Quick answer
To run an SEO audit, work in the order search engines do: confirm the site is crawlable and indexed (robots.txt, noindex, sitemap coverage), verify each important page has one canonical URL and a descriptive title, assess whether the content genuinely resolves its target query, then check internal links and external authority. Fix failures in that order, because downstream work is wasted while retrieval is broken.
What are the key takeaways?
- —Audit in pipeline order — crawl, index, relevance, authority — because each stage gates the next.
- —Search Console's coverage report is the ground truth; third-party crawlers only estimate what Google actually did.
- —A missing title costs clicks; a stray noindex costs the page entirely. Learn to tell urgent from cosmetic.
- —Score findings by impact and effort, then fix the top of the list — audits fail as projects when everything is treated as equal.
- —Re-run the audit after fixes ship; an audit is a loop, not a report.
What should you check first in an SEO audit?
Always start with indexation, because nothing else matters if the page is not in the index. Open Search Console's coverage report and read the exclusions before you read the valid pages. The exclusions are where audits earn their keep: pages excluded by a noindex tag that should never have shipped, pages blocked by robots.txt, duplicates Google canonicalised somewhere unexpected, and soft 404s where a thin page was judged not worth storing.
Cross-check with a site: search and with the sitemap report. If the sitemap lists two hundred URLs and the index holds ninety, the gap is the audit. Do not write a word of new content until you can explain every exclusion in that gap, because the same template or configuration flaw that excluded those pages will quietly exclude anything new you publish.
Our SEO Readiness Audit automates this first pass for a single URL: it fetches the page as a crawler would, checks the robots policy, the canonical, the sitemap membership and the indexability signals, and reports what it actually found rather than what should be there.
How do you audit titles, descriptions and headings?
Export every title and meta description — Screaming Frog or a simple crawl does this in minutes — and sort by the failures that matter. Missing titles, duplicated titles, and titles over sixty characters that truncate mid-word are the three worth fixing first, because they either hand control of your snippet to Google or make two pages compete as twins.
Then check that every page has exactly one H1 that names its topic, and that the H2s beneath it read as a sensible table of contents. The heading audit is really a content-structure audit in disguise: if you cannot tell from the headings alone what the page covers, neither can a parser, and the page will underperform on passage-level retrieval even if its prose is excellent.
How do you audit content quality without reading everything?
You cannot read everything, so triage with data. Pull the pages that get impressions but poor click-through — the content is ranking and losing the click, which points at titles and intent mismatch. Pull the pages that get no impressions at all — these have a retrieval or relevance problem. Pull the pages with high exit rates from organic landing — the content may be thin or misaligned with the query that brought the visitor.
For each flagged page, ask one question: if I searched the target query and landed here, would I need to go back to Google for anything? Every 'yes' is a missing section. The pages that win competitive queries are almost never the longest; they are the ones that leave the searcher with no follow-up question unanswered.
How do you audit internal links and site structure?
Crawl the site and count internal links pointing to each page. The pattern to find is orphaning: important pages reachable only through the sitemap or a deep paginated archive. Pages earn a share of the site's authority through the links that point at them, and an orphaned page is one its own publisher appears not to value — Google reads the signal exactly that way.
Check the shape as well as the counts. Every important page should be within two or three clicks of the homepage, linked with descriptive anchor text rather than 'read more', and should link onward to its siblings. A hub page per topic cluster, linking down to the detail pages and collecting links back, is the simplest structure that consistently works.
How do you turn audit findings into a fix list?
Score every finding on two axes: how much traffic the affected template or page type represents, and how expensive the fix is. A noindex on the blog template is one line of code affecting every post — top of the list. A slightly long title on a page with twelve impressions a month is real but belongs at the bottom.
Assign each fix an owner and a verification step. 'Fixed' means the new HTML is live and the specific evidence from the audit no longer appears — not that a ticket was closed. Re-crawl after deployment, and use Search Console's validate button on coverage issues so Google re-examines the affected URLs promptly.
Then schedule the next audit. Sites drift: templates change, new sections ship, redirects rot. A quarterly light audit against a fixed checklist catches drift while it is still cheap to fix.
What do people ask most about this topic?
How often should you run an SEO audit?
Run a full audit quarterly and a targeted one after any significant change: a redesign, a migration, a template update or a CMS upgrade. The failures that cost the most traffic are almost always introduced by deployments, not by gradual decay.
What tools do you need for an SEO audit?
Google Search Console for indexation and query data, a crawler such as Screaming Frog for titles, canonicals and internal links, and PageSpeed Insights for performance. A single-URL checker like our SEO Readiness Audit covers the on-page essentials for individual pages without a full crawl.
What is the difference between a technical audit and a content audit?
A technical audit verifies the machinery: crawlability, indexation, canonicals, speed, mobile rendering. A content audit judges the material: whether each page resolves its query, deserves its traffic, and should be kept, improved, merged or removed. A complete audit does both, technical first, because content work on an unindexable template is wasted.
Why does an audit show issues that do not seem to hurt traffic?
Because audits measure conformance, not outcomes. Some violations are latent: a missing canonical may be harmless until a parameterised URL duplicates the page next quarter. Treat zero-traffic findings as insurance work — worth fixing when cheap, safe to defer when not.
How long should an SEO audit take?
A disciplined audit of a small site takes a day: an hour in Search Console, an hour crawling, and the rest reading the pages that matter. What takes weeks is the follow-through — the audit that finds forty issues and fixes none is worse than a narrow audit that fixes five.
Written and reviewed by Jim Vernon, Editor, AI Intelligence International. Last reviewed 2026-08-25. Published by AI Answer Engine and checked against our editorial standards.