PageSpeed Insights

Why Page Speed Matters for SEO

Page speed is a confirmed Google ranking factor. Google first announced speed as a ranking signal for desktop in 2010, then extended it to mobile in 2018, and formalized it through Core Web Vitals as a ranking factor in 2021.

The business impact is clear: a 1-second delay in page load time reduces conversions by an average of 7% (Portent, 2024). Pages that load in 1 second have a conversion rate 3x higher than pages that load in 5 seconds.

Amazon found that every 100ms of added latency cost them 1% in revenue (Amazon Internal Study, cited by Google Web Fundamentals, 2024). Speed affects both rankings and revenue.

How to Use PageSpeed Insights

Using PSI is simple: go to pagespeed.web.dev, enter any URL, and click Analyze. Within seconds, you get a comprehensive performance report.

The results are divided into two sections:

• Field Data — real user experiences collected from the Chrome User Experience Report (CrUX) over the past 28 days
• Lab Data — simulated test results generated by Lighthouse in a controlled environment

Field data is what Google uses for rankings. Lab data is what you use for debugging. They often show different results, and that is expected.

Understanding Your Scores

PSI gives you a performance score from 0 to 100 for both mobile and desktop:

• 90-100 (Green) — Good. Your page performs well.
• 50-89 (Orange) — Needs improvement. There are optimization opportunities.
• 0-49 (Red) — Poor. Significant performance problems.

The global median Lighthouse performance score on mobile is 50 (HTTP Archive, 2025). Most websites are in the "needs improvement" range. Reaching a score of 90+ puts you in the top tier.

Important: the mobile score is almost always lower than desktop. This is by design. PSI simulates a mid-tier mobile device on a 4G connection, which is slower than most desktop setups. Focus on the mobile score since Google uses mobile-first indexing.

Core Web Vitals Breakdown

PSI reports the three Core Web Vitals metrics that directly affect rankings:

Largest Contentful Paint (LCP)

LCP measures how long it takes for the largest visible element (usually a hero image or heading) to fully render.

• Good: 2.5 seconds or less
• Needs improvement: 2.5 to 4 seconds
• Poor: more than 4 seconds

The most common cause of slow LCP is unoptimized images. Converting images to WebP or AVIF format and serving them at the correct size can improve LCP by 25-50% (web.dev, 2024).

Interaction to Next Paint (INP)

INP measures how quickly your page responds when a user interacts with it (clicking a button, typing in a field, expanding a dropdown).

• Good: 200 milliseconds or less
• Needs improvement: 200 to 500 milliseconds
• Poor: more than 500 milliseconds

INP replaced First Input Delay (FID) as a Core Web Vital in March 2024. It is a harder metric to pass because it measures all interactions, not just the first one. About 65% of websites pass the INP threshold on mobile (Chrome UX Report, 2025).

Cumulative Layout Shift (CLS)

CLS measures visual stability. When elements on the page move unexpectedly while loading (a banner pushes content down, an ad shifts the text), that creates layout shift.

• Good: 0.1 or less
• Needs improvement: 0.1 to 0.25
• Poor: more than 0.25

The main fix for CLS is setting explicit width and height attributes on images and video elements. This tells the browser how much space to reserve before the content loads.

Lab Data vs Field Data

Understanding the difference prevents confusion:

• Lab data runs a simulated test right now on a standardized device and network. It is consistent and reproducible, making it ideal for debugging specific issues.
• Field data comes from real Chrome users who visited your page over the past 28 days. It reflects actual user experience but varies based on real-world devices and connections.

A page might score 95 in lab data but only 70 in field data if real users are on slow devices. Or the reverse: field data might be better if most of your users have fast devices. Google uses field data (CrUX) for ranking signals, not lab data.

Pages need at least a threshold of real-user visits (approximately 1,000 page loads over 28 days) to have field data in CrUX (Google CrUX Documentation, 2025). New or low-traffic pages will only have lab data.

Optimization Suggestions

Below the scores, PSI lists specific opportunities and diagnostics:

• Opportunities — actions that will improve load time (reduce image sizes, eliminate render-blocking resources, defer offscreen images)
• Diagnostics — information about page structure that might cause issues (DOM size, JavaScript execution time, cache policy)

Each opportunity shows an estimated time savings. Prioritize the ones with the largest savings. The three most common issues across all websites are unoptimized images, render- blocking CSS/JavaScript, and excessive DOM size (HTTP Archive, 2025).

Mobile vs Desktop

Always check the mobile tab first. Mobile scores are what matter for rankings because Google uses mobile-first indexing for 100% of websites since October 2023 (Google Search Central, 2023).

Mobile devices account for over 60% of global web traffic (Statista, 2025). If your mobile score is poor, you are delivering a bad experience to the majority of your visitors and hurting your rankings simultaneously.