They are part of how pages get ranked. More usefully, they are the three things a visitor actually notices: whether the page shows up, whether it answers a tap, and whether it moves under your thumb while you are reading it. Google scores them. Visitors just leave.
- Largest Contentful Paint, loading: how long until the biggest thing on screen has drawn. Good is under 2.5 seconds. Usually your hero image or headline.
- Interaction to Next Paint, responsiveness: how long after a tap before the page visibly responds. Good is under 200 milliseconds.
- Cumulative Layout Shift, visual stability: how much the page jumps around while it loads. Good is under 0.1. This is the one where you reach for a button and something above it arrives and pushes it out of the way.
Fix layout shift first.
Page speed scores are not the whole story. Get into the 75 to 80 range, then stop arguing about them.
Layout shift is the worst of the three, the most irritating, and the cheapest to fix. Nearly all of it is images and embeds with no reserved space: the browser does not know how tall they are until they land, so everything below them jumps when they do.
Width and height on every image, a reserved box for every embed, fonts that do not resize when they swap in. That is most of the score, and it is an afternoon of work.
Before you chase a bad mobile score, re-run it with real throttling instead of the simulated default. Simulated runs charge third-party embeds to your first paint, and I have watched one report 8.8 seconds on a page that measured 1.6 for real. Check the number is real before you spend a week on it.
What I check on every launch.
- Every image carries width and height, plus a global rule setting height to auto.
- The largest image above the fold is preloaded or marked high priority. Everything below the fold is lazy loaded.
- Fonts self-hosted as WOFF2 with font-display swap, loaded from the head.
- Third-party scripts async, and embeds loaded after the page load event.
- Measured on a real device on mobile data, not only in a desktop browser.
The trade nobody mentions.
Analytics and session replay cost you performance. On Pigeon’s Mission, adding GA4 and Clarity took the performance score from 75 into the high 60s, Speed Index from 3.5 to 7.8 seconds, and best practices from 100 to 79. That can still be a perfectly good trade. Make it knowingly, by measuring before and after, so the client hears the price from you rather than finding it in a Core Web Vitals report six months later.
Core Web Vitals are section 10 of the 5Four Launch Checklist, and a real-throttled mobile measurement is part of every launch I run.
Related: Core Web Vitals in the glossary, the Site System, 5Four® Run.
Designer, developer and educator in Miami. I run strategy and creative direction at 5Four Digital, a web design and development agency that leans on AI. I have spoken at AIGA, Adobe MAX and Webflow Conf, and I put the whole process on YouTube.
