What is Core Web Vitals?
In 2020 Google took a long list of possible performance measures and chose three, named them Core Web Vitals and said they would count toward ranking. Largest Contentful Paint (LCP) measures how long until the biggest visible element, usually the hero image or headline, has rendered; the target is under 2.5 seconds. Interaction to Next Paint (INP), which replaced First Input Delay in 2024, measures how long the page takes to respond when someone taps or clicks; the target is under 200 milliseconds. Cumulative Layout Shift (CLS) measures how much the content jumps around while loading; the target is under 0.1.
They are measured from real visitors, collected by Chrome and reported in Search Console and PageSpeed Insights as field data, which is why a lab test can pass while the real-world score fails.
Why it matters
Why Core Web Vitals matter
They are a ranking signal, a modest one, applied as a tiebreaker among pages of similar relevance. More importantly they are a description of whether the site is pleasant to use on a phone with an ordinary connection. A brand site built around a full-screen film and heavy campaign imagery is exactly the kind of page that fails LCP, and a page that fails LCP is a page that a share of visitors leave before the film starts.
They also give marketing and engineering a shared language. "The site feels slow" is an argument; "LCP is 4.1 seconds on mobile and the hero image is the element" is a ticket.
In practice
How Core Web Vitals work in practice
- 01
Measure in the field first. Search Console's Core Web Vitals report groups URLs by pass or fail; PageSpeed Insights shows the field numbers for a single page alongside a lab run that explains what is slow. The usual causes are the usual suspects: oversized images and fonts for LCP, heavy third-party scripts and long JavaScript tasks for INP, images and embeds without reserved dimensions, late-loading fonts and injected banners for CLS.
- 02
Fix in order of impact. Serve the hero image in a modern format at the right size and preload it; defer or remove scripts that are not needed for the first screen; give every image, embed and ad slot explicit dimensions; load fonts in a way that does not swap the layout. Re-measure after each change, because the field data lags by a month and the lab is the fast feedback.
Watch out
What people get wrong about Core Web Vitals
Chasing a perfect lab score. Lighthouse's 100 is a synthetic result on a synthetic device, and Google ranks on the field data. A page can score 60 in the lab and pass in the field, or the reverse. The other mistake is expecting rankings to jump when the vitals pass: they are a tiebreaker, and the business case is bounce rate and conversion, which move far more.
Treating it as a one-time project. Every new plugin, tracking script, font and campaign embed is a chance to fail again, and sites that passed last year quietly fail this year because nobody is watching the report.
