Google PageSpeed Checker: How to Test Your Website’s Speed

A slow website can frustrate visitors, increase the chance they will leave and make it harder for them to complete tasks. A Google PageSpeed checker can help you understand how a page performs and identify changes that may improve its loading experience.

Google’s PageSpeed Insights is a free tool for checking individual web pages. It combines real-world data, where available, with tests run in a simulated environment. The results can help website owners, developers and content teams decide what to investigate next.

What is Google PageSpeed Insights?

PageSpeed Insights analyses a web page and reports on its performance for mobile and desktop devices. It uses Lighthouse to run laboratory tests and may also show data from the Chrome User Experience Report, often called CrUX. CrUX data is collected from eligible Chrome users who have opted in to sharing usage statistics.

These two types of data serve different purposes:

• Field data reflects the experience of real users over a recent period. It may not be available for every page or website.
• Lab data comes from a controlled Lighthouse test. It is useful for diagnosing potential issues, but it represents a simulated visit rather than every visitor’s experience.

How to use the checker

1. Open Google PageSpeed Insights.
2. Enter the full URL of the page you want to test.
3. Select Analyse and wait for the report.
4. Review the mobile and desktop results, then examine the recommendations and diagnostic details.

Test more than the home page. Product pages, articles, landing pages and other templates can load different images, scripts and features, so their results may vary.

Understanding the main performance metrics

PageSpeed Insights reports on several measures of loading speed and page responsiveness. The Core Web Vitals are particularly useful for understanding real-world user experience:

• Largest Contentful Paint (LCP) measures how long it takes for the largest visible content element to appear. A good result is 2.5 seconds or less.
• Interaction to Next Paint (INP) measures how quickly a page responds visually to user interactions. A good result is 200 milliseconds or less.
• Cumulative Layout Shift (CLS) measures unexpected movement of page content. A good result is 0.1 or less.

These thresholds apply to the 75th percentile of page visits, assessed separately for mobile and desktop field data. A single test result is a useful clue, but it should not be treated as a complete picture of every visitor’s experience.

The report may also include Lighthouse metrics such as First Contentful Paint and Speed Index. These help explain how a page behaves during the simulated test. Lighthouse also provides an overall performance score, but the score is a diagnostic summary, not a direct measure of business success or search ranking.

What do PageSpeed recommendations mean?

Recommendations depend on the page, but common opportunities include:

• Optimising images: resize images to suit their display dimensions, compress them and consider modern formats where appropriate.
• Reducing JavaScript: remove unused code, delay non-essential scripts and review third-party tools.
• Improving server response: check hosting performance, caching and content delivery arrangements.
• Optimising fonts: limit unnecessary font files and weights, and consider how fonts are loaded.
• Preventing layout shifts: reserve space for images, adverts and other elements that load after the initial page content.
• Using browser caching: configure suitable caching for static files so returning visitors may not need to download them again.

Some recommendations are estimates or technical diagnostics rather than instructions that must all be followed. Review each suggestion in context, particularly where a change could affect accessibility, tracking, security or important site functionality.

Why can scores change?

PageSpeed results can vary between tests because of network conditions, server load, browser versions, page content and the testing environment. Field data also changes as new user experiences are recorded. For a more reliable view, test several times and compare results over time rather than reacting to one score.

Mobile and desktop results may differ substantially. Mobile tests use a more constrained simulated environment, and real mobile visitors may have different devices and connections. Prioritise the experience of the audience that matters most to your website.

Does a high score guarantee good performance?

No. A high Lighthouse score can be encouraging, but it does not guarantee that every visitor will see a fast page. Equally, a lower score does not automatically mean that a page is unusable. Pay attention to real-user field data, the Core Web Vitals and the specific issues affecting your visitors.

Page speed is also only one part of a successful website. Clear content, accessible design, useful functionality and reliable service all contribute to a good experience.

Using PageSpeed Insights effectively

Start by testing important pages and identifying issues that affect the largest number of visitors. Make one set of changes at a time, then test again and check real-user data when it becomes available. Work with a developer when recommendations involve complex changes to scripts, hosting or site architecture.

Used thoughtfully, Google PageSpeed Insights is a practical way to investigate website performance. It can turn a general concern about speed into specific areas to review, helping teams make informed improvements to the experience they provide.

 

Optimising Website Performance: Top 5 Tips for Using Google Page Speed Checker

1. Test both mobile and desktop performance.
2. Check field data for real-user experience.
3. Review lab data to spot issues quickly.
4. Prioritise the recommendations with the biggest impact.
5. Retest after making changes.

Test both mobile and desktop performance.

Test your website on both mobile and desktop, as performance can vary between devices. Mobile visitors may have slower connections and less powerful hardware, while desktop pages can load different layouts or features. Use Google PageSpeed Insights to review each result separately, then prioritise improvements based on how your audience accesses the site most often.

Check field data for real-user experience.

Check field data to understand how real visitors experience your website. In Google PageSpeed Insights, this data—when available—comes from real Chrome users and shows how a page performs across different devices and network conditions. Pay particular attention to the Core Web Vitals, including loading speed, responsiveness and visual stability. Field data reflects a broader range of visits than a single lab test, making it a useful way to spot issues that may affect your audience in everyday use.

Review lab data to spot issues quickly.

Review lab data in Google PageSpeed Insights to quickly spot potential performance issues. Because it is generated under controlled test conditions, it can highlight problems such as large images, render-blocking resources or slow scripts without waiting for real-user data. Use its recommendations as a starting point for investigation, then check field data where available to understand how your site performs for visitors in everyday conditions.

Prioritise the recommendations with the biggest impact.

Prioritise the recommendations likely to make the biggest difference to visitors, rather than trying to fix every warning at once. Start with issues affecting key measures such as LCP, INP and CLS, and consider how many users or important pages are affected. Addressing oversized images or render-blocking scripts may bring greater benefits than minor technical tweaks. After each change, run the test again and check real-user data where available to see whether the experience has improved.

Retest after making changes.

After making changes to your website, retest the affected pages with the Google PageSpeed checker to see whether performance has improved and to spot any new issues. For a fair comparison, test the same URL under similar conditions, and avoid relying on a single result: scores can vary between runs. Where available, review real-user data as well as the lab results to understand how the changes affect visitors in practice.