TL;DR
Hreflang helps Google choose the appropriate URL when you publish equivalent pages for different languages or regions. It does not translate content, consolidate authority, or guarantee that Google will display a particular page. Successful implementation requires valid codes, indexable URLs, self-references, reciprocal return links, compatible canonicals, and validation across the entire page cluster.
What Is Hreflang, and What Is It Used For?

Hreflang—pronounced “href-lang”—is an annotation that connects equivalent URLs intended for different languages or regions. It appears as an hreflang attribute on an alternate URL reference and helps Google match eligible pages with the appropriate searcher.
For example, hreflang tags can distinguish a Spanish product page for US customers from its Spanish equivalent for customers in Mexico. They can influence which eligible URL Google displays, but they are not a direct ranking boost.
Hreflang is also not a duplicate-content fix. It does not replace canonical tags, redirects, localized content, or a coherent international site structure.
If you only need the terminology, see this concise hreflang definition.
What hreflang can and cannot tell search engines
Hreflang tells search engines that specified URLs are equivalent localized alternatives. Do not use it to connect unrelated translations, loosely similar articles, or pages serving different search intents.
It provides a language signal and, optionally, a regional signal. Google can still choose another URL when its indexing, canonicalization, or relevance signals conflict with your annotations.
Invalid or contradictory annotations may simply be ignored. They do not ordinarily trigger a manual penalty, but they can leave you with the exact problem you were trying to solve: the wrong page appearing for the wrong audience.
Do You Actually Need Hreflang?
Use hreflang when you maintain separate, indexable URLs containing equivalent content for different languages or regions. Valid cases include translated products, regional pricing, local inventory, market-specific legal terms, shipping options, and different conversion paths.
Do not add hreflang merely because one page mentions several countries. You also do not need it when no equivalent alternate URL exists.
Country targeting is frequently overbuilt. Separate en-US and en-GB pages may create more maintenance than search value when their copy, currency, availability, legal terms, and conversion experience are identical.
Hreflang should not replace intentional redirect logic. Avoid automatically forcing visitors elsewhere based only on an IP address or browser language; users may be travelling, using a VPN, or deliberately requesting another market’s page.
Hreflang decision test
Before creating another regional URL, confirm three things:
- Each alternate serves the same underlying search intent.
- Every version has a stable, indexable URL.
- The linguistic or regional difference is meaningful enough to maintain.
Then assess operational capacity. Your team must update the complete reciprocal cluster whenever a page is launched, unpublished, redirected, canonicalized, or removed.
This governance question matters when you outsource SEO without losing control. Ownership of localized URLs, redirects, and annotations should be explicit rather than split ambiguously between developers, translators, and marketers.
Hreflang versus redirects and canonicals
Use a redirect when a URL has permanently moved or should no longer remain independently accessible. Use a canonical tag to identify the preferred URL among duplicate or near-duplicate pages—not to assign visitors by country or language.
Localized pages that deserve independent indexing should normally self-canonicalize. An es-MX page that canonicalizes to a generic Spanish page sends conflicting signals: hreflang presents it as an eligible regional alternative, while the canonical suggests that Google should prefer another URL.
Hreflang Syntax, Language Codes, and Country Codes

A standard HTML annotation contains three important parts:
<link rel="alternate"
href="https://example.com/us-es/producto/"
hreflang="es-US">
rel="alternate"identifies an alternate version.hrefsupplies the alternate’s fully qualified URL.hreflang="es-US"defines Spanish-language content for the United States.
Use ISO 639-1 language codes and, where needed, optional ISO 3166-1 alpha-2 region codes. Language comes first: es-US is valid, while US-es and a country-only value such as US are invalid.
Destinations should be absolute, stable, indexable URLs returning HTTP 200. Each page must reference itself and contain the same reciprocal set of alternates.
Spanish hreflang example: es-US, es-MX, and es
Use:
es-USfor Spanish content tailored to people in the United States.es-MXfor Spanish content tailored to people in Mexico.esfor a genuine language-wide Spanish fallback, if one exists.
