SEO
Technical SEO launch checklist: what to verify before a website goes live
A practical pre-launch checklist for crawlability, indexation, canonicals, redirects, structured data, Core Web Vitals, JavaScript rendering and measurement.
Key takeaways
Protect discoverability first
A beautiful launch is a failure if priority pages cannot be crawled, rendered or indexed.
Preserve existing equity
When URLs change, map relevant old URLs to their true replacements and update internal links rather than relying on chains.
Keep indexation deliberate
Sitemaps should contain intended canonical URLs, while drafts, admin pages, filters, tests and thin variants stay excluded.
Measure immediately
A launch needs analytics, Search Console and conversion tracking ready so regressions are identified quickly.
1. Crawlability, indexation and HTTP behaviour
Access checks
Test priority URLs
Confirm important pages are accessible without authentication and return the intended page. Check the rendered content, not just whether the SPA returns HTTP 200.
Test invalid URLs
A nonexistent path should produce a useful not-found experience and, ideally, a true 404 status. Some SPA hosts return 200 for unknown routes; document that limitation and monitor soft-404 handling.
Review robots.txt
Allow public pages and assets required for rendering. Keep admin and authentication areas out of normal crawling. Name the sitemap.
Review robots meta
Important pages should be index,follow. Drafts, filters, account pages, test routes and incomplete location pages should be noindex where appropriate.
Check rendering
Use Search Console URL Inspection after deployment to see what Google renders, particularly on JavaScript-heavy platforms.
2. Canonicals, redirects and URL consistency
Choose one preferred URL for each piece of indexable content. The canonical, internal links, XML sitemap and redirects should all reinforce that choice. Avoid using canonical tags as a substitute for fixing a messy architecture. If two pages satisfy the same intent, decide whether they should be merged, differentiated or deliberately excluded.
For a redesign or migration, create the redirect map before launch. Redirect an old URL only when the destination is a genuine replacement. Do not send every removed page to the homepage. Avoid redirect chains by updating internal links and legacy redirects to point directly to the final destination. Google’s site-move guidance recommends preserving consistent signals and maintaining redirects long enough for users and search engines to adapt.
Normalise trailing slashes, protocol and hostname consistently. A site that mixes HTTP/HTTPS, www/non-www or parameterised duplicates can dilute signals and make debugging harder. Canonical URLs should be absolute and use the permanent production domain.
3. Sitemap, navigation and internal linking
The XML sitemap is an indexation inventory, not a list of every route the application can render. Include canonical, indexable URLs that are intended to return useful public content. Exclude drafts, noindex pages, filters, login routes, internal search, staging pages and duplicate states. If a content hub has no useful published content, there is no need to promote it through the sitemap yet.
Internal links should make the commercial and topical hierarchy obvious. Service hubs link to services. Service pages link to related services, proof and useful guides. Editorial content links back to the relevant commercial pages. Breadcrumbs help users and crawlers understand hierarchy. Avoid orphan pages that exist only because they were inserted into the sitemap.
Check links after the production build, not only in the editor. Broken anchors, relative path mistakes and legacy URLs often appear during template or routing changes. Prioritise links on navigation, footer, high-traffic pages, service clusters and conversion journeys.
4. Metadata, headings and structured data
Every priority page should have a descriptive title, a useful meta description, one clear H1 and a heading hierarchy that reflects the information structure. Titles and descriptions should be unique to the page purpose rather than generated by swapping a city or service term into a template.
Structured data should match visible facts. Use Organization and WebSite entities consistently, Service markup for genuine services, Article or BlogPosting for editorial content, Person for real authors, and BreadcrumbList where the breadcrumb is visible. Do not invent reviews, aggregate ratings, offices, prices or awards simply because schema supports those properties.
Validate JSON-LD and inspect the rendered page for duplicate entity IDs. When both global and page-level components emit schema, they should refer to the same organisation and website IDs instead of accidentally creating multiple versions of the business.
5. Performance, mobile and JavaScript
Core Web Vitals are useful user-experience thresholds rather than a standalone ranking strategy. Pay particular attention to the largest element above the fold, interaction responsiveness and layout shifts caused by late-loading images, fonts, banners or injected components. Optimise images to the rendered size, preserve dimensions, defer non-critical code and split admin or heavy application features away from marketing routes.
On JavaScript applications, reduce the amount of work required before meaningful content appears. Google can render JavaScript, but other bots may not, and users still pay the performance cost. If the hosting platform cannot provide server or static rendering, document that constraint and test live rendering regularly rather than claiming it has been solved.
Mobile QA needs to be intentional. Check navigation, forms, long headings, tables, accordions, consent controls, sticky elements, focus behaviour and tap targets on real viewport sizes. Do not assume desktop components automatically become usable when stacked.
6. Measurement and first-week monitoring
Before launch, confirm that analytics and advertising tags respect consent choices and that conversion events represent actions that matter. Growth-review completions, qualified contact submissions, phone clicks and meaningful CTA interactions are more useful than a large collection of low-value events.
After launch, submit or confirm the sitemap in Search Console and Bing Webmaster Tools, inspect the most important pages, monitor indexing and crawl errors, review structured-data reports, and compare traffic and conversions with the pre-launch baseline. If the site moved domains or changed large numbers of URLs, watch redirect coverage and legacy landing pages closely.
The first month should have a defined monitoring cadence. Check blockers immediately, indexation weekly, Core Web Vitals and technical health after material releases, and commercial performance often enough to spot a broken form or tracking gap before a month of data is lost.
Technical launch questions
Should noindex pages be in the XML sitemap?
Generally no. The sitemap should communicate the canonical URLs you want search engines to discover and index, not routes you are simultaneously asking them to exclude.
Should every old URL redirect to the homepage after a redesign?
No. Redirect only when there is a genuinely relevant replacement. A blanket homepage redirect can create poor user journeys and soft-404 behaviour.
Can JavaScript websites rank?
Yes. Google can render JavaScript, but rendering and performance still need to be tested, and not every crawler executes JavaScript in the same way.
When should Search Console be checked after launch?
Immediately verify critical URLs and sitemap access, then monitor index coverage, performance and technical reports frequently during the first few weeks.
Sources & references
Related services
Launching or rebuilding a website?
Use the Growth Review to identify the technical and commercial risks that deserve attention before more traffic is sent to the site.
Get Your Free Growth ReviewApply the thinking