Common International SEO Mistakes
Last updated
International SEO failures are usually predictable. The same set of mistakes appears across campaigns regardless of industry or market. Most of them are structural: they occur at the point when the international site is initially set up and persist until someone systematically audits the implementation.
1. Broken or missing hreflang
Hreflang tags are the most technically demanding part of international SEO and the most frequently misconfigured. The specific errors that cause the most damage:
Missing self-referential tags: every page in a hreflang set must include an annotation pointing to itself. Without this, the entire annotation set for that page may be ignored.
Non-reciprocal annotations: if page A declares page B as its French equivalent, page B must declare page A as its English equivalent. One-sided annotations are ignored by Google.
Wrong ISO codes: en-UK is not valid; the correct value is en-GB. The failure is more insidious than a straight rejection. Google states that where you use codes reserved for something else, it “ignores that part of the annotation”, giving EU, UN and UK as examples that have no effect.1 So en-UK does not vanish, it silently degrades to a generic en annotation. Related and equally common: the language code must come first, and “you can’t specify the country code by itself”,1 so de meaning Germany is invalid.
Relative URLs in annotations: alternate URLs “must be fully-qualified, including the transport method (http/https)”, so /foo and //example.com/foo are both invalid.1 Pointing to HTTP URLs from an HTTPS site is the same class of error.
Hreflang pointing to redirects or noindex pages: if a hreflang annotation points to a page that redirects elsewhere or carries a noindex directive, the annotation is weakened or ignored.
Canonicalising across language versions: this kills the annotation set and is the error most likely to undo otherwise correct work. Google states that canonical annotations carrying hreflang “are not used for canonicalization”,2 and advises that if you use hreflang, “make sure to specify a canonical page in the same language, or the best possible substitute language if a canonical page doesn’t exist for the same language”.2 Every page in a set should carry a self-referential canonical.
Note one thing that is not an error, despite appearing on most mistake lists: declaring hreflang in both the <head> and the XML sitemap. Google’s position is that “the three methods are equivalent” and that “while you can use all three methods at the same time, there’s no benefit in Search (in fact, it maybe be much harder to manage three implementations instead of just picking one)”.1 Pick one method for maintainability, not because combining them conflicts.
Where to audit. Not in Search Console: Google’s Page Indexing documentation states that “alternate language pages are not detected by Search Console”,3 and its URL Inspection documentation does not mention hreflang at all.4 The old International Targeting report was the hreflang monitor and nothing replaced it. Use third-party crawlers (Screaming Frog, Sitebulb), which crawl the full annotation graph, or the tools Google itself names for debugging.1
2. IP-based geo-redirects
Redirecting users to a regional version based on their IP address is common and creates a serious crawl problem. Google’s guidance is direct: “avoid automatically redirecting users from one language version of a site to a different language version of a site”, because “these redirections could prevent users (and search engines) from viewing all the versions of your site”.5 On the IP-based version specifically it cautions: “don’t use IP analysis to adapt your content. IP location analysis is difficult and generally not reliable. Furthermore, Google may not be able to crawl variations of your site properly.”5
The mechanism: Google “egresses primarily from IP addresses in the United States”,6 and states it does “not attempt to vary the crawler source used for a single site in order to find any possible variations in a page”.5 So a site redirecting US IPs to /en-us/ will tend to show Googlebot that version and little else, leaving other market pages undiscovered and unindexed. Don’t state this as an absolute, though: Google notes that “in case Google detects that a site is blocking requests from the United States, it may attempt to crawl from IP addresses located in other countries”.6 Treat it as a high-probability failure rather than a certainty.
The same mistake wearing a different hat. Sites that “fix” IP redirects by switching to Accept-Language header sniffing have the identical problem, and it is less obvious: “the crawler sends HTTP requests without setting Accept-Language in the request header”.5 There is no header to read, so the fallback path is what Google sees.
The correct approach is to give each version its own URL, use hreflang to signal which serves which audience, and let users choose. Google recommends “using different URLs for each language version of a page rather than using cookies or browser settings to adjust the content language on the page”,5 and suggests “adding hyperlinks to other language versions of a page”.5 A cookie may remember a user’s choice between distinct URLs; it should never swap the content served at a single URL.
If IP-based redirects already exist, test with a tool that simulates a crawl from a US IP address. If the redirects fire for Googlebot, fix them before any other international SEO work.
3. Cookie-based overlays that block crawling
A variation of the same crawl problem: showing a language or region selection modal to new users before serving any page content.
Be precise about the mechanism, because the common version of this warning is wrong. Content that is already in the rendered DOM is indexable even when a modal visually covers it. The failure is content that only loads after an interaction, or a redirect that fires when a user makes a selection. Google “does not interact with your page”,7 so anything gated behind a click is invisible to it.
Test by using Search Console’s URL Inspection tool (Test Live URL). If the rendered page shows only the modal rather than the content beneath it, the overlay is preventing indexation.
4. Machine translation without review
Relying entirely on machine translation for international content is one of the most common quality failures. The issues:
First, be clear about what the risk is not. Google’s policy line is about value, not production method: its scaled content abuse policy applies “no matter how it’s created”,8 and Google has stated that its “policies do not strictly define content that has been translated by AI as spam”.9 It removed its older advice to block automatically translated pages in June 2025. So unreviewed machine translation is a quality problem, not an automatic policy violation, and hand-translation is not a safe harbour if the output is thin.
The practical issues:
- The keyword targeting reflects the source page’s queries, not what local users actually search for. The translated page may rank for the wrong queries, or fail to rank at all. This is the strongest objection and better translation quality does not fix it.
