Multilingual SEO best practices, from ranking to being quoted
Your French and German pages may rank and still bring in few enquiries. More buyers now get their answer from an AI summary or a chatbot before they click, and the brand quoted in their language is the one they contact.
Ranking is still the foundation, and now the first half of the job. The work that gets you quoted is generative engine optimization (GEO). Below: the multilingual SEO best practices that get each language version found, and cited too.
Show search engines which version serves which market
When search engines can tell which version of your site is meant for which market, an excellent translation gets seen. Localized multilingual SEO puts each version in front of the people looking for it.
The groundwork, in every language:
- Language tags on each page.
- Region-specific sitemaps.
- Localized URL structures.
The same signals are now the foundation for GEO in each language, so a gap here costs you twice.
Keyword localization: start from how each market searches
Localized keywords are the terms buyers in each market actually type, which a translation of your English list rarely finds. Keyword localization starts from how buyers in each country search, often in words of their own.
Each market needs its own research, covering:
- Regional dialects.
- Colloquial phrasing.
- Local intent.
Google Keyword Planner, Semrush and Ahrefs all support it, but native-level knowledge is what separates real queries from assumptions carried over from English. Precise local wording also helps AI models associate your brand with the right topics in each language.
Write meta tags for each language
Your title and description are the first sales copy a searcher in each market reads, often before they know your name. Every language version needs its own:
- Title tags under 60 characters, focus keyword near the front.
- Meta descriptions under 160 characters, a clear summary and a localized call to action.
- Localized keywords used naturally, in the market's own phrasing.
Good metadata lifts click-through in classic results and makes it more likely an AI system picks your page as a reference.
Choose your multilingual URL structure once
Your URL structure is expensive to undo later. It is a core technical element of multilingual SEO, because search engines use it to identify language and regional targeting.
There are three main approaches:
- Subdomains: en.example.com.
- Subdirectories: example.com/en/, the most common choice for keeping authority under one domain.
- Country-code top-level domains: example.fr.
They differ in crawl management, authority consolidation and hosting complexity. Whichever you choose:
- Keep URLs clean, descriptive and consistent.
- Use the same structure for every language.
Get hreflang right with three rules
Correct tags send a buyer in France to your French page, and setting them is core to our technical SEO services. Hreflang tags tell search engines which language and region each page targets. Three rules:
- Place the tags in the HTML head or the XML sitemap.
- Make every tag bidirectional: if page A lists page B as an alternate, page B must list page A.
- Include an x-default for users outside your defined languages and regions.
In a 2023 study of 374,756 domains using hreflang, 67% had at least one hreflang issue. The most common, pages missing an x-default, affected 56.3% of the domains.
Hreflang is among the most common technical SEO issues to check on multilingual sites, so audit the tags regularly.
Earn links from local press and sector bodies
A link from the local press or a sector association in your target country is worth more than a generic international link, because search relevance is geographic as well as topical. Three ways to earn regional links:
- Localized guest posts.
- Partnerships with local businesses.
- Targeted PR.
Quality outweighs quantity, so earn every link: that keeps you clear of the penalties bought links and link schemes risk. For a market-specific example, see our guide to link building in Spain.
Authoritative local links also make it more likely that AI models treat your content as a source worth citing in that market.
Make each market clear with structured data
An AI system quotes the page whose market it can identify with confidence. Structured data now shapes how AI systems interpret your content as well as how it ranks. Every language version should carry accurate schema markup:
- WebPage.
- Organization.
- BreadcrumbList.
- Any relevant Product or Service types.
Entity clarity matters even more on multilingual sites. AI models need unambiguous signals about which version applies to which market, language and audience. With them, your content competes in AI answers on the same footing as it does in rankings.
Get cited in AI answers with GEO
A growing share of your buyers now compare suppliers inside Google’s AI Overviews and AI Mode, ChatGPT, Perplexity and Gemini, which all draw on structured, authoritative, entity-rich content. Content written for AI answers as well as classic results reaches that part of the decision.
In March 2025, 18% of the Google searches made by 900 US adults produced an AI summary. Users who saw one clicked a traditional search result link in 8% of visits, against 15% for those who did not.
GEO makes content citable by AI as well as indexable. For multilingual sites it asks for the same four things in every language you sell in:
- Factual density.
- Clear entity relationships.
- Correct language tagging.
- Authoritative sources.
Brands that give every language the same care as the primary one, through full content localization, gain ground in each market. Our view on the future of SEO goes further into where this is heading.
Run multilingual SEO and GEO together
| Layer | What it covers | What it wins |
|---|---|---|
| Multilingual SEO | hreflang, URL structure, localized keywords, metadata | Rankings and intent-driven traffic in each market |
| GEO | Entities, structured data, factual and sourced content | Citations in AI answers, often before a visit |
Multilingual SEO wins the ranking and GEO wins the citation, so together they reach the buyer at both moments. Global SEO now needs both, working together in every target language and market.
