aiwebpageseo / SEO Tools / Site Audit / Site Audit Guide

Site Audit: A Complete SEO Audit Walkthrough Guide

A site audit gives you a complete picture of your website's technical SEO health — every page, every issue, scored and prioritised so you know exactly what to fix first.

🔍 Run Site Audit All Audit Tools →

What is a site audit?

A site audit is a systematic technical review of every page on your website. It identifies issues that prevent Google from crawling and indexing your content effectively — missing title tags, broken links, slow pages, duplicate content and more. Without a regular audit, these issues accumulate silently and erode your rankings over time.

Key fact: Most websites have technical SEO issues they are completely unaware of. A site audit surfaces them all in one pass, giving you a clear action list prioritised by impact.

How to read your site audit results

Results are scored from 0 to 100. Pages scoring below 50 have critical issues that need immediate attention. Each issue is colour coded: red for errors that directly harm rankings, amber for warnings that should be fixed, and blue for recommendations that improve performance.

Fix in this order

  1. Pages returning 4xx or 5xx errors — fix or redirect immediately
  2. Missing title tags — every page must have a unique, descriptive title
  3. Missing H1 tags — one H1 per page, matching the page topic
  4. Duplicate title tags — each page needs a unique title
  5. Thin content — pages under 300 words risk being skipped in indexing
  6. Missing meta descriptions — Google writes its own if you do not provide one

Crawled, indexed, ranking — three gates, and an audit only sees the first

The most useful thing to understand about any audit is what it structurally cannot tell you. A page must pass three separate gates, and they fail for entirely different reasons:

GateQuestionWhere the truth lives
CrawledCan a bot fetch this page?Your audit — and your server logs
IndexedDid Google decide to keep it?Only Search Console
RankingDoes Google think it is the best answer?Search Console, and the SERP itself

An audit lives almost entirely in the first row. It will tell you a page is fast, valid, well-titled and perfectly crawlable — and be silent on the fact that Google looked at it, decided it was not worth keeping, and dropped it.

The pattern to watch for: a clean audit, and a Search Console report full of "Crawled — currently not indexed" or "Discovered — currently not indexed". That combination is diagnostic. It means the technical work is done and the problem is the content — Google fetched the pages, judged them insufficient, and declined. No technical fix addresses this, and every hour spent on one is wasted.

So: run the audit, and read it alongside the Page Indexing report. The audit tells you what is broken. Search Console tells you what Google actually did about it. They answer different questions, and only one of them is Google's opinion.

An audit is not an inventory of your site

A crawler starts at the homepage and follows links. It therefore finds only what is linked. That has a consequence people consistently miss: orphan pages are invisible to your audit and entirely present in Google's index.

Pages routinely become orphaned without anyone doing anything:

The only way to find them is to compare sources. Take the crawl, take your XML sitemap, take the list of pages Search Console has seen, and take your server logs. A URL appearing in any of the last three but not in the crawl is an orphan. Server logs in particular are the only place you can see what Googlebot actually fetched, as opposed to what you think exists.

The 1,200-issue report, and why it gets fixed backwards

A large audit produces an intimidating number, and the number is a trap. What happens next is predictable: somebody sorts by count, sees 400 missing alt attributes, spends two days adding them, and reports progress. Meanwhile the missing H1 on the single page that earns most of the revenue sits untouched, because it was one row.

Two corrections make the list actionable:

1. Issue counts are inflated by templates

One broken template is one bug, reported four hundred times. 400 pages missing a meta description is not four hundred problems — it is one problem in one template, and one fix. Before triaging anything, group the issues by the template that produced them. The list usually collapses from twelve hundred rows to about fifteen actual defects.

2. Severity is not impact

Every audit tool assigns severity in the abstract — a missing title is "critical" wherever it occurs. But impact is severity multiplied by the value of the page. A critical issue on a page nobody visits is not urgent. A warning on the page in your paid ads is.

So before touching the list, pull your top twenty pages by traffic, conversions and inbound links. Fix everything on those, however minor, before fixing anything anywhere else, however severe. It feels wrong and it is right.

On scores: the 0–100 number is a vendor heuristic, not a Google metric. It is a useful way to compare a page against itself over time and a poor way to compare it against anything else. Chasing 100 is not a strategy — a page can score 100 and rank for nothing at all.

