Diagnose this sequentially, ruling out the most common and most fixable causes first rather than assuming any single explanation by default. Start with a full audit of redirect mapping (crawl and log data checking for 404s, redirect loops, and one-to-many or many-to-one mapping errors), then check whether high-authority backlinks properly resolve through to their intended new destinations, then compare old versus new page content for genuine parity, and only after ruling out those three should the remaining decline be attributed to Google’s normal post-migration re-evaluation period, which Google itself has described as commonly taking weeks, and sometimes longer for large sites, before signals fully stabilize.
Why sequential elimination, not a default assumption
The tempting shortcut after a migration is to attribute any decline to “Google just needs time to re-evaluate everything,” since that’s true in a general sense and requires no further action while it plays out. But treating that as the default explanation without first ruling out actual errors risks leaving real, fixable problems unaddressed for weeks while waiting for a re-evaluation period that was never going to fully resolve a decline that had an actual technical cause underneath it. The diagnostic order matters specifically because the more common causes (mapping errors, particularly) are also the most straightforward to detect and fix, and should be cleared before falling back on the “it just needs time” explanation, which is harder to disprove and easier to hide behind.
Step one: redirect mapping audit
Run a full crawl of the new site’s URL set and cross-reference against a complete list of the old site’s URLs (from the pre-migration sitemap, historical crawl data, or server logs) to verify every old URL redirects somewhere, and that it redirects to the correct, genuinely equivalent new URL, not just to a live URL that happens to return a 200. Common mapping errors at this stage include: old URLs that were missed entirely and now return 404, redirect loops or chains introduced by the migration itself, and mass redirects to a single generic destination (like the homepage) for URLs that should have mapped to specific equivalent content, which is a frequent shortcut taken under migration time pressure that quietly discards the value of every individually mapped page.
Server log analysis is particularly useful here beyond a simple crawl, since it reveals what Googlebot is actually encountering when it revisits old URLs, including patterns a manual crawl might miss, like Googlebot repeatedly hitting old URLs that return errors rather than clean redirects.
Step two: link equity transfer check
Separately from mapping-correctness, check whether the site’s most valuable backlinks (highest-authority referring domains, or previously highest-converting inbound links) actually resolve cleanly through to their new destinations. A backlink pointing to an old URL that technically redirects correctly, but through an unnecessary multi-hop chain, or to a new URL that isn’t quite equivalent in content or intent to what the link was originally pointing to, represents a link equity transfer risk even where the mapping technically “works.” This check specifically targets whether the site’s most valuable external signals are transferring effectively, since a small number of high-authority backlinks resolving poorly can matter disproportionately compared to volume alone.
Step three: content parity comparison
Compare old and new versions of key pages for genuine content parity: whether sections were dropped during the migration (common when a redesign strips out content deemed less visually important without checking its SEO contribution), whether the new content is meaningfully thinner or restructured in a way that changes what the page is actually about, and whether structured data, headings, and internal links present on the old version carried over faithfully to the new one. A technically correct redirect to a page that no longer contains the content that originally earned it rankings is a content parity failure masquerading as a successful migration.
Step four: attribute remaining decline to re-evaluation period, with a defined expectation
Only after the first three steps come back clean should the remaining decline reasonably be attributed to Google’s documented re-evaluation period. Google’s own site-move guidance explicitly frames migrations as requiring time for recrawling, reprocessing, and signal consolidation across the entire migrated URL set, and normalizes temporary fluctuation during that window as expected rather than indicative of an error. This step should come with a defined expectation of typical duration (commonly weeks, longer for very large sites) so it functions as a genuine, time-bounded hypothesis being tested, not an open-ended excuse that avoids ever revisiting the diagnosis if the decline doesn’t recover within a reasonable window.
A hypothetical illustration
Imagine a hypothetical B2B software company that migrates to a new site platform and sees organic traffic drop by roughly a third within the first two weeks. The redirect-mapping audit might come back mostly clean, a small number of 404s, nothing systemic. The link-equity check might also look fine. Content parity comparison, however, could reveal that the redesign’s new template dropped the detailed technical-specification sections from every product page in favor of a shorter, more visual layout, sections that previously ranked well for specific long-tail technical queries. In that scenario, the decline isn’t Google “settling in,” it’s a genuine content-parity failure masquerading as a normal migration dip, and restoring the dropped sections would be the actual fix rather than waiting out a re-evaluation period that was never going to resolve it.
Practical implication
Treat this as a checklist to work through in order, documenting findings at each step, rather than jumping to whichever explanation is easiest to accept. If mapping, link transfer, and content parity are all confirmed clean and the decline still hasn’t meaningfully recovered well past a reasonable re-evaluation window for the site’s scale, that’s a signal worth revisiting the diagnosis rather than continuing to attribute an unusually persistent decline to “Google still settling in,” since genuine re-evaluation periods, while real, aren’t indefinite.
