Free Website Audit Tool: What to Look For (and What to Ignore)

Type "free website audit tool" into Google and you get roughly three flavours of result: lab-data performance scanners (PageSpeed Insights, GTmetrix), SEO crawlers gated behind a signup wall (Ahrefs, Semrush, Ubersuggest), and lead-gen funnels that grade your site A-F and then immediately email you about hiring an agency.

None of these are bad. Most are useful. But they measure very different things, and a Lighthouse score of 92 tells you nothing about whether the site converts visitors into customers — which is usually the actual question behind "audit my website."

This post separates the audits worth running from the ones that mostly waste your time, with concrete CLI commands for the technical ones and a frank note about what they don't catch. We also link our own free audit at the end — but the goal here is to help you actually understand what each tool measures, not push you down a funnel.

The five categories of website audit

Before picking a tool, decide which question you're answering. The five real categories:

The last row is where every popular "audit tool" silently fails. Lighthouse will not tell you that your hero says "Transform your workflow with AI-powered insights" when it should say "Cut your AWS bill 30% in 14 days." Ahrefs will not tell you that your pricing page hides the price behind "Contact sales." That's the work the rest of the article walks through.

Performance: Lighthouse and PageSpeed Insights

PageSpeed Insights is Lighthouse with two extra things bolted on: it pulls real-world Chrome User Experience Report (CrUX) data alongside the lab run, and it's the version Google Search Console links to. If you're going to run one performance tool, run this one.

What it actually measures (per the Lighthouse 12 scoring weights, current as of late 2025):

• First Contentful Paint — 10%
• Largest Contentful Paint — 25% (a Core Web Vital)
• Total Blocking Time — 30%
• Cumulative Layout Shift — 25% (a Core Web Vital)
• Speed Index — 10%

Run it locally instead of through the web UI when you can — you get a saveable HTML report and avoid the throttling skew of Google's data centres:

npm i -g lighthouse
lighthouse https://yourdomain.com \
  --output=html \
  --output-path=./report.html \
  --form-factor=mobile \
  --throttling-method=simulate \
  --chrome-flags="--headless"

What it gets right: the Core Web Vitals scoring is the same metric Google uses as a small-but-real ranking signal. If your LCP is 6 seconds on a Moto G Power simulation, mobile users on real 4G are having a worse experience and Google can see it via CrUX.

What it misses: a perfect 100 Lighthouse score on a homepage that says "Welcome to our website" tells you exactly nothing about whether the site is doing its job. We've seen 98-Lighthouse marketing sites with zero leads per month, and 67-Lighthouse local-business sites doing six figures a year. Performance is a hygiene factor, not a conversion driver.

GTmetrix: useful, but read the small print

GTmetrix wraps Lighthouse in their own grade alongside a YSlow-derived score. The free tier tests from a single Vancouver location on a desktop browser at full speed by default — which means your score will look noticeably better than what mobile users on cellular actually see. That's fine if you know it; misleading if you don't.

The waterfall view is its strongest feature: every request laid out by start time, transfer size, and timing breakdown. It's the fastest way to spot a render-blocking 800KB analytics script you didn't realise was loading.

SEO: Search Console first, third-party crawlers second

The single most underused free SEO tool is Google Search Console. It's free, it's the source of truth for how Google sees your site, and it tells you which queries are bringing traffic vs which queries you're nearly ranking for. Anyone selling you a "free SEO audit" without first asking whether you have GSC set up is selling you noise.

Practical first checks once you've verified the property:

1. Coverage report → look for "Discovered — currently not indexed" and "Crawled — currently not indexed" pages. These are pages Google found but chose not to rank, usually a content quality signal.
2. Performance report → sort by Impressions, filter by Position 5–20. Those are queries you're nearly ranking for. A 200-word content tweak on the page often moves them into the top 5.
3. Sitemaps report → check submitted vs indexed counts. A 30% gap is normal; a 70% gap is a problem.

When third-party crawlers add value

Screaming Frog SEO Spider is free up to 500 URLs and gives you the kind of crawl-level inventory Search Console hides. Duplicate title tags, missing H1s, internal links to 404s, redirect chains, orphaned pages — all in one CSV.

Ahrefs Webmaster Tools (free for verified site owners) gives you a real backlink index — the part Search Console deliberately undercounts. Useful for spotting toxic links you didn't know about, or measuring whether a guest post actually delivered the link.

