Free website audit tools: what they actually check

A free website audit checks what a machine can verify. That is not the same as what makes people buy, and the gap has been measured.

You ran three of them last night. PageSpeed gave you 89. The SEO grader gave you a B minus. The accessibility checker found four things and you fixed two.

Then you went to bed with the same feeling you woke up with. Something is wrong, and none of it told you what.

You were right.

What does a free website audit actually check?

Nearly all of them check three things: load speed against Core Web Vitals, machine-readable SEO markup such as titles and meta descriptions, and automated accessibility rules like colour contrast and missing alt text. Every one of those is a rule check. None of them reads your offer, your price, or your headline.

Google says so about its own tool. The post announcing the SEO category in Lighthouse states the list of checks is not exhaustive and makes no ranking guarantees. It was built to validate SEO basics.

Basics is the right word. It is also the ceiling.

The fourth free option is pasting your URL into an AI chat, and that is not a website audit either. It never renders the page.

Is this happening on your site?

Two minutes. Run PageSpeed Insights on your highest-traffic page and write the number down. Then open that same page on your own phone, thumb only, and answer four questions out loud: what is this, who is it for, why is it better than the alternative, what do I do next.

If the score came back green and you hesitated on any of the four, you have found the gap. The tools passed a page that does not explain itself.

Most people hesitate on "why is it better". That one is almost never on the page.

Stuck on more than one? The traffic but no sales diagnostic walks the other eight checks in order.

Where the money actually goes

A slow page loses some people. An unclear page loses the ones who stayed, and that second loss is larger. No free tool prices it, because the price needs three numbers the tool cannot see: your traffic, your average order, and what your conversion rate should be.

Here is our formula, published and checkable by hand: traffic × CVR gap × AOV × mobile weight. Which of those four we measure and which one we assume is set out in are website audit revenue estimates real.

Picture this. 12,700 sessions a month at 1.1%, where 1.7% is ordinary for that traffic mix. The gap is 0.6 points. At a €71 average order, 12,700 × 0.006 × 71 comes to €5,410 a month. Weight by the 0.68 of sessions on a phone and €3,679 sits on one mobile page.

PageSpeed will not show you that number. It was never trying to.

Most sites do not have one leak either. Across our audits the count runs nearer forty, which is why we keep a conversion leak index and not a top ten list.

Run this check on your own site, free.

What the data says

The UK Government Digital Service planted 142 accessibility barriers into one page and ran it through thirteen automated tools. The best of them found 40% of the barriers. The worst found 13%. Every free tool in that test landed under a third.

Here is how the tools scored.

That test ran in 2018 and the tools have improved. Worth saying plainly: the boundary moved less than the marketing did. The people running the biggest scanner on the web agree.

WebAIM scans the top million home pages every year. Its 2026 report found 95.9% failing an automated WCAG check, at 56.1 errors on the average page. Then it puts this in the methodology, about its own engine:

Absence of detected errors does not indicate that a page is accessible or conformant.

Read that again with your green score in mind.

Why does the free audit tool want your email?

Because for a lot of them, the audit is not the product. You are. Agencies buy embeddable audit widgets to collect a URL and an email, then route the report to a salesperson. SEOptimer says on its own site that over 2,000 agencies use its lead generation widget.

A tool that runs in your browser and shows you the result, like Lighthouse or WAVE, is a tool. One that emails a PDF after a form is a first sales meeting. Not a scandal, but worth knowing which you are using.

See what a full crawl finds on your site.

What to change, in order

Speed comes first only when the field data is genuinely poor. Otherwise start with the four questions, because a page nobody understands sells badly at any load time. This order has held up across the 134 sites we have audited.

1. Run the free tools once, then stop. Fix anything red in Core Web Vitals field data, plus anything on WebAIM's six most common errors. A morning of work, then finished.
2. Answer the four questions on the page. Read the hero aloud with your logo covered. If it could belong to a competitor, rewrite it.
3. Split your conversion rate by device. Mobile at half of desktop gives the leak a location. Recordings help here, though heatmaps have a blind spot of their own.
4. Price the gap before you fix anything. A €200 problem and a €4,000 problem look identical in a tool report and should not get the same afternoon.
5. Change one thing, then wait three weeks. Most sites under 20,000 sessions a month never reach significance on a split test. Compare matched periods instead.

When this does not apply

If your site takes nine seconds to load on a phone, none of the above matters yet. Fix that first. Free tools are excellent at speed and will find it faster than any person. Same goes for a checkout that breaks on Safari.

Now the awkward part, since we sell an audit.

Revslip is an automated crawl too, and sits under the same ceiling on anything needing judgment. We check about 200 signals across copy, offer, trust, UX and technical, and still cannot tell you whether your product is worth €71.

The 134 figure carries a bias as well. Sites that get audited are sites whose owners already suspected something. Not a random sample of the web. A waiting room.

Common questions

Is a free website audit worth running at all?

Yes, once. Lighthouse and WAVE find real problems in ten minutes and cost nothing. Treat the output as a floor, not a verdict. A clean report means no machine-detectable rule was broken, a far smaller claim than it looks.

Which free website audit tool is best?

For speed, PageSpeed Insights, because it shows real Chrome user data next to the lab test. For accessibility, WAVE scored highest of the free tools in the government audit at 30%. For SEO basics, Lighthouse. All three take fifteen minutes.

Why do two free audit tools give different scores?

They check different rules and weight them differently. No shared standard exists for what a website audit covers, so a score describes the tool as much as your site.

Does a paid audit fix this?

Only when the arithmetic works. We published what a CRO audit actually costs, including the traffic threshold below which none of the paid options pay back. Under that threshold, free tools plus the four questions is the honest answer.