Site speed gets treated like a developer's pet project, something to tidy up after the real work is done. That framing is backwards. A slow page is a leak in the funnel: people who would have bought, signed up, or filled out the form get tired of waiting and leave before they ever see your offer. Core Web Vitals are Google's attempt to measure that experience in numbers, and they line up closely with the things that actually move revenue.
The useful part is that performance is an engineering problem with known causes and known fixes. You don't need to guess. You measure, you find the bottleneck, you fix it, and the page gets faster. Here is how the metrics work and what to do about them.
What the three metrics actually measure
Core Web Vitals are three numbers, each tied to a moment in the page-loading experience.
- LCP (Largest Contentful Paint), how long until the biggest visible thing finishes loading. Usually a hero image, a headline block, or a banner. This is the user's gut sense of "is this page up yet." Google considers under 2.5 seconds good.
- INP (Interaction to Next Paint), when someone taps or clicks, how long before the page visibly responds. It replaced the older FID metric and is stricter, because it looks at interactions across the whole visit, not just the first. Under 200 milliseconds is the target.
- CLS (Cumulative Layout Shift), how much the page jumps around while loading. You've felt this: you go to tap a button, an ad or image loads above it, and the whole layout shoves down so you hit the wrong thing. Under 0.1 is good.
LCP is about waiting. INP is about responsiveness. CLS is about stability. A page can pass one and fail the others, so you have to look at all three.
Why this is a revenue conversation, not a tech one
There are two channels through which speed touches the money.
The first is rankings. Core Web Vitals are a confirmed Google ranking signal. They are not the biggest signal, relevant content and good links matter more, but they're a real tiebreaker. When two pages are close on everything else, the faster one tends to win the better position. Better position means more clicks, and more clicks mean more people in the funnel.
The second is conversion, and this one is bigger. Every second a user waits is a chance for them to lose interest and close the tab. This is intuitive and well documented across the industry: slower pages bounce more and convert less. The relationship isn't always linear, but the direction is consistent, faster pages keep more people, and more people kept means more conversions from the same traffic.
That last point is what makes performance work pay for itself. You are already paying to bring people to the page through ads, SEO, and content. Speed determines how many of them survive long enough to convert. Fixing it is one of the few growth levers that improves results without spending more on traffic.
The fixes that move the numbers
Most performance problems trace back to a short list of causes. These are the ones worth checking first.
- Images. Images are usually the heaviest thing on the page and the most common LCP culprit. Serve modern formats like WebP or AVIF, size them to the space they actually occupy instead of shipping a 4000px photo into a 600px slot, and set explicit width and height so the browser reserves space and the layout doesn't jump (that fixes CLS too). Lazy-load images below the fold, but never lazy-load your LCP image, that just makes it slower.
- Fonts. Custom fonts block text from showing or cause a flash and a layout shift when they swap in. Use
font-display: swapso text appears immediately, preload the one or two fonts that matter, and self-host instead of pulling from a third party so you're not waiting on someone else's server. - JavaScript. Heavy JS is the main driver of bad INP. Every script the browser has to parse and run is time the page can't respond to a tap. Cut unused dependencies, split bundles so a page only loads the code it needs, and defer non-critical scripts (analytics, chat widgets, A/B tools) so they don't block the first interaction.
- Rendering strategy. How the page is built decides how much work lands on the user's device. Server-rendering or pre-rendering the initial HTML means the content shows up fast without waiting for a pile of client-side JavaScript to execute first. For marketing and content pages, static generation is usually the right default, it's the fastest thing you can serve.
- Caching and delivery. Serve assets from a CDN so they come from a server near the user, set long cache lifetimes on things that rarely change, and compress everything in transit. This is low-effort and broadly effective, and it's often the first thing missing on a site that "feels slow."
Measure with real users, not just lab tools
A common trap is to run a single test on a fast laptop with good wifi, see green scores, and call it done. Your users aren't on that laptop. They're on mid-range phones and patchy mobile connections, and that's where slow pages actually hurt.
Test on throttled mobile to simulate real conditions. Then watch field data, the numbers Google collects from actual visitors, because that's what feeds the ranking signal and reflects what people genuinely experience. Lab tools like Lighthouse are great for catching specific issues and confirming a fix, but field data tells you whether it mattered in the real world. Use both: lab to diagnose, field to verify.
One more thing worth saying plainly: performance regresses. A site you tuned six months ago picks up a new tracking script, a heavier hero image, an extra font, and slides back into the red. Treat it as something you monitor, not a one-time cleanup. Building speed budgets into how you ship, and into how you measure growth, keeps the gains from quietly eroding.
Takeaway
Core Web Vitals are a clear, honest way to measure something users already feel: whether your site respects their time. LCP, INP, and CLS each map to a real moment of frustration, and each has concrete fixes, lighter images, smarter font loading, less JavaScript, the right rendering strategy, and proper caching. Done well, this is one of the rare improvements that lifts both your rankings and your conversion rate without spending another dollar on traffic. If you're not sure where your bottleneck is, start by measuring on a real phone, and if you'd rather hand the engineering to someone who does this often, let's talk.