The three metrics, in plain language
Core Web Vitals are Google's attempt to measure whether a page feels fast to a person, rather than whether it is technically fast.
- LCP — how long until the biggest thing on screen appears. Usually a hero image or headline
- INP — how quickly the page responds when someone taps or clicks something
- CLS — how much the layout jumps around while loading, the reason you tap the wrong thing
Lab scores versus field data
Running a page speed test gives you a lab score: one simulated load on a simulated connection. It is useful for diagnosis and close to useless as a report card.
What Google uses is field data — real Chrome users on real devices and networks. A site can score well in the lab and fail in the field, particularly when a large share of visitors are on mobile data rather than fibre.
Look at the field data section of a PageSpeed report, or the Core Web Vitals report in Search Console. If there is not enough traffic for field data yet, that is worth knowing too.
What usually causes a bad LCP
In practice it is almost always one of a short list:
- An enormous hero image served at full resolution to a phone
- Web fonts that block text from rendering until they download
- A carousel or video that has to load before anything appears
- Render-blocking scripts in the head — analytics and chat widgets are the usual suspects
- Slow server response, which no front-end work can compensate for
CLS is almost always images without dimensions
If an image has no width and height, the browser does not reserve space for it, so everything below jumps when it loads. Add the attributes and most CLS problems disappear.
The other frequent cause is a banner or notification injected at the top of the page after render, pushing the content down as the reader starts reading.
INP, and why third-party scripts dominate it
INP measures responsiveness to interaction. The usual culprit is not your own code but the accumulated weight of tag managers, chat widgets, heatmap tools and pixels all competing for the main thread.
Audit what is actually loading. Most sites carry at least one script nobody remembers adding and nobody looks at the output of.
How much does this affect ranking
Less than the amount of attention it receives. Page experience is a real but modest signal, and it will not lift a page above a materially better result.
It matters most as a tiebreaker between comparable pages — and considerably more for conversion than for ranking. A slow page loses people before Google is involved at all.
Common questions
Do I need a perfect 100 score?
No. Passing the three thresholds on field data is the goal. Chasing the last few lab points is usually the least valuable work available to you.
Will a faster site rank better immediately?
Not by itself. Fix speed because it improves conversion and stops it being a handicap, not as a ranking tactic.