Does a clean 301 redirect really transfer 100% of PageRank without any loss?

What does a 'complete and immediate' 301 redirect mean according to Google?

Mueller emphasizes three specific criteria: the redirect must be instantaneous (no delay between the migration announcement and the implementation), page-by-page (no bulk redirects to the homepage), and without orphaned content on the old domain. This last point is often overlooked: if pages remain accessible on the old URL during the transition, Google considers there are two competing sites and must reevaluate the signals.

The transfer of signals explicitly includes PageRank, but also thematic authority, inbound links, and probably the accumulated behavioral signals. No dilution is applied if the conditions are met. This statement contradicts some older beliefs that a 301 would result in a 10-15% loss of 'juice'.

Why does a late redirect force Google to recalculate everything?

The specific case of a one-year delay is revealing. If you migrate a domain but keep the old site active for 12 months, Google treats each URL as an autonomous entity. Backlinks continue to arrive on the old domain, the new one accumulates its own signals... and when you finally decide to redirect, Google must merge two divergent histories.

The result becomes 'unpredictable' because the algorithm must arbitrate: which content is authoritative? Which link profile do we keep? The Core Web Vitals of which site? This late merging creates structural inconsistencies that Google cannot resolve deterministically. Hence the insistence on speed.

What mistakes break the transfer of signals?

The most common mistake: redirecting all pages to the root of the new domain. This saves time mapping, but Google loses the semantic match between the old and new versions. The result: it cannot transfer signals in a granular way.

Another pitfall: leaving a 200 content page on the old site 'just in case'. Even if it’s just a transition page with a manual link to the new domain, Google sees valid content and does not trigger the automatic transfer. The 301 directive must not return any exploitable HTML.

• The redirect must be page-by-page, with precise mapping from the old URLs to the new ones.
• No content should remain accessible with a 200 status on the old domain after migration.
• Timing matters: the longer you delay the redirect, the more Google must recalculate from divergent signals.
• PageRank transfers completely if the three conditions (complete, immediate, page-by-page) are met.
• A hybrid migration (two sites active simultaneously) generates unpredictable results during late consolidation.

Is this statement consistent with field observations?

Yes and no. Well-executed migrations do show a near-complete transfer of positions within 4-8 weeks following the 301. No massive loss, no sudden drop. This validates the 'all signals' part.

However, on sites with hundreds of thousands of pages, we sometimes observe persistent fluctuations for 2-3 months, even with perfect mapping. Google seems to reindex the entire architecture, recalculate the internal linking, and reevaluate orphaned pages created by the restructuring. This is not a loss of signals; it's a structural adaptation phase. Mueller does not mention this gray area.

What concrete cases contradict the claim of '100% transfer'?

International migrations with changes to hreflang structure. You migrate example.com to example.fr/fr, example.de/de, etc. Even with clean 301s, Google must rebuild the language clusters. Positions fluctuate by country for several weeks.

Another problematic case: migrations with radical changes in CMS or loading times. If your new site is twice as slow as the old one, the behavioral signals (bounce rate, time on page) will diverge immediately. Google may transfer PageRank, but degraded Core Web Vitals cancel out some of the benefits. [To be verified]: Mueller speaks of 'all signals', but real-time UX metrics cannot be transferred—they must be rebuilt.

In what scenarios can a late redirect still work?

If the old domain received no new backlinks during the waiting period and the content remained strictly identical (static archive), Google can still merge the histories. But this is a risky bet.

The real danger: sites doing domain A/B testing. They migrate 50% of the traffic to the new domain for six months and then redirect everything. Google sees two sites with divergent traffic profiles and must choose which one is authoritative. The result: often a weighted average that satisfies no one.

What should you do concretely before launching the redirect?

First step: map every URL from the old site to its exact counterpart on the new domain. No shortcuts, no 'logical' groupings. A spreadsheet with columns old_url | new_url | status_code | priority. Pages without a direct equivalent should point to the closest parent category, never to the homepage.

Second imperative: ensure that the old site returns 301s, not 302s or meta refreshes. Test with curl -I or a crawler. A single page with a 302 can delay the transfer of signals by several weeks, as Google needs to revisit and reassess.

What common mistakes break the transfer and how to avoid them?

The classic mistake: keeping a 'This site has moved' page with a 200 status on the old domain, with a manual link to the new one. Google indexes this transition page, sees it as valid content, and does not trigger the automatic transfer. If you need to inform users, do it server-side with a 301 + a robots.txt file blocking the old domain.

Another pitfall: redirecting too late after the public announcement. You tweet 'We're migrating to new-domain.com', the media picks it up, backlinks start coming to the new domain…but the old one is still active. Google sees two competing sites with divergent backlink profiles. When you finally activate the 301, it must merge two incompatible link graphs. Result: unpredictable.

How to verify that the transfer of signals was successful?

In Search Console, monitor the evolution of impressions and clicks on the new domain. If everything goes well, you should see a stair-step curve: a sharp increase in the 7-14 days following the migration, then stabilization at the level of the old organic traffic. A flat or jagged curve indicates a mapping issue.

On the crawling side, use Screaming Frog or Oncrawl on the old domain to check that 100% of URLs return a 301. Zero tolerance: any single remaining 200 status must be corrected immediately. Finally, compare the positions on a sample of 50-100 strategic queries before/after. If you lose more than 10% of visibility after 4 weeks, it means the transfer is incomplete.

• Create a comprehensive mapping old_url → new_url (no bulk redirects to the homepage)
• Configure 301s server-side (Apache, Nginx, CDN), never in JavaScript or meta refresh
• Remove all content accessible with a 200 status on the old domain (including transition pages)
• Launch the migration on a low-traffic day and monitor the logs in real-time
• Submit the new sitemap in Search Console immediately after the switch
• Monitor positions and traffic daily for 6 weeks