SiteVerdict — hreflang Validator & Return-Tag Matrix

Find out exactly why Google shows the wrong country version of your pages. The validator crawls your international site and reads hreflang from all three places Google accepts: <head> link tags, HTTP Link headers and XML sitemaps. It then fetches every alternate, including those on other domains, and checks each pair. The result is a return-tag matrix with an issue list that says what to fix and where.

Who it is for: international SEO teams, agencies running multi-country sites, developers of localised templates, and AI agents (MCP-ready).

Three ways people use it

1. International audit. Get one matrix per hreflang cluster (pages × language codes) with ✓ return / ✗ no return per cell. MATRIX-CSV drops straight into a spreadsheet.
2. Release guard. Schedule it with monitoring on and get a Slack/webhook alert when a deploy breaks return tags, adds an invalid code or points hreflang at redirects/404s.
3. ccTLD / subdomain setups. Alternates on other domains (example.de, fr.example.com) are fetched politely to verify return tags across properties.

What it checks

Issue	Severity
Invalid code: not ISO 639-1 / ISO 3166-1, en-UK, en_GB, jp, gb, es-419, en-EU, unknown scripts. Includes plain-English explanations ("UK is not an ISO country code; use en-GB")	high
Missing return tag (the alternate does not link back), including across domains	high
Same code pointing to different URLs	high
hreflang target is 4xx/5xx/unreachable, noindex, or canonicalises elsewhere	high
hreflang <link> inside <body> (ignored by Google)	high
Relative or invalid hreflang URLs	high
hreflang target redirects	medium
Page with hreflang is not canonical	medium
hreflang target blocked by robots.txt	medium
HTML/header and sitemap annotations disagree	medium
Missing self-reference	medium
Missing x-default	low
<html lang> does not match the hreflang language	low
Cluster pages list different sets	low
Return tag could not be verified (alternate not fetched; listed, never counted as an error)	info

Output

• Dataset views: Issues (one row per problem, with fix), Clusters (matrix) and Summary.
• Key-value store: REPORT (HTML with issues + matrix per cluster), REPORT-PDF, MATRIX-CSV (cluster, page, hreflang, target, returnTag), ISSUES-CSV, SUMMARY.

Pricing (pay per event)

Event	Price	When
page-checked	$0.002	Each page or alternate URL fetched (200) and analysed
site-report	$0.02	Once per site: matrix, HTML/PDF/CSV reports, monitoring diff, alert

Examples: 5 languages × 40 pages (200 URLs) ≈ $0.42. 2,000 URLs ≈ $4.02. Robots-blocked or failed URLs and sample mode are free.

FAQ

Do I need hreflang in both HTML and the sitemap? No. Google accepts either. If you use both, they must agree, and we flag pages where they don't.

Why is "uk" valid? uk is Ukrainian. If you meant the United Kingdom, use en-GB. We add a note so the mistake is visible.

Does it render JavaScript? No. Google strongly recommends hreflang in the server HTML, headers or sitemap, which is what this tool reads.

Limits

• 5,000 pages and 5,000 alternate fetches per site; 20 sites per run.
• Sitemaps up to 20,000 URLs are read.
• Return tags for alternates beyond the limit are reported as "unverified".

Sources & compliance

• Rules follow Google Search Central "Tell Google about localized versions of your page" (developers.google.com/search/docs/specialty/international/localized-versions, read 11 Oct 2026). Code lists are ISO 639-1, ISO 15924 and ISO 3166-1 alpha-2 (facts).
• Data comes from the public pages, headers and sitemaps of the sites you check. They are fetched with the honest user agent SiteVerdictBot/1.0 (+aman.thakran@bharatfare.com), and robots.txt is obeyed on every host. The tool never bypasses bot protection. Fetching is SSRF-safe.
• No personal data is collected. Under UK GDPR/PECR you are the controller of your audit, so only check sites you own or may audit. The webhook URL is a secret input.