The value es-US does not mean English users in the United States. The language always appears before the optional region.
Suppose the cluster contains these URLs:
https://example.com/us-es/producto/ → es-US
https://example.com/mx/producto/ → es-MX
https://example.com/es/producto/ → es
Every one of those pages must list all three URLs, including itself. Do not add the generic es page merely to complete a pattern; it should provide a real experience for Spanish speakers who do not match a regional version.
A copy-ready HTML hreflang cluster
Place the following elements in the <head> of the es-US page:
<!-- Self-reference -->
<link rel="alternate" hreflang="es-US"
href="https://example.com/us-es/producto/">
<link rel="alternate" hreflang="es-MX"
href="https://example.com/mx/producto/">
<link rel="alternate" hreflang="es"
href="https://example.com/es/producto/">
The es-MX page uses the same set, with its own URL acting as the self-reference:
<link rel="alternate" hreflang="es-US"
href="https://example.com/us-es/producto/">
<!-- Self-reference -->
<link rel="alternate" hreflang="es-MX"
href="https://example.com/mx/producto/">
<link rel="alternate" hreflang="es"
href="https://example.com/es/producto/">
The generic Spanish page must also return the relationship:
<link rel="alternate" hreflang="es-US"
href="https://example.com/us-es/producto/">
<link rel="alternate" hreflang="es-MX"
href="https://example.com/mx/producto/">
<!-- Self-reference -->
<link rel="alternate" hreflang="es"
href="https://example.com/es/producto/">
The order is not the critical part. Completeness, valid destinations, self-references, and return links are.
How to Add Hreflang Using HTML, HTTP Headers, or XML Sitemaps
Google supports three implementation methods: HTML link elements, HTTP headers, and XML sitemap annotations. Their rules for codes, destinations, self-references, and reciprocity are the same.
Choose one maintainable source of truth. Publishing annotations in both HTML and XML sitemaps offers no known ranking advantage and creates another place for clusters to drift out of sync.
That principle also applies when evaluating AI-powered SEO content platforms. Automated WordPress, Webflow, or Shopify output remains unverified until you crawl the rendered implementation.
Method 1 — HTML link elements
Add <link> elements to the <head> of every HTML page in the cluster. This is usually the most visible option for small and moderately sized sites with dependable template control.
Avoid relying on scripts that do not consistently place annotations in rendered output. Compare the raw source and rendered HTML if JavaScript or a plugin injects them.
Method 2 — HTTP Link headers
HTTP headers are useful for non-HTML files, including localized PDFs that cannot contain HTML link elements. A response can declare multiple alternates in one comma-separated header:
Link: <https://example.com/us-es/guia.pdf>; rel="alternate"; hreflang="es-US",
<https://example.com/mx/guia.pdf>; rel="alternate"; hreflang="es-MX",
<https://example.com/es/guia.pdf>; rel="alternate"; hreflang="es"
Headers can be harder for editorial teams to inspect than visible HTML. Confirm that your server or CDN returns them consistently for every cluster member.
Method 3 — XML sitemap annotations
XML sitemaps are practical when centralized management is more reliable than page-template changes. Each localized URL entry must list the complete set, including itself:
<url>
<loc>https://example.com/us-es/producto/</loc>
<xhtml:link rel="alternate" hreflang="es-US"
href="https://example.com/us-es/producto/"/>
<xhtml:link rel="alternate" hreflang="es-MX"
href="https://example.com/mx/producto/"/>
<xhtml:link rel="alternate" hreflang="es"
href="https://example.com/es/producto/"/>
</url>
Repeat the complete annotation set in the entries for the es-MX and es URLs. Sitemap clusters commonly become inconsistent when teams publish a translation without regenerating every related entry or redirect an old URL without updating its destinations.
Choosing the right method for your stack
Prefer the method your team can update atomically whenever an alternate changes. Consider template access, deployment ownership, non-HTML assets, and whether your CMS or plugin reliably reflects publishing status.
