Website Performance Audit: How to Use Lighthouse, PageSpeed Insights, and WebPageTest

You know your site is slow, but you do not know why. Or you think it is fast because it loads quickly on your office fiber connection, which tells you almost nothing about what real visitors experience. The answer is in the data, and three free tools extract it in minutes: Google Lighthouse, PageSpeed Insights, and WebPageTest. Used together, they give a complete picture of your performance, pinpoint the bottlenecks, and rank the fixes by impact.

Performance is not just a technical concern. Slower pages lose visitors before they convert, and the relationship between load time and lost customers is well documented across major published studies. The first step toward fixing it is measuring it accurately, so here is how to run a proper audit.

Which free tools should you use for a performance audit?

Use Google Lighthouse, PageSpeed Insights, and WebPageTest together, because each answers a different question. Lighthouse gives a broad lab audit, PageSpeed Insights adds real-user field data, and WebPageTest provides the deepest diagnostics. No single tool tells the whole story, which is why the audit uses all three.

How does Google Lighthouse work?

Lighthouse is built into Chrome DevTools and runs a thorough audit that scores your site on Performance, Accessibility, Best Practices, and SEO. Each category scores from 0 to 100, with specific, prioritized recommendations attached to the numbers.

For business impact, three performance metrics matter most. Largest Contentful Paint (LCP) measures when the main content becomes visible, with a target under 2.5 seconds. Interaction to Next Paint (INP) measures responsiveness to user input, with a target under 200 milliseconds. Cumulative Layout Shift (CLS) measures visual stability, with a target under 0.1. These three are Google's Core Web Vitals, and they directly influence search rankings.

To run an audit, open Chrome DevTools, go to the Lighthouse tab, select "Performance" and "Mobile" since mobile is what Google ranks on, and click "Analyze page load." The audit takes under a minute and produces a report with scores, metric values, and ranked improvement opportunities. Run it a few times and use the median, because network and CPU variability can make any single run misleading.

Why is PageSpeed Insights field data important?

PageSpeed Insights field data is important because it reflects real measurements from actual Chrome users who visit your site, not a simulated test on a single machine. The tool gives you two datasets: lab data, which is a Lighthouse test run on Google's servers, and field data, which is collected from real visitors over time. When the two disagree, trust the field data.

This matters because lab and field results often diverge. If field data shows an LCP of 4.2 seconds while lab data shows 2.8 seconds, your real visitors are having a worse experience than your test suggests, usually because they are on slower connections or older devices than your test environment assumes. Always prioritize field data when it is available, because it is the experience Google actually measures and ranks.

What does WebPageTest show that the others don't?

WebPageTest offers the most detailed analysis of any free tool, including views the others do not provide. It shows filmstrips of exactly what a visitor sees at each moment during load, waterfall charts of every network request in sequence, connection views with server response times and DNS and TLS timing, and repeat-visit tests that reveal how much your caching helps.

The waterfall chart is the most diagnostic view of all. Each horizontal bar represents one resource, such as an HTML file, a stylesheet, a script, an image, or a font, plotted against its loading timeline. Long bars mark slow resources. Bars that do not start until others finish reveal render-blocking dependencies. The waterfall turns a vague goal like "make it faster" into a specific list of which files are causing the delay and in what order, which is exactly what you need to fix the problem rather than guess at it.

What are the most common performance issues, and how do you fix them?

After running all three tools, a familiar pattern usually emerges. The most common issues on business websites are predictable, and so are the fixes:

• Unoptimized images. Compress them and serve modern formats like WebP or AVIF.
• Too many third-party scripts. Analytics, chat widgets, and social embeds each add load time, so audit them and remove what you do not need.
• Render-blocking CSS and JavaScript. Defer non-critical resources so the page can render before they finish.
• No caching. Set appropriate cache headers on static assets so repeat visits are faster.
• Slow server response time. Upgrade hosting or add a CDN to bring content closer to visitors.

Addressing the top few issues usually produces most of the improvement, and faster pages tend to convert better, which is the whole point of the exercise. For why that link between speed and revenue holds, see our guide on how load time affects conversions.

How do you set a performance budget?

A performance budget sets maximum thresholds for things like page weight, number of requests, and load time, so the site cannot quietly bloat over time. For example: a page weight cap of 1.5MB, no more than 50 HTTP requests, and an LCP under 2.5 seconds on a 4G connection.

The discipline is what makes it work. When a new feature or content change pushes the page past a threshold, something else has to be optimized or removed to compensate. Without a budget, a fast site slowly accumulates scripts, images, and embeds until it is a slow site, and no one can point to the change that did it. The budget forces the trade-off into the open while it is still easy to fix.

How often should you run a performance audit?

Run a performance audit at least monthly to catch regressions, and again after any significant change such as a new plugin, a design update, or a large content addition. A performance audit is not a one-time event; sites drift slower over time, and regular checks catch the drift early.

For ongoing coverage, set up automated monitoring with a tool like SpeedCurve or Calibre so you get continuous tracking rather than occasional spot checks. The sites that stay consistently fast are not lucky. They measure on a schedule and fix small issues before they compound into a slow page nobody enjoys.

The gap between a site that loads in a couple of seconds and one that takes several is the gap between keeping visitors and losing them, and it shows up in both bounce rate and ranking. Performance optimization is one of the highest-return improvements most sites can make.

If you would rather have speed built in from the architecture up than audited and patched later, book a call with Studio Aurora and we will tell you where your pages stand.