What Is a Technical SEO Audit? (Checklist Inside)

Most of the sites I audit have the same problem: the owners know their rankings are not where they should be, they have been told the content is good, and they cannot work out what is holding them back. In the majority of cases, the answer is in the technical infrastructure, not in the content itself. A technical SEO audit finds what Google sees when it crawls your site, and it is usually different from what you see in your browser.

A proper technical audit covers every layer of how your site is built, served, and indexed. It is not a Semrush report. Those are useful as a starting point for identifying issues at volume, but they miss the most important problems and generate so much noise that most people do not know where to start. A real audit involves reading what Googlebot actually encounters and understanding why specific pages are not indexing, ranking, or converting the way they should.

What a technical SEO audit covers

Crawlability and indexation. This is the foundation. If Google cannot crawl a page, it cannot index it. If it cannot index it, it cannot rank it. I start every audit by mapping what Googlebot can and cannot reach, checking robots.txt for over-restrictive rules, identifying noindex tags that have been applied incorrectly, looking for crawl traps that consume budget without producing indexable pages, and reviewing the XML sitemap for accuracy. The number of sites where the sitemap contains 404 pages or URLs that redirect is consistently higher than it should be.

Site architecture and internal linking. How your pages link to each other determines which ones Google treats as important. A technically sound site architecture distributes PageRank efficiently across the pages that matter most. I look at whether key pages are reachable within three clicks from the homepage, whether orphaned pages exist (indexed pages with no internal links pointing to them), and whether internal anchor text is helping or hurting keyword targeting across the site.

URL structure and duplicate content. Duplicate content is one of the most underestimated technical problems on mid-sized sites. It occurs more often than most site owners realise, multiple URLs serving the same or near-identical content due to faceted navigation, parameter handling, www versus non-www variations, HTTP versus HTTPS inconsistencies, or trailing slash inconsistencies. Each variation splits ranking signals rather than consolidating them. The fix is usually canonicalisation, but only when applied correctly.

Core Web Vitals and page experience. LCP, CLS, and INP are now confirmed ranking signals. I assess both the field data in Google Search Console (real user experience) and lab data from PageSpeed Insights (what Google measures during crawls). The issues I most commonly find: unoptimised hero images that slow LCP on mobile, layout shift caused by fonts loading after initial paint, and third-party scripts that delay interaction response times beyond the INP threshold.

Structured data and schema markup. Schema does not directly improve rankings but it expands SERP real estate and eligibility for rich results, FAQ dropdowns, review stars, article carousels, and event listings. More importantly, correctly implemented schema helps Google understand the relationship between entities on your site, which matters increasingly as the algorithm becomes more entity-based. I check for implementation errors (the number of sites with schema validation errors they do not know about is striking), coverage gaps on key page types, and opportunities for additional markup types that competitors are using.

Redirect chains and broken links. Every redirect in a chain loses a small amount of link equity. A chain of three redirects, which is extremely common on sites that have been migrated or restructured, can lose 15-20% of the link value that should be reaching the destination page. I map all redirects, identify chains, and flag 404 pages that have inbound links which should be redirected. On sites with history, this is often one of the highest-ROI fixes available.

Rendering and JavaScript SEO. Sites that rely heavily on JavaScript to render content create crawling complexity that many developers do not account for. Googlebot can execute JavaScript, but it does so in a second pass that can delay indexation by days or weeks. Content rendered client-side that is important for SEO should either be server-side rendered or have a static HTML fallback. I check what Googlebot actually sees by comparing cached page source against the rendered DOM.

What the audit output looks like

A technical audit is only useful if you know what to do with it. The output I provide is a prioritised action plan, not a list of 200 issues in alphabetical order. Issues are ranked by their likely impact on rankings and traffic, with clear implementation instructions for each one. Critical issues (things that are actively suppressing indexation or splitting ranking signals) are separated from important fixes (things that will improve performance over time) and housekeeping items (things that are worth fixing but will not move the needle immediately).

For most sites, the critical issues list is short, three to eight items that, once fixed, account for the majority of the ranking opportunity. The rest is a roadmap for continuous improvement rather than urgent action.

Tools used in a technical audit

Screaming Frog is the primary crawl tool, it maps what a bot sees when it visits your site sequentially, collecting status codes, redirect chains, meta data, heading structure, and canonical tags at scale. Google Search Console provides the complement: real data on what Google has actually indexed, coverage errors it has found, and Core Web Vitals field data from real users.

PageSpeed Insights and Chrome DevTools provide lab performance data. Ahrefs shows backlink profile health and identifies broken inbound links worth redirecting. For JavaScript-heavy sites, I use Google’s URL Inspection tool in Search Console to compare the rendered versus source HTML as Googlebot sees it.

The tools are the starting point, not the analysis. What a tool flags as an issue and what is actually causing a ranking problem are not always the same thing. That distinction is what an experienced auditor brings that a generated report cannot.

Frequently asked questions