Multilingual customer support for Shopify stores with AI

Why does multilingual customer support matter for a Shopify store?

Shoppers buy more comfortably in their own language, and many ask questions before they buy. If your storefront is translated but your answers arrive only in English, the conversation breaks at the moment of decision. Multilingual support keeps the language consistent from product page to answer to checkout.

The demand is measurable. In a 2020 survey of 8,709 consumers in 29 countries, CSA Research found that 76% of online shoppers prefer to buy products with information in their native language, and 40% will never buy from websites in other languages.

It matters at home, too. The U.S. Census Bureau reports that 22% of people age 5 and older in the United States spoke a language other than English at home in 2017 to 2021. A US-only store can still have a meaningful share of shoppers who would rather ask in Spanish, Chinese or another language.

What do Shopify Markets and Translate & Adapt cover, and what do they leave out?

Shopify Markets lets you set up countries or regions, local currencies and the languages each market uses. Translate & Adapt translates storefront content such as products, collections, blog posts, policies and pages. Neither one answers a shopper's question in a chat or email. That support layer is yours to add.

According to Shopify's international selling overview, you can create a market for one country or group several, and tailor the shopping experience for each. Shopify's language settings let you choose which languages each market uses, and each published language gets its own URLs.

Translate & Adapt can translate content manually or automatically with AI translations from Shopify, and lets you customize content for specific markets. Shopify's changelog notes that auto-translation for Arabic and Hebrew arrived in March 2024, with the reminder that your theme must support right-to-left text.

For pricing, Shopify's local currency documentation explains that customers can see prices, pay and receive refunds in their local currency, with automatic or manual exchange rates, rounding and fixed prices per market through catalogs. Keep that in mind: the price a shopper sees depends on their market, which matters for every support answer.

How does an AI assistant decide which language to reply in?

A good assistant replies in the language the shopper writes in, not the language of the page or the browser. It starts in the store's default language, switches when the shopper writes in another, and stays there. Short or mixed messages need care, so it should not flip languages on a single borrowed word.

Language signals can disagree. A shopper may browse your English storefront, have a French browser and type in Spanish. The message itself is the strongest signal of what they want to read. Practical rules for detection:

• Greet in the store's default language, or the language of the market the shopper is in, until they write something.
• Switch on a full sentence, not a single word. "Hola, does this come in blue?" is probably an English conversation with a friendly greeting.
• Keep product names as they appear in the store. Do not translate a product called "Midnight Bloom" into another language unless the store did.
• Mirror the store's formality. Languages with formal and informal "you" make this visible immediately; take the cue from the store's own copy.

What changes for right-to-left languages like Arabic and Hebrew?

Right-to-left languages flip the reading direction of the whole interface: text aligns right, bubbles and icons mirror, and the input field starts on the right. Numbers, prices, SKUs and Latin product names still run left to right inside the sentence, so the widget must handle mixed direction cleanly or prices appear scrambled.

Our glossary entry on RTL explains the basics. In a support chat the common failures are small and very visible: a currency symbol jumps to the wrong side of the number, a parenthesis flips, a phone number reads backward, or a product link with an English name breaks the line order. Each one makes the answer look untrustworthy.

1. Confirm your theme supports right-to-left text, as Shopify's changelog advises for Arabic and Hebrew.
2. Check that the chat window itself mirrors (alignment, close button, input direction), not only the text inside it.
3. Test messages that mix scripts: a Hebrew or Arabic sentence containing a price, an order number and an English product name.
4. Test on a phone, where line wrapping exposes direction bugs that a wide desktop window hides.

How do you keep prices, units and sizes correct across languages?

Never let the model translate or convert a price. Read the price from the store for the shopper's market at the moment of the answer, and display it as the store does. Treat sizes and units the same way: quote the store's own size chart and measurements rather than converting them in the reply.

Language models are good at language but can be unreliable with arithmetic and have no access to live data on their own. Asked in German about a product priced in US dollars, a model may helpfully "convert" it with an exchange rate it remembers. That figure will not match checkout, and the customer will notice. The fix is architectural: prices come from the store, not from the model. Read more about how AI assistants get prices and stock wrong and the grounding that prevents it.

When does a multilingual conversation need a human?

Hand off to a person when the shopper is upset, when the question involves a decision (refund, exception, damaged item), when legal or safety questions come up, or when the assistant cannot answer confidently in that language. The shopper should be told who will reply and roughly when, in their own language.

Our support automation page shows how routine questions and hand-offs fit together. The awkward case is a hand-off to a team that does not speak the shopper's language. Plan for it rather than discovering it live:

• Pass the conversation with a translation attached, so your agent sees both the original and an English version.
• Let your agent reply in their own language and translate the reply, with a short note that a translation tool was used.
• Keep policy text in its approved translation. If a return window is in your translated policy, quote that text, not a fresh paraphrase.
• Set expectations about response time honestly, especially across time zones.

How does Tycoon handle multilingual support?

Tycoon detects the store's main language from the site on install and works in it, answers each shopper in the language they write in, and lets the merchant change the default language anytime. It supports 16 languages, with right-to-left layouts for Arabic, Urdu and Hebrew, and every language is included on every plan.

The full list is English, Simplified Chinese, Hindi, Spanish, Arabic, French, Bengali, Portuguese, Russian, Urdu, Indonesian, German, Japanese, Turkish, Korean and Hebrew. You can see how it looks on the languages page.

Whatever the language, price, compare-at price and stock are read live from the store for every answer, and the model is never the source of a price or a factual claim. Tycoon picks up the store's voice from its own pages, and it hands off to the merchant's team when a person is needed. To see it reply in your shoppers' languages, book a demo on your own store.

Frequently asked questions

Sources

1. CSA Research: Consumers Prefer their Own Language (July 7, 2020) · checked October 20, 2026
2. U.S. Census Bureau: 2017 to 2021 ACS language use tables (June 3, 2025) · checked October 20, 2026
3. Shopify Help Center: International selling and Shopify Markets · checked October 20, 2026
4. Shopify Help Center: Languages · checked October 20, 2026
5. Shopify Help Center: Translate & Adapt app · checked October 20, 2026
6. Shopify Changelog: Auto-translating Arabic and Hebrew in Translate & Adapt · checked October 20, 2026
7. Shopify Help Center: Pricing in local currencies · checked October 20, 2026