Lab vs field data: Lighthouse LCP was up to 16.7x the CrUX p75 on 7 sites
You run PageSpeed Insights and get a 34 on mobile. Panic? Not yet. That prominent Performance score comes from a Lighthouse lab test under controlled conditions. Real visits span different devices, networks, and cache states; some are faster than the lab test, and some are slower. At straffesites.com , a small web studio in Belgium, we run a recurring Lighthouse benchmark of web development…
The story explores the significant gap between lab and field data for Largest Contentful Paint (LCP) metrics on seven websites. Lab data comes from controlled tests, while field data represents real-user experiences. For the September 2026 edition of an Lighthouse benchmark, seven Belgian web development agencies were analyzed. Lab LCP values ranged from 3.2 to 21.4 seconds, while field LCP (p75) values ranged from 1.09 to 2.51 seconds.
The median ratio between lab and field LCP was 8.1x, meaning lab LCP values were up to 16.7 times higher than field LCP values. The large discrepancy arises due to factors like cache state, device and network conditions, user behavior, and timing. Field data is a distribution of numbers, not a single value, and LCP percentiles represent the experience of measured page loads, not every visitor's experience.
The gap between lab and field data highlights the limitations of lab tests in capturing the full range of real visits.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
