What is PageSpeed Insights?
PageSpeed Insights is Google's free tool for measuring how fast a page loads. It shows two very different kinds of data on one screen, and mixing them up causes most of the confusion:
- Field data (the top section) comes from real Chrome users who visited your site over the last 28 days, through the Chrome User Experience Report. This is what Google uses for Core Web Vitals in search.
- Lab data (the score out of 100) comes from Lighthouse running one simulated test, right now, on an emulated mid-range phone with a throttled connection.
What is a good PageSpeed Insights score?
- 90 to 100: good (green).
- 50 to 89: needs improvement (orange).
- 0 to 49: poor (red).
But the score out of 100 is a lab number. Google's ranking signal uses your field Core Web Vitals, not the Lighthouse score. A site can score 65 in the lab and still pass Core Web Vitals for real users.
Why does my PageSpeed score change every time I test?
Each lab test is a fresh run, and lots of things vary between runs:
- How quickly your server responds at that moment.
- Third-party scripts, adverts and chat widgets that load differently each time.
- Whether your CDN cache was warm or cold.
- Load on Google's own test machines.
- Updates to Lighthouse, which can shift scores across the board overnight.
A swing of a few points either way is normal. Run the test three to five times and use the middle result, and compare before and after with the same tool and settings.
Why is my mobile score so much lower than desktop?
The mobile test deliberately simulates a slower phone on a slower connection, because that's what many real visitors use. Your page has to do the same work with less processing power and less bandwidth, so heavy JavaScript and large images hurt far more. A low mobile score with a good desktop score is very common.
Lab data vs field data: which matters more?
Field data tells you what real visitors experience and is what Google uses for ranking. Lab data is a repeatable test that's best for diagnosing problems and checking fixes. When they disagree, trust the field data for 'how are we doing?' and use the lab data for 'what should we fix?'
If the field section says there isn't enough data, your site doesn't get enough Chrome traffic to be included yet. That's normal for smaller sites. Use the lab data as your guide.
What does 'Core Web Vitals Assessment: Failed' mean?
At least one of the three Core Web Vitals, measured at the 75th percentile of real visits, is outside the 'good' range. The thresholds are LCP (loading) at 2.5 seconds or less, INP (responsiveness) at 200 milliseconds or less, and CLS (visual stability) at 0.1 or less. Read Core Web Vitals explained for what each one means.
How is the performance score calculated?
The lab score is a weighted mix of five metrics. In current versions of Lighthouse, Total Blocking Time and Largest Contentful Paint carry the most weight, followed by Cumulative Layout Shift, with First Contentful Paint and Speed Index making up the rest. That's why heavy JavaScript, which drives Total Blocking Time, drags mobile scores down so much.
How to improve your PageSpeed score
- Resize and compress images, and serve WebP or AVIF. See our image optimisation guide.
- Lazy-load images below the fold, but never the main hero image. See lazy loading.
- Turn on Gzip or Brotli compression. See enabling compression.
- Cut server response time with caching and better hosting. See slow server response.
- Remove third-party scripts you don't need, and delay the rest until after the page loads.
- Set width and height on images and embeds so the layout doesn't jump.
Frequently asked questions
Is a PageSpeed score of 50 bad?
It means there's room to improve, especially on mobile. But the score is a lab test. What matters for Google rankings is whether your real-user Core Web Vitals pass.
Does PageSpeed Insights score affect SEO?
Not directly. Google uses Core Web Vitals from real users as a ranking signal, not the Lighthouse score. Improving the score usually improves real-user speed too, though.
Why is my field data different from my lab data?
Field data reflects many real visitors on different devices and networks over 28 days. Lab data is one simulated test on a slow phone. Real visitors are often faster than the lab simulation.
Why does PageSpeed Insights say there isn't enough real-world data?
Your site doesn't yet have enough Chrome visitors to be included in the Chrome User Experience Report. Use the lab data instead, or measure real users with your own monitoring.
Can I get a 100 PageSpeed score on mobile?
Yes, on a simple, lightweight page. Sites with adverts, chat widgets and lots of tracking scripts rarely reach it, and usually don't need to.