Website SEO Audit: How to Run One Step by Step

A website SEO audit is a full review of the technical, on-page and off-page signals that decide whether your site can rank. This guide sets out what a credible website SEO audit checks, the order to check it in, and where the free tier stops being useful. A complete website seo audit works through five layers in sequence - crawlability, technical health, on-page content, internal and external links, and a competitor benchmark - because a content problem found before a crawl problem is usually a symptom rather than the cause. Most free audits cover the first two layers well, sample the third, and stop before the fourth.

What is a website SEO audit?

A website SEO audit is an analysis of your website's search engine optimisation carried out at no cost, usually by an automated crawler, a consultant offering a sample of paid work, or both. It evaluates keyword analysis, on-page SEO components, meta tags and backlinks to identify issues that are holding back your search rankings. The output is normally a report listing technical audit findings and on-page issues, often produced with ai-powered tools such as Semrush, Ahrefs or Screaming Frog.

The distinction that matters is not free versus paid. It is diagnostic versus prescriptive. A free audit is good at telling you what is broken. It rarely tells you which broken thing is costing you the most, because that requires knowing which pages have earning potential - and that is a judgement about your market, not a value a crawler can read off your HTML.

What a website SEO audit actually checks

Free tools differ in depth, but the credible ones in this free SEO audit guide's checklist all cover the same five groups. If a report skips a group entirely, that is the gap to fill manually.

1. Crawlability and indexing

• Response codes. 404s, 301 and 302 redirects, redirect chains, and 5xx server errors. Chains matter more than people expect: each hop leaks a little authority and adds latency.
• robots.txt and sitemap. Whether the sitemap exists, resolves, lists live canonical URLs only, and is referenced from robots.txt.
• Index coverage. Which URLs Google has actually indexed, read from Google Search Console rather than inferred from a crawl.

2. Technical health

• Core Web Vitals - Largest Contentful Paint, Interaction to Next Paint and Cumulative Layout Shift, measured on mobile first because that is the crawl Google uses.
• Mobile rendering. Whether the mobile page contains the same content and links as the desktop page, not merely whether it fits the screen.
• HTTPS and certificates. A valid certificate, no mixed content, and a single canonical protocol and host.

3. On-page content

• Title and meta description length. Roughly 50-60 characters for titles and 150-160 for descriptions, and unique on every URL.
• Duplicate content. Identical titles, descriptions or body copy across URLs - the most common cause of a page that ranks below a weaker competitor.
• Heading structure. One H1 per page, and H2s that describe sections rather than restate the H1.

4. Links

• Internal links. Orphan pages, dead internal links, and whether your most valuable pages are reachable in a few clicks from the homepage.
• Backlink profile. Referring domains, anchor text distribution, and any obviously toxic patterns.

5. Competitor benchmark

• The sites currently outranking you on your target terms, and what they have that you do not - depth, original data, tooling, or simply more relevant page types.

How to run a website SEO audit in seven steps

Order matters. Running these steps out of sequence is how teams spend a quarter rewriting copy on pages Google was never able to crawl. This is the working sequence behind the free SEO audit guide above.

1. Crawl the site first. Use a crawler - Screaming Frog, Semrush Site Audit or Ahrefs Site Audit - and export the full URL list with status codes before you look at anything else. Every later step depends on knowing which URLs exist.
2. Reconcile the crawl against Search Console. Compare crawlable URLs to indexed URLs. The difference is your real indexing problem, and it is invisible to a crawl alone.
3. Fix technical blockers. Redirect chains, 404s in the sitemap, noindex tags on pages you want ranked, and Core Web Vitals failures on templates rather than individual pages.
4. Audit on-page content by template. Duplicate titles and descriptions cluster by template, so fixing the template fixes hundreds of URLs at once.
5. Review internal and external links. Repair dead internal links, then link your commercially important pages from the pages that already have authority.
6. Run a competitor SEO analysis. Take your five most valuable target terms and read the pages currently ranking for each. Record the page type that wins, not just the domain.
7. Re-measure, then prioritise by expected gain. Rank the fixes by the traffic each is plausibly holding back rather than by how easy they are, and record the before state so the after state means something.

For a deeper walk-through of the individual steps, see our guide on how to do an SEO audit.

Competitor SEO analysis: benchmarking against the pages beating you

Competitor SEO analysis is the part of an audit that free tools handle worst, because it is comparative rather than diagnostic. A crawler can tell you your page loads in 3.1 seconds. It cannot tell you the page outranking you loads in 3.4 seconds and wins anyway, because it answers the question better.

A useful competitor SEO analysis records four things per target term: the domain ranking first, the page type that ranks (guide, tool, category, product), the specific asset it carries that you lack, and whether an AI Overview appears and who it cites. That last one increasingly decides whether a click happens at all. Our opportunity telemetry module does this scoring continuously, but a manual pass across five terms is enough to see the pattern.

What a website SEO audit cannot tell you

Being honest about the ceiling is more useful than overselling the floor. A free audit - automated or human - reliably surfaces broken things. Three things it consistently misses:

• Which fix is worth doing. A list of 340 issues sorted by severity is not a plan. Severity is a property of the issue; value is a property of the page. Only the second one decides where to start.
• Intent mismatch. A page can pass every technical check and still never rank, because the results Google shows for its target term are a different page type. No crawler flags this; it requires reading the live results.
• Cannibalisation between your own pages. Two of your URLs competing for one term looks healthy per-page and is a real ranking loss in aggregate.

Those three are judgement work, which is why the free tier stops there.

Common mistakes when running a website SEO audit

Every mistake below survives a clean crawl report, which is why a free SEO audit guide is worth more than the tool output it is built on.

