Official statement
Other statements from this video 12 ▾
- 17:46 Les textes en pied de page nuisent-ils vraiment au référencement de votre site ?
- 18:30 Combien de temps faut-il vraiment pour qu'un changement de métadonnées impacte vos positions ?
- 21:11 Googlebot indexe-t-il vraiment les images en lazy loading ?
- 25:45 Les pop-ups intrusifs détruisent-ils vraiment votre SEO ?
- 27:25 Les menus burger pénalisent-ils vraiment le référencement de vos liens internes ?
- 29:20 Le Data Highlighter vaut-il encore le coup face au JSON-LD ?
- 42:00 Pourquoi Google réécrit-il vos balises title et meta description sans vous demander votre avis ?
- 46:00 Le masquage de contenu en mobile est-il vraiment sans risque pour le SEO ?
- 53:02 Le code 503 est-il vraiment l'ami du SEO en cas de surcharge serveur ?
- 54:20 Les erreurs 410 nuisent-elles vraiment au référencement de votre site ?
- 55:44 Hreflang et sous-domaines multilingues : contenu dupliqué ou non ?
- 57:30 Pourquoi diviser ou fusionner des domaines ralentit-il votre visibilité SEO ?
Google confirms that backlinks acquired by an AMP page pass on to the normal HTML version through the rel=canonical tag. Essentially, a link gained on your AMP does not go to waste: it directly strengthens the authority of your standard page. Just ensure that your canonical implementation is correct; otherwise, this transmission fails silently.
What you need to understand
How does this statement change the perception of AMP?
For a long time, many SEOs viewed AMP pages as parallel versions that could dilute link juice. The concern was that a backlink pointing to the AMP version might get "stuck" on that technical URL, failing to benefit the main page.
John Mueller dismisses this worry. The rel=canonical tag placed in the AMP code explicitly indicates to Google what the reference version is. Popularity signals — backlinks, mentions, shares — therefore flow back to the classic canonical HTML URL.
How does this transmission technically work?
Each AMP page includes in its <head> a rel="canonical" tag pointing to the standard HTML URL. Google crawls the AMP, detects this directive, and consolidates the ranking signals on the canonical version.
This mechanism is the same as for any technical duplicate: pagination, separate mobile versions, URL parameter variations. The difference here is that the AMP is hosted on a Google domain (google.com/amp/) or in AMP cache, which has caused confusion.
What implementation errors break this consolidation?
Three classic scenarios sabotage the transmission. The first case: the canonical tag points to an HTTP URL while the normal version is in HTTPS. Google then considers these two as distinct entities.
The second trap: circular or self-referential canonicals on the AMP itself. A third common mistake: poorly managed 301/302 redirects between the AMP and the HTML version, creating redirect chains that weaken signal consolidation.
- The rel=canonical tag on the AMP must point to the normal HTML URL, not to the AMP itself
- The canonical URL must be absolute, in HTTPS, and exactly match the reference version
- Google does not guarantee 100% transfer if intermediate redirects are present between AMP and canonical
- Backlinks acquired on the Google AMP cache (
google.com/amp/) follow the same consolidation logic - An AMP page without a valid rel=canonical is treated as a distinct SEO entity, diluting signals
SEO Expert opinion
Is this statement consistent with field observations?
Yes, and tests have confirmed this for years. Sites that have massively deployed AMP — media, e-commerce — have not noted any loss of domain authority despite the proliferation of technical URLs.
Yet, Google never publishes precise figures on the consolidation rate. To say that benefits "transfer" does not mean it is instantaneous or at 100%. Crawling delays, exploration budgets, and implementation anomalies create delays we observe in production. [To be verified]: Google has never communicated about the speed of this consolidation.
In what cases does this rule not fully apply?
The first problematic case: sites that serve the AMP in standalone version, without a corresponding standard HTML version. There, there is no canonical to which signals can revert — the AMP effectively becomes the reference page.
The second limitation: backlinks featuring specific AMP anchor text. If a link points to your AMP with anchor text like "fast AMP version", Google may interpret a different usage context than the normal HTML page, which weakens the relevance of the thematic transfer.
What nuances should be added to Mueller's statement?
Mueller talks about the “benefits of backlinks,” a deliberately vague formulation. Does it include raw PageRank, thematic authority, trust, link freshness? Probably all of these, but Google never details the weighting.
Another point: Mueller does not mention behavioral signals. A super-fast AMP page often generates a better click-through rate, less bounce. Do these signals also flow to the canonical? Nothing in this statement explicitly guarantees this. [To be verified]: does the consolidation only concern backlinks or also UX metrics?
Practical impact and recommendations
What should you prioritize checking on your AMP pages?
First action: audit all your rel=canonical tags on the AMP side. Use a crawler (Screaming Frog, Oncrawl) in "AMP only" mode and extract the value of the canonical tag. Compare with the reference HTML URL: they must match exactly, protocol HTTPS included.
Second check: test your AMP in the Search Console, on the "Page Experience" tab or via the AMP testing tool. Google flags canonical errors, redirect chains, and markup inconsistencies there.
How can you maximize the consolidation of AMP backlinks to HTML?
If you are engaged in active link building, prioritize links to the classic HTML version rather than to the AMP. Sure, Google consolidates, but why add an intermediate step that lengthens crawling and might dilute the signal?
For backlinks already acquired on AMP (press mentions, social shares, aggregators), ensure that the canonical is in place and stable. A change in canonical URL during an AMP page's lifetime can create a temporary ranking loss while Google re-crawls and re-consolidates.
What mistakes should absolutely be avoided in your AMP strategy?
Never deploy AMP without a corresponding HTML version if you aim for sustainable SEO. The standalone AMP locks you into a format controlled by Google, with no means to recover your signals if you decide to migrate.
Avoid relative canonicals (href="/article"): Google sometimes misinterprets them, especially when the AMP is served from Google's cache. Always use absolute URLs (https://yourdomain.com/article).
- Audit all canonical tags of AMP pages to ensure they point to the HTML version in HTTPS
- Test each AMP template in the Search Console and fix reported errors
- Prioritize backlinks to the classic HTML version during your link-building campaigns
- Avoid redirect chains between AMP and HTML: one canonical consolidation step only
- Monitor the indexing of your AMPs in the Search Console to detect any URL cannibalization
- Never deploy standalone AMP without an HTML version if you want to maintain control over your SEO
❓ Frequently Asked Questions
Un backlink vers une page AMP a-t-il la même valeur qu'un backlink direct vers la version HTML ?
Que se passe-t-il si ma page AMP n'a pas de balise rel=canonical ?
Les backlinks depuis le cache AMP Google (google.com/amp/) comptent-ils aussi ?
Faut-il encore déployer de l'AMP si la priorité est le SEO backlinks ?
Comment vérifier que mes canoniques AMP sont correctement interprétées par Google ?
🎥 From the same video 12
Other SEO insights extracted from this same Google Search Central video · duration 59 min · published on 09/02/2017
🎥 Watch the full video on YouTube →
💬 Comments (0)
Be the first to comment.