Somewhere in a Slack channel, a founder is staring at a green Lighthouse score like it is a participation trophy. Nice. Meanwhile the landing page still feels like it woke up slowly and asked for coffee before accepting the first click.
Core Web Vitals matter because humans matter. Google’s LCP, INP, and CLS are just names for three very human questions: Did it appear? Did it respond? Did it jump around like a startled cat?
What each vital is secretly asking
LCP: “Show me the thing I came for”
Largest Contentful Paint is usually your hero headline, product shot, or big card. If that takes forever, people bounce before your value prop gets a hearing. On SaaS and Web3 marketing pages, LCP problems are often image weight, late fonts, or a giant JS bundle escorting a single paragraph onto the screen.
Fix first: compress and size the hero media, preload the critical font, and stop making the headline wait on a chat widget.
INP: “Stop ignoring my thumb”
Interaction to Next Paint replaced FID for a reason. Real users click menus, tabs, and CTAs while your main thread is still sorting out seventeen analytics scripts and a Carousel Of Ambition.
Fix first: defer non-critical JS, split heavy client components, and keep the primary CTA path embarrassingly light.
CLS: “Do not move the button I was about to press”
Cumulative Layout Shift is the silent conversion thief. Banner loads, font swaps, late images: the book now button becomes the footer. Magical if you enjoy rage. Terrible if you enjoy money.
Fix first: reserve space with width/height or aspect-ratio, use font-display with care, and never inject late UI above the fold without a reservation.
A practical triage for busy founders
- Measure on a real mid-range phone, not only your MacBook in airplane mode optimism.
- Open Search Console field data if you have it. Lab scores flatter. Field data tattles.
- Fix LCP and CLS on the landing page first. That is where ad spend lands.
- Then trim main-thread work around the primary interaction: connect wallet, book call, start trial.
What I build for
When I ship frontends for startups and product teams, I treat Core Web Vitals as product requirements, not SEO seasoning. Pretty pages that stutter lose trust in the same breath they ask for it.
If your ads are healthy and your page feels sluggish, that gap is usually cheaper to close than another creative test.