Your PageSpeed Score Is Green. Your Site Still Feels Slow. Here's Why.

A green PageSpeed Insights score measures lab data: one simulated page load on a mid-tier phone (a Moto G4) over a fixed, throttled connection. It says nothing about how your actual visitors, on their actual phones and networks, experience your actual pages. That's field data, a different report living further down the same page, and it's the one that decides whether you pass Core Web Vitals for real.

Lab data and field data are not the same test

Google's own documentation for PageSpeed Insights is direct about this: lab data is "useful for debugging issues, as it is collected in a controlled environment," but "may not capture real-world bottlenecks." Field data, by contrast, comes from the Chrome User Experience Report (CrUX), anonymised real-world data from actual Chrome users on whatever device and network they happen to be using. The same page explains plainly why the two disagree: "the lab data is based on a simulated load of a page on a single device and fixed set of network conditions," while field data is a rolling historical report of how real visits actually went.

Four concrete reasons the two disagree

• Real devices and networks vary; the lab test doesn't. If a meaningful share of your visitors are on entry-level phones or patchy mobile data, and the lab test isn't, your field data will fail even while your lab score looks clean.
• You probably tested the wrong page. A single lab run is usually the homepage. Real traffic hits your product pages, your category pages, your blog posts, often heavier, more cluttered pages that were never the one you screenshotted.
• Field data is a rolling 28-day average. A fix shipped this morning will not move your field data today. Give it time before deciding it didn't work, and don't re-test the lab score obsessively as a proxy for something it can't measure.
• The 0-100 score is a weighted composite, not a Core Web Vitals pass/fail. It blends several lab metrics with different weightings, so a page can score comfortably in the green while still individually failing a specific real-world Core Web Vital. INP in particular can't be measured in a lab run at all: it needs a real interaction from a real visitor.

What to check instead of chasing the number

Open the Core Web Vitals report in Google Search Console. It reads directly from CrUX field data and groups your actual URLs as good, needs improvement, or poor, which is the same signal that factors into ranking. The lab-only 0-100 number itself is not a ranking signal; the Core Web Vitals thresholds measured from field data are. If Search Console shows a chunk of URLs sitting in "poor" while your last PageSpeed run was all green, trust Search Console. It's reading real visits, not a single simulated one.

The takeaway

A green PageSpeed Insights score is a useful debugging signal, not proof your site is fast for the people actually using it. It's testing one device, one network, one page, once. Your visitors are testing all of it, every day, on whatever they happen to be holding. If your Search Console field data and your lab score disagree, that gap is usually where the real story is, and it's exactly what a Core Web Vitals audit is built to find.