Core Web Vitals: field data vs lab scores (what UK sites should trust)
Why PageSpeed Insights shows two stories, how CrUX differs from Lighthouse, and which fixes move real user metrics on mobile.
PageSpeed Insights shows two stories. The lab test is a controlled run, useful for finding what is slow. The field data is what Chrome users actually experienced, and it is the one that matches the Core Web Vitals report in Search Console. Lead with the field trend when you are deciding if the site has a problem. Use the lab test to debug it. A quick pass is included in the website audit; the field report tells you whether the issue shows up for customers.
Common false positives
- Perfect Lighthouse on Wi-Fi does not disprove slow 4G field LCP.
- CLS spikes often come from cookie banners, late fonts or embeds — test with the same conditions your customers have, not only a desktop on office Wi-Fi.
- A score that jumps around between runs is a hint to look at third-party scripts, not to chase the last five points.
Fixes that usually move the mobile field data
Make the largest element on the first screen easy to find: compress the hero image, give it width and height, and do not hide the headline behind a font that arrives late. Stop banners, chat bubbles and ad slots from shoving the page down after load. Delay scripts that are not needed to read the page or send the form. Then recheck field data after a few weeks — it lags the lab test, because it is an average of real visits.
Related on this site
Questions people ask
Should I trust Lighthouse or the field data?
Use field data — the Chrome User Experience Report in PageSpeed Insights and Search Console — to decide if real visitors are having a slow or unstable page. Use the lab score to reproduce a problem and test a fix. A green lab score on a fast laptop does not cancel a poor mobile field result.
Which Core Web Vital should a small business fix first?
The one failing for real users on the pages that receive enquiries, usually the homepage or the main service page on mobile. Largest Contentful Paint is often the hero image or a late font. Layout shift is often a banner or embed. Interaction delay is often too much JavaScript.
Do Core Web Vitals guarantee better rankings?
They are one signal among many, and a tie-breaker more than a magic lever. They do affect whether a visitor stays long enough to enquire. Fix them because the page feels broken, and treat any ranking movement as a bonus.
Thanks for reading. If this was useful, we'd love to help with the next step.
Get in touch
