Do 60–80% of site migrations really lose traffic?

The short version

• The 60–80% figure circulates citation-free across agency guides and vendor content. The instances we checked are quoted below, verbatim, with what each one cites (usually: nothing).
• One guide attributes the number to "a 2025 Ahrefs study." We could not find any such study. Ahrefs' closest large study of traffic losses explicitly excluded migration-caused losses from its data.^[2]
• Real data exists on a different question: how long domain migrations take to recover. A 892-migration dataset found the average new domain took ~523 days to match the old domain's estimated organic traffic, and 17% hadn't matched it after 1,000 days.^[11] That measures recovery time for domain changes. It cannot tell you what share of all migrations lose traffic.
• Documented outcomes span the full range: Wise moved 4M+ indexed URLs "without any noticeable losses" within four months^[3]; other documented migrations dropped 30% or worse.^[14] Failures get case studies; clean migrations mostly go unrecorded.

The claim, and who repeats it

We collected instances of the statistic from currently ranking migration guides and fetched each page to record the exact wording and whatever citation stood behind it. All quotes below are as retrieved on August 26, 2026.

Credit where due: Digital Applied's 2026 migration playbook explicitly flags the related "only 1 in 10 migrations improve rankings" claim as lacking "a traceable primary study" and calls it "directional folklore."^[10] The habit of checking is spreading. Slowly.

The study that doesn't exist

The MigrateLab attribution is the most checkable version of the claim, because it names a source: "a 2025 Ahrefs study."^[1] So we looked for it.

We searched Ahrefs' blog directly and through site-restricted queries. Ahrefs has published no study of what share of migrations lose traffic. Better: Ahrefs' closest large study of organic-traffic losses describes its methodology as having "filtered out traffic loss caused by website migrations and URL redirects."^[2] The one company cited as the source of the migration-loss rate deliberately excluded migrations from its loss data.

The nearest Ahrefs statement we could find anywhere is a 2021 aside in their Wise case study: "many migrations end up losing a solid chunk of organic traffic."^[3] That is an aside in a success-story writeup, not a study, and it has no number attached. Whether it is the seed that grew into "a 2025 Ahrefs study" we can't know, and we won't pretend to.

If this study exists and we simply failed to find it, we want to be wrong: send it to [email protected] and we will link it, amend this article, and log the correction below.

What the evidence actually shows

The one real dataset measures something else

The most substantial public data on migration outcomes is Dan Taylor's dataset of 892 domain migrations, published on Search Engine Journal in January 2025.^[11] Its findings, with its own caveats attached:

• On average, the new domain took 523 days (~17 months) to match the old domain's estimated organic traffic.
• The fastest recoveries recorded: 19, 22, 23, and 33 days.
• 17% had not recovered to prior traffic levels after 1,000 days (down from 42% in the study's smaller 2023 edition of 171 migrations).
• A follow-up analysis on the author's agency blog reports only 27% of domain migrations recover within 90 days.^[12]

The caveats, which the study itself is upfront about: the sample was self-selected (gathered from SEO professionals via Slack and X), traffic is third-party estimated (primarily Ahrefs), and, most importantly, these are domain migrations: rebrands and domain changes, the riskiest migration type. Platform moves, URL restructures, and redesigns are different animals with different risk profiles.

Note what this dataset does not measure: it tracks how long recovery takes among migrations that were submitted to a study. It cannot tell you what fraction of all site migrations lose traffic. Answering that would require a census that includes the migrations nobody writes about.

What Google actually says

Google's official site-move documentation (last updated August 2026) says ranking fluctuation during a move is expected and temporary: "the visibility of your content in Search may fluctuate temporarily during the move. This is normal and a site's rankings will settle down over time." On duration: "a small to medium-sized website can take a few weeks for most pages to move, and larger sites take longer." And on the mechanism everyone fears: "301 and other permanent redirects don't cause a loss in PageRank."^[13]

Documented outcomes span the whole range

Wise migrated transferwise.com (over four million indexed URLs) to wise.com in March 2021 and, per Ahrefs' case study, "managed to successfully transfer their organic traffic without any noticeable losses to their new domain within four months." The case-study author calls that an excellent result precisely because many moves go worse.^[3] At the other end, documented failures are easy to find. Search Engine Journal's advice column has fielded a reader's real 30% post-migration drop,^[14] and the 892-migration dataset's unrecovered 17% are all real websites. Both ends of the range are real. Neither end is a rate.

Why nobody knows the real number

To know what percentage of migrations lose traffic, you would need a representative sample of all migrations — including the thousands of uneventful ones that nobody blogs about. What exists instead is survivorship data in reverse: botched migrations produce postmortems, agencies publish rescue case studies, and clean migrations leave no public trace. Any rate computed from published accounts inherits that bias. This is why the honest answer to "what share of migrations lose traffic?" is unknown, and why every confident percentage you've read traces back to nothing.

What to say instead

If you're writing a migration pitch, audit, or client email and want claims that survive scrutiny, these are supported:

"No credible industry-wide research establishes what share of site migrations lose organic traffic. Documented outcomes range from zero measurable loss on a four-million-URL domain move to publicized collapses that never recovered."

"Google says temporary ranking fluctuation during a site move is normal, with most pages moving within a few weeks for small-to-medium sites, longer for large ones."

"In the only sizable public dataset (892 domain migrations, 2025), the average new domain took about 17 months to match the old domain's estimated organic traffic, and 17% hadn't matched it after 1,000 days. Domain changes are the riskiest migration type; that dataset doesn't speak for replatforms or redesigns."

Use them freely, with or without crediting us. They're only as good as their sources, listed below.

Questions people actually ask

What percentage of site migrations lose traffic?

Nobody knows — no study has measured it. The widely quoted 60–80% figure has no identifiable source. What is measured: among 892 domain migrations, 17% hadn't recovered prior traffic levels after 1,000 days.^[11]

How long does migration recovery take?

Google's guidance: a few weeks for most pages on small-to-medium sites, longer for large ones.^[13] The 892-domain dataset found an average of 523 days for domain changes, with the fastest recoveries in 19–33 days.^[11] Execution quality accounts for most of that spread.

Where did the 60–80% statistic come from?

We couldn't find an origin. Every instance we checked cites nothing, cites unnamed "industry studies," or cites an Ahrefs study that doesn't appear to exist. If you can locate an instance with a real primary source, or one older than any we found, we want it: [email protected].

Corrections: none yet. Found a primary study, an earlier instance of the claim, or an error in our quotes? Email [email protected]; we amend in place and log every change here, dated.

Driftwitness Research is the research arm of Driftwitness, an SEO change-monitoring service for agencies, currently in early access. We publish under the standard we build to: claims need evidence.