Core Web Vitals Impact Conversion Rates: Data from 200+ SMBs

We pulled performance data from 200 of our SMB clients across ecommerce, services, and B2B verticals. The question we needed answered: does fixing Core Web Vitals actually move the needle on revenue? The answer is yes—and the correlation is stronger than most agencies claim. Sites that improved Largest Contentful Paint (LCP) from 4.2 seconds to under 2.5 seconds saw average conversion lift of 18%. For Cumulative Layout Shift (CLS), reducing score from 0.15 to under 0.1 added 12% to checkout completion. Interaction to Next Paint (INP) improvements showed similar gains. But here's what matters: these aren't theoretical benchmarks. These are real revenue changes we tracked over 90 days.

The 15-23% Conversion Loss We Keep Seeing

Mobile conversion rates drop noticeably when LCP exceeds 3 seconds. We compared two furniture retailers with identical traffic: one had LCP at 2.1 seconds, the other at 4.8 seconds. Same product, same pricing, same traffic source. The slower site lost 19% of mobile conversions. That's not a guess—that's Google Analytics 4 data over three months. The faster site converted 3.2% of mobile visitors. The slower site converted 2.6%. Over 50,000 monthly mobile visitors, that difference equals roughly $12,000 in lost monthly revenue for an average order value of $185.

CLS matters more than most people realize because it directly causes accidental clicks. We tracked a B2B SaaS demo request form where CLS was 0.28—terrible. Users clicked "Request Demo" but the button shifted as images loaded, and they hit a text link instead. Form abandonment was 34%. After reducing CLS to 0.08 by lazy-loading images and reserving space for ads, abandonment dropped to 18%. That's a 47% improvement in form completion from one metric.

Which Vital Hits Revenue First

• LCP (Largest Contentful Paint) — biggest impact on initial perception and mobile bounce. Improving from 4+ seconds to under 2.5 seconds typically adds 15-22% conversion lift.
• CLS (Cumulative Layout Shift) — directly causes accidental clicks and user frustration. Every 0.1 point reduction in CLS we've seen adds 2-4% conversion improvement.
• INP (Interaction to Next Paint) — slowness here kills form fills and checkout flow. Improvements from 300ms to under 100ms add 8-14% to multi-step conversions.
• First Input Delay (FID) — less critical now that INP replaced it, but still impacts perceived responsiveness on older devices.

We had a 2.8-second LCP and blamed traffic quality. After we fixed it to 1.9 seconds, same traffic source converted 21% better. The traffic was fine. Our page was the problem.

How to Prioritize Your Fixes

Start with LCP because it has the broadest impact. Use PageSpeed Insights to identify what's slow—usually large images, render-blocking JavaScript, or slow server response. For most SMBs, image optimization alone (using WebP format, lazy loading, proper sizing) drops LCP by 1-2 seconds. We saw a landscaping company reduce LCP from 3.7 to 2.1 seconds just by compressing hero images and lazy-loading gallery photos. Their mobile conversion rate jumped from 2.1% to 2.7%.

For CLS, audit layout shifts using Chrome DevTools. Common culprits: ads loading after content, web fonts changing size mid-render, embedded videos without aspect ratio. Reserve space for known content before it loads. A tax preparation firm had CLS 0.22 because their client testimonial images loaded progressively. By adding `aspect-ratio: 4/3` to the image container and lazy-loading with placeholder color, CLS dropped to 0.06. Mobile form completion improved 9 percentage points.

The Tools We Use to Track This

• Google PageSpeed Insights — free, real-world data from Chrome UX Report, pinpoints exact issues
• Vercel Speed Insights or similar monitoring tool — tracks Core Web Vitals over time, alerts on degradation
• Google Analytics 4 — segment conversion rate by device and compare to Core Web Vitals in Search Console
• Chrome DevTools Lighthouse — local testing before deployment, shows filmstrip of page load
• WebPageTest.org — advanced waterfall analysis, real browser testing from multiple locations

We set up automated monitoring for all clients now. When LCP creeps above 2.5 seconds, we get alerted. When a code deployment tanks CLS, we know within hours. One client pushed an ad network update that raised CLS to 0.19—we caught it the same day, rolled it back, and implemented a solution that kept CLS under 0.08. Without monitoring, they would have lost 8-10% conversion for days.

← Back to all articles