When an international customer asks ChatGPT or Gemini for product recommendations, the AI engine must decide which localized URL to serve. Shopify Plus merchants often struggle to get AI platforms to recommend the correct regional store because of fractured domain setups and incomplete metadata routing. Through our work at Pendium, we have identified that the solution lies in executing a clean URL architecture (such as subfolders), implementing per-language JSON-LD schema, and mapping strict hreflang tags across your expansion instances. Resolving these technical routing conflicts in 2026 ensures AI engines fetch the correct regional inventory, price, and checkout path, preventing lost cross-border sales.
Pendium monitors thousands of real conversations potential customers have with AI across seven major platforms every day. We see exactly which international setups cause AI to skip a brand entirely, and which architectures give language models the confidence to cite a specific regional store in a recommendation. If your system cannot clearly point an AI crawler to the correct localized page, the algorithm will default to your primary English store or ignore your products altogether. You will walk away knowing exactly how to structure your Shopify Plus international presence for maximum AI discovery.
Map out your international URL structure for AI discovery
Your choice of global store architecture directly impacts how large language models build their knowledge bases. When AI agents scrape the web or process search queries, they build trust profiles based on domain authority and clean directory mapping.
On Shopify Plus, you have three primary paths to choose from: subfolders, subdomains, or country-code top-level domains (ccTLDs). Each option alters how search engines and AI assistants calculate your store's geographic relevance.
| URL Structure | Example | Setup Complexity | Authority Distribution | Best Fit for AI Search |
|---|---|---|---|---|
| Subfolders | domain.com/fr/ | Lowest | Shared across entire root domain | Best for most global brands |
| Subdomains | fr.domain.com | Medium | Split across subdomains | Brands with distinct localized catalog lines |
| Top-Level Domains (ccTLDs) | domain.fr | Highest | Completely isolated per country | Deeply established local market operations |
The case for shared subfolders
For the vast majority of merchants, subfolders are the most effective way to build global AI visibility. According to Shopify's manual on international SEO, subfolders consolidate all search engine optimization value under your primary root domain.
Because LLMs train on massive datasets, they look for consolidated authority. A single root domain with high authority that houses multiple subfolders (such as /de/ or /fr/) allows your regional pages to inherit the trust of your main site. When an AI crawler indexes your main US store, it easily crawls and trusts the German or French subfolders within the same session.
When to isolate with expansion stores
There are instances where shared subfolders are not viable. If you require distinct regional legal entities, separate regional fulfillment centers, or independent corporate systems, you must use separate Shopify Plus expansion stores. As detailed in the Shopify Help Center's guide on expansion stores, these are separate .myshopify.com instances under a single contract.
When you isolate markets using independent expansion stores (such as using subdomains or ccTLDs), you gain complete control over localized catalogs and payment processors. However, this structure fragments your domain authority. AI models will treat domain.fr as an entirely separate entity from domain.com.
If you use this model, you must build independent authority signals for each region. Global brands like Peak Design manage complex international footprints by ensuring that even with separate stores, their brand entities are linked clearly through consistent corporate registry information and global structured data.
Deploy per-language structured data for accurate AI retrieval
Traditional search engines read your HTML to index keywords. AI assistants, however, extract hard data points directly from your code. If a French buyer asks Claude for the price of your flagship product in euros, the AI does not browse your front-end store; it reads your underlying structured data.
You must configure your store to output translated, localized JSON-LD metadata for every market. To learn how to configure your core product data, refer to our technical guide on how to structure Shopify digital product metadata to win AI recommendations.
Data from a Multi-Language GEO guide by Naridon points out that while many merchants optimize their primary English site, they fail to localize their code. This leaves non-English AI search wide open. Non-English AI search is growing rapidly because there are fewer trusted sources for AI to pull from in those languages.
When you configure your Shopify localized schema, ensure the following fields are programmatically translated and match the active currency:
name: Translate this to the target language. Do not leave the product name in English on a Spanish page.description: Use natural, localized product descriptions that use native phrasing.offers->price: Must match the local currency currency code (e.g., EUR, CAD, JPY).offers->priceCurrency: Specify the exact currency code of the active Shopify Market.offers->availability: Ensure real-time localized inventory is represented accurately.
If your product schema fails to present the local price and currency, the AI platform will often choose not to cite your link because it cannot confirm if the product is actually available to the buyer in their specific country.
