Website redesign SEO checklist: protect what already works

Keep useful pages reachable through the redesign

To reduce avoidable search disruption, preserve URLs that still answer the same question, retain useful content and map genuinely changed URLs to their closest replacement. Check indexing rules, redirects and measurement before release, then monitor the live site. A redesign is not a reason to rename every page.

No checklist guarantees unchanged rankings. Search engines need to recrawl and process changes, and visibility can fluctuate. The practical goal is to avoid broken routes, lost answers and contradictory indexing signals while making the site better for its readers.

Inventory URLs and content before designing

Build a worksheet from a crawl, the existing sitemap, analytics and Search Console. Include important landing pages, articles, images and downloads, not just the pages currently shown in navigation. Account for externally linked URLs and old campaign destinations that visitors still use.

For each URL, record:

• Current status code, canonical URL and whether it is indexable.
• Page purpose, title, headings and useful content to retain.
• Search and visitor evidence available for that page, with the date range.
• Internal links, important external references and conversion actions.
• Decision: keep, improve, merge, move or deliberately remove.
• Destination URL, redirect requirement and named review owner.

Keep the URL when the page still answers the same question. A refreshed layout or better wording usually does not need a new address. If merging pages, verify that the destination actually covers the old page's useful answer. Do not delete content merely because it does not fit the new template.

Map permanent redirects before launch

Create an explicit old-to-new mapping for URLs that must change. Use server-side permanent redirects, normally 301 or 308, to the relevant replacement. Update internal links to point directly to the final URL rather than depending on those redirects. Test for chains, loops and destinations that return errors.

Avoid redirecting every removed page to the homepage. If no relevant replacement exists, return an appropriate 404 or 410 rather than pretending the missing page has equivalent content elsewhere. Check path variants and valuable query-string URLs against the actual old site, not an assumed pattern.

Check staging and production separately

Keep staging out of search, preferably behind access controls. Where staging uses a noindex directive, track it as a release hazard: it must not carry over to public pages that should be indexed. A robots.txt disallow is not equivalent to noindex and can prevent a crawler from seeing that directive.

Before switching traffic, test the production configuration:

• Important live pages return 200 and contain the expected visible content.
• Changed URLs redirect once to the intended live destination.
• Canonicals use the correct public domain and intended final URL, not staging.
• Indexable pages have no accidental noindex meta tag or HTTP header.
• Robots rules allow intended public content and required rendering resources.
• Retired URLs return the chosen status, and private routes remain protected.

Also inspect titles, headings, structured data and mobile navigation. Keep their meaning aligned with the page instead of importing old metadata blindly. Use our guide to what an AI-ready website means for the related question of making each page understandable, rather than treating extra discovery files as a substitute for migration checks.

Preserve measurement and publish the sitemap

Record a pre-launch baseline using comparable periods and the same definitions. Keep the appropriate analytics property, consent behaviour and conversion events working through the redesign. Test important forms and confirmation journeys; otherwise a tracking change can look like a business change. Note the launch date so later comparisons have context.

Publish a sitemap containing intended canonical, indexable URLs that return 200, not redirect sources or staging addresses. Check its domain and robots.txt reference. Submit it through your authorised Search Console property, and use URL Inspection for representative pages. Submission is not proof of indexing.

Monitor and respond after launch

Immediately check key landing pages, redirects, forms and analytics on the live domain. Over the following days and weeks, review Search Console indexing, reported crawl problems, server errors and search performance. Compare like-for- like periods and account for seasonality before attributing every movement to the redesign.

Prioritise accidental blocking, widespread errors and broken high-value routes. Keep a named owner and a rollback plan for serious release faults. Inspect individual affected pages before making speculative site-wide changes.

For actual studio proof, our Dossier site foundation field report describes how page purpose, canonical URLs, redirects and discovery surfaces fit together. It is an implementation example, not evidence of guaranteed traffic gains. Our search-ready website service covers this foundation alongside the visible redesign.

Related capability

AI Search & SEO-Ready Websites

Websites built for people, search engines, AI assistants, and conversion from the same foundation.