Technical quality
Is Your Trade Website Technically Healthy? Compare Its Lighthouse Scores With 30,000 UK Sites
More than 30,000 UK trades websites were measured across Performance, Accessibility, Best Practices and technical SEO. The useful comparison is not a perfect score. It is understanding where your site sits and what the score can actually tell you.
Website health is often reduced to one score. That is convenient, but misleading. A site can be fast and difficult to use. It can score well for technical SEO and still rank poorly. It can pass automated accessibility checks while creating problems for real people.
A 2026 study of UK trades websites gives us a more useful reference point because it measured four Lighthouse categories across more than 30,000 sites using the same automated process.
The four mobile benchmarks
Across the measured mobile websites, the median Performance score was 67 from 30,397 sites. The median Accessibility score was 89 from 30,557. Best Practices had a median of 81 from 30,538. The technical SEO category had a median of 92 from 30,545.
The distribution matters too. Only 15.7% of measured mobile sites reached Lighthouse's good band for Performance. 49.0% reached the good band for Accessibility. 38.0% reached it for Best Practices. 68.6% reached it for the SEO category.
Run your own four score check
Use Google PageSpeed Insights on your homepage with the mobile test. Record all four category scores rather than only Performance.
Performance: compare your score with the sample median of 67, then look at the actual loading metrics rather than celebrating or panicking over the headline number.
Accessibility: compare with the median of 89, but remember that automated checks only cover part of accessibility.
Best Practices: compare with the median of 81. This category checks a mixture of technical issues such as HTTPS, browser errors and deprecated features.
SEO: compare with the median of 92, but do not call it a ranking score. Lighthouse checks a limited set of technical and on-page basics. It does not know whether your page deserves to rank for 'roofer near me' or any other commercial search.
The SEO score is the easiest one to misunderstand
In the study, 68.6% of mobile sites reached Lighthouse's good SEO band. That does not mean nearly seven in ten of those businesses rank well. A website can pass technical basics and still have weak content, little authority, poor local relevance or no meaningful service structure.
This is why SEO is no longer the whole job, and why a technical score is even less of the whole job. It is a diagnostic checklist, not a commercial outcome.
Look for imbalance rather than perfection
The most useful pattern may be a gap between categories. A site with a strong SEO score but weak Performance may have the technical basics in place while giving mobile users a poor experience. A visually attractive site with weak Accessibility may be excluding customers. A site with strong Performance but weak business information may be fast but unconvincing.
That is why one number should not determine whether a website is good.
Use the scores to decide what deserves attention
If Performance is the obvious outlier, investigate images, scripts, fonts and page weight. If Accessibility is weak, inspect contrast, labels, headings and keyboard use. If Best Practices is weak, review the technical issues flagged by the audit. If the SEO category is weak, fix the basics but do not assume the job is finished once the score rises.
The dataset is a benchmark, not a national league table
The source study is careful about its limitations. The sample was seeded from business directory listings and cities rather than randomly drawn from every UK trades business. Lighthouse was run once per page under laboratory conditions. Failed measurements were retained rather than silently removed.
Those limitations mean the figures should be used as a reference point for similar trade websites, not as a claim about the entire UK market.
Technical health should support the customer journey
A technically healthy site should load reasonably, avoid obvious browser problems, be easier to use and give search engines clean basic information. But the final test remains human: can the customer understand the business, trust it and take the next step?
The scores help identify friction. They do not replace the judgement needed to decide whether the site is doing its actual job.
Sources and method
These sources support the factual claims and benchmarks used above. Where a study is vendor produced, its sample and stated limitations are treated as part of the finding rather than as a national estimate.
The State of UK Trades Websites 2026, Websites for Tradespeople