Translating your website will not open a new market for you. It is the most expensive illusion in online international growth: run the site through a translator, add a language switcher, and wait for the foreign customers who never come. Multilingual SEO is a different craft from translation. It brings together three things we too often handle separately: language, technical setup, and the real search behaviour of each market. This guide lays out the strategy, and the one mistake you must never make.<\/p>\n\n\n\n
We underestimate how much language shapes trust, and therefore sales. The benchmark study by CSA Research, conducted with 8,709 consumers across 29 countries, is unambiguous: 76% of online shoppers prefer to buy a product whose information is in their own language, and 40% never buy from a site in another language (verified on 15 June 2026). Among people who are not fluent in English, the preference for their own language rises to 89%.<\/p>\n\n\n\n
The strategic consequence is clear. You do not reach a market by being technically accessible, but by speaking the buyer’s language naturally. A page that is understood yet foreign converts less than a native page. Language is not a surface layer, it is a condition of purchase.<\/p>\n\n\n\n
Here is the distinction that changes everything. Translating means making content understandable in another language. Multilingual SEO means making that content findable, judged relevant and preferred by the search engines and users of a given market.<\/p>\n\n\n\n
And people do not search with the same words from one country to another, even within a shared language. A term that dominates in France can be marginal in Quebec or Belgium. A literal translation of a high-performing keyword may target a phrase nobody types. Multilingual SEO therefore starts with keyword research specific to each market, not with translating the keywords of the home market. That is what separates a presence that captures traffic from a presence that exists for appearance only.<\/p>\n\n\n\n
On the technical side, a few foundations govern everything else.<\/p>\n\n\n\n
These settings are invisible, but they decide whether your multilingual content efforts pay off or cancel each other out. A sound multilingual architecture is designed before publishing, not after noticing that the versions are cannibalising one another.<\/p>\n\n\n\n
There is one error that ruins everything else: visible machine translation. Content that has obviously been run through a machine, without proofreading by a native speaker, discredits a brand as surely as a spelling mistake on a shopfront. The visitor senses it within a sentence, and trust collapses, especially among the 40% who already do not buy in a foreign language.<\/p>\n\n\n\n
The rule we hold to is simple: every commercial language demands native quality, or we hold off. Better to delay publishing a version than to release it in visible translation. A language spoken badly does more harm than a language that is absent.<\/p>\n\n\n\n
The brief before the brief.<\/strong> Before launching a multilingual site, ask two questions: which market are we genuinely targeting, and who guarantees native quality. If the answer is \u201cwe will machine-translate it and see\u201d, the project is already compromised. Multilingual work is scoped market by market, with a native proofreader for each language.<\/p>\n\n\n\n The temptation is to open everything at once. Discipline means doing the opposite: one market at a time, validated, before the next. You pick the language where demand is most likely, do the local keyword research, produce native content, set up hreflang properly, and measure. Only then do you replicate on the next market. Covering three markets halfway pays less than covering one properly.<\/p>\n\n\n\n Are multilingual SEO and translation the same thing?<\/strong> No. Translation makes content understandable; multilingual SEO makes it findable and preferred in a given market, with keyword research specific to each country and a dedicated technical setup.<\/p>\n\n\n\n What are hreflang tags for?<\/strong> They tell Google which language version to serve to which user. Set up badly, they make your own versions compete against each other, which destroys the benefit of going multilingual.<\/p>\n\n\n\n Subfolders or separate domains per language?<\/strong> Subfolders per language are generally preferable: they consolidate the domain’s authority and simplify management. Separate domains fragment it.<\/p>\n\n\n\n Can you use machine translation?<\/strong> As a starting point, yes; as published content, no. A text that is obviously machine-made discredits the brand. Every language requires native proofreading, otherwise it is better to wait.<\/p>\n\n\n\n Should you translate the whole site at once?<\/strong> No. It is better to cover one market properly, validate it, then replicate. Three markets done halfway pay less than a single one done seriously.<\/p>\n\n\n\n Targeting several markets online is not a translation matter, it is a strategy matter: the right language, local research, hreflang set up properly, and non-negotiable native quality. Start with one market, do it well, measure, replicate.<\/p>\n\n\n\n In practice, we scope the priority market, its keyword research and its multilingual architecture during a workshop, then roll out a native, technically clean presence, market after market.<\/p>\n\n\n\nWhere to start without spreading yourself thin<\/h2>\n\n\n\n
FAQ: multilingual SEO<\/h2>\n\n\n\n
Before you ask for a quote<\/h2>\n\n\n\n