Do not maintain HTML and sitemap versions in parallel unless a specific operational requirement justifies it and you can keep them identical. Tool expenses also accumulate quickly, so evaluate implementation and monitoring needs alongside the real cost of common SEO software plans, not as an isolated technical decision.
Self-References, Return Links, Canonicals, and x-default
Every localized page should reference itself and all equivalent alternatives. If page A names page B, page B must name page A; this is a reciprocal return link.
Keep every destination aligned with a final, canonical, indexable URL returning HTTP 200. Localized pages intended for search should normally use self-referencing canonicals.
The special x-default value identifies a fallback for users who do not match another language or region annotation. It does not mean English, your primary market, or whichever country page receives the most traffic.
When x-default is useful
Use x-default for a genuine global homepage, language selector, or neutral fallback experience:
<link rel="alternate" hreflang="x-default"
href="https://example.com/choose-location/">
Include that URL reciprocally across the cluster. If it is intended to appear in search, verify that it is indexable and does not automatically redirect every crawler or visitor.
When x-default is unnecessary
Do not add x-default solely because a plugin treats it as mandatory. Omit it when you have no genuine selector, global page, or neutral fallback.
Never point it to an unrelated page or use it to conceal incomplete regional targeting. A generic homepage is not automatically a suitable fallback for every product or article cluster.
Hreflang Validator Walkthrough: Test the Entire Page Cluster
A validator result for one URL is not enough. Hreflang validity depends on the relationships among every member of the cluster.
For the Spanish example, test es-US, es-MX, and es together. Different checkers may classify warnings differently, so validate the underlying URLs and relationships instead of trusting a single pass/fail label.
Treat every plugin- or CMS-generated cluster as unverified until it has been crawled.
Step 1 — Discover and crawl every alternate
Start with one representative URL and collect every destination in its annotations. Crawl the complete set, then repeat discovery from each alternate to expose destinations absent from the first page.
Enable JavaScript rendering when tags may be injected client-side. Compare source HTML with rendered HTML to establish where and how consistently the annotations appear.
Step 2 — Confirm status and indexability
Verify that every destination returns HTTP 200 without a redirect chain or soft-404 behavior. Check robots.txt, meta robots directives, authentication requirements, and X-Robots-Tag headers.
Each intended alternate must be crawlable and indexable. Otherwise, Google cannot reliably retain it as an eligible result.
Step 3 — Compare canonicals and URL destinations
Confirm that each localized URL normally canonicalizes to itself. Hreflang references should use final, fully qualified URLs with consistent protocols, hostnames, paths, and trailing-slash conventions.
Flag redirected, obsolete, parameterized, and noncanonical destinations. These conflicts often explain why valid-looking tags fail to produce the expected selection.
Step 4 — Verify self-references and return links
Check that every page includes itself. Then confirm that every relationship is reciprocal across all cluster members.
Compare the complete language-region set on each page. A three-page cluster should not contain three annotations on one page, two on another, and four on the third.
Step 5 — Test codes and x-default
Validate language values against ISO 639-1 and regions against ISO 3166-1 alpha-2. Confirm language-first formatting, such as es-US, and use consistent capitalization.
If x-default is present, verify that it leads to a genuine fallback and appears consistently throughout the reciprocal cluster.
The 7 Hreflang Errors That Break Localized Clusters
Google Search Console no longer provides a dedicated international targeting report or a comprehensive list of hreflang errors. Any workflow promising otherwise is outdated.
Use GSC to identify symptoms: unexpected landing-page impressions, the wrong regional URL appearing in a country, indexing exclusions, or Google selecting a different canonical. Use Screaming Frog to crawl annotations and export implementation details for cluster-level diagnosis.
Include these checks within a broader technical SEO priority checklist. Hreflang matters, but an inaccessible or noncanonical page has a more fundamental indexing problem.
1. Missing reciprocal return links
This occurs when one alternate references another page that does not link back. In Screaming Frog, review missing return-link errors and export the source-destination pairs so you can correct the full cluster.
