Most traffic losses after a website migration were preventable. The SEO migration process we use to protect rankings: benchmarking, redirect mapping, launch-day checks and recovery.
Most websites lose traffic after a migration. Not because migrations are inherently destructive, but because the SEO work is treated as a launch-day checklist item instead of a project that starts months earlier. The pattern is depressingly consistent: a beautiful new site goes live, organic traffic drops 30 to 60 percent, and everyone discovers that the old URLs, the ones Google spent years learning to trust, now lead nowhere.
Almost all of that loss is preventable. Here is the process we use.
Before anything moves: benchmark and map
You cannot protect what you have not measured. Before a single template changes, capture the full picture of what currently earns traffic: every URL with impressions or clicks in Search Console, every page with external links pointing at it, every ranking that drives revenue rather than vanity. This benchmark is what recovery gets measured against, and it settles arguments later about whether a drop is migration damage or seasonality.
Then build the redirect map: every old URL matched to its closest new equivalent, one to one wherever possible. This is the single highest-value document in the entire migration. Redirecting everything to the homepage, the lazy default, tells Google the specific page it trusted no longer exists, and the authority that page earned quietly evaporates.
The decisions that quietly change your SEO
Migrations rarely fail on redirects alone. They fail on the accumulated small decisions nobody flagged: templates that render headings differently, content that got trimmed because the new design wanted less text, internal links that vanished with the old navigation, or a new platform that handles URLs, pagination or structured data differently. A proper technical SEO audit of the staging site before launch catches these while they are still cheap to fix.
Page speed deserves its own line. A redesign is the single most common moment for performance to regress, and as we covered in our work on web performance as a revenue lever, that regression costs conversions as well as rankings.
Launch day and the fortnight after
On launch day the checklist is short but unforgiving: redirects live and returning 301s, the new site crawlable with no stray noindex tags left over from staging, XML sitemaps updated and submitted, and Search Console watching the new structure. Then comes the part most teams skip: two weeks of daily monitoring. Crawl errors, redirect chains, rankings on your benchmark keywords, and indexation counts. Small problems found in week one are corrections; the same problems found in month three are recoveries.
What a good migration looks like in the numbers
Expect some turbulence: a few weeks of fluctuation while Google re-crawls and re-evaluates is normal even in a perfect migration. What should not happen is a step change downward that persists past a month. When we manage migrations for clients, the benchmark from step one becomes a weekly scorecard until every revenue-driving ranking is confirmed stable or better. That discipline reflects how we run everything: measured against commercial outcomes, not task completion.
If you are planning a migration now
The best time to involve SEO is when the new site is still wireframes, because that is when URL structure, templates and content decisions are still cheap to change. The second best time is before launch, when a staging audit can still catch the expensive mistakes. The worst time, and the most common, is after the traffic has already dropped.
If a redesign or replatform is on your roadmap, speak to us before it goes live. Protecting rankings through a migration costs a fraction of winning them back afterwards, and our SEO team has managed enough migrations to know exactly where the bodies get buried.
