Discussion
  • International SEO

Is x-default worth the effort in practice? Have you ever seen a measurable change from adding or removing it?

Published 1 reply

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.

Disclosure: Niraj Raut, who posted this answer, runs the SEO consultancy linked at the end of it.

1 reply

  • admin
    Admin
    Original poster

    x-default is cheap to add and hard to measure, because it only changes anything for searchers whose language and location match none of your hreflang values. If your hreflang set already covers the markets your search demand comes from, there is little left for x-default to do, and adding or removing it will not show up in sitewide numbers. It earns the effort in two situations: when you have a genuine fallback (a language or country selector, or a root URL that redirects by location), and when a meaningful share of your impressions comes from markets you don’t target and landing those searchers on the wrong regional page costs you sales.

    Public evidence is thin in both directions. Google’s hreflang documentation recommends x-default as the fallback for users whose language settings don’t match any of your localised versions, and adds: “While you can use the x-default value for any page, it was designed for language selector pages and so it will work best with those.” In a 2026 article, Eoghan Henn of Rebelytics argued that most websites should not use x-default, keeping it for selector pages with no content and for URLs that redirect by language or location. He also wrote that he was not aware of any proven case of x-default changing which existing version Google picks as a global fallback. No controlled public test of adding or removing it exists that I can point to.

    Who actually falls through to x-default

    x-default applies only when none of your hreflang values match the searcher, so who it affects comes down to how you have written your codes.

    Google’s own example uses three English regional versions (en-GB, en-US and en-AU) with a country selector as x-default. With that set, English speakers in Ireland, India or Singapore are not explicitly covered, and the selector is the intended destination for anyone who matches nothing. Add a language-only en version and English speakers everywhere match it instead. x-default is then left with people searching in languages you don’t serve, who may not see your pages in the first place.

    Whether Google treats a same-language searcher in an unlisted country as a match for one of your regional versions, or sends them to x-default, is not something I would assume. Check what those searchers actually see in live results from those countries before deciding anything.

    Your setup Who can fall through to x-default Worth the effort?
    Language-region codes only (en-US, en-GB, en-AU), with real demand from other English-speaking markets Possibly English speakers in unlisted countries, plus speakers of other languages Worth testing if you have a genuine fallback. A language-only en version may be the more precise fix for the English speakers.
    Language-only codes for every language you serve Only speakers of languages you don’t serve Low value. Cheap to keep if it is correct.
    Selector page, or a root URL that redirects by location or browser language Unmatched users, plus whatever Googlebot sees when it crawls that URL Yes. Google says x-default was designed for selector pages, and redirecting URLs are Henn’s other exception.
    x-default pointing to the same URL as your main market (for example en-US) Unmatched users land on that page Harmless, but if that page already wins for them there is nothing to measure.
    Every page’s x-default pointing to the home page In effect nobody: the home page’s own annotations don’t list the deep page, so the pair is not confirmed Fix or remove it. Google ignores the tags when “two pages don’t both point to each other”.

    Why adding or removing it barely registers in reports

    Even where x-default does something, the signal is easy to miss:

    • The affected audience is a slice. Only impressions from unmatched markets can move. If those are a few per cent of the total, sitewide clicks will not show a change.
    • The effect is a URL swap, not a ranking change. hreflang decides which of your equivalent URLs is shown; it is not a way to make the cluster rank higher. The expected result is a different URL mix within unmatched countries, and any business effect shows up in conversion rather than traffic.
    • Reporting can hide the swap. Where Google treats same-language versions as duplicates and picks one canonical, Search Console can credit impressions to that canonical even when an alternate was displayed. Pair it with rank checks from the countries you care about.
    • Changes roll in unevenly. Because each pair needs return links, Google needs to have crawled both sides, including the x-default target, before an annotation can count. A sitewide deploy takes effect cluster by cluster.

    Important distinction:

    “No measurable change” in total clicks is the expected result even when x-default works. The question is whether unmatched searchers now land on a page that serves them better, and that only shows up when you segment by country and landing page.

    Where the effort actually goes

    If your hreflang is generated centrally, the markup is one extra line per page. The work is elsewhere:

    1. Choosing a real fallback. An international English page with a country selector and clear shipping and currency information is a fallback. A copy of the US store is only a default.
    2. Keeping the target valid. The x-default URL needs to return a 200, be indexable, carry a self-referencing canonical and appear in the same annotation set as every alternate. Google’s documentation says each language version “must list itself as well as all other language versions”.
    3. Checking what Googlebot sees at that URL. Googlebot mostly crawls from US IP addresses and without an Accept-Language header, so a root URL that redirects by IP or browser language is likely to be crawled as whatever it serves a US visitor with no language preference. If the x-default target redirects Googlebot, the annotation points at a redirect.

    How to get a real answer for your site

    If unmatched markets matter commercially, test it rather than argue about it:

    1. Size the opportunity first. List the countries and languages your hreflang set doesn’t cover and pull their impressions, clicks and conversions. If they are a sliver and you can’t sell there anyway, stop here.
    2. Split comparable clusters. Add x-default (or a language-only version, if that is the better fix) to one group of categories or templates and leave a similar group unchanged. A sitewide change is hard to read against seasonality.
    3. Start the clock at recrawl, not at deploy. Use your logs to confirm Googlebot has refetched both the pages and the fallback target.
    4. Primary metric: for unmatched countries, the share of impressions and clicks going to the intended fallback URL (Search Console filtered by country, with the page dimension), and the share of sessions landing on it (analytics landing pages by user country).
    5. Secondary metrics: conversion rate and revenue per session for those users, and use of the country selector or currency switcher.
    6. Guardrail: the URL mix and clicks in your target markets should not move.
    7. Length: set by recrawl time plus enough unmatched-market sessions to compare conversion rates. On a busy section that can be a few weeks; on a small one it may take a quarter.

    If the URL mix shifts towards the fallback and those sessions convert better, roll it out. If nothing moves after both sides have been recrawled, x-default isn’t doing anything for your setup, and removing a correct implementation is unlikely to show anything either.

    Hypothetical example:

    A retailer has en-US, en-GB, en-AU and de-DE versions and no language-only English version. Search Console shows around 10% of impressions coming from Ireland, India, Singapore and the UAE, and those users land on a mix of US and UK pages with the wrong currency and shipping rules. That is worth testing: an international English page with a country selector, set as x-default (with or without a language-only en annotation), could route those searchers better. If the same markets were 1% of impressions and the retailer couldn’t ship there, the effort would be better spent elsewhere.

    My rule: add x-default when you have a real fallback page or a redirecting root, or when unmatched markets are a material share of impressions and the wrong regional page hurts conversion. Point it at a valid page in the cluster, never at the home page from every URL. Otherwise, don’t spend a sprint adding it, and don’t spend one removing a correct implementation either. Fix or remove broken x-default annotations, leave working ones alone, and put the effort into the hreflang errors that affect the markets you actually target.

    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.

    See SEO services

Add a reply

Replies are open to approved members. Log in or apply to join.