The technical SEO checklist we run on every site

This is the list we work through on a client audit, in roughly the order that finds the expensive problems first. It assumes you have Search Console access and a crawler.

1. Indexing and crawl control

Start here, because nothing else matters if Google cannot reach the pages.

• Search Console, Pages report. Compare indexed pages against the number you expect. A large gap in either direction is the first thing to explain.
• robots.txt. Check nothing important is blocked. Check AI crawlers are allowed if you want AI citations.
• noindex tags. Crawl for them. A noindex left behind after a staging launch is the most expensive single mistake in this list.
• Canonical tags. Every page should have a self-referencing canonical unless it genuinely is a duplicate. Check for canonicals pointing at the wrong protocol, the wrong domain, or a staging URL.
• XML sitemap. It should contain only indexable, 200-status pages. No redirects, no 404s, no noindex pages.
• Parameter URLs. On any store, check how many URLs filters and sorting create. This is covered in detail in faceted navigation is eating your crawl budget.

2. Status codes and redirects

• 404s with inbound links. Every one of these is wasted link value. Redirect them to the closest equivalent page.
• Redirect chains. A to B to C should be flattened to A to C. Chains slow crawling and lose a little value at each hop.
• Redirect loops. Rare and fatal where they exist.
• Soft 404s. Pages returning 200 with no useful content. Google treats these as 404s and they clutter your index.
• HTTP to HTTPS. One redirect, direct, no chain through a www variant.
• www and non-www. Pick one, 301 the other, and make sure canonicals agree.

3. Site structure and internal links

• Click depth. Every important page should be reachable in three clicks from the home page. Anything deeper gets crawled less.
• Orphan pages. Pages with no inbound internal links. They exist in the sitemap and nowhere else, and they rarely rank.
• Internal link counts. Your most valuable commercial pages should have the most inbound internal links. Check that this is actually true rather than assumed.
• Anchor text. Descriptive and varied. A site where every internal link says "read more" is telling Google nothing.
• Breadcrumbs. On every page below the home page, with BreadcrumbList schema.

4. Duplicate and thin content

• Duplicate titles and meta descriptions. Crawl for them. Duplicates here usually mean duplicate pages behind them.
• Near-duplicate pages. Location pages spun from a template, product variants as separate pages, paginated pages repeating the same intro copy.
• Thin pages. Pages with almost no content that exist because a CMS created them. Tag archives are the usual culprit.

5. Core Web Vitals

Use field data from Search Console or PageSpeed Insights, not lab scores. Lab scores measure a simulation. Field data measures your actual visitors.

• LCP under 2.5 seconds. Usually the hero image or a web font. Preload the LCP resource and stop lazy loading it.
• INP under 200 milliseconds. Usually third-party JavaScript. Audit your tag manager. Most sites carry scripts nobody has used in two years.
• CLS under 0.1. Usually images without explicit width and height, or content injected above existing content after load.

Fix mobile first. Most sites pass on desktop and fail on mobile, and mobile is what is measured.

6. Rendering

• Check the rendered HTML, not the source. Use the URL Inspection tool in Search Console and read what Google actually sees.
• Client-side rendered content. If your main content only appears after JavaScript runs, confirm it is present in the rendered output.
• Links as real anchors. Navigation built from click handlers rather than <a href> elements is not crawlable.

7. Structured data

• Organization sitewide.
• Service on service pages, Product on products, Article on posts.
• FAQPage where you genuinely have questions and answers.
• BreadcrumbList everywhere below the home page.
• Validate it. Most theme-generated schema is missing a required property.

8. International, if it applies

• hreflang tags that are reciprocal. A points to B, B must point back to A.
• A self-referencing hreflang on every page in the set.
• x-default for the fallback.

9. Server logs

The most useful data source and the hardest to get.

Logs tell you what Googlebot actually requested, how often, and what it got back. That answers questions a crawl cannot: which pages are never visited, how much budget goes to parameter URLs, and whether crawl rate changed after your last release.

Ask your host. On a VPS it is a file. On managed hosting it may take a support ticket.

10. What to do with the findings

Sort by traffic at risk, not by severity label. A crawler flags hundreds of issues and most of them cost you nothing.

Our own audits produce three tiers:

1. Costing you traffic now. Fix this month.
2. Preventing future growth. Fix this quarter.
3. Worth knowing. Fix when convenient.

If your audit does not tell you which is which, it is a crawler export rather than an audit.

We run this whole list as technical SEO, and it is step one of the Rapid Rank Method on every engagement.