Core Web Vitals and conversion rate: what actually moves the number
Performance work is easy to sell and easy to overdo. The first second of improvement is worth far more than the fourth, and two of the three Core Web Vitals matter more for revenue than the one teams usually obsess over.
Largest Contentful Paint is the commercial one
On paid landing pages, LCP is where money leaks. A visitor arriving from an ad has no loyalty and a back button. Most LCP problems come down to four causes: an unoptimised hero image, render-blocking fonts, a slow server response, or the hero being lazy-loaded when it should be preloaded.
- Serve the hero as a modern format at the size it actually displays
- Preload the LCP image and never lazy-load it
- Self-host fonts with font-display: swap and subset the character set
- Cache the HTML response at the edge where the page allows it
Interaction to Next Paint is the frustration one
INP measures how long the interface stays frozen after a tap. Poor INP is usually a hydration cost or a third-party script monopolising the main thread. Audit the tag manager before rewriting your own code — chat widgets, session recorders and heat maps are common offenders and are frequently forgotten trials.
Cumulative Layout Shift is the trust one
Shifting layout causes mis-taps on mobile, and a mis-tap on a checkout page is a lost order. Reserve space for images, ads and embeds with explicit dimensions or aspect ratios, and never inject a banner above existing content after load.
Where to stop
Once LCP is comfortably under 2.5 seconds on real mobile data, INP under 200 milliseconds and CLS under 0.1, further optimisation is engineering pride rather than revenue work. Move the budget to offer clarity, proof placement and form design, which at that point have far more upside.