Discussion starter: an international SEO question, posted with an answer from Niraj Raut to open the thread. If you have dealt with this on a site, reply with what you saw, especially where it differs.
- International SEO
Aleyda Solis’s May 2026 analysis showed UK results shifting from .com retailers to local-market domains. Is market fit now a stronger lever than hreflang?
Disclosure: Niraj Raut, who posted this answer, runs the SEO consultancy linked at the end of it.
1 reply
-
adminAdmin
For winning UK visibility against local competitors, yes: market fit is the stronger lever, but it always was. hreflang has never been a ranking lever. It helps Google choose which of your own alternate URLs to show in a given market, and Google’s representatives have said many times that it does not give a ranking boost. Market fit decides whether any of your URLs deserve to compete with a local retailer in the first place.
So the useful reading of Aleyda Solis’s May 2026 data is not “hreflang matters less”. It is that Google got stricter about showing UK searchers a result that fits the UK, and that shows up in two different ways: swapping in a brand’s own local site, and replacing foreign sites that have no local equivalent. hreflang can help with the first. Only a genuine UK offer helps with the second.
What the data shows, and what it cannot show
Aleyda’s May 2026 core update analysis used Sistrix visibility data for the US and UK, measured from 26 May to 2 June 2026 (by her dates the rollout ran 21 May to 2 June; Google has confirmed it complete). In the UK she reports that local-market retail rose about 20% while US.com marketplaces fell about 55%. The domain-level figures are worth reading closely:
Domain UK visibility change What kind of shift it is amazon.com -54.6% Same brand, local sibling gained: a market-version swap amazon.co.uk +21.3% ebay.com -53.7% Same pattern ebay.co.uk +22.6% walmart.com -59.5% No UK consumer storefront to swap to: a market-fit loss alibaba.com -55.3% Largely B2B and not a natural UK retail destination screwfix.com +25.2% A UK retailer on a.com domain gained Two things stand out. First, half of the losers had a local sibling that gained, which is exactly the outcome hreflang and country targeting exist to produce. Second, Screwfix gained on a.com. The top-level domain was not the deciding factor; being a UK business serving UK demand was.
The limits are just as important. Sistrix visibility is a weighted score across a tracked keyword set, not traffic. The figures are domain-level, cover the week after rollout, and describe some of the largest retail domains on the web. They do not tell you whether the losing.com pages carried hreflang, whether individual queries switched to the sibling domain, or how a mid-sized retailer with a UK subfolder fared. The pattern is consistent with a market-fit preference; it does not measure hreflang’s contribution at all. Aleyda frames it as an actionable pattern within a broader recalibration, not a universal rule.
Two levers doing two different jobs
Think of hreflang as the signage inside a chain store’s network: it tells a visitor which of your branches serves their country. It cannot make a visitor choose your branch over the independent shop down the road. That choice depends on what each shop actually offers.
Technically, hreflang annotations tell Google that a set of URLs are alternates for different languages or regions, so it can show the right one in each market. The swap only happens between URLs that are already eligible to rank for the query. If your UK page is weaker than a local competitor’s page for a UK searcher, hreflang has nothing to swap in.
Google’s guidance on multi-regional sites lists the signals it uses to work out a page’s intended audience: ccTLDs, hreflang (in tags, headers or sitemaps), server location, and “local addresses and phone numbers on the pages, the use of local language and currency, links from other local sites, or signals from your Business Profile.” hreflang is one signal in that list, and the rest are the documented part of what people now call market fit.
Documented locale signals
ccTLD, hreflang, server location, local address and phone, local language and currency, links from local sites, Business Profile signals.
Commercial fit
Aleyda’s list: inventory, currency, pricing, shipping, returns, local reviews and support. These shape whether the page satisfies a UK shopper.
Hygiene
hreflang and canonical consistency, no automatic redirects by IP or language, and local entity clarity so Google knows which site is the UK business.
