For most small businesses, the best website speed test is Google’s PageSpeed Insights, read for its real-visitor data rather than its headline score. GTmetrix and WebPageTest are better for finding what’s slow, because they let you test from Sydney and show every file the page loads. A red score on its own isn’t an emergency: it’s a lab estimate, and the numbers that matter are the ones your real visitors experience.
Key takeaways
- PageSpeed Insights shows two things: real Chrome visitors over 28 days, and a one-off lab test. Judge your site on the first.
- The 0–100 score is a lab result that moves from test to test. Don’t panic over a few points.
- PageSpeed Insights tests from overseas data centres. GTmetrix and WebPageTest let you pick an Australian location.
- Test the same pages, the same way, three times, and write the results down before you change anything.
Why does my website speed test score look so bad?
Usually because the mobile test is deliberately tough. PageSpeed Insights simulates a mid-tier phone (a Moto G4) on a mobile network, according to Google’s documentation. Many real visitors have faster phones and connections than that, so the lab score is a stress test, not a typical visit.
The score also moves on its own. Google’s Lighthouse team says much of the variation isn’t caused by the tool: network routing, ads, A/B tests, browser extensions and antivirus software can all shift it (Chrome for Developers). Two runs a minute apart can differ by several points.
So a red or orange number is a prompt to look closer, not proof that customers are leaving. What you need to know is whether real people get a slow page.
What’s the difference between lab data and field data?
Lab data comes from one simulated visit in a controlled setting. It’s repeatable and good for debugging, but Google notes it “may not capture real-world bottlenecks”.
Field data comes from real visitors. In PageSpeed Insights, it’s drawn from the Chrome User Experience Report and covers the previous 28 days. Smaller sites often don’t have enough visits for page-level field data, so the tool falls back to data for the whole site, or shows none at all.
Google’s Core Web Vitals are field measures. A page is “good” when 75% of visits load the main content within 2.5 seconds (LCP), respond to taps within 200 milliseconds (INP) and don’t jump around much (CLS of 0.1 or less), per web.dev. Our guide to core web vitals WordPress fixes covers what to change when one fails.
PageSpeed Insights, GTmetrix and WebPageTest compared
All three are free to start with, and all three run a lab test of a page. The differences are in where they test from, what extra data they show and who they suit.
PageSpeed Insights: the one Google uses
PageSpeed Insights is the place to check whether real visitors pass Core Web Vitals. Its 0–100 performance score comes from a Lighthouse lab test: 90 or more is good, 50–89 needs improvement and under 50 is poor.
That score is weighted. Total Blocking Time counts for 30%, Largest Contentful Paint and Cumulative Layout Shift 25% each, and First Contentful Paint and Speed Index 10% each (Lighthouse scoring). One heavy script can drag the score down even when the page looks fast.
The catch for Australian sites: the lab test runs in a Google data centre in North America, Europe or Asia. If your server is in Brisbane, the lab result includes a long trip overseas that your local customers never make.
GTmetrix: a clearer view of what loads
GTmetrix gives a letter grade, but its real value is the waterfall chart. It lists every image, script and font in the order they load, and how long each one took. That’s where you spot a 3 MB hero photo or a slow chat widget.
GTmetrix lists 28 test locations, including Sydney, and you set your location in the analysis options (GTmetrix). Choose Sydney so the test reflects what Australian visitors see.
WebPageTest: the most detail
WebPageTest is the most technical of the three. You choose the test location, the device or browser and the connection speed. It can also run the page several times and load it twice per run, so you can compare a first visit with a return visit (Catchpoint documentation).
Its filmstrip shows screenshots as the page builds, which makes “it feels slow” visible. It’s ideal for proving whether a change helped, but it’s more than most owners need week to week.
Which numbers actually matter?
Ignore most of the report and watch these:
- Core Web Vitals field result. Pass or fail, from real visitors. This is the result Search Console reports on too.
- Largest Contentful Paint. When the main content appears. Aim for 2.5 seconds or less.
- Time to First Byte (TTFB). How long the server takes to start replying. web.dev calls 0.8 seconds or less good and over 1.8 seconds poor (web.dev). A slow TTFB points to hosting, caching or the database rather than images.
- Total page weight and the slowest files. From the GTmetrix or WebPageTest waterfall.
Search Console’s Core Web Vitals report is also worth a look. It uses the same real-user data and groups similar pages together, so one template problem shows up across many URLs (Google Search Console Help).
Speed matters because it changes what visitors do. Deloitte found a 0.1-second mobile speed gain lifted retail conversions by 8.4% (web.dev case study). Speed is one lever; our guide on how to improve website conversion covers the others.
How do I test website speed properly from Australia?
- Pick three pages. Your home page, your busiest service or product page, and a contact or checkout page.
- Check field data first. Run each page through PageSpeed Insights and note whether Core Web Vitals pass, on mobile and desktop.
- Run a Sydney lab test. Use GTmetrix or WebPageTest with Sydney as the location, and note LCP, TTFB and page weight.
- Test three times. Use the middle result. One run is too noisy to trust.
- Test logged out. Use a private window, so you see the cached page visitors get, not your admin view.
- Write it down. Keep a simple table with the date, page, tool and results, so you can tell whether a change helped.
Don’t skip mobile. Phones made up 48.6% of Australian web traffic in October 2026 (StatCounter). A page can pass speed tests and still be awkward to use, which our article on making a mobile friendly website covers.
I’ve tested. What should I fix first?
Let the waterfall tell you. A long TTFB means the server is the bottleneck, so caching or better WordPress hosting in Australia will help more than image tweaks. A fast TTFB with a slow LCP usually means large images, heavy themes or too many scripts.
We’ve ranked the usual fixes in how to speed up a WordPress website, and the common causes in why a WordPress site is slow. If you’d rather hand it over, our website speed and improvement work starts with these same tests and ends with a before-and-after report.
Speed also slips over time as plugins, scripts and content pile up. Re-test monthly, or let one of our website maintenance plans keep watch.
Frequently asked questions
What is a good PageSpeed Insights score?
Google rates 90 to 100 as good, 50 to 89 as needing improvement and below 50 as poor. The score is a lab estimate, though. Passing Core Web Vitals with real visitors matters more than chasing 100.
Why do GTmetrix and PageSpeed Insights give different results?
They test from different places, with different hardware, connections and settings. PageSpeed Insights runs its lab test overseas, while GTmetrix can test from Sydney. Compare each tool with itself over time rather than against the other.
Why does my score change every time I test?
Network routes, server load, ads, A/B tests and browser extensions all vary between runs. Test three times and use the middle result. A swing of a few points is normal and doesn’t mean anything changed.
Why is there no real-user data for my site?
Field data needs enough Chrome visitors over 28 days. Smaller sites may only get whole-site data, or none. In that case, rely on careful lab tests from Sydney and watch your enquiries and sales for changes.
How often should I run a website speed test?
Monthly is enough for most small business sites, plus after any big change such as a new plugin, theme or page builder update. Keep a record so you can spot slow creep before customers notice it.
Sources
- About PageSpeed Insights, Google for Developers
- Lighthouse performance scoring, Chrome for Developers
- Web Vitals, web.dev
- Time to First Byte (TTFB), web.dev
- Test server locations, GTmetrix
- WebPageTest Instant Test, Catchpoint documentation
- Core Web Vitals report, Google Search Console Help
- Milliseconds make millions, web.dev (Deloitte research, 2020)
- Desktop vs mobile vs tablet market share, Australia, StatCounter, October 2026