A brand launches a beautiful new website, and within three weeks the phone goes quiet. Traffic halved, rankings gone, and the designer says the site is better than ever, which it is, visually. This is one of the most common disasters small brands bring me, and the cause is almost never mysterious.

Google did not punish your redesign. It lost the pages it had ranked. Rankings belong to specific page addresses with specific content, and a redesign that changes addresses and rewrites pages without a migration plan quietly deletes years of accumulated evidence.

The Recovery Plan, In Order

If the Redesign Has Not Happened Yet

Handled this way, a redesign costs a small dip at most. The crash version of this story is entirely a planning failure, which is the good news, because plans are cheap.

The Migration Checklist, Expanded for Non-Technical Owners

If a redesign is coming, hand this to whoever builds it, and do not accept a launch until each line is confirmed in writing. Before launch: export every URL from your analytics and Search Console with its traffic and rankings, that list is the treasure map. Insist the new site keeps those addresses where possible, and produce a one-to-one redirect map where not, old address to closest new equivalent, never everything to the homepage, which Google treats as a soft deletion. Confirm the content on your top twenty pages carries over substantially intact: same topics, same depth, same answers, restyled but not gutted.

At launch: check robots.txt allows crawling, check five key pages for stray noindex tags, resubmit the sitemap, and request indexing on the money pages the same day. After launch: watch Search Console daily for two weeks, the Pages report for a spike in not-found errors, the performance report for which queries slip. A dip of some traffic for a few weeks is normal turbulence; a halving that keeps falling means a redirect or content gap the daily check will catch while it is still cheap to fix. None of this needs technical skill, only the insistence that someone shows you each box ticked.

When the Crash Is Weeks Old, Triage Order

Speed beats completeness in triage: each week a formerly-ranking address returns nothing, Google trusts it less, and past a few months some equity stops being recoverable at all. A crash caught in week two usually recovers nearly whole. A crash discovered in month six recovers most, not all, of what it lost, which is the politest argument that exists for watching Search Console after any launch.

The Two Dangerous Variants: Platform Moves and Domain Changes

What To Salvage When Redirects Were Never Possible

Occasionally the old site is simply gone, the platform died, the developer vanished with access, the export never existed. Salvage still beats surrender. Your analytics history names the pages that earned traffic; web archives usually hold their content; and rebuilding the ten most valuable pages at their old addresses, or redirecting those addresses to faithful rebuilds, recovers a meaningful share even months later. External links are the other map: ask Search Console which sites link to which dead pages, and make those addresses resolve first, because each one is a recommendation currently pointing at a wall.

Choosing a Builder Who Will Not Repeat This

A builder who answers those five fluently has done real migrations; one who waves at plugins and modern platforms has done real damage somewhere and not noticed. The craft is unglamorous list-keeping, which is why it gets skipped, and why the businesses that insist on it sail through launches that sink their competitors.

The closing thought for anyone mid-crash: this is the most mechanical recovery in SEO. No algorithm forgiveness required, no authority to rebuild, just forwarding restored and substance returned to where Google last saw it. Move fast, work the order, and the traffic that took years to earn usually proves loyal to the addresses that treated it well.

Reading the Post-Launch Data Like a Professional

For the fortnight after any launch, three views in Search Console tell the whole story. The Pages report shows the not-found count: a small bump is normal as old addresses retire, a climbing curve means redirects are missing, and the report lists exactly which URLs, which is your work queue. The Performance report, compared week over week, shows which queries slipped: if the losses cluster around one section, blog, categories, a location, the migration failed in that one place, not everywhere, and the fix is sectional, not general. And the sitemaps view confirms the new sitemap was fetched and its pages are entering the index, the heartbeat of the new site being adopted.

Analytics adds the commercial lens: compare organic landings by page for the four weeks before and after. Pages that held their traffic prove their redirects and content survived; pages that cratered name themselves for triage. This is thirty minutes a week of looking, and it converts a redesign from an act of faith into a monitored operation, which is all professional migration management has ever been.

The Questions Owners Ask Mid-Crash

The Recap for the Owner Mid-Crisis

A redesign crash is forwarding and substance, nothing more mysterious. Get the old URL list from analytics history today. Redirect the twenty most valuable addresses one-to-one before the weekend. Restore the gutted content on the five pages that earned the most. Sweep for staging leftovers in one Search Console pass, resubmit the sitemap, and then watch the three views weekly, Pages for not-found spikes, Performance for slipping queries, sitemaps for adoption. Expect the turbulence to settle within weeks and the true losses to return as the redirects and content do their work over one to two months. And when the recovery is banked, write the migration checklist into your company’s permanent memory, because the cheapest crash is the one the next redesign never has.