Free tool
What is your site actually costing you?
Paste the address. About twenty seconds later you get the same numbers Google uses, in plain language, with the three fixes worth doing first.
- Fetching page
- Running Lighthouse
- Scoring
- Writing report
This tool needs JavaScript, because the measurement runs in your browser rather than on this server. That is deliberate: a run takes up to thirty seconds and holding a server worker open for that long is how a shared host falls over. Send the address with the quote form and you will get the numbers back by email instead.
That address is too long
Paste the page address on its own, without the tracking parameters on the end.
That does not look like an address
Try it as a plain domain, like yourbusiness.com.
Addresses with a login cannot be measured
Paste the public address, without the username and password before the @.
That port cannot be measured
The tool only reaches sites on the standard web ports.
That is an IP address
Use the domain name instead, so the measurement matches what a visitor actually gets.
That address is not reachable from the internet
It has to be a live public site. A staging URL behind a login will not work either.
That domain did not resolve
Check the spelling. If it is right, the domain may not be pointing anywhere yet.
The site did not answer
It may be down, or it may be blocking automated visitors. Worth trying again in a minute.
The page rendered nothing measurable
Sites that block Google or sit behind a login cannot be scored. That is the tool being honest, not your site being broken.
The measurement service is at its limit
That is our daily quota, not a problem with your site. Try again later today.
That took longer than 45 seconds
Very heavy sites time out here. That is itself a finding, and it is the kind of thing worth fixing.
The measurement came back broken
This one is on Google's end rather than yours. Worth another go in a minute.
Measured
--
Time until the main thing appears. Good is under 2.5 seconds.
How much the page moves while loading. Good is under 0.1.
How fast it answers a tap. Needs real visitor data to report.
How much a visitor downloads. Under 1MB is a reasonable target.
Want the full breakdown and a fix plan?
The score above is yours to keep, and it is the same number anyone else would measure. What it cannot tell you is where the remaining time is going, or whether getting it back is an afternoon of work or a rebuild. That part takes a person looking at it.
Method
Five things, and nothing invented.
The measurement is Google PageSpeed Insights, running Lighthouse from Google’s own infrastructure against a simulated mid-range phone on a slow connection. Every number below comes straight out of it. Nothing is estimated, weighted, or made worse to sell you something.
- Largest contentful paint
- How long somebody waits before the main thing on the page appears. Google treats anything over 2.5 seconds as a problem, and visitors treated it as a problem long before Google did.
- Cumulative layout shift
- How much the page moves around while it loads. This is the one that makes people tap the wrong thing.
- Interaction to next paint
- How quickly the page answers a tap. Only shows a number if enough real visitors have been measured on your site.
- Page weight
- How many kilobytes a visitor has to download. Almost always dominated by images that were never resized.
- The specific fixes
- What is actually costing you the time, taken from the same measurement, with the estimated saving on each.
FAQ
Is this really free?
Yes, and there is no email gate on the score. The number is the part that earns your trust, so putting it behind a form would defeat the point. The form is at the end, for people who want the fix plan as well.
What does it actually run?
Google PageSpeed Insights, which runs Lighthouse against your site from Google's own infrastructure. It is the same engine behind the Core Web Vitals report in Search Console. Nothing here is a proprietary score.
Do you keep my URL?
The result is cached for 24 hours so a second look is instant, then it is deleted. Beyond that only anonymous aggregates are kept: how many audits ran and the distribution of scores. No URL, no identity, nothing joined to anything.
Why is my score different from last time?
Lighthouse runs on shared infrastructure and simulates a mid-range phone on a slow connection. Two runs minutes apart can differ by several points. Treat a single score as a range, and treat a change of twenty as real.
My site scores badly. Is it broken?
Usually not broken, just heavy. The most common cause by a distance is images served at full resolution to phones. That is a fixable afternoon, not a rebuild.
Next
A number is not a plan.
Knowing the score is the easy half. The useful half is knowing which fix is worth doing, which one is a morning, and which one means the site needs rebuilding rather than patching. That is the conversation the quote form starts.