new template, which means page experiments and any other type that duplicates a template.

### Why this is needed

Translate and Adapt stores translations against a specific template. When Elevate A/B Testing duplicates a template for a variation, the copy is a new file, and a new file has no translations.

Skip this and visitors browsing in any language other than your default see untranslated content on the variation and translated content on the control. That is a second, unintended difference between your variations, and it will move your numbers as much as the change you meant to test.

### Identifying the variation template

Elevate A/B Testing names duplicated templates with an `eab-` identifier. If your control template is `templates/index.json`, the variation is named something like `templates/index.eab-XXXXXXXX.json`, where the suffix is generated for that variation.

That name is how you find the variation inside Translate and Adapt. Note it when you create the experiment.

### Translate a variation

**Prerequisites:** the experiment created, and the variation finished in the Shopify theme editor.

1. **Finish your changes first.** Complete every edit to the variation template before translating. Translating first means retranslating anything you change afterwards.
2. **Open Translate and Adapt** in your Shopify admin.
3. **Find the variation template** in the sidebar using its `eab-` identifier. Use the search field if the list is long.
4. **Translate every language your store supports.** Cover everything visible on the page: headings, body copy, button labels, section content, and any text inside the sections you changed.
5. **Match the control.** Translations on the variation should say the same thing as the control, except for the change you are deliberately testing. An untranslated or differently worded string is an extra variable.
6. **Review in each language before launching.** Use Shopify's market preview, or a VPN to view the store as a visitor in each market would.

Then launch the experiment.

### What to watch for

**Longer and shorter text.** German and French commonly run longer than English. A layout that holds together in your default language can break in another. Check the variation at mobile width in the longest language your store supports.

**Locale prefixed URLs.** If you use Shopify Markets with locale prefixes such as `/fr/` or `/de/`, experiments run across them. During QA, visit the experiment page through each prefix to confirm.

**Newly added sections.** A section you added to the variation and not the control has no equivalent on the control to copy translations from. It needs translating from scratch in every language.

### Troubleshooting

**I cannot find the variation template in Translate and Adapt.** Look for the template name containing the `eab-` identifier. If your control is `templates/index.json`, the variation is `templates/index.eab-` followed by a generated suffix. Search rather than scrolling.

**The variation shows untranslated content in some languages.** Parts of the variation have not been translated for those languages. Open the variation in Translate and Adapt and check every section, not just the ones you changed.

**The layout breaks in certain languages.** Translated text ran longer or shorter than your default language. Adjust the variation template so it handles the longest translation, and check the control for the same problem while you are there.

**The experiment does not seem to run on some locales.** Visit the page through each locale prefix directly. If the experiment is genuinely absent on one locale, check the experiment's audience targeting for a country or market condition you did not intend.

### Build it into the workflow

Translation is easy to leave until after launch, at which point the experiment is already collecting contaminated data. Fold it into setup instead:

1. Create the variation.
2. Customise it in the Shopify theme editor.
3. Translate it in Translate and Adapt.
4. Review it in every language.
5. QA and launch.

### Next steps

* Preview and QA a Page Experiment
* Troubleshooting Page Experiments