What Is Core Web Vitals? LCP, INP and CLS Explained

Aug 05, 2026 · SEO Glossary

Open Chrome DevTools on any page, run a performance trace, and you'll see hundreds of measurable events: paints, layout passes, script tasks, input handlers. Google's problem was picking a handful of numbers from that firehose that predict whether a real person finds the page pleasant or infuriating. Core Web Vitals is that handful.

So what is Core Web Vitals? It's a set of three user-experience metrics — Largest Contentful Paint, Interaction to Next Paint, and Cumulative Layout Shift — measured from real Chrome users and used by Google both as a public health standard and as a lightweight ranking input. One update matters more than any other: in March 2024, INP replaced First Input Delay. If a guide still lists FID as a Core Web Vital, it's out of date.

The three metrics and their thresholds

A URL passes the assessment when the 75th percentile of real-user measurements is in the "good" range for all three. That percentile choice is deliberate: your page has to be fast for three out of four visitors, including the ones on mid-range Android phones over patchy 4G, not just for you on office fibre.

Each metric has its own failure modes, covered in depth in the dedicated guides to LCP, INP, and CLS.

Field data vs lab data — why the numbers disagree

This distinction causes more confusion than anything else in performance work.

Field data comes from the Chrome User Experience Report (CrUX): anonymized measurements from actual Chrome users who visited your page over the trailing 28 days. This is what the Core Web Vitals assessment uses. It requires enough traffic per URL — low-traffic pages fall back to origin-level data or show nothing.

Lab data comes from a tool like Lighthouse loading your page once under simulated throttling. It's reproducible and great for debugging, but it can't measure INP at all (no real user is interacting) and its LCP often differs wildly from field LCP.

So when PageSpeed Insights shows a Lighthouse score of 95 but a failing field assessment, believe the field data. The lab run tested one synthetic load; the field data averaged thousands of real ones. The reverse also happens — a scary lab score with passing field data usually means your real audience has better devices than Lighthouse's simulated Moto G.

How much Core Web Vitals affect rankings

Honest answer: modestly. Google has described the page experience signals as a tiebreaker-weight factor — relevance and content quality dominate. A brilliant page with a 3.1s LCP will outrank a thin page with a 1.2s LCP on almost any query.

Then why care? Three reasons that show up in real numbers:

• Conversions. The correlation between speed and revenue is far stronger than between speed and rankings. Users abandon slow pages regardless of where they rank.
• Close contests. In competitive SERPs where the top results are all strong on relevance, experience signals can separate positions 3 and 5.
• Crawling. A faster site lets Googlebot fetch more URLs within the same host-load limits — a real effect on large sites.

Diagnosing a failing metric

Start in Search Console's Core Web Vitals report, which groups URLs by shared template — usually your product pages fail together, or your blog posts do, because they share code. Then:

1. Failing LCP: find the LCP element in DevTools (Performance panel labels it). It's usually a hero image or heading. Check whether the image is lazy-loaded (it shouldn't be), whether the server response is slow, and whether render-blocking CSS delays it.
2. Failing INP: field tooling matters here. Long JavaScript tasks blocking the main thread are the usual cause — third-party scripts, heavy hydration, oversized event handlers.
3. Failing CLS: reproduce with DevTools' Layout Shift regions. Images without dimensions, late-loading ads, and web fonts are the usual suspects.

Because failures cluster by template, one fix typically repairs hundreds of URLs. And because slow templates correlate with other technical debt, it's worth running a site-wide audit that surfaces slow pages alongside other crawl issues — the pages failing CWV are often the same ones carrying redirect chains and oversized images.

What Core Web Vitals don't measure

Worth keeping perspective: no CWV metric knows whether your content answers the query, whether your navigation makes sense, or whether your checkout works. A site can pass all three metrics and still be a bad site. Treat the thresholds as a floor for user experience, not a strategy.

Also worth knowing what sits just outside the three: Time to First Byte and First Contentful Paint are officially "supporting" metrics — not part of the assessment, but upstream of it. A 1.8-second TTFB makes a 2.5-second LCP arithmetically impossible, so when all three metrics look mediocre at once, check the server response time first. It's the one number that taxes everything downstream, and it's often the cheapest to fix — caching, a CDN, or a hosting upgrade rather than weeks of front-end work.