Technical SEO Audit Template

A crawler shows every URL your links reach and what each returns. Search Console shows which ones Google indexes and what each earns in clicks.

1. 1

   Crawl the whole site

   Use a crawler such as Screaming Frog or Sitebulb. Set it to crawl as a smartphone, the way Google does, with JavaScript rendering on if pages build content in the browser. Where the crawler supports it, connect Search Console to find orphan pages.

2. 2

   Export Search Console

   Download the Performance report by page, plus the Page indexing and Sitemaps reports. The Performance table stops at 1,000 rows, so use the API on larger sites.

3. 3

   List templates and releases

   Note which template builds each URL pattern and when recent releases shipped. Tickets name templates, and release dates explain sudden drops.

Google's minimum technical requirements¹ are short: Googlebot isn't blocked, the page returns an HTTP 200 status code, and it has indexable content. Anything that breaks one of them comes first.

Read the Page indexing report next. Google says not to expect every URL to be indexed², just to make sure your key pages are. Look for important pages in the wrong state.

While you're in robots.txt, check which AI crawlers it lets in. An AI search visibility report shows whether assistants can reach and cite you.

Next, check what Google sees once a page loads. Spot-check each template in the URL Inspection live test, which shows the page as Google renders it, and confirm the main content and links appear.

Navigation and pagination need real links with an href, since Google's crawlers don't click "Load more" buttons that fetch products by script. If deep pages are indexed far less often than shallow ones on the same template, fix the linking.

Then check Search Console's Core Web Vitals report, which uses real-user data. Google's targets are loading (LCP) within 2.5 seconds, responsiveness (INP) of 200 milliseconds or less and layout shift (CLS) of 0.1 or less. Each is measured at the 75th percentile of page loads⁴. Treat them as pass marks, not a score to max out.

Google wrote its crawl budget guide mainly for sites with a million or more pages that change weekly, or 10,000 or more that change daily⁵. It also lists sites with many URLs marked "Discovered - currently not indexed". For the rest, it says, a current sitemap and regular Page indexing checks are enough.

Your crawler ranks its findings, but don't adopt its order. Screaming Frog, which makes the SEO Spider crawler, says its issue priorities are not hard rules for what to fix, because they lack context⁶. The missing context is whether Google indexes the affected URLs and what they earn.

Sort every finding into one of three classes first. The class decides which number ranks it.

Presentational findings can't remove a page. Google says it primarily uses the page's own content to create snippets⁷, and may use the meta description when that describes the page better. So a missing description can earn a ticket, but never a click figure.

Pages Google doesn't index put no traffic at risk, unless they should be indexed. Those that dropped out after a recent release go first: count the clicks they earned before the drop, because that return is easiest to predict.

Confirm the rest are excluded on purpose, as filters and variants usually are, before discarding their findings. A big template with nothing indexed is either deliberate or your biggest problem. Log each discard with its reason, so nobody thinks it was missed. Thin or near-duplicate pages aren't code fixes; settle them in a content audit.

Developers change templates, not URLs, so write one ticket per template change, even when it clears several crawler warnings. Google notes that error spikes can come from a template change², so match each spike to a release.

Give every ticket the six parts in the example below, and leave out severity labels and long URL exports.

Check every fix twice. On release day, fetch the example URLs and run the URL Inspection live test to confirm the change shipped. Weeks later, check Search Console: at 14 and 28 days for pages that dropped out, 28 and 56 days for presentational fixes, and monthly for structural ones.

If a URL still shows the old issue, check its last crawl date, since Google notes you may have fixed it after the last crawl². Compare with similar pages you didn't change, so you know the fix caused the movement.

The template at the top of this page holds the register, template map, tickets and retest plan, with a worked example. If you serve customers in one area, pair it with a local SEO audit.