Semrush and Ubersuggest both offer free tiers but heavily rate-limit them and push you toward a paid plan. Use them for one-off competitive research, not as a recurring audit.

Accessibility: axe and WAVE catch most automatable issues

Automated accessibility scanners catch roughly 30-57% of WCAG violations depending on which study you read — Deque's own analysis sits at the higher end of that range. The remaining issues require manual review (keyboard navigation, screen reader flow, focus order, semantic structure of complex widgets). We covered the gap in detail in The 43% of WCAG issues automated scanners miss.

For the automatable share, two tools cover it:

• axe DevTools — Chrome extension, runs in the DevTools panel, ranks issues by severity. Same engine that powers Lighthouse's accessibility audit but with more detailed remediation guidance.
• WAVE — annotates your live page in the browser with icons showing exactly where issues sit. Better for non-developers because the issues are visible on the page rather than in a flat list.

For a CI-friendly version, axe-core runs in any Node script:

npm i -D @axe-core/cli
npx axe https://yourdomain.com --tags wcag2a,wcag2aa

The audit category nobody runs: conversion

Every tool above measures whether the site is technically sound. None of them measure whether the site sells. The conversion review is the one that actually moves revenue, and it's the one that requires a human (or an LLM with the right prompt and context) to actually look at the page.

The minimum viable conversion review answers seven questions:

1. What is the product or service, in eight words or fewer? (If you can't tell from the hero in 5 seconds, the headline is broken.)
2. What is the primary CTA, and is it singular? Two equally-weighted CTAs typically reduce clickthrough on both.
3. Is there visible social proof above the fold — logos, testimonials with names, or specific numeric outcomes?
4. Is the price visible without filling out a form, or is it gated behind "Contact sales"? (For sub-$10K products, gating is usually a mistake.)
5. Are objections addressed? An FAQ section sitting above the final CTA does measurable work.
6. Is there a trust band — a clear physical address, a real founder photo, the company registration, security badges if relevant?
7. What's the page weight in JavaScript? On mobile, every 100KB of JS over budget delays time-to-interactive enough to drop conversion measurably.

You can run this manually in 15 minutes. We covered the playbook in Why your website gets traffic but no customers and The 2026 SaaS landing page checklist.

Where our free audit fits in

We built a free mini-audit at /free-audit for one specific reason: every free tool listed above grades the wrong thing. Lighthouse grades milliseconds. Ahrefs grades backlinks. None of them grade the headline, the CTA, the trust signals, or the offer.

The mini audit takes a URL, fetches the page, and runs an LLM-grounded review against the seven-question conversion list above. It returns 3-5 specific, page-anchored findings — not a generic checklist, not "improve your meta descriptions," but actual line-level rewrites tied to copy on your page. It runs in 20-30 seconds, requires no email, and there's no upsell pressure on the result page.

You should buy the paid audit if…

The free mini audit returns 3-5 findings. The $49 AI Website Revenue Audit returns 15 sections — every page on the site, the full funnel, mobile-specific issues, copywriting rewrites for every weak section, the technical performance pass, and a prioritised remediation plan ranked by estimated revenue impact.

It's the right purchase if:

• Your site has more than three pages and you want every one reviewed, not just the homepage.
• You're spending money on ads or SEO and want to know whether the landing page is the bottleneck before spending more.
• You've already implemented the obvious fixes from the free tools above and want a deeper read.
• You want a written deliverable you can hand to a developer or copywriter to act on, rather than figuring out the implementation yourself.

It's the wrong purchase if you haven't yet run Lighthouse on your homepage, set up Search Console, or fixed the obvious accessibility issues axe flags. Do those first — they're free.

The honest comparison

If you're auditing a site for the first time and want to spend zero dollars, this is the order we'd suggest:

1. PageSpeed Insights — performance baseline + Core Web Vitals.
2. Search Console — verify the property, look at Coverage and Performance reports.
3. axe DevTools — accessibility scan of the homepage and the highest-traffic landing page.
4. WAVE — visual second opinion on accessibility for the same pages.
5. Screaming Frog (free 500-URL crawl) — full site crawl for technical SEO inventory.
6. Our free mini audit — the conversion read the others can't give you.

That stack costs nothing, takes about an hour, and gives you a more complete picture than any single tool. The paid audit ($49) is the right next step if the conversion findings warrant a deeper investment than 30 seconds of LLM time.

Related reading: Why your website gets traffic but no customers · Website trust signals that convert · The 43% of WCAG issues automated scanners miss