Core Web Vitals

Each metric isolates a different felt problem. LCP measures how long the largest visible element, usually a hero image or headline, takes to render. INP measures how quickly the page responds to taps, clicks, and key presses across the whole visit, replacing the older First Input Delay because it captures every interaction rather than just the first. CLS measures how much the layout jumps as assets load. Google publishes target thresholds (LCP under 2.5 seconds, INP under 200 milliseconds, CLS under 0.1) and judges a page on the 75th percentile of field data from real Chrome users, so a handful of fast sessions cannot mask a slow experience for the majority.

The field-data point matters because lab tools such as Lighthouse run on a single simulated device and rarely match what shoppers on mid-range phones and patchy mobile connections actually feel. The Chrome User Experience Report (CrUX) is the source Google trusts, and it lags by roughly 28 days, so a fix made today will not show as passing for several weeks. Operators who optimise only against the lab score often see no movement in Search Console, then assume the work failed.

Core Web Vitals are a confirmed ranking signal, but a modest one. They act mainly as a tie-breaker between pages of similar relevance, so a fast page with thin content will not outrank a slower page that answers the query better. The honest framing is that good vitals protect rankings and conversion rather than win them outright.

Consider a Shopify store on a heavily customised theme whose product pages score a CLS of 0.28, well into the failing band. The cause is a star-rating widget and a recommended-products row that both load after first paint and shove the add-to-cart button down the screen just as a shopper reaches for it. Reserving fixed height for both blocks with a min-height container, and rendering the aggregate rating server-side into the initial HTML, brings CLS under 0.05 without removing any social proof. The add-to-cart button stops moving, mis-taps fall, and INP usually improves too because the main thread is no longer scrambling to reflow the page.

The answer-engine angle is indirect but real. ChatGPT, Perplexity, and Google AI Overviews lean on content their crawlers can fetch and parse cleanly, and a page that renders its key content and structured data quickly and stably is easier to retrieve and quote than one that paints late or shifts under a headless fetch. Vitals are not a direct citation factor, but the same discipline that earns a passing LCP, fast server response, light scripts, content present in the initial markup, also makes a page legible to the systems that decide which sources to surface.