Google Pagespeed Insights

How PageSpeed Insights Combines Field and Lab Data

PageSpeed Insights pulls real-user data from the Chrome User Experience Report (CrUX). This field data shows how actual visitors experienced the page over the last 28 days. It is only available for pages that meet a minimum traffic threshold.

The tool also runs a Lighthouse audit in a controlled environment. This lab data simulates a mid-range mobile device on a slow 3G connection. Lab results help diagnose issues that field data may not reveal, such as render-blocking resources or unused JavaScript.

By showing both views, PSI lets you see whether a performance problem affects real users or only appears under test conditions. This distinction is critical for prioritising fixes.

For example, a page might have a good LCP in the lab but poor field LCP because real users have slower connections. Conversely, lab data might flag a large image that field data does not, because real users have already cached it. Combining both views helps you decide what to fix first.

Understanding the Score and Its Limitations

The PageSpeed Insights score is a weighted composite of lab metrics. It ranges from 0 to 100, with 90 or above considered good. However, the score is not a ranking factor in Google Search.

A high score does not guarantee fast load times for every visitor. Network conditions, device capabilities, and geographic location all affect real-world performance. The score is a diagnostic aid, not a performance guarantee.

For a broader view of site speed, combine PSI with other tools like Screaming Frog to crawl multiple pages. What page speed means in practice depends on user context, not just a single number.

Another limitation is that the score can change based on the device and network simulated. A page that scores 95 on a fast connection might drop to 70 on a slower one. This variability means you should test under consistent conditions and focus on the underlying metrics rather than the number itself.

• The score is a composite of lab metrics, not a direct measure of real-user experience.
• Field data from CrUX may be missing for low-traffic pages, leaving you with only lab results.
• The score can vary between tests due to network conditions and device simulation settings.

Key Metrics Reported by PageSpeed Insights

PSI reports three Core Web Vitals as field metrics: Largest Contentful Paint (LCP), Cumulative Layout Shift (CLS), and Interaction to Next Paint (INP). These are the metrics Google uses to assess user experience. The tool also shows First Contentful Paint (FCP) and Time to First Byte (TTFB).

Lab metrics include many of the same plus additional diagnostics like Total Blocking Time and Speed Index. Each metric has a threshold that determines whether it is Good, Needs Improvement, or Poor.

The table below shows the current thresholds for the three Core Web Vitals. These values are defined by Google and updated occasionally.

Core Web Vitals are central to PSI's field report. Understanding these thresholds helps you interpret whether your page meets Google's user experience standards.

How to Use PageSpeed Insights for Core Web Vitals

To improve Core Web Vitals, start by running a PSI report on your most important pages. Look at the field data first: if LCP, CLS, or INP are Poor, those pages need attention. Then review the lab recommendations for specific fixes.

Common fixes include optimising images, reducing server response time, removing render-blocking resources, and ensuring stable layout dimensions. PSI provides a list of opportunities with estimated savings.

For WordPress sites, plugins like diy seo wordpress can help implement many of these changes. amp seo is another option for mobile performance, though it is not required for good Core Web Vitals.

You can also use the PSI API to automate testing for multiple pages. This is useful for monitoring changes over time without manually entering each URL.

• Run a PSI report on your most important pages.
• Check field data for LCP, CLS, and INP.
• Review lab recommendations for specific fixes.
• Implement fixes and re-test.

Common Misconceptions About PageSpeed Insights

Many site owners treat the PSI score as the main goal. This leads to chasing a number without fixing the actual user experience. The underlying metrics matter more than the composite score.

Another mistake is assuming a good result on one page means the whole site is fast. PSI works at page level, so other templates may perform differently. You need to test each template type.

Ignoring the split between field and lab data is also common. Lab improvements may not immediately change real-user data, and vice versa. Both sources are valuable but serve different purposes.

Finally, some users still rely on the deprecated Chrome extension. That version is no longer supported and may give outdated results. Always use the web-based tool at pagespeed.web.dev.

PageSpeed Insights vs Other Performance Tools

PSI is one of several free performance tools. Lighthouse is the underlying engine for PSI's lab tests but does not include field data. GTmetrix and WebPageTest offer more detailed lab diagnostics but lack CrUX integration.

For SEO professionals, PSI is often the starting point because it ties directly to Core Web Vitals. Tools like seopress can automate PSI checks across multiple pages and track changes over time. SEO Press Pro offers similar automation for broader site audits.

SEO Tools complement PSI by providing broader site audits. No single tool covers everything; using PSI alongside others gives a fuller picture of performance.