What technical audits systematically cannot see

This is the section most audit guides omit, and it is the one that explains why so many sites are technically immaculate and commercially invisible.

Technical SEO is a floor, not a lever. It removes the reasons a good page might fail. It cannot make a page good. Once the errors are cleared, the remaining work is content — and continuing to run audits at that point is a way of feeling productive while avoiding the harder job.

The issues that are usually mis-diagnosed

Closing the loop, which almost nobody does

An audit is not finished when the fixes ship. It is finished when the fixes are confirmed — and the confirmation step is where most audit programmes quietly stop.

  1. Re-crawl after deploying. Confirm the issue is gone in the built output, not just in the source. A fix applied to a template and not deployed to the CDN cache is a fix that did not happen.
  2. Validate in Search Console. For indexing issues, use the Validate Fix flow. It tells Google to look again, and it tells you whether Google agreed.
  3. Wait for a recrawl before judging. Two to four weeks. Google must refetch and re-evaluate. Nothing meaningful can be concluded in five days, and reverting a good fix because it "did not work" after a week is a common and expensive mistake.
  4. Track the count by issue type, not the overall score. The score moves for reasons you did not cause. The number of pages missing an H1 is a fact about your work.
  5. Re-audit on change, not on a calendar. Monthly is a sensible default, but the audits that pay are the ones triggered by an event: a migration, a replatform, a CMS upgrade, a redesign, a plugin update. Those are when issues are introduced, and they are introduced in bulk.

Frequently asked questions

What does a site audit check?

A site audit checks every page it can reach from your homepage. For each page it records the HTTP status code, title tag, meta description, heading structure, page speed, canonical tags, robots directives, structured data, and internal and external links. The result is a prioritised list of issues ranked by their impact on your rankings.

How often should I run a site audit?

Monthly for small sites. Weekly for sites that publish frequently or have regular content updates. Always run a full audit after any site migration, CMS update, redesign or significant structural change.

What is the most important issue to fix first?

Fix pages returning 4xx and 5xx errors first — these waste crawl budget and harm user experience. Then fix missing title tags, which Google replaces with its own often-inferior version. Then address duplicate content and thin pages that may be diluting your site's overall quality score.

My audit is clean but my pages are not indexed. Why?

Because an audit measures whether a page can be crawled, not whether Google chose to keep it. Those are separate gates and they fail for different reasons. A clean audit alongside a Search Console report full of "Crawled — currently not indexed" is diagnostic: the technical work is done and the problem is the content. Google fetched the pages, judged them insufficient, and declined. No technical fix addresses that.

Will a site audit find every page on my site?

No. A crawler starts at the homepage and follows links, so it finds only what is linked — and orphan pages are invisible to it while remaining entirely present in Google's index. Products that dropped out of their category listings, campaign pages linked only from an advert that stopped running, and sections removed from the navigation all become orphans. Find them by comparing the crawl against your sitemap, Search Console and your server logs.

How should I prioritise a report with over a thousand issues?

First group the issues by the template that produced them: one broken template is one bug reported four hundred times, so twelve hundred rows usually collapse to about fifteen real defects. Then remember that severity is not impact — impact is severity multiplied by the value of the page. Pull your top twenty pages by traffic, conversions and inbound links, fix everything on those however minor, and only then address anything else however severe.

Are all 404 errors a problem?

No. A page you deliberately removed should return 404, and a 404 with nothing pointing at it is an accurate report rather than an error. What matters is whether something still points at it: fix the internal link, or redirect an externally-linked URL to a genuine equivalent. Do not redirect it to the homepage to make the number go down — Google treats a redirect to an irrelevant destination as a soft 404, so it passes no value and merely hides the row.

🔍 Run Site Audit Now

Run the Site Audit and get actionable results in minutes. Pay as you go — no subscription needed.

Run Site Audit →

Related tools

About aiwebpageseo

aiwebpageseo.com is a data-driven SEO and AEO (Answer Engine Optimisation) platform providing a free suite of technical website tools. Rather than relying on AI-theorised assumptions, the platform analyses live URL performance, delivering objective diagnostics, page speed metrics, CLS debugging, and site crawl data alongside actionable technical tutorials.