Core Web Vitals Explained: LCP, INP and CLS

Core Web Vitals are Google’s three user-experience metrics: LCP for loading, INP for responsiveness, and CLS for visual stability. They are measured on real users, at the 75th percentile, and they matter for ranking as a tiebreaker rather than a primary lever. They also matter for AI visibility, for a reason most guides miss.

The three metrics

LCP: Largest Contentful Paint

How long until the largest visible element, usually a hero image, heading or main text block, finishes rendering. It is the closest proxy for “the page looks loaded”.

INP: Interaction to Next Paint

How quickly the page responds to interactions, measured across the whole visit and reported near the worst case. It replaced First Input Delay in 2024 and is stricter, because a page can feel fine on the first click and terrible on the fifth.

CLS: Cumulative Layout Shift

How much visible content jumps around during load. It is the metric behind tapping a button just as an ad loads and hitting something else entirely.

Common causes and fixes

Fixing LCP

• Slow server response. Reduce time to first byte with caching, a CDN, or static generation.
• Render-blocking CSS and JavaScript. Inline critical CSS, defer the rest.
• Unoptimized hero images. Modern formats, correct dimensions, and preload the LCP image.
• Client-side rendering of the main content. The fix is server-side rendering, which also solves an AI crawler problem covered below.

Fixing INP

• Long JavaScript tasks. Break them up, and defer anything not needed for first interaction.
• Heavy third-party scripts. Tag managers and chat widgets are frequent culprits; audit what each one costs.
• Excessive event handlers firing on scroll or input.

Fixing CLS

• Images and video without dimensions. Always set width and height, or use aspect-ratio.
• Ads and embeds without reserved space.
• Web fonts swapping late. Use font-display sensibly and preload key faces.
• Content injected above existing content after load.

Why speed matters for SEO and crawl

Three separate mechanisms, often conflated.

Ranking. Core Web Vitals are part of Google’s page experience signals. Real, but modest, and easily outweighed by relevance. Do not expect a fast site to outrank a more useful one.

Crawl efficiency. Slow responses reduce how much of your site any crawler fetches in a given visit. That applies to Googlebot and to AI retrieval crawlers alike, and it compounds on large sites.

Conversion. The least discussed and usually the largest effect. Faster, stabler pages convert better, which is worth more than the ranking tiebreaker for most businesses.

The AI angle most guides miss

Core Web Vitals are not an AI citation signal. No engine has said otherwise, and you should be suspicious of anyone claiming they are.

What is true is more useful: the architectural fix for poor LCP is the same fix that makes your content visible to AI crawlers. Sending complete, server-rendered HTML quickly improves LCP and ensures crawlers that do not execute JavaScript actually receive your content. One decision, two payoffs. The crawler side is in JavaScript rendering and AI crawlers.

So treat vitals work as part of the same programme rather than a separate chore, alongside the rest of the technical SEO checklist for AI search.

Measure with field data, not just lab scores

Field data is what real users experienced, collected in the Chrome User Experience Report and surfaced in Search Console’s Core Web Vitals report and PageSpeed Insights. This is what counts.

Lab data, from Lighthouse or WebPageTest, simulates a load on specified hardware and network conditions. Invaluable for diagnosis, useless as a scorecard, because it does not reflect your actual audience’s devices.

A common trap: a perfect Lighthouse score alongside failing field data. That usually means your real users are on slower devices or connections than your test, and the field number is the one that matters.

Two practical notes. The 75th percentile means the metric passes only if roughly three in four real loads pass, so a fast median does not save you. And field data lags by weeks, so a fix made today will not appear in Search Console immediately.

The bottom line on Core Web Vitals

Fix them because they make the site better for the people you are trying to convert, and because the underlying work, fast server responses and real HTML, happens to be exactly what AI crawlers need too. Do not fix them expecting a ranking jump, and do not chase a perfect lab score while your field data fails.

Want to know where your vitals stand with real users, and what is dragging them down? Request an AI visibility audit: performance is part of how we assess whether engines and users can actually use your site, and our web development service builds sites that pass by default.