Last year I wrote about the same store launching into seven empty indexes: a skincare brand with storefronts in the UK, the US, Australia, Singapore, Canada, New Zealand and the EU, each one starting from zero because Google does not transfer trust between domains. That post ends where the launches end. This one is about what the account looks like two years in, because the launches turned out to be the easy part.
The hard part is that seven indexes never stay in sync on their own. Every guide I have read on international Shopify SEO is a setup guide: pick ccTLDs or subfolders, wire the hreflang, publish the translations, done. Nobody writes about month fourteen, when the German catalogue quietly slides while the UK one holds, nothing turns red anywhere, and the owner asks why EU revenue is soft.
Can a page be indexed in one country and not another?
Yes, and on a multi-store brand it is the normal state, not the exception. Each storefront is its own domain, its own Search Console property, and its own index record. Google judges each one on its own history, so the same product page can be indexed for a year in the UK store and sit in “Discovered, currently not indexed” on the Singapore one.
People often mean something else by this question, though. If your one site simply ranks differently in different countries, that is country-specific ranking, not indexing, and I covered it in why Google results differ by country. Indexing drift is specifically this: the same URL set, duplicated across market domains, with a different fraction of it indexed in each property.
The comparison sheet: how the drift becomes visible
Nothing inside a single Search Console property shows you drift, because drift is a difference between properties. So the account runs on one habit: on the same day each month, I pull two numbers from every property and put them in one sheet.
- Page indexing: indexed count and the top not-indexed reasons.
- Sitemaps: discovered URLs versus indexed URLs for that market’s sitemap.
Then I track one figure per market: indexed divided by submitted. Raw counts are useless here because the catalogues are not the same size in every market. Ratios are comparable. When six stores sit between 85 and 95 percent and one has drifted to 70, that store has a problem months before its traffic graph admits it. Across this account, more than 330 URLs are tracked per market and the last snapshot averaged roughly 88 percent indexed, so a market in the low seventies stands out immediately.
The routine and the account numbers, from the same ledger described in the skincare case study.
What actually causes one market to drift
Two years of this sheet keeps pointing at the same three causes.
Broken hreflang clusters, silently. Shopify Markets only emits hreflang alternates for translations that are published, and country domains outside Markets get no automatic hreflang at all. One unpublished translation or one bad language code drops the whole alternate cluster for that page. The weaker store then loses its only vouch from the strong one, and its copy of the page collapses into “Duplicate, Google chose different canonical”, usually pointing at the UK version. That status is grey, not red, so nothing alerts you. The cost of these breaks is the subject of my hreflang errors post.
Same-language markets are duplicates by default. The UK, US, Australian and Canadian stores are all English with the same product copy. Google’s own multi-regional site documentation is honest about this: near-identical versions get folded together and the strongest one wins. The markets that drift on this account are almost always the same-language ones, and the fix is localisation that is real (currency, delivery, spelling, market-specific ranges), not a rewritten sentence.
Theme deploys. The stores share a theme, so a deploy touches all seven at once, and twice now a deploy has broken the hreflang block on some stores but not others. I now run the hreflang check against the deploy date, not the monthly cycle. Waiting for the monthly audit means the recrawl damage is already weeks deep.
What to check if one of your stores will not index
If you run one brand on multiple country domains, in Shopify or anywhere else, this is the short version of the routine:
- Give every market domain its own Search Console property and submit that market’s sitemap inside it. The sitemap report’s discovered-versus-indexed split is only readable per market, and a sitemap that is submitted but not indexed tells you which market is stuck.
- Compare ratios across properties on the same day, monthly. One low ratio is a market problem; all ratios low is a brand problem.
- Check the lagging market’s hreflang cluster on a few money pages first, including whether the translations behind it are actually published.
- Check the geo-redirect. Automatic country redirection that fires on every visitor will fire on Googlebot too, which mostly crawls from the US. Google’s locale-adaptive page documentation is clear that this makes other markets hard to crawl; suggest, do not force.
- Ration manual submissions per property. The URL Inspection tool allows roughly a dozen requests a day per property in my experience, which across seven properties is a real budget, but spend it on money pages. The Indexing API is not a way round this: Google limits it to job postings and broadcast events, and product pages do not qualify.
Drift is not a launch problem, so it never announces itself. It is the slow divergence of seven records that nobody is comparing. The comparison is the job, and it takes twenty minutes a month once the sheet exists. If one of your markets is the one sitting low and you cannot see why, send me the domains and I will tell you whether it is hreflang, duplication or a redirect, usually within a day.
Sources
- Managing multi-regional and multilingual sites, Google Search Central
- Localized versions of your pages, Google Search Central
- How Google crawls locale-adaptive pages, Google Search Central
- International SEO for Shopify Markets, Shopify Help Center
- Indexing API quota and eligibility, Google
- Account numbers (7 stores, 330+ URLs per market, ~88 percent indexed) are my own ledger, published in the skincare case study