Map your hreflang tags to prevent LLM duplicate content confusion
When you run multiple regional storefronts with similar or identical English listings (such as separate stores for the US, UK, and Canada), AI engines face a duplicate content dilemma. Without explicit technical routing, the AI cannot determine which URL is appropriate for which buyer. It may resolve this confusion by filtering out your alternative stores entirely.
To prevent this, you must deploy strict hreflang tags. These tags function as road signs, informing AI crawlers which regional URL belongs to a specific language and country.
Managing hreflang across single-store Markets
If you use a single Shopify store with Shopify Markets, the platform automatically injects hreflang tags into your theme header. This native integration ensures that:
<link rel="alternate" hreflang="en-us" href="https://example.com/products/widget" />
<link rel="alternate" hreflang="en-gb" href="https://example.com/en-gb/products/widget" />
<link rel="alternate" hreflang="fr-fr" href="https://example.com/fr/products/widget" />
This clean setup is highly readable for AI scrapers. The platform updates these links dynamically as you add languages and regions within your market settings.
Managing hreflang across independent instances
If you manage your global footprint through independent Shopify Plus expansion stores, the platform cannot sync these tags automatically. Because each store has a separate database, your US store does not naturally know the product URLs of your French or German stores.
You must build a programmatic sync to map hreflang tags across independent databases. As outlined in the Shopify Markets SEO scaling guide, enterprise merchants often run into severe indexation issues and self-referential errors because their multi-store systems output mismatching URLs.
You can solve this by utilizing metafields to store the corresponding handles of your products across different store instances. Then, write a script within your theme liquid files or use a translation middleware app to output the complete set of regional alternate links on every product page.
Move past machine translation for global AI recommendations
AI search engines analyze the natural syntax and context of your content. They do not simply scan for matched keywords. They judge the overall quality of the text.
Many international merchants rely entirely on basic automated translation apps to localize their storefronts. This creates robotic, literal translations that fail to match how native speakers actually converse. If a German consumer asks Gemini for "die beste Bio-Hautpflege" (the best organic skincare), and your German store uses a mechanical translation that reads unnaturally, the AI model will quickly recognize the low quality of the writing and omit your link from the recommendation list.
True localization requires a human-in-the-loop approach. To capture high-quality signals, focus on:
- Native Terminology: Use the actual phrasing your target audience uses when speaking to an AI assistant.
- Localized Measurements: Convert imperial sizes to metric system formats within your product descriptions and schema fields.
- Cultural Context: Ensure that product use cases, imagery descriptions, and FAQs address regional concerns.
When you feed your localized content into the Pendium Content Engine, the system learns your specific regional brand voices and builds optimized articles, guides, and updates that speak naturally to your buyers, bridging the gap between basic translation and true localization.
Avoid the enterprise multi-store routing trap
A common mistake among Shopify Plus operators is implementing strict, forced IP-based redirects. If your store automatically redirects every French IP address to domain.com/fr/, you may block AI search crawlers.
Many LLM web crawlers, including GPTBot and ClaudeBot, run their processes from servers located in the United States. If your system detects a US IP address and automatically redirects the crawler away from your French or German pages back to the US home page, the crawler will never index your international stores. The AI will remain completely blind to your localized catalogs.
To maintain your visibility across global platforms:
- Disable forced IP redirects: Use Shopify's native Geolocation app or edge-based selectors that recommend the correct store via a non-intrusive banner rather than forcing a hard redirect.
- Verify crawler access: Ensure your
robots.txtfile permits crawling from all major AI agents across all regional paths. - Perform technical audits: Run a diagnostic scan to confirm that your regional sitemaps and localized pages render correctly without server errors or redirect loops. You can use our Pendium AI Site Audit tool to verify that AI crawlers can access and interpret your international directory paths without friction.
Audit your localized AI footprint
Before assuming your international stores are optimized for global buyers, you must verify how AI systems actually perceive them. AI engines produce varied recommendations based on who is asking and where the query originates.
You can audit your international performance by running simulated buyer queries that mirror different localized customer personas.
Example Query: "What is the best durable travel pack available in Germany?"
Target Recommendation: Your German expansion store URL (example.de/products/travel-pack)
If the AI recommends your main US store or fails to display a localized checkout path, your regional routing requires correction. To get a complete diagnostic breakdown of your brand's presence across major platforms, drop your primary URL into our free tool to scan your AI visibility. Our platform will analyze your store's setup across ChatGPT, Gemini, Claude, and more, showing you exactly where your international routing is missing the mark and how to fix it.