Google PageSpeed Insights

Free tool from Google that tells you how fast your site loads and where the bottlenecks are. Paste any URL into pagespeed.web.dev, and it returns a performance report with real-world data and specific recommendations.

It matters for SEO because Google uses page speed as a ranking signal. Slow pages rank lower, bounce higher, and convert worse.

Core Web Vitals

These are the three metrics Google actually uses for rankings:

Largest Contentful Paint (LCP) — How long until the biggest visible element loads. Under 2.5 seconds is good. This is usually your hero image, headline, or featured video thumbnail. If your LCP is slow, your images are probably too large or your server is too slow.

Interaction to Next Paint (INP) — How quickly the page responds when someone clicks, taps, or types. Under 200 milliseconds is good. Sluggish buttons and laggy forms kill conversions. Heavy JavaScript is usually the culprit.

Cumulative Layout Shift (CLS) — How much the page jumps around while loading. Under 0.1 is good. You know that thing where you're about to tap a button and an ad loads above it, pushing everything down? That's layout shift. It's infuriating, and Google penalizes it.

Two Types of Data

PageSpeed Insights gives you two separate reports, and they measure different things:

Field Data (Real Users)

Data from actual people visiting your site over the past 28 days, pulled from the Chrome User Experience Report (CrUX). This is what Google actually uses for rankings. It reflects real devices, real connections, real behavior.

If your site doesn't have enough traffic, this section will be empty — you need a minimum threshold of visits before Google has enough data to report.

Lab Data (Lighthouse)

A simulated test run by Google's Lighthouse engine in a controlled environment. It uses a fixed device profile and network speed, so it's consistent and reproducible — but it doesn't reflect how real users experience your site.

Lab data is useful for debugging. Field data is what matters for rankings.

Reading the Scores

PageSpeed Insights scores each metric with a color:

Green (Good) — Meeting Google's threshold. Don't touch it.
Yellow (Needs Improvement) — Not hurting you badly yet, but worth fixing.
Red (Poor) — Actively hurting your rankings and user experience. Fix this first.

The overall performance score (0-100) is a weighted composite from Lighthouse. It's a useful gut check, but the individual Core Web Vitals matter more for SEO.

What to Fix First

PageSpeed Insights doesn't just diagnose — it tells you what to do. The "Opportunities" section lists specific fixes ranked by estimated impact:

Serve images in next-gen formats — Switch to WebP or AVIF. This alone can shave seconds off load time.
Eliminate render-blocking resources — CSS and JavaScript files that prevent the page from displaying until they load.
Reduce unused JavaScript — Bloated scripts that load but never run.
Properly size images — Serving a 4000px image in a 400px container wastes bandwidth.
Enable text compression — Gzip or Brotli compression on your server.

The "Diagnostics" section goes deeper with technical details for developers.

Mobile vs Desktop

Always check both. Mobile scores are almost always lower because PageSpeed Insights simulates a mid-range phone on a 4G connection — not the MacBook Pro on fiber you're testing from. Google uses mobile-first indexing, so your mobile score is the one that matters for rankings.

When to Use It

Before launching a new page — Catch performance issues before they go live
After making changes — Verify that your optimizations actually worked
Competitor analysis — Paste a competitor's URL and see how you compare
Client audits — Quick way to show a prospect where their site is underperforming

Pair it with Google Search Console for the full picture — Search Console shows which pages have Core Web Vitals issues at scale, while PageSpeed Insights gives you the deep dive on a single URL.

For ongoing monitoring beyond spot checks, Rybbit tracks Core Web Vitals continuously across all your pages so you catch regressions before they hurt rankings.