In GSC, compare landing pages by country and query to spot an unintended version receiving visibility. Do not expect GSC to enumerate every missing return link.
2. Invalid language or region codes
Common mistakes include unsupported codes, reversed order, country-only values, and invented market abbreviations. Use Screaming Frog’s invalid-language and invalid-region reports, then compare the values with the relevant ISO standards.
GSC may show mismatched landing-page impressions as a symptom. It will not reliably provide a complete invalid-code report.
3. Missing self-references
A page can list every other alternate while omitting itself. Filter for non-self-referencing hreflang URLs in Screaming Frog, then inspect the responsible template, plugin output, or sitemap row.
Use GSC URL Inspection to review indexing and canonical selection. It will not necessarily issue a specific missing-self-reference warning.
4. Alternates point to redirects or error URLs
Flag destinations returning 3xx, 4xx, or 5xx responses, plus suspected soft 404s. Use Screaming Frog’s response-code and redirect-chain exports to replace them with final HTTP 200 destinations.
GSC’s indexing reports and URL Inspection can help confirm whether unavailable URLs or their final destinations are indexed.
5. Alternates are noindex or blocked
Look for pages excluded through meta robots, X-Robots-Tag, robots.txt, authentication, or another crawl restriction. Screaming Frog’s indexability and directives reports help connect those destinations to the source pages naming them.
Use GSC Page Indexing and URL Inspection to confirm exclusion reasons where possible. Remember that blocked crawling can limit the evidence Google can report.
6. Canonical tags conflict with hreflang
A regional page may incorrectly canonicalize to a global page, redirect target, parameter variant, or URL outside the intended cluster. In Screaming Frog, compare hreflang destinations with declared canonicals and isolate non-self-canonical alternates.
In GSC URL Inspection, compare the user-declared and Google-selected canonical. Pay particular attention when the wrong regional URL is receiving impressions.
7. URL clusters are incomplete or inconsistent
Templates, publishing states, plugins, and sitemap generators can produce different alternate sets for pages that should belong to one cluster. Export annotations from Screaming Frog, group URLs by cluster, and compare the language-region values attached to each member.
In GSC, segment performance by country, query, and landing page. Investigate markets where an unexpected URL has replaced the intended alternate.
This kind of verification should remain part of your workflow even as SEO trends and search interfaces change. New discovery channels do not remove the need for clean crawl, indexation, canonical, and localization signals.
Hreflang Pre-Publish Checklist and Post-Launch Monitoring
Before publishing, confirm that:
- Every URL contains equivalent content for a defined language or regional audience.
- Destinations are fully qualified, stable, indexable, and return HTTP 200.
- One implementation method acts as the maintained source of truth.
- Every page contains a self-reference and complete reciprocal set.
- Language and region codes are valid and correctly ordered.
- Each independently indexable localized page self-canonicalizes.
x-defaultexists only when there is a genuine fallback.- WordPress, Webflow, or Shopify output has been crawled rather than assumed correct.
After launch, monitor GSC performance by country, language-related queries, and landing page. Investigate when the wrong regional URL receives impressions or clicks intended for another market, but do not wait for a dedicated hreflang report that no longer exists.
Repeat crawls after migrations, template changes, plugin updates, redirects, and translation launches. The same verification discipline should apply to your wider publishing system, from selecting blog post ideas with ranking and conversion potential to using verified-input ChatGPT prompts for SEO.
If your workflow combines separate research, crawling, publishing, and reporting products, assess whether that fragmentation is helping. A practical Semrush workflow review can help you evaluate where manual checks still remain, while link acquisition should be handled separately through tactics such as these HARO alternatives.
Final answers to common hreflang questions
- What is hreflang used for? It helps Google choose among equivalent language or regional URLs.
- How do you add hreflang? Choose HTML elements, HTTP Link headers, or XML sitemap annotations, then implement a complete reciprocal cluster.
- What are the most important best practices? Use valid codes, final indexable URLs, self-references, return links, and compatible canonicals.
