What Is PageSpeed Insights And How To Improve Your Score
A fast website gives visitors a smoother experience and helps search engines evaluate page quality. Google PageSpeed Insights is a free tool that checks how a web page performs on mobile and desktop, then reports practical opportunities for improving loading speed, responsiveness, and visual stability.
The score is useful, but it is not the whole story. A page can receive a high Lighthouse score while real visitors experience delays on a busy mobile network. For Australian websites, this matters because users may browse through variable NBN connections, crowded 4G networks in Sydney or Melbourne, or slower links in regional areas.
What PageSpeed Insights measures
PageSpeed Insights, often called PSI, combines two types of performance data. The first is lab data generated by Lighthouse under a simulated device and network connection. The second is field data collected from real Chrome users through the Chrome User Experience Report, commonly known as CrUX.
The report assesses loading speed, interactivity, and layout stability. Its main Core Web Vitals are Largest Contentful Paint, Interaction to Next Paint, and Cumulative Layout Shift. These metrics show whether the main content appears quickly, whether the page responds promptly to actions, and whether elements jump around as the page loads.
PSI also provides a performance score from 0 to 100. A score from 90 to 100 is considered good, 50 to 89 needs improvement, and below 50 is poor. Treat this number as a diagnostic signal rather than a direct ranking guarantee.
How to read the report
Enter a complete page address into the PageSpeed Insights website, then review both the mobile and desktop results. Mobile testing should receive priority because Google primarily uses mobile-first indexing and many visitors reach websites from smartphones.
The field data section can be especially valuable. It reflects the 75th percentile of real user experiences over a rolling period, so it gives a broader view than a single test. A new page, low-traffic website, or page with insufficient Chrome user data may have no field results available.
Study the diagnostics beneath the score. Items such as “Eliminate render-blocking resources”, “Reduce unused JavaScript”, and “Properly size images” point towards causes rather than merely reporting symptoms. Open each recommendation to see its estimated saving and the affected files or page elements.
Why your score can change
A PSI result is not fixed. Server load, third-party scripts, advertising auctions, image delivery, caching status, and test conditions can all affect a single run. Running the same URL several times may produce different results, particularly for pages that depend on external services.
Lab data and field data can also disagree. Lighthouse uses a controlled simulation, while real visitors use different phones, browsers, locations, and internet connections. A page may perform well in a developer’s office on fast Wi-Fi but feel sluggish for someone in regional Queensland using a mid-range Android phone.
Use the score to identify patterns. If the same templates repeatedly show a slow LCP or poor INP, the underlying issue deserves attention. Avoid chasing small score changes when real-user Core Web Vitals are already healthy.
| Performance metric | Good | Needs improvement | Poor |
|---|---|---|---|
| Largest Contentful Paint | Up to 2.5 seconds | 2.5–4 seconds | Over 4 seconds |
| Interaction to Next Paint | Up to 200 milliseconds | 200–500 milliseconds | Over 500 milliseconds |
| Cumulative Layout Shift | Up to 0.10 | 0.10–0.25 | Over 0.25 |
Fast improvements with a clear payoff
Start with images because they are often the largest files on a blog or online shop. Resize images to the dimensions actually displayed, compress them, and use modern formats such as WebP or AVIF where supported. Avoid uploading a 4,000-pixel photograph when a 900-pixel version is enough for the content area.
Set width and height attributes for images and embedded media. These dimensions reserve space before the files load and help prevent layout movement. Lazy-load below-the-fold images, but do not lazy-load the main hero image if it is the largest visible element.
Browser caching can reduce repeat load times, while a content delivery network can serve static files from locations closer to visitors. For an Australian audience, choose a CDN with useful points of presence in Australia or nearby regions, then verify that caching works correctly for CSS, JavaScript, fonts, and images.
Improve Core Web Vitals
A slow Largest Contentful Paint often comes from a delayed server response, an oversized hero image, render-blocking CSS, or a web font that loads too late. Improve time to first byte with reliable hosting, page caching, database optimisation, and a CDN. Then prioritise the main content instead of loading every asset at once.
Interaction to Next Paint measures how quickly the page responds after a visitor clicks, taps, or types. Excessive JavaScript is a common cause of poor responsiveness. Remove unused plugins, postpone non-essential scripts, split large bundles, and reduce complex animations. A WordPress site with several marketing, analytics, social sharing, and pop-up plugins can accumulate significant browser work.
Cumulative Layout Shift is usually linked to images without reserved dimensions, dynamically inserted advertisements, late-loading banners, and fonts that change text size. Define space for these components before they appear. Review cookie notices and promotional bars carefully, especially when they push the main content down after the first render.
Make technical changes safely
Before editing a live website, create a backup and test changes on a staging copy when possible. Change one major factor at a time so you can identify which action improved or damaged the result. Record the URL, device type, score, and Core Web Vital values before making changes.
Review hosting performance as well as front-end code. Cheap shared hosting may struggle during traffic spikes, and a slow origin server can undermine excellent image compression. Australian publishers serving readers across Perth, Adelaide, Brisbane, and smaller regional towns should check whether their host and CDN deliver consistently across different locations.
Third-party services deserve particular scrutiny. Chat widgets, heat maps, video embeds, social feeds, consent platforms, and advertising scripts can add requests and main-thread work. Keep essential tools, delay the rest, and remove services that provide little value to visitors or your publishing workflow.
Build a faster publishing routine
Run PageSpeed Insights on important templates rather than testing only the homepage. Include a blog post, category page, contact page, product page, and any landing page used in Google Ads or social campaigns. Performance problems often affect a template or plugin across hundreds of URLs.
Test after publishing major changes, installing a new theme, adding advertising, or migrating hosts. Pair PSI with Google Search Console’s Core Web Vitals report and real-user monitoring when available. This combination helps distinguish a one-off lab result from a problem affecting visitors over time.
Make performance part of your regular content process. Compress media before upload, limit unnecessary plugins, reserve space for page elements, and check mobile usability on an actual phone. Run your key URLs through PageSpeed Insights today, fix the highest-impact issues first, and measure the results again before moving to smaller refinements.