PageSpeed Insights (PSI) is Google's free tool for measuring how fast a page loads and how stable it feels to real visitors. Enter a URL at pagespeed.web.dev, run the test, and you get two things at once: real world field data from actual Chrome users, and lab data from a controlled simulation. Reading both correctly is the key to acting on the results instead of chasing a single number.
Field data vs lab data
PSI reports are split into two data sources, and confusing them is the most common mistake.
- Field data (Core Web Vitals) comes from the Chrome User Experience Report (CrUX). It is the 75th percentile of real user sessions over the past 28 days. This is what Google actually uses for its page experience signals, so it is the number that matters for search.
- Lab data (Lighthouse) is a single, simulated load on a throttled mobile device in a Google data center. It is reproducible and great for debugging, but it is one synthetic run, not your real audience.
A brand new page, or a low traffic page, may show no field data at all because CrUX needs enough real visits to build a sample. In that case, lean on the lab section while the field data accumulates.
Real Chrome users Field data (CrUX) Simulated load Lab data (Lighthouse) PageSpeed Insights report Fix and re-measureRunning a test step by step
- Open
https://pagespeed.web.dev/in your browser. - Paste the full URL of the specific page you want to test, not just the domain root. Test the pages that matter, such as your top landing pages and product pages.
- Click Analyze and wait for the report to build.
- Toggle between the Mobile and Desktop tabs. Mobile is throttled harder and is usually the tougher, more representative score.
- Read the field data first, then scroll to the Lighthouse diagnostics for the specific fixes.
The metrics that matter
Three Core Web Vitals define the field score. Largest Contentful Paint (LCP) should be 2.5s or less and tracks when the main content appears. Interaction to Next Paint (INP) should be 200ms or less and measures responsiveness to clicks and taps. Cumulative Layout Shift (CLS) should be 0.1 or less and captures unexpected layout jumps. Green means good, and you want all three in the green band. For the definitions and current thresholds, Google's own guidance at web.dev/articles/vitals is the authoritative reference.
Turning scores into fixes
Below the metrics, the Opportunities and Diagnostics sections list concrete actions with estimated savings: serve images in next gen formats, defer offscreen resources, reduce unused CSS and JavaScript, and set an efficient cache policy. Work top down, since the items are ordered by impact. After each change, re-run the test to confirm the lab number moved, then wait for the 28 day field window to reflect the improvement for real users.
Fast measurements are easier to act on when the underlying platform is quick and consistent. Our SEO web hosting runs on LiteSpeed and CloudLinux, which helps server response time and Time to First Byte, one of the inputs to LCP. For a deeper walkthrough of tuning these metrics, see the Performance and Core Web Vitals guides, or contact us if you want help reading your report.
