What does an automated UX audit actually measure?
Four proxies, totalling weight 4.5: the Lighthouse best-practices score (1.5), the Lighthouse accessibility score reused as a usability signal (1), internal link count with full credit from 5 to 60 links (1) and whether the page has exactly one H1 (1). No user is watched and no click is recorded, so this is a structural read of the page, not a usability study.
Internal link count is the closest thing here to a navigation measure. Fewer than 5 internal links suggests a dead end with nowhere to go next, and more than 60 suggests a page where nothing is prioritised. Credit falls off over a 55 link taper on either side, which is forgiving: 2 internal links still scores about 0.95, while 100 links scores about 0.27.
The accessibility score appears here as well as in its own category, at weight 1 rather than 2.5. That is intentional. Element names, contrast and focusable controls are usability facts as much as accessibility facts, and a page that fails them is hard to use for everybody.
The H1 check in this category is scored differently from the one in SEO. Here, exactly one H1 scores 1.0 and anything else, including zero H1s, scores 0.4. SEO distinguishes between having none and having several. UX does not, because from a reader's point of view an unclear page heading is an unclear page heading either way.
Both Lighthouse numbers are averaged across the mobile and desktop runs when both exist. UX carries a weight of 1.0 in the composite score, deliberately below CTA, SEO and Performance because it is assembled from proxies rather than from direct measurement.
What is measured, and how it is scored
4 checks make up the UX / Usability score. Weights are relative within the category and sum to 4.50.
| Check | Rule | Weight |
|---|---|---|
| Lighthouse best-practices score, averaged across the mobile and desktop runs | The 0 to 100 score divided by 100. Covers engineering hygiene such as console errors, deprecated APIs and insecure requests. | 1.5 |
| Lighthouse accessibility score, reused here as a usability signal | The same 0 to 100 score used in the Accessibility category, divided by 100, but weighted 1 here instead of 2.5. | 1 |
| Number of internal links on the page | Full credit from 5 to 60 internal links. Outside that range credit falls off over a 55 link taper, so 2 links scores about 0.95 and 100 links scores about 0.27. | 1 |
| Whether the page has exactly one H1 | Exactly one H1 scores 1.0. Any other count, including zero, scores 0.4. This differs from the SEO category, which scores zero H1s at 0.0 and several at 0.5. | 1 |
UX / Usability questions
Does this audit watch real users?+
No. All four UX checks read either the page's HTML or a Lighthouse category score. There is no session recording, click tracking, heatmap or usability testing behind the number.
How many internal links should a page have?+
Between 5 and 60 earns full credit. The taper outside that range is 55 links wide, so both a page with 2 internal links and a page with 100 still earn partial credit, at roughly 0.95 and 0.27 respectively.
Why does the accessibility score count twice?+
It is weighted 2.5 in the Accessibility category and 1 here in UX. Missing button names, poor contrast and unreachable controls are usability problems for every visitor, not only for people using assistive technology, so the rubric counts them in both places at different strengths.
Why is the H1 rule different here than in SEO?+
UX scores any count other than exactly one H1 at 0.4, while SEO gives 0.0 for no H1 and 0.5 for several. Search engines treat a missing H1 as worse than a duplicated one. A reader just experiences an unclear page heading either way.
See how your own page scores on UX / Usability and the other eight categories. One full audit every 7 days is free, no card needed.
Run a free audit