Google PageSpeed Insights vs. Lighthouse: Which Measures Site Performance Best?

Google PageSpeed Insights and Google Lighthouse both help you measure site performance, but they draw on different data sources and can return scores that differ by 10 to 15 points for the same page. Understanding what each tool actually measures, and where those differences come from, is the starting point for using them correctly.

If you’ve run the same URL through both tools and found the scores don’t match, you haven’t done anything wrong. That discrepancy is expected and stems from differences in how each tool runs its tests: where it routes traffic, how it throttles the network, and what it measures. Once you understand those differences, both tools become more useful.

In this article, I’ll answer:

• What is page speed?
• What is Google PageSpeed Insights?
• What is Google Lighthouse?
• Should I use Google PageSpeed Insights or Google Lighthouse?

What Is Page Speed?

Page speed measures the time between a browser request and content rendering. Pages contain many components with varying load times, and Google breaks speed into specific metrics tracking loading milestones like First Contentful Paint and Time to Interactive. Speed fluctuates as sites grow and technology evolves, affecting visitor experience and conversion rates. Thus, tracking page speeds should be a fundamental part of your regular site health monitoring process.

But what’s the best page speed tracking tool? Google offers two options, both of which are free: Google PageSpeed Insights and Lighthouse. 

Let’s compare PageSpeed Insights vs. Lighthouse and discuss what you need to know about tracking and improving page speed.

What Is Google PageSpeed Insights?

Google PageSpeed Insights (PSI) is a free tool measuring page speed on desktop and mobile using real-world data from the Chrome User Experience Report (CrUX) and lab data via the Lighthouse application programming interface (API). Simply visit pagespeed.web.dev, enter a URL, and receive a Performance Score weighted across several metrics.

CrUX collects performance data from real Chrome users who have visited your site, aggregated over a rolling 28-day window. PSI reports this field data at the 75th percentile, meaning the score reflects the experience of a user with below-average performance rather than the median visitor. That’s a more conservative benchmark than a controlled lab test, which explains why PSI scores often look worse than what you see in a Lighthouse run.

The Performance Score is weighted across these metrics:

• First Contentful Paint: (10%) The time between a user landing on a page and the browser rendering the first piece of content.
• Speed Index: (10%) The average time until all page elements are visible.
• Largest Contentful Paint: (25%) The time it takes for the largest image or text block to fully load.
• Time to Interactive: (10%) The time until a page is fully interactive and responding to user input within 50 milliseconds.
• Total Blocking Time: (30%) The time difference between First Contentful Paint and Time to Interactive during which a page is blocked from user input. (In other words, the time difference between when something first appears on-screen vs. when the user can actually click on it.)
• Cumulative Layout Shift (CLS): (25%) The instability of a page (content loading out of order, page elements that move or jump unexpectedly, etc.)

PSI also reports Interaction to Next Paint (INP), which replaced First Input Delay (FID) as a Core Web Vitals metric in March 2024. INP measures the delay between a user’s interaction and the next visual response. Because INP requires real user interaction data to measure accurately, it appears in PSI’s CrUX field data rather than the lab-based performance score.

Don’t be too concerned if your website’s Performance Rating isn’t perfect or if some of your metrics seem low. Speed metrics can fluctuate depending on a few factors outside your control. Remember that these metrics are an aggregate of data and actual experiences will vary from user to user. 

That said, you can maximize your Performance Rating and improve page speed by focusing on the ‘Opportunities and Diagnostics’ sections near the bottom of the report. These sections contain recommendations for specific actions that can improve speed metrics, though they may require help from an SEO web developer to implement.

What Is Google Lighthouse?

Google Lighthouse is a free, open-source tool that can perform site audits to measure many different metrics. In addition to assessing page speed, Lighthouse can evaluate how well a site fits modern web development standards, a site’s level of accessibility, crawlability, and more.

Lighthouse exclusively uses lab data, which means audits are run in a controlled, pre-defined environment, not gathered from actual user data.

You can access the Google Lighthouse speed test in a few ways. If you’re not a developer, the easiest way is to access Lighthouse through the Chrome browser extension, although the extension can’t audit pages with cookie authentication on its own. Developers can use Chrome DevTools for complete access to Lighthouse.

Your Lighthouse report will include information about:

• Performance audits that measure the same speed metrics as PSI, plus provide more detailed development data.
• Best Practices Audits that assess a site’s overall “code health” compared to modern development standards, including security standards.
• Accessibility Audits that highlight opportunities for improvement in design, user interfaces, and development.
• SEO Audits that test whether search engines can easily crawl your page. 
• Progressive Web App Audits that measure the general speed, reliability, and installability of web apps.

Lighthouse can’t directly measure INP, which matters for any Core Web Vitals assessment after March 2024. It uses Total Blocking Time (TBT) as a proxy for interaction responsiveness. TBT tracks main-thread blockage during initial load, which correlates with INP but doesn’t capture actual user interactions. For post-March 2024 Core Web Vitals assessments, PSI’s CrUX field data is the only tool in this comparison that reflects actual INP performance without requiring separate Real User Monitoring.

What Are Lighthouse Stack Packs?

Lighthouse Stack Packs provide platform-specific recommendations tailored to your site’s technology stack, offering more precision than general suggestions. Supported platforms include WordPress, React, Angular, Next.js, and AMP, among others. Rather than surfacing a generic warning to eliminate render-blocking resources, a Stack Pack identifies the framework-specific cause, such as flagging a particular Angular module or a WordPress plugin creating the bottleneck.

