← all checklists

Search Engines

38 items · source

Copy as markdown

Filter items

Before you launch — the ones that actually bite

• Verify noindex is removed from production. A staging <meta name="robots" content="noindex"> or X-Robots-Tag that shipped is the most common launch mistake there is, and it can cost weeks before anyone notices.
• Verify robots.txt in production is not the staging one that disallows everything.
• Verify staging and preview environments are blocked from indexing and password protected — robots.txt is a request, not access control, and a disallowed URL can still be indexed if something links to it.
• Verify the canonical host is decided and enforced: one of www or apex, one of http or https, one trailing-slash convention.
• Verify every URL that changed at launch has a 301 to its replacement, and that the map was built from the old site's real URL list rather than from memory.
• Verify the new site does not 301 everything to the homepage; that reads as a soft 404 and loses the ranking outright.

Search Console and Webmaster Tools

• Verify a Google Search Console domain property is set up, not only a URL-prefix property — a domain property covers every subdomain and protocol at once.
• Verify Bing Webmaster Tools is set up; it can import the Search Console configuration in a couple of clicks, and it feeds more than Bing.
• Verify Yandex or Naver or Baidu are set up if you actually serve those markets, and not if you do not.
• Verify ownership is held by an account that survives a person leaving — a shared or role account, not someone's personal login.
• Verify at least two people have access.
• Verify email alerts are enabled and go somewhere a human reads.
• Verify the verification method will not break: a DNS TXT record survives a redeploy, an uploaded HTML file may not.
• Verify Search Console data is exported or archived; it only keeps 16 months.

Sitemaps

• Verify a sitemap exists, is generated from the live routes rather than hand-maintained, and updates when content does.
• Verify the sitemap is submitted in Search Console and Bing, and referenced from robots.txt.
• Verify the sitemap contains only canonical, indexable, 200-status URLs — no redirects, no noindex, no parameters.
• Verify lastmod reflects a real change, not the build timestamp on every entry; a sitemap that claims everything changed today is ignored.
• Verify a sitemap index is used if you exceed 50,000 URLs or 50MB uncompressed.
• Verify image and video sitemaps exist if those are a meaningful part of what you offer.
• Verify the sitemap URL count roughly matches the number of pages you expect to be indexed; a large gap either way is a signal.

Crawlability

• Verify robots.txt does not block CSS, JavaScript or images — a crawler that cannot render the page cannot judge it.
• Verify robots.txt and meta robots do not contradict each other; a page blocked in robots.txt can never be seen as noindex, so it stays indexed without a snippet.
• Verify X-Robots-Tag headers are checked as well as meta tags — a header set at the CDN is easy to forget.
• Verify the URL Inspection tool shows the rendered HTML you expect on your key templates, including content that JavaScript adds.
• Verify content that only appears after interaction — tabs, accordions, infinite scroll — is present in the HTML or reachable by a crawlable link.
• Verify pagination is crawlable by real links, not only by a button that calls an API.
• Verify faceted navigation cannot generate an unbounded crawl space of parameter combinations.
• Verify crawl stats show the crawler reaching the pages you care about, and not spending its budget on parameters and redirects.

Fast indexing

• Verify IndexNow is wired up if you publish frequently; it pushes changes to Bing, Yandex and others immediately.
• Verify the Google Indexing API is not being misused — it is only for job postings and livestream markup, and other uses are ignored.
• Verify new content is discoverable by an internal link from an already-indexed page, which is still the most reliable path.

Watching it

• Read the index coverage report rather than assuming; the difference between submitted and indexed is the whole story.
• Verify pages excluded as duplicate, crawled-not-indexed or discovered-not-indexed have been looked at at least once.
• Verify no manual action or security issue is open.
• Verify the Core Web Vitals report is green, or that its failures match what performance/ already told you.
• Verify 404s reported by the crawler are either fixed, redirected, or deliberately left to die.
• Verify someone checks this on a cadence rather than only when traffic drops.