People looking for the best free website grader usually want a single score they can trust. That score does not exist. What exists is a set of automated checks, a way of weighting them, and a test run under particular conditions. Change any of the three and the number changes.
This post explains what graders in general can and cannot tell you, using Lighthouse as the worked example. It is free and open source, and unlike most graders it documents what each check does. If you want the narrower question of whether a Lighthouse SEO score affects rankings, our post on the Lighthouse SEO score covers that in depth.
What a website grader actually does
Chrome for Developers’ introduction to Lighthouse describes it as “an open-source, automated tool to help you improve the quality of web pages. You can run it on any web page, public or requiring authentication. It has audits for performance, accessibility, SEO, and more.”
The workflow is the same shape for every grader: “Give Lighthouse a URL to audit, it runs a series of audits against the page, and then it generates a report on how well the page performed.” Then comes the line that matters most: “Use the failed audits as indicators for how to improve the page. Each audit has a reference that explains why the audit is important, as well as how to fix it.”
That is the honest description of any website grader. It is a list of automated tests with a summary score on top. The tests are the useful part. The score is a convenience for people who do not have time to read the tests.
Presence checks, not quality checks
The clearest illustration is Chrome’s documentation for the Lighthouse audit “Document does not have a meta description”. The audit fails if “Your page doesn’t have a <meta name=description> element” or the content attribute is empty. And then: “Lighthouse doesn’t evaluate the quality of your description.”
So a page with a meta description that reads “page” passes, and a page with a carefully written one passes too. The grader confirms the element exists. Whether it earns a click is a separate job, covered in our guide to writing meta descriptions. The same documentation notes that “Each SEO audit is weighted equally in the Lighthouse SEO Score, except for the manual Structured data is valid audit.” A missing tag and a deeper technical problem can therefore move the score by the same amount.
| A grader checks well | A grader can’t judge |
|---|---|
| Whether a meta description exists and is not empty | Whether the description is any good |
| Failed audits, each with a reference on how to fix it | Which fixes matter most for your business |
| Performance in a controlled lab run | What your real users on real devices experience |
| Pages behind a login, when run from DevTools | Content quality, links or whether a page matches search intent |
| Regressions between releases, via Lighthouse CI | How a page ranks, which the score does not measure |
Why scores differ between runs and tools
Run the same page twice and the performance score can change. Chrome’s documentation on performance scoring says “A lot of the variability in your overall Performance score and metric values is not due to Lighthouse.” The causes it lists include “A/B tests or changes in ads being served”, “Internet traffic routing changes”, “Testing on different devices”, “Browser extensions” and “Antivirus software”.
That is why a grader run on your laptop with a dozen extensions installed can disagree with the same grader run on a server. It is also why two different graders rarely agree: they run different checks under different conditions and weight them in their own way. Comparing scores across tools tells you very little.
The colour bands are the more stable signal. The documentation gives “0 to 49 (red): Poor”, “50 to 89 (orange): Needs Improvement” and “90 to 100 (green): Good”, and adds that “A ‘perfect’ score of 100 is extremely challenging to achieve and not expected.” Chasing the last few points is usually poor value. Our post on the PageSpeed Insights score covers that tool’s version of the number.
Lab data is not your users
Most graders produce lab data. web.dev defines it as data “determined by loading a web page in a controlled environment with a predefined set of network and device conditions”, and notes that “Chrome tools that report lab data are generally running Lighthouse.” Field data is different: it “reflects the actual devices, network conditions, and geographic locations of your users.”
A grader can give a page a green performance score while real visitors on slow phones have a poor time, and the reverse happens too. If Search Console says your Core Web Vitals assessment failed, a passing lab score does not overrule it; see Core Web Vitals assessment failed and lab vs field data for how to reconcile the two.
What makes the best free website grader
Since every grader’s score is shaped by its own checks and conditions, the best free website grader is the one that shows its working. You want to see each check, why it matters and how to fix it. Lighthouse meets that bar, which is why it is the example here; this post does not rank other graders, because their scoring was not reviewed.
Lighthouse runs “in Chrome DevTools, from the command line, or as a Node module”, and through PageSpeed Insights, where you enter a URL and click Analyze. In DevTools, “After 30 to 60 seconds, Lighthouse gives you a report on the page.” The DevTools route “allows for testing local sites and authenticated pages, while the extension does not”, which matters for staging sites and members’ areas. Teams shipping often can use Lighthouse CI to “prevent regressions on your sites.” For in-browser options beyond Lighthouse, see our round-up of SEO Chrome extensions.
How to use a grader without being misled
- Run it on pages that matter. Your templates and money pages, not just the homepage.
- Read the failed checks, not the score. Fix the ones that affect real pages and real users first.
- Run it more than once. Use a clean browser profile, since extensions skew results.
- Confirm with field data and Search Console. The grader shows a lab run; your users and Google’s reports show the rest.
Treat a free on-page SEO checker as one input to a wider review, not a verdict. Our SEO audit checklist covers what a grader skips, including content, links and intent.
The limitation is built in. A grader is automated, so it can only test what a script can detect: tags, attributes, timings in a lab. It cannot tell you whether your content answers the query, whether anyone links to you, or how your pages rank. The score is not a Google metric and a higher one does not by itself raise rankings. Read it as a list of things to check, which is exactly what Lighthouse’s own documentation says it is.
Frequently asked questions
What is the best free website grader?
One that documents every check and explains how to fix it. Lighthouse, free and open source, does this and runs in Chrome DevTools, from the command line, as a Node module, and through PageSpeed Insights.
Why does my website grader score change every time?
Chrome’s documentation says much of the variability is not due to Lighthouse, citing causes such as changing ads, traffic routing, different devices, browser extensions and antivirus software.
Does a website grader check content quality?
No. Lighthouse’s meta description audit checks only that the element exists and is not empty; its documentation says it doesn’t evaluate the quality of your description.
Is a score of 100 necessary?
No. Chrome’s documentation treats 90 to 100 as good and says a perfect 100 is extremely challenging to achieve and not expected.
The takeaway Pick a grader that shows its working, ignore small score swings, fix the failed checks that affect real pages, and confirm with field data and Search Console.
