Web Performance

Core Web Vitals, Explained Without the Jargon

Core Web Vitals are the three performance metrics Google uses as a ranking signal and, more importantly, as a proxy for whether a real visitor had a good experience on your page. There are three of them, and each measures something different: Largest Contentful Paint (LCP), Interaction to Next Paint (INP), and Cumulative Layout Shift (CLS).

LCP measures how long it takes the largest visible element — usually a hero image, a heading, or a banner — to render. Google considers 2.5 seconds or faster 'good.' The most common cause of a slow LCP isn't server speed, it's an unoptimized hero image: a 4MB PNG served at full resolution to a phone on a mobile connection will blow the budget by itself, regardless of how fast your server responds.

INP replaced First Input Delay as the official responsiveness metric in March 2024. It measures the time between a user interaction — a click, a tap, a keypress — and the next time the browser can actually paint the result on screen, sampled across every interaction on the page, not just the first one. Good is 200 milliseconds or under. Heavy client-side JavaScript that blocks the main thread is the usual culprit, which is one reason we default to server-rendered pages and only add client-side interactivity where the page actually needs it.

CLS measures visual stability: how much content jumps around as a page loads. It's caused by images without reserved dimensions, web fonts that swap in and reflow text, or ads and embeds injected above existing content. Good is a CLS score under 0.1. It's usually the cheapest of the three to fix — reserving space for images and embeds before they load eliminates most of it — and one of the most visible to a real visitor, since a shifting page is what makes a site feel cheap even when it 'works.'

None of these numbers are pass/fail gates that flip a switch in rankings. They're one input among many, and a page can rank well with mediocre vitals if everything else about it is strong. But they're also the rare SEO factor you can measure directly, for free, in Chrome's own DevTools or via Google's PageSpeed Insights — so there's no excuse for not knowing where a site actually stands before guessing at fixes.