Comparison
PageSpeed Insights grades speed.
A website is judged on more than that.
PageSpeed Insights is an excellent tool — free, maintained by Google, and we use it as our performance reference. It says nothing about the certificate, the security headers, DNS, trackers set before the banner, or the legal notice. CheckWeb measures those pillars, in a real browser, then proposes the fix.
What each tool covers
| Pillar | CheckWeb | PageSpeed Insights | Typical online scanner |
|---|---|---|---|
| Performance | Real Chromium, mobile and desktop, CPU and network throttled. | Yes — that is its job. Throttling simulated arithmetically (default mode). | Sometimes, often through the PageSpeed API. |
| Security | Mozilla Observatory scale, grade A+ to F. | No. | Rarely the full scale. |
| Infrastructure | DNS, IPv6, certificate, CAA, SPF, DKIM, DMARC. | No. | Exceptional. |
| Accessibility | axe-core, Lighthouse’s engine, plus a subset of the RGAA criteria. | An axe pass, with no RGAA plan or conformity statement. | Often the only subject, or absent. |
| SEO | Tags, canonical, sitemap, Open Graph, checks specific to your sector. | A few technical signals. | Varies. |
| GDPR | Trackers before consent, legal notice, privacy policy — without ever clicking the banner. | No. | Sometimes the only pillar sold. |
| And then | A fix proposed, applied in the built-in editor, and the gap measured on the next run. | A list of complaints. | A list of complaints. |
“Typical online scanner” means the wave of online audit tools that appeared on the same searches (GDPR, accessibility, a single score). We do not compare against a named competitor: their scope changes, and a frozen page would become untrue.
Why the speed scores diverge
PageSpeed Insights, by default, simulates a low-end phone and a slow connection. We apply both to a real Chromium: throttled CPU, throttled network, cold cache. On a heavy site the gap reaches tens of points — and our score matches what a real visitor gets.
Against a Lighthouse run with the same applied throttling, the gap falls back to a few points. We publish those gaps, including the ones that do not flatter us, on the methodology page.
When to stay with PageSpeed
To track Core Web Vitals alone inside Google’s ecosystem, or for a measurement your client already knows by name. CheckWeb is not meant to replace it on that ground.
Use CheckWeb when you also want to know whether the certificate will renew, whether a tracker fires before consent, whether the headers would hold up at Mozilla — and above all: when you want to fix, not just to observe.
What the trial costs
One full audit is free, with no account and no card. The report is not cut down. A free account then gives 3 audits a day, with the same measurements as the paid plans. Prices are on /tarifs.
The simplest thing is to look at your own score.
A full audit, right now. Two minutes will tell you whether the gap with PageSpeed Insights changes anything for you.
Audit my site