Visit the official Github repository to learn more about Stack Packs.

Should I Use Google PageSpeed Insights vs. Lighthouse?

Google PageSpeed Insights combines lab and field data to provide a thorough, easy-to-understand overview of a page’s speed. Lighthouse is developer-focused and collects lab data from audits that cover everything from page load speed and web security to accessibility and best practices. Both offer improvement recommendations.

Google PageSpeed Insights vs. Lighthouse: Overview

When To Use Google PageSpeed Insights

Use PSI if you want a clear understanding of a URL’s speed and specific improvement recommendations to implement yourself or relay to a developer.

When To Use Google Lighthouse

Use Lighthouse if you’re a developer (or non-developer with some dev knowledge) and want a deep dive into many different site metrics beyond page speed. Stack Packs can help provide more tailored recommendations than the standard ones offered by Lighthouse.

Why Do My PageSpeed Insights and Lighthouse Scores Look Different?

A 10- to 15-point gap between PSI and Lighthouse is normal. Larger gaps usually trace back to one or more of these causes.

• Geographic routing adds latency that doesn’t show up in local tests. PSI routes tests from the Google data center nearest to your IP address, not your web server. For a US-based tester working on a site hosted in Europe without a content delivery network, that routing adds real latency before a single asset loads.
• Throttling method creates the most consistent score difference between the tools. PSI uses simulated throttling, which loads your page at full speed and then calculates what performance would have looked like on a slower connection. Lighthouse in Chrome DevTools supports applied throttling, which directly slows the network during the test. Applied throttling generally produces lower, more realistic scores.
• Cache state affects what each tool sees on first load. PSI always runs against a cold cache. If you’re testing locally with warm cached resources, you’re comparing different conditions, not the same page.
• Hardware variance matters on JavaScript-heavy pages. PSI runs on Google’s servers, and your local machine’s processor profile affects JavaScript execution times differently than Google’s test hardware.
• Browser extensions add overhead in local Lighthouse runs. Lighthouse in Chrome DevTools shares your browser environment, where extensions run in the background. PSI operates in isolation. Running Lighthouse in an Incognito window without extensions closes a significant portion of this gap.

For persistent discrepancies, run Lighthouse in Incognito without extensions and use applied throttling. If a gap with PSI remains after that, the difference most likely comes from geographic routing or server hardware rather than anything misconfigured on your end.

Which Tool’s Data Does Google Use for Rankings?

Google’s page experience ranking signal draws from CrUX field data (the real-user measurements PSI surfaces), not from Lighthouse lab scores. A page can earn a 95 in Lighthouse and still fail Core Web Vitals if real users experience slow load times.

For SEO, that means PSI’s CrUX data is what counts toward Google’s page experience assessment. Lighthouse diagnoses why performance is poor, but it doesn’t feed the signal Google uses for ranking decisions. If your CrUX data shows a failing INP score, improving your Lighthouse score won’t change your Core Web Vitals standing until real users experience faster interactions.

Page performance is also a factor for search systems beyond traditional rankings. AI-powered features like Google’s AI Overviews rely on pages that load reliably and consistently, making CrUX performance data relevant to an integrated SEO and answer engine optimization (AEO) strategy, not just traditional search visibility.

When To Use Each Tool

Google PageSpeed Insights works best when you want a clear read on speed and specific recommendations to implement or share with developers. It offers realistic user performance data.

Use PSI to understand how real users experience your site, to build stakeholder performance reports, and to verify that improvements made in development are actually reaching users in CrUX data. PSI also offers 25,000 free API requests daily, which makes automated monitoring across a large number of pages practical without manual effort. Google has flagged that it plans to drop CrUX field data from this API specifically, so that automated-monitoring use case may need revisiting down the line.

Google Lighthouse suits developers wanting deep dives into multiple metrics beyond speed, especially with Stack Packs for tailored guidance.

Use Lighthouse during development for fast feedback on performance regressions before code deploys. Running Lighthouse in a continuous integration/continuous deployment (CI/CD) pipeline catches regressions automatically before they reach production. Because Lighthouse doesn’t require accumulated CrUX data, it’s also the right tool for new pages, recently redesigned URLs, or any page that doesn’t yet receive enough traffic to generate reliable field data.

Want the Best Data? Use Both PageSpeed Insights & Lighthouse

PSI and Lighthouse each have their strengths and limitations — but they’re most effective when used together to build a complete picture of your site. In fact, a reputable SEO agency would probably use both.

While Lighthouse provides a broader view of data, PageSpeed Insights gives you a more realistic glimpse at how web pages perform for actual users. Use both to get recommendations for improving your site’s front- and back-ends. The more data you have at your disposal, the more effective your site adjustments will be.

Make the Most of Google Tools With Our Free Resources

If you take away one thing from this article, it should be that page speed is important — and much more complex than it seems on the surface. Just like there’s not just one way to boost a site’s speed, there isn’t only one way to boost your site’s search engine performance — or even one best Google tool to track data.

Optimizing your site involves juggling a lot of data and navigating complicated tools. To make things less complicated, check out our other Google tools resources.

Want help improving your site performance? Our website maintenance services can help you improve your page speed and UX so your site visitors can enjoy your offerings. Sign up for a free consultation.