International SEO one brand, the right page in every market

A buyer in Manchester lands on your US page. Prices in dollars, shipping from Ohio, a phone number starting +1, American spelling throughout. Nothing is broken, exactly. They just conclude this shop is not for them and go back to the results.

Google has the same problem from the other side. It sees four pages of nearly identical English on one domain, cannot tell which market each is for, and picks one to show everybody. Usually the wrong one.

International SEO prevents both, and most of it is architecture: which market gets which URL, how search engines are told those pages are alternates rather than duplicates, and what actually changes between them beyond the currency symbol. Our published work spans Canada, the US and India, with client leadership in Brampton and engineering in Patiala, so multi-market is our normal case, not a special request.

What international SEO means in our hands

Hreflang is generated, never hand maintained. Hand-written tags rot within a quarter: someone adds a page in one language, forgets the return tag in the other four, and the cluster quietly stops working. We generate the whole set from the routing layer, self-referencing and bidirectional, with correct language and region codes and a deliberate x-default, then validate it, so a broken cluster fails a build instead of costing you a market.

The domain decision is made once, with the trade-offs in writing. A ccTLD gives the strongest local signal and the most in-market trust, but splits your authority and multiplies the operational load. Subfolders consolidate authority, cost least to run, and suit most businesses below a certain size. The recommendation and the reasoning go in writing, because reversing it later is a migration, not an edit.

Localisation, not translation. Translated copy ranks for the words a translator chose, not the words a market types. Keyword research is redone per market instead of run through a dictionary, and the page changes with it: pricing, shipping promises, payment methods, in-country proof, units, date formats, and the legal pages each jurisdiction expects.

The approach and the stack

Markets are routed in Next.js, so the hreflang cluster, canonicals, sitemaps, and language switcher all come from one route configuration. Nothing is maintained twice. Currency, units, and dates format at render for the active locale, and each market's content lives in the CMS as its own record, not a translation field on the original.

The usual standards apply: fast on a real phone (we measure it before launch), WCAG 2.1 AA as the accessibility target, and no third-party translation widget dragging down the pages you are trying to rank. The technical foundation is the one described under SEO, catalogue-level work sits under ecommerce SEO, and a site that cannot support markets properly needs web design first.

What this looks like in practice

Two typical engagement shapes:

An English-language expansion. A business selling in one country adds two or three more that speak the same language. Near-duplicate English is the whole problem: the fix is a correct hreflang cluster plus real differences between the pages, so pricing, shipping, guarantees, proof, and compliance give each one a reason to exist.

A multi-language market entry. A company entering a market where it has no footprint. Research happens in-market first, pages are localised rather than translated, local trust signals get built, and markets launch one at a time so the first funds the next.

Pricing

International SEO is scoped per project. The number of markets, whether you are migrating country sites, and how much content genuinely needs localising change the work by an order of magnitude, so we publish no rate card. Tell us which markets you are chasing and you get a written scope and a fixed price. A free audit of your current setup is the usual first step, and the case studies show how we work.