A site owner on Reddit says Google swapped a migrated domain back to the old one four months after traffic had recovered, and Google’s John Mueller replied that a domain move can take a very long time to finish. Search Engine Journal’s Roger Montti covered the thread on October 13, 2026. The poster’s account is unverified, and Google has not confirmed any detail of it.
The case adds a wrinkle to our earlier report on Gary Illyes’s timings, where Illyes gave one to three months as typical for a site move and six months to a year or more as the slowest outcome. A migration that looks finished is not necessarily settled.
According to the Reddit post, the owner renamed the site and switched domains in early 2026, set up redirects, used Search Console’s address-change tool, and updated internal names. Traffic returned by April, and the new domain drew roughly 5,000 impressions a day. Then, the owner writes, Google reverted. Over 5,000 URLs under the new domain moved to “Crawled, currently not indexed,” impressions fell to about 20 a week, and the old domain’s pages reappeared in Search Console with the clicks and impressions. Visitors who click the old listing are still redirected to the new address.
Mueller offered two points. Sometimes, he wrote, processing takes “a particularly long time,” and he invited the poster to send the domains by direct message. He also said Google noticed, and later changed its documentation, that the move should be registered for the www host and the bare host alike, so Google can complete the migration. Mueller did not say this caused the reversion.
Montti points out what the post leaves out. The owner does not say whether the new site was crawled for internal links hardcoded to the old domain, how the redirects were built, or whether the XML sitemap was updated. Montti notes that a registrar-level redirect may not preserve URL paths and that HTTPS forwarding can fail when the old domain lacks a certificate. Chained redirects and a changed navigation structure can also break redirects. A sitemap that still lists the old domain would send conflicting canonical signals.
Montti frames the diagnosis as canonicalization, meaning Google’s choice of which URL to index and rank. External links pointing at one www variant or the other could push Google toward the old domain as canonical, which is why the www and non-www advice fits that reading.
Search Engine Roundtable’s Barry Schwartz, who also flagged the thread, was skeptical that the www fix explains this case. He added that readers have emailed him about migrations going wrong months afterward, and that Glenn Gabe and Marie Haynes report the same. Haynes offered a theory, not a finding: that a migration triggers a fresh quality assessment, and that Google is stricter about what it indexes. Haynes says affected sites often show every page as crawled but not indexed, and that the content offers little original compared with what is already indexed. Google has not endorsed that explanation.
Treat that quality theory with caution. It rests on anecdotes from consultants, and no data has been shared. Still, it suggests a second check beyond redirects: whether the pages would earn indexing if Google evaluated them from scratch.
For practitioners, the case works as a migration checklist with a longer tail than most plans allow. Before launch, crawl the old site to find absolute internal links, confirm redirects map URL to URL over HTTPS with no chains, and publish a sitemap that lists only the new domain. At launch, file change of address for every host variant. Afterward, keep reviewing coverage reports and canonical choices well past the point where traffic looks recovered, and tell stakeholders the project is open until Google’s indexing matches the plan.
Based on Search Engine Journal’s October 13, 2026 report by Roger Montti, with additional context from Search Engine Roundtable’s coverage by Barry Schwartz.
