What it is
Core Web Vitals are Google's page-experience metrics. They are a real ranking signal, but more importantly they correlate with whether visitors stay or leave. This capability measures them against Google's current thresholds and tracks them over time.
The metrics
| Metric | What it measures and the good threshold |
|---|---|
| LCP — Largest Contentful Paint | How long until the main content is visible. Good is 2.5s or less. |
| INP — Interaction to Next Paint | How quickly the page responds to input. Good is 200ms or less. Replaced FID in March 2024. |
| CLS — Cumulative Layout Shift | How much the layout jumps while loading. Good is 0.1 or less. |
| TTFB — Time to First Byte | Server responsiveness. Good is 800ms or less. |
INP, not FID. A surprising number of tools still report FID, which was retired in March 2024. Reporting a dead metric is worse than reporting none, because it looks like coverage. This platform tracks INP.
How it works
The crawler captures lightweight render timing while it fetches each page, so vitals are estimated without needing a real browser, a field-data API or a paid key. That keeps the whole feature working offline — useful in CI, on an air-gapped install, or when you simply do not want a third-party script on your site.
These are estimates, not measurements. The values are proxies derived
from crawl data — render time, document weight and DOM complexity — not real-user
field data (CrUX) and not a Lighthouse run. They are labelled with
source: "lab" in the API. Use them to spot pages that are getting
heavier and to compare pages against each other; do not present them as the
Core Web Vitals scores Google reports for your URLs. For the authoritative figure,
read the CrUX or Search Console data for the same URL.
Because the crawler already stores a history of crawls, vitals become a trend rather than a snapshot. You can see the day a deploy made LCP worse, which is the only way to actually prevent regressions.
What it does for your site
Speed problems are usually concentrated: a handful of templates account for most of the slow pages. Measuring per template turns "the site feels slow" into "the product template needs its hero image optimised and this script deferred", which is a fix someone can schedule and verify.
And because a regression is caught by a scheduled micro-audit, you find out from an alert rather than from a drop in conversions.
Where the results go
Vitals feed the on-page audit score, and breaches raise alerts. The timing data also becomes a ranking-factor input for the SERP simulator.
Related capabilities
See it on your own site
Point the crawler at your domain and you will have a full audit, a health score and a prioritised fix list in minutes.
Open the dashboard All features