A practical, step-by-step walkthrough of every technical SEO check that matters - what it is, why it affects rankings, and exactly how to fix it. The same process we run on 10+ production sites.
A systematic review of the technical factors that decide whether search engines can crawl, index and rank your pages - and a prioritised list of what to fix.
Technical SEO is the foundation everything else sits on. You can have the best content in your niche, but if a page is blocked by robots.txt, carries a stray noindex, canonicalises to the wrong URL, or takes eight seconds to load, it won't rank. A technical audit finds those barriers. This guide walks through the audit in the order we actually run it - fixing root causes first so you don't chase symptoms.
Work top to bottom. Each step has a free tool to check it instantly.
Start with robots.txt. A single over-broad Disallow: can block more than you intend, and blocking a URL stops Google re-crawling it (so it can't even see a noindex you add later). Confirm your important paths aren't blocked and that your sitemap is declared. Then check for broken internal links and redirect chains that waste crawl budget. Broken Link Checker →
A page is indexable only if it has no noindex tag or X-Robots-Tag header, isn't blocked in robots.txt, has a self-referencing canonical, and returns HTTP 200. The classic outage is a stray noindex on a shared template that ships to production and deindexes a whole section. Search Console's Pages report is the source of truth for what's indexed and why the rest isn't. Indexability Checker →
Every indexable page should self-reference its canonical. The common bug is a canonical pointing to a redirect or 404 (Google ignores it and picks its own), or every paginated page canonicalising to page 1 (which deindexes the rest). Keep canonical, hreflang and internal links all pointing at the same URL. Canonical Checker →
Your XML sitemap should contain only canonical, indexable, 200-status URLs - no redirects, 404s or noindex pages. Declare it in robots.txt and submit it in Search Console. For large sites, use a sitemap index pointing to child sitemaps (50,000 URLs max each). Also flag indexable pages missing from the sitemap - they may be under-crawled. Sitemap Validator →
Collapse redirect chains (A→B→C should be A→C) and fix redirects that land on 404s. Use 301s for permanent moves so backlink equity transfers, and update internal links to point at the final URL directly. After HTTP→HTTPS or trailing-slash migrations, re-check that canonicals and sitemaps reference the new URLs.
Each page needs a unique, descriptive title (~50-60 chars), a meta description under ~160 chars, one H1, and a sane heading hierarchy. These are quick wins that affect click-through and clarity. Meta Tag Checker →
Validate any JSON-LD - invalid schema can throw errors in Search Console and block rich results (a frequent one is aggregateRating on an unsupported type). If you're multilingual, hreflang is the most error-prone area: it needs reciprocal return tags, a self-reference, valid language codes, and targets that return 200. Hreflang Checker →
Speed is a ranking factor and the first thing users feel. Optimise LCP (compress the hero image, preload it, cut TTFB), CLS (set image dimensions, reserve space, font-display: swap), and INP (break up long JS tasks, defer non-critical scripts). Test your key templates on mobile, not just the homepage. PageSpeed Test →
A technical audit isn't a one-time task. Run it, fix root causes, re-crawl to confirm (and to catch issues the first fix revealed), then re-audit after every deploy that touches templates, redirects or the sitemap. Tracking your health score over time turns SEO from guesswork into a measurable trend.
For a small-to-medium site, budget a few hours. The crawl itself finishes in minutes; reviewing the findings, deciding what is a real problem and what is intentional, and prioritising the fixes is the bulk of the work. Large sites take proportionally longer on the review, and re-audits after a round of fixes are much faster because you already know the shape of the report.
Audit after any deploy that touches templates, redirects, canonicals or the sitemap, and at least monthly regardless. Regressions are introduced by changes, so auditing immediately after a release catches them while the responsible commit is still recent and the fix is a one-line reversal rather than an investigation.
Google Search Console and PageSpeed Insights, both free, cover indexing status and performance measurement. Beyond those you need a crawler to see the whole site at once. Screaming Frog's free tier handles 500 URLs, which suits a small site; Kalenux SEO Audit crawls your whole site against 180+ checks plus PageSpeed for one-time pricing.
Technical SEO does not guarantee rankings. What it does is remove the barriers stopping pages from ranking: it ensures your content can be crawled, indexed, and loaded quickly enough that users do not leave first.
Content quality and relevance still decide where you place among the pages that can compete. Think of the audit as qualifying for the race rather than winning it, which is why it belongs before content investment, not after.
Free to start. Crawl your whole site against every check in this guide.
Start my free audit