Core Web Vitals for non-technical founders

If you've ever sat in a meeting where someone said "we need to fix Core Web Vitals" and you nodded politely without knowing what they meant — this is the explainer I wish existed for that moment. No jargon, no Lighthouse screenshots. Just the three numbers that matter, what they cost you when they're bad, and the handful of decisions that fix them.

The three numbers

Google measures three things on every public page. They're called Largest Contentful Paint, Interaction to Next Paint, and Cumulative Layout Shift — but you can think of them as: how fast does it look done, how fast does it react, and does anything jump around.

LCP — does it look done? This is the moment the biggest visible thing on the page (usually a hero image or headline) finishes loading. Under 2.5 seconds is "good." Above 4 seconds is "poor." On mobile networks, this is what separates a site that feels instant from one that feels broken.

INP — does it react? When a user taps a button, how long until something happens? Under 200 milliseconds feels instant. Above 500 milliseconds and people start tapping again, thinking the site is frozen. Then they bounce.

CLS — does it jump? You've felt this. You're about to tap a link, then an ad loads above it, the link moves, and you tap the wrong thing. CLS measures how much your layout shifts after it first appears. Under 0.1 is good. Above 0.25 is what makes people physically angry at your site.

The conversion data

This isn't theoretical. There is now a decade of public data showing what these numbers cost.

• Walmart published that every 100 ms of load time improvement was worth roughly 1% in incremental revenue.
• Pinterest cut perceived wait times by 40% and saw a 15% lift in sign-ups and search engine traffic.
• Vodafone improved LCP by 31% and saw an 8% increase in sales.
• A study across 2,000 sites by Google in 2023 found a 24% reduction in page abandonment when sites met all three Core Web Vitals thresholds.

If you sell anything online, treat speed as a revenue lever, not a technical chore. The math is consistent.

What actually causes bad scores

Almost every poor Core Web Vitals score on a WordPress site comes from one of five sources:

1. **Hero images that aren't optimized.** A 4 MB photo where 200 KB would do. Modern themes serve WebP or AVIF automatically — older themes serve the original JPEG.
2. **Fonts loaded synchronously.** When the browser waits for a custom font to download before rendering text, LCP suffers. Themes that use `font-display: swap` and self-host their fonts avoid this.
3. **Ad and chat scripts loading too early.** Intercom, Drift, ad networks — they all add hundreds of milliseconds to INP. The fix is loading them after the page is interactive, not during.
4. **Layout that depends on JavaScript.** If your hero section is built and positioned by JavaScript instead of CSS, the page jumps when the JS finishes. CLS spikes.
5. **Plugins doing too much on every page.** A page builder that loads its entire library on the contact page, even though the contact page is a form, is a common offender.

Notice that none of these are "bad code." They're all design and architecture decisions. Which is why your theme matters more than your hosting.

What our themes do differently

We treat Core Web Vitals as a release blocker, not a marketing checkbox. Every theme we ship has to pass these gates on the demo site, on a 4G mobile simulation, with no caching, on the second visit (not the first):

• LCP under 1.8 s
• INP under 150 ms
• CLS under 0.05

Three things make that possible:

We ship lean. Each theme has its own bundle, not a shared "framework" loaded across thirty themes. If a feature isn't used, the code isn't loaded. The default homepage of every theme is under 90 KB of JavaScript.

Images are deferred properly. Above-the-fold images get `fetchpriority="high"` and explicit dimensions. Below-the-fold images get native lazy loading with intersection observers as a fallback. The browser knows what to download first.

Fonts are self-hosted with size adjustments. No Google Fonts CDN call. No FOUT (flash of unstyled text). The first paint shows the right typography because the font is already there.

The five-minute audit you can do today

Open your site in an incognito window on your phone, on cellular (not Wi-Fi). Visit a typical product or article page. Time three things by feel:

1. How long until you see the headline? If it's longer than two breaths, your LCP is bad.
2. Tap the menu. Does it open instantly? If not, your INP is bad.
3. Scroll past the hero. Does anything below it shift? If it does, your CLS is bad.

You don't need a tool to know whether your site has a problem. You need it to confirm what your gut already told you. The good news: the fixes are almost always at the theme level. Pick a theme that takes performance seriously, and you'll spend the rest of your time on the parts of the site that are actually your job.