What Page Speed Insights measures
PageSpeed Insights loads your page in a simulated phone and a simulated desktop browser, then scores what it saw from 0 to 100. The score is built almost entirely from loading speed measurements: how long the biggest thing on screen takes to appear, how much the layout jumps around while it loads, and how quickly the page responds to a tap. When Google has enough real visitors on record for your site, the report also shows field data at the top, which is what actual people experienced over the past 28 days.
How to read your result
Read the field data first if it is there, because that is real and the score below it is a simulation. Green is good, orange is borderline, red needs work. A score in the 90s on desktop and the 50s on mobile is normal and does not mean something is broken; phones are slower and Google tests on a throttled connection. The list of opportunities under the score is ordered by how much time each fix would save, and the estimate is usually optimistic.
What to change first
Start with the Largest Contentful Paint element, which the report names for you. Nine times out of ten it is a hero image that is far larger than it needs to be, loading lazily when it should load immediately, or served as a JPEG when WebP would be half the size. Fixing that one image moves the score more than anything else on the list.
Common questions
- Why is my mobile score so much lower than desktop?
- Google tests mobile on a throttled connection and a mid range phone, so the same page takes several times longer to finish. A 40 point gap between the two scores is common on sites that are perfectly healthy.
- Does the PageSpeed score affect my Google rankings?
- The score itself does not. The field data does, because Core Web Vitals are a ranking signal. A site can score 68 in the lab and still pass Core Web Vitals for real visitors.
- Why does my score change every time I run it?
- The lab test runs on shared Google infrastructure, so a few points of movement between runs is normal. Run it a few times and use the middle result rather than the best one.
Related tests
Run this alongside every other test in Website Speed Tests, or start from the full list of website testing tools.