How to Read PageSpeed Insights
Updated 30 August 2026
Jump to section
To read PageSpeed Insights, start with the field data at the top, not the big colored score. The field data shows how real Chrome users experienced your page across the three Core Web Vitals: LCP, INP, and CLS, and whether the page passed. The 0-to-100 performance score below it is a lab estimate from Lighthouse, useful for fixing problems but not the final verdict. Read them in that order.
Field data vs lab data: which to trust
Read the field data first, because it is what real visitors actually lived through. PageSpeed Insights reports two kinds of data, and they answer different questions.
Field data comes from real users. It is "powered by the Chrome User Experience Report (CrUX) dataset" (Google), gathered from real Chrome visits over the previous 28 days.
Lab data comes from a single simulated test. Google runs Lighthouse in a controlled environment, which is useful for debugging but is not how real users experience the page.
A brand-new page has no field data yet, so it shows lab estimates only until real traffic arrives. That is normal, not a fault.
The Core Web Vitals assessment: passed or failed
The line that matters most is the Core Web Vitals Assessment, marked Passed or Failed. It grades three real-user metrics against fixed thresholds.
| Metric | Measures | Good | Poor |
|---|---|---|---|
| LCP (Largest Contentful Paint) | Loading | ≤ 2.5 s | > 4.0 s |
| INP (Interaction to Next Paint) | Responsiveness | ≤ 200 ms | > 500 ms |
| CLS (Cumulative Layout Shift) | Visual stability | ≤ 0.1 | > 0.25 |
Source: thresholds from web.dev, Google's performance guide; each metric is assessed at the 75th percentile of visits. The middle band between good and poor is labeled "needs improvement."
A page passes only when all three hit the good threshold for at least 75% of visits. Miss one, and the whole page is marked failed. Our Core Web Vitals guide explains what each metric means.
The performance score, 0 to 100
The big colored number is a lab score, not the field verdict. It comes from Lighthouse, so treat it as a diagnostic, not your grade from real users.
The same PageSpeed Insights documentation puts 90 or above as good, 50 to 89 as needs improvement, and below 50 as poor. Green is 90-plus, orange is the middle, and red is poor.
Do not chase a perfect 100. A page can score 95 in the lab yet still fail its field data (an illustrative example, not a client result). Real phones on mobile data often load a page far slower than the test machine does.
Opportunities and diagnostics: what to fix first
Below the scores sits a list of specific fixes, ranked by how much load time each could save. Start at the top of that list.
In our experience auditing Malaysian SME sites, the biggest offender is almost always an uncompressed hero image, a 2 MB photo where a 150 KB one would look identical. After images come render-blocking scripts, late-loading fonts, and content that pops in and shifts the layout.
These are the usual causes of a slow site, and PageSpeed Insights just puts numbers on them.
Fix the top item, then re-test. Field data moves on a rolling 28-day window, so a real improvement takes about four weeks to show in full.
Speed also feeds discovery beyond Google. Crawlers behind AI Overviews fetch fast, stable pages more reliably, so the work pays off in more than one place.
Frequently asked questions
Should I aim for a 100 score?
No. A score of 90 or above already counts as good, and the climb from 90 to 100 often means chasing tiny gains that no visitor would notice. Your effort is better spent making sure the real-user field data passes, since that reflects how actual customers experience the page. Treat 100 as a nice extra, never the goal.
Why do I have no field data?
Your page has not yet had enough real Chrome visitors for Google to report on. Field data comes from the Chrome User Experience Report, which needs a certain volume of real traffic before it will show. New pages, low-traffic pages, and pages behind a login often show lab data only. Once real visits build up, the field section appears on its own.
Does a low PageSpeed score hurt my Google ranking?
The lab score itself is not a ranking factor, but the real-user Core Web Vitals behind it are a signal. Google uses page experience as one input among many, so a page that genuinely frustrates visitors can be at a slight disadvantage against an equally useful, faster page. Content relevance still decides most of how you rank in Malaysia. A poor score is worth fixing for your visitors first, and for the organic traffic you already have.
How often does the data update?
Field data updates on a rolling 28-day window, so a fix you ship today takes roughly four weeks to show in full. The lab score, by contrast, is recalculated every time you run the test. This is why a repair can look instant in the lab while the field assessment still reports the old experience for a few weeks.
Is PageSpeed Insights the same as Lighthouse?
Not quite. PageSpeed Insights runs Lighthouse to produce its lab score, then adds the real-user field data from CrUX that Lighthouse alone does not have. Lighthouse also lives inside Chrome DevTools for local testing. Think of PageSpeed Insights as Lighthouse plus real-world data, in one web page you can run against any URL.
Should I test on mobile or desktop?
Test mobile first. Google indexes most sites from their mobile version, and for Malaysian businesses the bulk of traffic arrives on phones, often on mobile data. A page that passes comfortably on a fast office desktop can still fail on a mid-range Android phone. PageSpeed Insights lets you switch between the two tabs, so check both, but treat mobile as the one that counts.
Reading the report so it leads to a fix
Storming Solutions builds and maintains websites for Malaysian businesses. We read PageSpeed Insights the way it is meant to be read: field data first, score second, then the fix list in order. We would rather cut a page's load time for real visitors than polish a lab number nobody feels.
Staring at a red score and unsure what it is telling you? Send us the URL on WhatsApp, or check the broader signs your site needs work. Our web development service treats speed as build-time work, so the scores start green and stay there.