By Mehdi HassanWhat is website migration recovery?
Website migration recovery is the work of finding and fixing what broke during a site move so that search traffic returns. It's triage: you rank problems by how much traffic they're costing and how quickly they can be fixed, then work down the list in a fixed order.
This plan is for teams who migrated with someone else, launched, and watched clicks fall. If you haven't confirmed the drop is real, read our guide to traffic drops after a website migration first, because broken tracking can look like lost traffic. The full prevention list lives in our website migration SEO checklist, and it doubles as an audit checklist after the fact.
How do you recover SEO after a failed website migration?
You recover SEO after a failed migration by fixing problems in order of impact: crawlability first, then redirects, then content, then everything else. Working out of order wastes the first week, which is when fixes pay off most.
Priority | Problem | Typical symptom | Fix |
|---|---|---|---|
1 | Pages blocked from indexing | Sudden site-wide drop | Remove noindex, fix robots.txt and canonicals |
2 | Content not in the HTML | Pages indexed with little text | Server-render or pre-render main copy |
3 | Missing redirects | 404s on old URLs that had clicks | Page-level 301s from a redirect map |
4 | Bad redirects | Chains, 302s, homepage redirects | Point each URL straight to its match |
5 | Lost content | Gradual slide on specific pages | Restore copy, headings, FAQs and links |
6 | Metadata and schema loss | Worse titles in results, rich results gone | Rebuild titles, descriptions and schema |
7 | Speed and UX regressions | Lower engagement, slow pages | Fix images, scripts and layout shifts |
What happens in the first week of recovery?
The first week of recovery is an audit and the fixes that can't wait. By the end of it you should know every cause of the drop and have fixed the ones that block crawling and the redirects that matter most.
Days 1–2: Gather evidence
Export Search Console performance comparing before and after launch, the pages report, and any pre-migration crawl. If no old crawl exists, rebuild the old URL list from Search Console, analytics and backlink data.
Days 2–3: Fix crawl blockers
Check live HTML for noindex and wrong canonicals, review robots.txt, and confirm the sitemap lists new URLs only. Fix and resubmit.
Days 3–5: Rebuild redirects
Crawl the old URL list, flag 404s, chains and homepage redirects, and fix them in order of past clicks and links.
Days 5–7: Compare top pages
For the pages that lost the most clicks, compare old and new versions. Note missing copy, headings, FAQs, links and schema.
If the old site is gone, archived copies and your old CMS export are the best record of what each page said. Redirect fixes depend on the platform: see Webflow 301 redirects, Framer redirects or redirects in Next.js.
What happens in weeks two to four?
Weeks two to four restore what the new site lost and fix the slower technical issues. This is where most of the remaining traffic comes back, page by page, as Google recrawls the fixed versions.
- Week 2: content parity. Restore the copy, H2s, FAQs and comparison tables that ranked on the top pages. Keep the new design; put the substance back into it.
- Week 2: internal links. Rebuild links from navigation, footers and related posts that the new templates dropped.
- Week 3: metadata and schema. Restore titles and descriptions that earned clicks, and re-add Article, FAQ and Organization schema where it existed.
- Week 3: rendering and speed. Fix pages that rely on client-side JavaScript for main content and pages that got much slower.
- Week 4: long tail. Work through remaining 404s from Search Console, orphan pages and stray chains.
Hypothetical example: a botched Webflow to Next.js launch
For example, suppose a B2B startup moves from Webflow to Next.js with a contractor, and clicks fall sharply in the first two weeks. The audit finds three causes: the blog moved from /blog/slug to /resources/slug with no redirects, a staging noindex is still set on the resources template, and the new pricing page lost its FAQ.
Week one removes the noindex and adds one pattern redirect for the blog folder. Week two restores the pricing FAQ as real HTML with FAQ schema. Weeks three and four clean up the long tail of 404s from old campaign URLs. Each fix is logged with its date, so the team can see in Search Console which change moved which pages.
How do you know the recovery is working?
You know the recovery is working when impressions recover on fixed pages first, then clicks follow. Impressions react sooner because Google shows the page again before it ranks as well as before.
Track a short list of pages that matter, not the whole site. Compare each week with the same week before launch, and annotate your fix log. If a fixed page hasn't moved after a few weeks of recrawling, look again at its content and internal links. Our guide on what to do after an SEO audit covers turning findings into a prioritized backlog.
Should you fix it yourself or bring in help?
Fix it yourself if you have someone who can edit redirects, templates and metadata on the new platform and a week of their focused time. Bring in help if nobody owns the new stack, the original builder is gone, or the drop is costing pipeline every week.
Recovery is usually cheaper than the migration was. Agency quotes for migrations themselves range widely, roughly $1.5k to $25k by site size in the ranges we've seen published, and they vary. A recovery sprint covers less ground because the site already exists. What it needs is someone who understands SEO, content and code at once, which is the gap many botched launches fell into.
- Trial week: one week of design and development, before any commitment
- $1,000
- Retainer: ongoing design and development, no hourly tracking
- $6,000/mo
- Free intro call to review your Search Console data
- 30 min
What should you do differently next time?
Next time, benchmark before launch, write the redirect map before building, and check the live site within hours of going live. Most recoveries trace back to one of those three being skipped.