The second card is partly inference. Google does not publish a list of commercial signals, but a page that shows US dollars, US delivery times and no UK returns policy is visibly a poor answer for a UK shopper, whatever the ranking systems measure directly. Aleyda’s own recommendation for international ecommerce puts both sides together: “audit wrong-market ranking, hreflang and canonical consistency, local entity clarity, and country specific signals: inventory, currency, pricing, shipping, returns, local reviews and support.”
Which lever applies to your situation
Your situation What is probably happening Main lever Evidence to inspect A.com only, ships to the UK, no UK version You look like a foreign store to UK searchers Market fit: a genuine UK offer, possibly a UK section UK impressions by query type; which local sites replaced you You have a UK version, but US URLs get UK impressions Wrong-market ranking: Google is not mapping the alternates hreflang and canonical consistency Search Console country filter by page; hreflang return links Your UK version ranks, but lost ground to UK retailers The UK page is present but a weaker fit than local competitors Market fit on the UK version itself Price, delivery, returns and reviews versus the winners Your UK pages copy the US pages with the currency swapped Near-duplicates; Google may fold them together Both: differentiate content and fix clustering “Duplicate, Google chose different canonical than user” on UK URLs For the wrong-market case, the checks that catch most problems are: UK pages that canonicalise to the US equivalent (which undermines the hreflang pairing), missing return links, hreflang pointing at redirected or non-indexable URLs, and automatic redirects by IP or language, which Google’s guidance says to avoid. Googlebot mainly crawls from US addresses, so an IP redirect can stop it from ever seeing the UK version.
Should you move to a.co.uk?
Google’s own comparison says a ccTLD offers “clear geotargeting” but is expensive, needs more infrastructure and targets a single country. Subfolders are easy and low maintenance, with the drawback that users might not recognise the geotargeting from the URL alone. The May data does not justify a domain migration by itself: Screwfix won on a.com, and the.co.uk winners were already established UK businesses.
Option Cost and timeline Main risk When it makes sense Improve the UK subfolder’s commercial fit Low to moderate; weeks to months Operational: UK pricing, stock and returns must be real Default first step for most retailers Fix hreflang and canonical clustering Low; days to weeks Little, if implemented from a single source of truth Whenever wrong-market ranking shows up Launch or migrate to a ccTLD High; months, plus a migration dip Split authority, migration losses, duplicate operations UK is a major market with its own entity, stock and support There is a supporting signal from a different surface: Google’s February 2026 Discover core update said it would show more locally relevant content from sites based in the user’s country (US English first). That is Discover, not web search, so it is directional context rather than proof about the May update.
A fast experiment to test market fit before bigger decisions
- Hypothesis. Strengthening UK commercial fit on existing UK category pages raises UK non-brand clicks more than hreflang fixes alone.
- Segmentation. Split comparable UK categories into two groups by traffic and seasonality. Fix hreflang and canonical issues on both, so hygiene is not the variable.
- Treatment. On one group only: GBP pricing in the page and markup, UK delivery times and costs, a UK returns policy, UK reviews and local stock messaging.
- Metrics. UK non-brand clicks and impressions (Search Console country filter plus the branded queries filter), share of UK impressions going to UK URLs, and UK conversion rate.
- Window. Eight to twelve weeks, with each change logged as a Search Console annotation. A core update inside the window affects both groups, which is why the control group matters.
- Decision. If the treated group clearly outperforms, fund market fit across the UK catalogue. If both groups move together, the hygiene fixes were doing the work, and wrong-market ranking was the real problem.
The decision rule I would use
Stop framing it as market fit versus hreflang. Check first whether Google is showing UK searchers your wrong URL. If it is, fix hreflang and canonical consistency, because that is cheap and fast. If it is showing the right URL and still losing to local retailers, no annotation will help; the UK page needs to be a better UK answer on price, delivery, returns and reviews. If you have no UK version at all, the May data is a clear warning that a US store shipping internationally is being treated as a poor fit for UK shoppers, and the business decision is whether the UK is worth a real local offer.
Need help with this on your own site?
Niraj Raut works with businesses in Nepal, Australia, the UK, Europe and the US on technical SEO, ecommerce SEO, local SEO and AI search.
Add a reply Cancel reply
Replies are open to approved members. Log in or apply to join.