How to Redesign a Website Without Losing SEO Traffic

Most websites do not lose their traffic to an algorithm update. They lose it to their own redesign. A new site goes live, everyone celebrates the visuals, and three weeks later organic traffic is down 40% because nobody mapped the old URLs, the metadata was regenerated by the CMS, and half the internal links now point at 404s. Approaching a website redesign for SEO is not optional polish; it is the difference between launching an upgrade and launching an expensive rollback.

The good news: protecting traffic through a redesign is a known, mechanical process. Here is the exact sequence we run, in order.

Why redesigns destroy traffic

Google does not rank your website. It ranks individual URLs, each with its own accumulated history: content relevance, internal links, external backlinks, engagement signals. A redesign typically changes URLs, deletes pages, rewrites content and restructures navigation all at once. Every one of those changes discards part of that accumulated equity unless you deliberately carry it across. The process below exists to carry it across.

Step 1: Crawl the current website

Before anyone designs anything, crawl the existing site with a tool such as Screaming Frog or Sitebulb. Export every URL with its status code, title, meta description, H1, canonical tag and internal link count. This crawl is your inventory: the definitive record of what exists today. Without it, every later step is guesswork.

Crawl before the old site is touched, and keep the export somewhere safe. Once the old site is gone, this spreadsheet is the only complete record of what you had.

Step 2: Export your ranking pages

The crawl tells you what exists; Search Console tells you what earns. Export twelve to sixteen months of page-level data (clicks, impressions, average position) from the Performance report. Then pull your top pages by backlinks from Ahrefs, Semrush or similar, because a page with strong external links is valuable even if it converts nothing directly.

Now you have a ranked list of the pages doing the work. In most cases a minority of URLs drive the large majority of organic traffic. Those pages get handled with care; everything else gets a decision, not an accident.

Step 3: Preserve valuable URLs

The single cheapest way to protect rankings is to not change URLs at all. Where a page’s content survives the redesign, keep its address identical: same path, same slug. Resist the temptation to “tidy” URL structures for aesthetics. A restructure from /services/web-design to /what-we-do/websites gains you nothing and forces Google to re-evaluate a page it had already decided to rank.

Change a URL only when there is a real reason: consolidating duplicates, fixing genuine structural problems, or retiring content. Every changed URL goes into the redirect map.

Step 4: Create a redirect map

The redirect map is a spreadsheet with three columns: old URL, new URL, redirect type. Every URL from your Step 1 crawl that will not exist at the same address after launch needs a row. The rules:

• Use 301 (permanent) redirects, not 302s, so equity transfers.
• Redirect to the closest equivalent page, not the homepage. Blanket homepage redirects are treated as soft 404s and pass little value.
• One hop only. No chains, no loops. If an old URL already redirects, update it to point at the final destination.
• Retired content with no equivalent can return 410 or redirect to its parent category. Decide per page, do not default.

This document is the single highest-stakes artefact in the entire redesign, and it needs a named owner. If your launch date is fixed and nobody on the project owns the redirect map, book a migration review and bring your URL list. We will go through it with you before it becomes a traffic problem.

Step 5: Transfer titles and descriptions

Title tags are among the strongest on-page signals you control, and they are routinely vaporised during rebuilds because the new CMS generates defaults (“Home | BrandName”) over metadata that took years to refine. Take the titles and meta descriptions from your Step 1 crawl and migrate them field by field for every kept page. Rewrite them where they were weak, but deliberately, page by page, never by template default.

Step 6: Retain internal links

Internal links tell Google which of your pages matter most. Redesigns break them in two ways: navigation changes silently remove sitewide links to key pages, and body-copy links get stripped when content is migrated. Check that your most valuable pages keep at least the internal link coverage they had, and update every internal link to point directly at final URLs. An internal link that bounces through a redirect wastes crawl budget and dilutes the signal.

Step 7: Check canonicals and indexing rules

Staging sites are built noindexed, as they should be. The most catastrophic redesign error is launching with that setting intact. Before and immediately after launch, verify three things: the noindex meta tag is gone from production, robots.txt is not disallowing the site, and canonical tags point to the live domain rather than the staging URL. Each of these has quietly deindexed entire websites. Check all three even though someone assures you they are fine.

Step 8: Submit the new sitemap

Generate a clean XML sitemap containing only live, indexable, canonical URLs (no redirects, no 404s) and submit it in Search Console on launch day. One useful trick: temporarily leave the old sitemap submitted as well. Google will recrawl the old URLs listed in it, hit your 301s, and discover the new addresses faster. Remove it once the redirects have been processed.

Step 9: Monitor Search Console after launch

Launch day starts the monitoring window; it does not end the project. For the first two weeks, check daily: the Pages (index coverage) report for spikes in 404s or “Crawled – currently not indexed”, the Performance report against your Step 2 benchmark, and a fresh crawl of the live site to catch broken links and redirect chains. Some ranking turbulence for two to four weeks is normal while Google re-evaluates. What you are watching for is a sustained slide, and page speed regressions are a frequent hidden cause, which is why we treat site speed as an SEO discipline in its own right.

Turbulence is normal. Collapse is not.

Expect movement of a few percent either way for up to a month after launch. A drop that exceeds 20% and keeps deepening after the second week is not turbulence; it is a defect in the migration, and it is findable.

What to do when traffic drops

If traffic falls off a cliff post-launch, work the list in this order (it is sequenced by likelihood):

1. Check indexing controls. Noindex tags, robots.txt, canonicals to staging. Minutes to check, catastrophic if missed.
2. Test the redirect map. Crawl your old URL list. Every old URL should 301 once to a live 200 page. Fix 404s and chains first.
3. Compare metadata. Spot-check your biggest former traffic pages against the old crawl: are titles, H1s and content substantially intact?
4. Audit internal links. Did the pages that lost rankings also lose internal link coverage?
5. Measure page speed. If the new site is heavier and slower than the old one, that shows up in rankings.

Most post-launch drops trace to one of the first three. Structured recovery works (we documented the full playbook in our case study on recovering 40% of lost traffic), but recovery always costs more than prevention. Deep-dive measurement, migration audits and technical SEO of this kind are exactly what our tech engine exists for.

Website redesign for SEO: opportunity, not just risk

Run properly, a redesign is not just risk management; it is the best moment you will ever get to consolidate cannibalising pages, fix years of structural drift and ship a faster site. Every step above also appears in our full 27-point website redesign checklist, alongside the conversion, tracking and DNS checks that sit around the SEO work.

Redesign your website without throwing away the traffic you already earned.