Hreflang and international SEO

1Overview

Sites that serve several languages or regions often have the same content in different versions: a page in English for the UK and the US, and in German for Germany. Hreflang lets the site tell Google that these pages are alternates of one another, so Google can show the right version in each market [1].

Google says hreflang is used to serve the appropriate version; it is not a way to consolidate ranking signals between versions [1][2].

2History

Google announced the rel="alternate" hreflang markup in December 2011 as a way to indicate multilingual and multi-regional versions of content [4]. In April 2013 it added the x-default value for pages that serve no specific locale, such as language selectors [5].

3Language and region codes

The hreflang value starts with a language code in ISO 639-1 format, optionally followed by a hyphen and a region in ISO 3166-1 alpha-2 format [1][7][8]. These conventions follow the wider BCP 47 standard for language tags [6][9].

4Three ways to declare

The three methods are equivalent; Google recommends choosing one and using it consistently [1].

5Rules that make it work

• Self-reference: each version includes itself in its list [1].
• Reciprocity: if page A lists page B, B must list A; otherwise the annotations may be ignored [1].
• Absolute URLs: each href is a fully qualified URL [1].
• Indexable targets: listed URLs should return 200 and be canonical [1][11].
• Self-canonicals: each version should name itself as canonical, not another language [1][11].

6Site structure

Google’s guidance compares four ways to organise international sites: country-code top-level domains, subdomains, subdirectories and URL parameters. Country-code domains give the clearest signal but cost more to run; subdirectories are easiest to maintain; URL parameters are not recommended [2].

Whatever the structure, Google recommends separate URLs for each language version rather than changing content on one URL based on cookies or browser settings [2].

7How Google detects language

Google determines the language of a page from its visible content. It does not rely on code-level indicators such as the HTML lang attribute for this purpose [2], although lang remains important for accessibility and browsers [10].

Mixing languages on one page, such as translated navigation with untranslated body text, makes the language harder to determine [2].

8Locale-adaptive pages and redirects

Some sites change content or redirect users based on their IP address or Accept-Language header. Google warns that Googlebot’s crawlers mostly originate from the United States and usually send no Accept-Language header, so other versions may never be crawled [3][14].

Google recommends letting users choose their version, for example with a language switcher, and using separate URLs with hreflang rather than automatic redirection [2][3].

9Common misconceptions