Core Web Vitals Foundations
Google turned "the page feels fast" into three numbers you can measure and fix. Core Web Vitals are those numbers. They are not the biggest ranking factor, but they are one of the few Google publishes exact thresholds for, so they are unusually concrete to work on.
The three metrics
Each Core Web Vital captures one part of the loading experience.
- LCP (Largest Contentful Paint) measures loading: when the largest image or text block in the viewport finishes rendering. It marks the moment the page feels loaded.
- INP (Interaction to Next Paint) measures interactivity: how quickly the page responds across nearly all clicks, taps, and keypresses over its life. It reports a value close to the worst interaction, not the average.
- CLS (Cumulative Layout Shift) measures visual stability: how much content jumps around unexpectedly while the page loads.
One important update: INP replaced First Input Delay (FID) on March 12, 2024. FID only timed the delay before the first interaction. INP watches almost all of them and reports close to the worst, which is a far truer read on how responsive a page feels. FID is gone; do not report it.
Thresholds and how Google grades them
| Metric | Measures | Good | Needs work | Poor |
|---|---|---|---|---|
| LCP | Loading | 2.5 s or less | 2.5 to 4.0 s | over 4.0 s |
| INP | Interactivity | 200 ms or less | 200 to 500 ms | over 500 ms |
| CLS | Stability | 0.1 or less | 0.1 to 0.25 | over 0.25 |
All three must be "good" to pass. The grading detail trips people up: Google scores each metric at the 75th percentile of real page loads over a rolling 28-day window, split by mobile and desktop. So a page passes a metric when at least 75% of real visits meet the good threshold. Your fastest loads do not count; the experience of your slower quarter does.
Google grades Core Web Vitals at the 75th percentile over a rolling 28-day window, split mobile and desktop. You are judged by your slower visits, not your best-case demo.
Lab versus field
Two kinds of data measure these metrics, and confusing them wastes hours. Lab data comes from a controlled synthetic test like Lighthouse: repeatable, great for debugging, but one machine on one network. Field data comes from real visitors on real devices; Google's public field source is the Chrome UX Report (CrUX), and it is what the 75th-percentile grade is built from. INP and CLS in particular accumulate over a real session, so lab tools can only estimate them. The working rule: fix in the lab, confirm in the field. Because field data lags 28 days, patience is part of the job. The measurement tooling gets its own chapter. Performance measurement →
Where it sits as a ranking signal
Be honest about weight. Page experience, which includes Core Web Vitals, is a Google ranking signal, but a tiebreaker rather than a headline factor. Great content on a slightly slow page still beats thin content on a fast one. What poor vitals do is cap an otherwise strong page and cost you users who bounce before the content loads. Fixing them is worth doing, in proportion: it will not rescue weak content, and it will stop good content from being held back. Quality as the bigger lever →
The rest of the performance stage traces a page from first byte to largest paint, then to interactivity and stability, and finally to delivery and tooling. This chapter is the map you grade everything against. The loading story next →