Lab data
A controlled measurement under fixed conditions. Reproducible, which makes it good for comparing one change against another.
It is not what your visitors experienced.
PageSpeed
One number is a poor description of a page.
Performance scores are compelling precisely because they compress something complicated into a single figure. PageSpeed gives you the diagnostics underneath — per URL, in both lab and field conditions, and comparable across runs.
No card required.
- Lighthouse and PSI
- Lab and field data
- Comparable across runs
PageSpeed Per-URL diagnostics
Two kinds of data
Lab and field measure different things, and they will disagree.
Treating these as interchangeable is the single most common way performance evidence gets misread. They answer different questions and both are worth having.
Field data
What was actually observed for real sessions, across real devices and connections, where available.
It reflects reality, but moves for reasons that have nothing to do with your code.
A lab improvement that never shows up in field data is worth investigating rather than celebrating.
What gets measured
Core Web Vitals, plus the diagnostics that explain them.
The vitals describe symptoms. The diagnostics underneath are where the cause usually is, which is why the score alone rarely tells you what to change.
-
LCP
Largest Contentful Paint
How long before the main content is actually there.
-
INP
Interaction to Next Paint
How quickly the page responds when someone tries to use it.
-
CLS
Cumulative Layout Shift
How much the page moves around while it is settling.
One run is a snapshot
Read the trend, not the reading.
Any individual run is subject to conditions that have nothing to do with your site — network variance, server load, what else was happening at that moment. Comparing across runs is what separates an actual change from noise, which is why runs are kept rather than overwritten.
What performance does and does not do
A faster page is a better page. It is not a visibility strategy.
Performance affects whether a page can be retrieved efficiently and whether people stay on it. It does not follow that improving a score will change how answer engines treat you, and we are not going to suggest it does. Treat performance as worth doing on its own terms, and address visibility on its own terms too.
What PageSpeed connects to
Performance is one technical condition among several that affect how a site is treated.
Get the diagnostics, not just the score.
Run Lighthouse and PSI against the URLs that matter and compare them over time.
Want help reading lab against field? Talk to us