- Unnatural phrasing puts local readers off. Note that the cost is lost readers and conversions rather than a ranking penalty: Google has consistently denied using analytics engagement metrics as ranking signals, with John Mueller stating there is “a bit of misconception here that we’re looking at things like the analytics bounce rate when it comes to ranking websites, and that’s definitely not the case”.10
- Machine translation produces predictable sentence structures that experienced readers recognise as non-native.
Machine translation with native-speaker post-editing is a reasonable default for most content. Publishing raw output at scale, without local keyword work, is the mistake.
5. Identical content across regional variants
Publishing the same content for both British and American English without meaningful differentiation is the case hreflang cannot rescue. Note that this is not a penalty: duplicate content generally isn’t one, and Bing states plainly that “duplicate content doesn’t trigger search penalties on its own”.11 What happens instead is consolidation. Google commonly picks one near-identical same-language variant as canonical and serves that, so the variants stop appearing separately.
Bing frames the same problem and prescribes the same fix: “localization creates duplicate content when regional or language pages are nearly identical and do not provide meaningful differences for users in each market”, and the remedy is to “localize with meaningful changes such as terminology, examples, regulations, or product details” while avoiding “multiple pages in the same language that serve the same purpose”.11
If you are targeting multiple English-speaking markets, the regional variants need genuine differentiation: different pricing and currency, different legal and regulatory context, different availability or shipping terms, different examples. Where none of those differ, one page serving both markets is the better answer.
There is an AI dimension here too, and Bing is the only platform to have documented it: “LLMs group near-duplicate URLs into a single cluster and then choose one page to represent the set. If the differences between pages are minimal, the model may select a version that is outdated or not the one you intended to highlight.”11 Google has published nothing equivalent, so treat this as Bing’s account of its own systems rather than a general rule.
6. Misapplied geo-targeting in Search Console
Google removed the country targeting setting from Search Console in September 2022.12 It no longer exists and cannot be misconfigured. If you encounter documentation advising you to set country targeting in Search Console, it predates the removal.
The mistake now is assuming hreflang is all that is left. Google documents four signal groups for target locale: ccTLDs; hreflang “whether in tags, headers, or sitemaps”; server location, which it notes “is not a definitive signal” given CDNs; and other signals including “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”.5 That last group is cheap to get right and frequently ignored. Note also that Google “ignores locational meta tags (like geo.position or distribution) or geotargeting HTML attributes”,5 so there is no markup substitute for the removed setting. See Geotargeting in Google Search Console for details.
7. Forgetting x-default hreflang
When no specific hreflang matches a user’s language or country, Google selects a URL based on other signals, which may not be the one you want. Adding hreflang="x-default" provides an explicit fallback. Google notes that although “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”,1 so a selector page is the stronger choice where you have one and a global homepage is a reasonable substitute rather than an equal alternative.
This is most important for large international sites where users from markets not covered by any specific hreflang annotation may still arrive.
8. Neglecting local keyword research
Assuming translated keywords have equivalent search volume and intent in the target market is a persistent mistake. Each market requires independent keyword research. The same product may be searched differently in Germany than in Switzerland, even though both markets use German.
Translating the home-market keyword list produces a starting point, not a finished strategy. Local keyword research, done in the target language with local SERP analysis, is the only reliable approach.
9. Ignoring local competitors
Competitors differ by market. The dominant players in your UK market are rarely the dominant players in Germany or France. International campaigns that benchmark against UK competitors and apply UK content standards often find they are under-competing in new markets because local competitors have more relevant content, stronger local links, and better local signals.
Before entering a new market, audit who ranks for your target keywords in that market, what content they produce, and what their link profiles look like.
10. Building links only from global domains
Earning links from high-authority global domains is valuable but not sufficient for strong international performance. Local link building, earning links from locally relevant domains in each target market, is a distinct activity requiring separate effort. Sites that rely on their global domain authority without building local link profiles consistently underperform against locally established competitors in each market.
Frequently asked questions
How do I know if IP-based redirects are affecting Googlebot?
Use Search Console’s URL Inspection tool (Test Live URL) and compare the rendered result with a direct URL fetch in an incognito browser. If they differ, a redirect or geolocation-based serving mechanism is likely affecting crawling.
What is the fastest way to audit hreflang errors across a large site?
Run a full-site crawl with Screaming Frog or Sitebulb and export the hreflang data. These crawlers map the full annotation graph, which is the only practical way to audit a large set. For individual pages, use the Merkle hreflang tags testing tool or Aleyda Solis’s hreflang tags generator, the two tools Google’s own documentation points to.1 Don’t expect Search Console to help: “alternate language pages are not detected by Search Console”.3
Is there a lower-risk way to test international SEO before committing to a full implementation?
Start with a single high-priority market: implement the URL structure, hreflang, localised keyword research, and localised content for one country. Measure the results over three to six months before extending to additional markets. This limits the initial investment and gives you data to inform subsequent decisions.
Footnotes
-
Tell Google about localized versions of your page — Google Search Central ↩ ↩2 ↩3 ↩4 ↩5 ↩6 ↩7
-
How to specify a canonical page — Google Search Central ↩ ↩2
-
Managing multi-regional and multilingual sites — Google Search Central ↩ ↩2 ↩3 ↩4 ↩5 ↩6 ↩7 ↩8
-
Overview of Google crawlers and fetchers — Google Search Central ↩ ↩2
-
Spam policies for Google web search — Google Search Central ↩
-
Google Softens Its Stance On Automated Translations — Search Engine Roundtable ↩
-
Is bounce rate a Google ranking factor? — Search Engine Journal ↩
-
Does duplicate content hurt SEO and AI search visibility? — Bing Webmaster Blog ↩ ↩2 ↩3
-
The International Targeting report is deprecated — Google Search Console Help ↩