- Can a generator guarantee correctness? No. Generated syntax must still be checked against response status, indexability, canonicals, rendered output, and the complete cluster.
FAQ
What is hreflang used for?
Hreflang helps Google identify equivalent pages created for different languages or regions and select the most appropriate eligible URL for a searcher. It is useful for translated pages and meaningful regional variations, but it does not provide a direct ranking boost or replace localization, canonical tags, redirects, and sound international architecture.
How do you pronounce “hreflang”?
Pronounce hreflang as “href-lang,” referring to the href URL reference and its language annotation. In implementation, the hreflang attribute contains a language code and, optionally, a region code—for example, es for Spanish or es-US for Spanish content tailored to users in the United States.
How do I add hreflang?
Choose one supported method: HTML link elements, HTTP Link headers, or XML sitemap annotations. Add every equivalent localized URL to each cluster member, including a self-reference, and ensure all pages return the relationship. Use absolute, canonical, indexable HTTP 200 URLs and validate the complete cluster after deployment.
What are the best practices for implementing hreflang tags?
Use valid language-first codes, final URLs, self-references, reciprocal return links, and self-referencing canonicals for independently indexable pages. Maintain one annotation method where possible, add x-default only for a genuine fallback, and crawl generated output after launches, redirects, migrations, plugin changes, or translation updates.
Does hreflang fix duplicate content?
No. Hreflang helps Google select among equivalent localized URLs, but it does not determine the canonical version or repair weak site architecture. Localized pages should normally self-canonicalize, remain independently indexable, and offer a legitimate language or regional experience. Use redirects or canonical tags for their intended purposes.
Make International Publishing Easier to Verify
Use seosorted.ai to bring research, SEO content creation, publishing, and performance tracking into one workflow—then keep hreflang validation in your technical release process. Start by mapping each localized URL cluster, assigning ownership, and crawling the final output before it reaches search results.
FREE BACKLINK PLAN
Build authority with backlinks that matter
Get a shortlist of sites that already link to competitors in your niche, ranked by authority, without doing the outreach research yourself.
- 20 link opportunities matched to your niche
- Discover competitor backlink sources
- Get actionable link-building opportunities

We'll email you the results and occasional SEO tips. Unsubscribe anytime.
Common questions
The core research-and-write loop is similar - both start from real SERP data and match your brand voice. The practical difference shows up in volume and what happens after publishing: SEO Sorted includes 30 articles a month at the $99 price point with add-on tiers if you need more, and it feeds Search Console performance data back in to improve future articles. Outrank currently has a live backlink exchange; SEO Sorted's is on the roadmap.
Typically 3–6 months before articles start climbing Google significantly — that's the nature of SEO, not something specific to seosorted. The compounding effect is the point: publishing consistently from day one means you're months ahead of starting later. Sites that publish 8–12 articles in the first 60 days consistently see their first rankings within 90 days. Every week without content is a week of organic traffic going to a competitor who is publishing.
Yes, always. Every article lands in a READY state first — you can read it, edit it, approve it, or reject it before anything goes live. If you'd rather not touch it, you can turn on auto-publish and articles go straight to your CMS on your chosen schedule. Full control or full autopilot — your call per article.
WordPress, Webflow, and Shopify natively — the article, featured image, meta description, and schema markup all land in your CMS automatically. For everything else — Ghost, Sanity, Strapi, Notion, or any custom headless setup — we publish via Webhook. We POST a JSON payload with the full article, SEO fields, images, and schema. If your CMS accepts a POST request, we can publish to it.
Related Blogs

Best HARO Alternatives for Link Building (2026)
Compare the best HARO alternatives for link building in 2026, including platform features, pricing, policies, and expert recommendations. Find your fit.
Updated on
Nov 13, 2026

Content Repurposing: A Practical Strategy for More Value, Not More Noise
Learn how to turn proven assets into channel-native content, avoid search-intent overlap, and measure real business value. Build a smarter strategy today.
Updated on
Nov 13, 2026
18 min read


