A technical SEO checklist before a site relaunch must preserve evidence from the old site, test the new site as a complete system and define what happens when the cutover fails. Export the current URL inventory before changing it, map every valuable destination, crawl staging under production-like conditions and monitor both old and new paths after launch. A successful deployment is not yet a successful migration.
Freeze the old site as a baseline
Capture every discoverable URL from crawls, sitemaps, analytics, search reporting, internal links and known campaigns. For each address, store its response, canonical, indexability, title, main heading, inbound-link context and intended outcome. This baseline lets the team distinguish a new defect from an old condition.
Do not migrate all pages by default. Decide whether each URL stays, moves, consolidates or retires. When an address changes, map it to the closest equivalent destination rather than a generic page. Preserve useful content and metadata where they still serve the same intent. The website asset and font audit adds a complementary check for licenses and hosted dependencies that can break during a redesign.
Record business-critical journeys separately. Forms, checkout steps, account entry, contact routes and analytics events may remain outside a crawler’s view. Name the owner and acceptance evidence for each one.
Freeze the measurement definitions as well. A relaunch can change event names, consent behavior and reporting filters while the visible pages remain intact. Without a comparable baseline, the team may mistake an instrumentation break for a change in user demand.
Crawl staging as if it were public
Keep staging out of public indexing while allowing the authorized audit tools to inspect it. Crawl with rendered output and compare the results to the baseline. Verify response behavior, redirects, canonicals, internal links, hreflang where used, structured data, images, page titles, headings and XML sitemaps. Search for production URLs embedded in the wrong context and staging addresses leaking into final tags.
- Test redirect chains and loops from representative old URLs.
- Confirm that blocked staging rules will not ship to production.
- Verify analytics, consent and conversion destinations.
- Check mobile rendering, performance and accessibility regressions.
- Prepare a current crawl and monitoring query for launch day.
Operators need a dashboard that explains failures, not only a launch checklist. The analysis of portals that overlook operator needs is relevant: the people handling a migration need filters, ownership and repair actions for exceptions that appear after release.
Control cutover and the observation window
At launch, remove temporary blocks, publish the correct sitemaps, test selected old and new URLs from outside the deployment environment and confirm that analytics receives real events. Save the deployed build, redirect rules and configuration so the team can explain exactly what changed.
Recrawl immediately and continue monitoring according to the site’s risk and change volume. Inspect server logs, search reporting, index coverage, redirect errors, not-found requests and conversion paths. Triage by business and search importance rather than fixing the easiest rows first. Keep a rollback or forward-fix decision owner available while recovery is still practical.
The staged go-to-market playbook supplies a useful governance pattern: a phase opens only when its evidence is accepted. For a relaunch, the release record should state passed checks, accepted limitations, open defects, owners and the next observation time.
AI4SALE’s public evidence supports only the general statement that it provides AI product development and has digital-product delivery evidence plus maintained open-source AI tooling. It does not guarantee stable rankings, conversion, adoption or a client outcome after a relaunch.
Frequently Asked Questions
Capture it before redirects, navigation and templates change. Combine crawl, sitemap, analytics, search and campaign sources so important URLs are not missed.
No. Redirect a changed URL to the closest relevant equivalent, consolidate deliberately, or retire it with an appropriate response when no replacement exists.
Keep it outside public indexing while giving authorized crawlers access. Before launch, verify that temporary blocks and staging addresses are absent from production configuration.
Completion requires post-launch crawling, monitoring, journey tests, accepted defects and a clear repair or rollback decision, not merely a successful deployment.
If a product relaunch needs URL mapping, staging evidence, cutover control and a repair plan, scope the delivery through AI4SALE AI product development. The objective is a migration the team can observe and correct, not a checklist completed before anyone sees the result.