The store localisation playbook
Translation is the smallest part of localisation. The work that decides success is structural — URLs, hreflang, per-market vocabulary — and operational: keeping six locales true as the catalogue changes weekly.
Stores internationalise in a predictable arc: machine-translate everything, launch, watch the localised stores underperform, and quietly stop maintaining them. This playbook runs the version that compounds instead: choose markets where your data shows demand, build the technical structure correctly once, localise the pages that convert against genuinely local vocabulary, and install the sync discipline that keeps every locale current from then on.
Stage 1: Choose markets on evidence, structure once
Your analytics already show where demand leaks in: traffic, conversion, and order attempts by geography. Enter markets where evidence exists, not where ambition points. Then the technical layer, done correctly the first time: subdirectory URLs for consolidated authority (barring a strong legal or brand reason otherwise), complete and symmetric hreflang clusters, and per-locale sitemaps.
Market shortlist from existing geographic traffic, conversion, and payment-attempt data
Subdirectory structure (/de/, /fr/) unless a specific reason dictates otherwise
hreflang clusters complete and symmetric from day one — retrofits are misery
Currency, payment methods, and delivery expectations mapped per market before launch
Stage 2: Localise where money changes hands
Localisation effort follows revenue: product and collection pages first, checkout and policy pages next, blog content last if at all. Written against local search vocabulary — what German buyers type is not the dictionary translation of what British buyers type — and local objections: sizing conventions, delivery anxieties, and payment preferences differ per market and belong in the copy.
Stage 3: Sync as a standing rule
Drift is the killer: new products shipping only in the primary language, price changes missing a locale, a template update dropping half the hreflang tags. Make the primary catalogue the source of truth, propagate every change to every locale as part of shipping it, and verify hreflang integrity continuously. A locale that cannot be kept in sync is better retired than left rotting — a stale store in a buyer’s language is worse than none.
How it works
Pick markets, build structure
Evidence-based market choice; URLs, hreflang, and payment rails set up correctly once.
Localise revenue pages first
Products and collections against local vocabulary and objections, then checkout and policies.
Install the sync rule
Every catalogue change propagates to every locale as part of shipping; integrity verified continuously.
What you get
Markets entered on demand evidence rather than ambition
A technical structure that never needs re-architecting
Localised pages that answer local objections in local vocabulary
Locales that stay true as the catalogue changes
Handled by these agents
Frequently asked questions
Machine translation first, or proper localisation from day one?
For revenue pages, localise properly from the start — that is where the return lives. Machine translation can bootstrap long-tail pages, clearly reviewed before anything customer-critical depends on them.
Same-language markets — do I need separate locales for the US, UK, and Australia?
Separate regional locales earn their keep when currency, catalogue, spelling, or delivery messaging genuinely differ. hreflang regional targeting handles the search side; do not multiply locales beyond what you will maintain.
What breaks hreflang most often?
Asymmetry (page A points to B, B forgets A), new templates shipping without the tags, and retired pages leaving dangling references. Continuous verification exists because all three happen silently.
Which agents run this?
Analyst supplies the market evidence; Builder owns structure and propagation; Visibility owns per-market vocabulary and hreflang integrity. The sync rule is enforced in the shared workflow, not in anyone’s memory.
Keep reading
Store localisation
Sell properly in more than one language — localised pages, correct hreflang, per-market SEO.
E-commerce SEO at scale
Metadata, structured data, and content across a whole catalogue — not just the top twenty pages.
Write collection copy
Collection page copy that helps buyers choose — and gives the page a reason to rank.
Build a B2B quote funnel
From spec-complete pages to fast structured follow-up — the wholesale conversion machine.
Put the team to work on your store
Connect your storefront, analytics, and email stack, set a goal, and let the agents run the work end to end. Start on the free plan with your own model key.