Domain migration mistake cuts traffic 90 percent, recovery took a year

A migration that quietly broke revenue

A 90 percent traffic collapse is not a routine post-launch wobble. It is the kind of failure that turns a domain migration from a technical project into a risk-management problem, especially when the site supports international search traffic and business across 13 country-specific domains.

The case study follows the Brazilian localization of a multinational media organization as it moved from a country-specific domain to a subdomain structure. The migration finished in January 2022, but the site never behaved like a healthy move should. Comparing December 2021 with December 2022, sessions and pageviews were down by roughly 90 percent, and Google Search Console showed the old domain pulling about 15,000 to 25,000 clicks per day before the switch, while the new subdomain settled at only 2,000 to 4,000 clicks per day after it.

Why the migration failed to consolidate authority

What makes this story useful is not just the size of the drop, but the mechanics behind it. The old domain kept getting crawled long after the migration, redirects and migration protocols were not fully implemented, and crawl budget got split between two domain structures instead of being funneled cleanly into the new one. That is the kind of failure that looks small in a staging checklist and huge in a production search profile.

For a growing agency, that is the first warning sign: a migration is not complete when the DNS flips or the new templates go live. If search engines can still spend meaningful time on the old structure, the authority you expected to consolidate stays fragmented. On a site with international reach, the damage compounds fast because every country-specific property depends on clear signals, clean redirects, and a crawl path that points in one direction only.

Do not blame algorithm noise for an execution problem

The migration also overlapped with Google’s June 2021 core update, spam update, and page experience update, which created the kind of volatility that can muddy the first read on a launch. But the case study makes the bigger point clearly: update noise does not explain a year-long failure to recover. When traffic does not normalize after the usual post-migration turbulence, the default assumption should be execution, not bad luck.

That distinction matters in agency work. If you treat every drop as a ranking fluctuation, you miss the operational clues buried in crawl behavior, index status, and redirect hygiene. If you treat it as a systems issue, you start asking the right questions immediately: Are the old URLs still being discovered? Are the redirects complete? Are the new pages actually getting indexed, or are they just returning successful status codes with thin, empty, or error-like content?

The real culprit was hiding in indexing signals

The team eventually found soft 404 issues and other indexing problems at the center of the collapse. That is where Google’s documentation becomes a practical guide instead of abstract theory. Google says soft 404 pages continue to be crawled and waste crawl budget, and it also notes that an HTTP 2xx response does not guarantee indexing if the content looks like an error or empty page.

That is exactly why the Search Console Page indexing report matters. It is built to show the indexing status of URLs and the reasons pages could not be indexed, which makes it one of the first places to look when a migration looks healthy on paper but dead in search. If you only check server response codes, you can miss the fact that the crawler sees a page as effectively useless. If you only check rankings, you can miss the crawl-budget drain that keeps dead URLs alive in the index pipeline.

Recovery started only when SEO and IT worked together

The turnaround did not come from a single tweak. In mid-August 2022, after the SEO and IT teams worked through the migration problems, the site finally showed a small uptick in search performance. The clearest signal was modest, not dramatic, including a peak of 12 clicks and 37 impressions on August 29, 2022.

That detail matters because it shows how slow recovery can be after a bad migration. The organization was not dealing with a one-day outage. It was dealing with a system where the old domain, the new subdomain, and the indexing layer were all sending mixed signals for months. The fact that recovery stretched across 13 country-specific domains makes the lesson even sharper: when a client’s revenue depends on international search visibility, technical remediation has to be coordinated across every market, not just the loudest one.

What agencies should build into higher-ticket SEO engagements

This is the part agencies should take personally. If you handle migrations at scale, pre-launch and post-launch checklists are not extras. They are the deliverable. A higher-ticket SEO engagement should assume that redirects, crawl behavior, and indexation all need formal QA, not a quick sign-off after deployment.

A practical migration checklist should include:

• A complete URL inventory before launch, so every important old URL has a mapped destination.

• Redirect testing on the live environment, with attention to chains, loops, and any old URLs that still resolve.

• Soft 404 checks on templates and high-value pages, especially where localization or dynamic content can produce thin outputs.

• Search Console validation after launch, including the Page indexing report and ongoing checks for whether important URLs are actually being indexed.

• Crawl-budget monitoring, so you can see whether the old structure is still consuming resources that should be spent on the new one.

• A cleanup plan that extends beyond redirects, because redirects alone do not solve indexing problems, duplicate discovery, or weak page content.

The main lesson is blunt: migrations are never just a dev task, and they are never just an SEO task either. If the site depends on search revenue, the agency has to own the technical governance that keeps crawl paths, indexation, and business outcomes aligned. That is how you keep a migration from becoming a year-long revenue leak.