Most people run their website through Google PageSpeed Insights once, see a red or orange number, feel briefly alarmed, and close the tab without acting. The report shows a lot of numbers, colors, and technical terms without explaining which parts actually matter.
PageSpeed Insights is free, fast, and powerful — once you know how to read it.
Step 1: Run a Proper Test
Go to `pagespeed.web.dev`, enter your page's full URL (including `https://`), and click Analyze.
Note: PageSpeed Insights cannot analyze password-protected staging sites or `localhost`. Always test key landing pages, blog posts, and checkout pages individually, rather than testing only the homepage.
Step 2: Field Data vs. Lab Data (The Most Misunderstood Distinction)
The tool presents two completely different sets of numbers:
| Data Type | Source | Purpose for SEO & Rankings |
|---|---|---|
| Field Data (CrUX) | Real-world Chrome visitors over a rolling 28-day window | This is what Google actually uses to evaluate search rankings |
| Lab Data (Lighthouse) | Simulated test run in a controlled environment | Used for debugging specific technical issues during development |
Site owners often panic over a single lab score, when real-world Chrome UX field data is what determines ranking performance (read what are core web vitals).
Step 3: Focus on the Three Core Web Vitals
Scroll past the top-line score to the Core Web Vitals field data metrics: - LCP (Largest Contentful Paint): Target under 2.5s (loading speed of main visible block). - INP (Interaction to Next Paint): Target under 200ms (responsiveness to clicks and taps). - CLS (Cumulative Layout Shift): Target under 0.1 (visual layout stability).
Step 4: Act on "Opportunities" and "Diagnostics"
PageSpeed Insights lists specific files and exact kilobyte savings: - Image Optimization: Identifies specific uncompressed images and potential WebP savings (read WebP vs AVIF vs JPEG). - Main Thread Execution: Flags heavy JavaScript files competing for processor execution time (read hidden performance killers). - Server Response Time (TTFB): Flags slow hosting responses over 0.8 seconds (read how fast should a website load).
Decision Framework
| Report Finding | Actionable Fix |
|---|---|
| Image weight opportunities | Compress assets and convert to WebP/AVIF formats |
| Missing caching/compression | Install caching plugin (WP Rocket or LiteSpeed) |
| Heavy JavaScript / long tasks | Defer third-party scripts or refactor script execution |
| Slow TTFB (>0.8s) | Upgrade server hosting or configure server-level caching |
Brandywebs builds custom websites starting from $999, pre-audited to pass PageSpeed benchmarks (read custom website cost breakdown).
Get a free quote from Brandywebs →
FAQs
Is Google PageSpeed Insights free? Yes. PageSpeed Insights is completely free with no account or subscription required.
Why does my PageSpeed score change between tests? Lab scores fluctuate based on instantaneous server load and connection speeds. Field data (based on real user traffic over 28 days) remains stable.
What is the difference between Field Data and Lab Data? Field data reflects real Chrome user visits over 28 days and drives rankings. Lab data is a single simulated audit for debugging.
How long after making a fix will field data improve? Field data updates continuously across a rolling 28-day window, so fixes typically reflect in Google Search Console within 4 to 6 weeks.