• Treating the issue count as the score. Crawlers report thousands of "issues" that include every image without alt text on a paginated archive. Volume of findings is not severity.
• Auditing pages instead of templates. On a templated site most issues are one bug repeated. Fix it once at the template.
• Checking desktop only. Google crawls mobile first, and mobile is where layout shift and render-blocking scripts usually bite.
• Acting on the crawl without Search Console. A crawler tells you what is reachable; only Search Console tells you what is indexed and what earns impressions.
• Auditing once. An audit is a measurement, not a project. Without a repeat measurement you cannot tell whether a fix worked or the market moved.

Free vs paid SEO audit: what the paid version actually buys

The free vs paid SEO audit question is usually framed as depth, and that is only half right. Free crawlers and paid crawlers read the same HTML and the same response headers, so the factual findings are broadly the same. What a paid audit buys is the ranking of those findings against your commercial goals, and the second measurement that tells you whether a fix worked.

• What free covers well. Crawlability, broken links, missing or duplicate titles and meta descriptions, redirect chains, image weight, Core Web Vitals field data. These are objective and machine-checkable.
• What free covers badly. Intent matching — whether the page you have is the page type the results reward for that query. A crawler holds no opinion on that.
• What free does not cover at all. Prioritisation against revenue, and the recorded before state. A list of 340 issues in no particular order is a list, not a plan.

The practical test: if the output is a score and a list, it is a free audit. If the output is a ranked sequence with a recorded starting position you can measure against later, it is a paid one — and you can build that yourself out of the free layers, provided you are willing to do the ranking by hand.

A website SEO audit checklist you can run in an afternoon

Everything in this free SEO audit guide condenses into a free SEO audit checklist. Work it top to bottom; every line is answerable with a free tool or a browser.

1. Does a site: search for your domain return roughly the number of pages you expect?
2. Does the Search Console Pages report show anything valuable sitting under Not indexed?
3. Does every important page return 200, with one canonical pointing at itself?
4. Is there exactly one H1 per page, and does it match what the title tag promises?
5. Does every title tag contain the term that page is actually built around?
6. Do any two of your pages target the same term? If so, decide which one should own it before you touch anything else.
7. Does PageSpeed Insights report Core Web Vitals passing on mobile, not only on desktop?
8. Are images sized, compressed, and lazy-loaded below the fold?
9. Does every page carry at least one internal link in and one out, to a genuinely related page?
10. Does your structured data validate, and does it describe copy a visitor can actually see on the page?
11. For your top five terms, what page type ranks — guide, tool, category, product — and is yours that type?
12. Record today's position and impressions for those five terms, so the next audit has something to measure against.

Line twelve is the one most checklists leave out, and the one that turns a list of issues into a measurement you can repeat.

Website SEO audit FAQ

How to get a free SEO audit?

Three routes, in ascending order of usefulness. Run a free automated tool such as Semrush Site Audit, Ahrefs Site Audit or PageSpeed Insights - instant, shallow, no intent analysis. Combine a free crawl with your own Google Search Console data, which is what this free SEO audit guide describes and costs nothing but time. Or request a reviewed audit from an agency, where a human reads the live results for your target terms; ours is the free SEO site audit form.

How long does an SEO audit take?

An automated crawl of a small site finishes in minutes. A reviewed audit that includes intent and competitor work takes one to three working days for a site under a few thousand URLs, and longer where the crawl itself is slow or the site has many templates. The crawl is never the bottleneck - reconciling it against Search Console and reading the live results is.

How often should you do an SEO audit?

A full audit once a year is the floor for a small stable site. Quarterly suits larger or fast-moving sites. Beyond the calendar, run a targeted audit after any of these: a migration or replatform, a template change, a redesign, a ranking or traffic drop, or a confirmed Google update that hit your sector.

What tools do you need for an SEO audit?

A crawler, a performance tool, and your own search data. Screaming Frog, Semrush or Ahrefs will crawl; PageSpeed Insights covers Core Web Vitals; Google Search Console supplies indexing, impressions and position data no third-party tool can see. The free tiers of those cover a small site adequately. What no tool supplies is the judgement about which finding matters.

Can I do an SEO audit myself?

Yes, for the technical and on-page layers. A crawler plus Search Console will find most broken things without any specialist knowledge, and the sequence in this free SEO audit guide is deliberately followable without one. The layers that reward experience are intent matching and prioritisation - deciding which of 340 findings is actually costing you traffic.

How accurate are free SEO audit tools?

Accurate on facts, unreliable on judgement. A free crawler reads the same HTML a paid one does, so its factual findings — status codes, missing tags, redirect chains, thin titles — are as accurate as any tool's. Accuracy drops on severity: most free tools grade issues against a fixed rubric rather than against what the issue costs your particular site, so a missing alt attribute and a noindexed money page can arrive with similar weight. Treat the findings as true and the priority order as a first draft.

Do I need an SEO audit?

If you have never had one, yes — a first audit almost always turns up something structural, and structural problems are cheap to fix once and expensive to leave in place. After that it is conditional rather than annual: run one when something changed (a migration, a redesign, a platform move, a template rewrite), or when traffic moved and you cannot explain why.

What happens after an SEO audit?

Three things, in order, or the audit was wasted. Rank the findings by what they plausibly cost you rather than by the tool's severity colour. Fix in that order, one change at a time, so the result is attributable to the change. Then re-measure the same terms against the same baseline — an audit with no second measurement cannot tell you whether any of it worked.

Worked end to end, this guide gives you a ranked list of fixes with a recorded before state - which is the only form of audit output you can later prove something from.

If you would rather not run it yourself, we will. The reviewed audit is a $200 piece of work and we run it at no cost, because pointing the engine at a real site is the fastest way to show what it does. You give us the domain and what you are trying to move; a person reads the output and ranks it, and the findings are yours whether or not you ever work with us. Request the audit here.