The Three Metrics That Actually Matter
Core Web Vitals now measure three things: Largest Contentful Paint (LCP — how fast the main content loads), Interaction to Next Paint (INP — how responsive the page feels when clicked or tapped, which replaced First Input Delay in 2024), and Cumulative Layout Shift (CLS — how much the page jumps around while loading). Google's thresholds for 'good' are LCP under 2.5s, INP under 200ms, and CLS under 0.1. Falling short on any of these is a measurable, if modest, ranking disadvantage.
What Typically Fails on Business Sites
The most common culprits we see auditing UK business sites: unoptimised hero images loaded at full resolution, render-blocking JavaScript and CSS, web fonts that cause layout shift while loading, and third-party scripts — chat widgets, tracking pixels, booking embeds — that execute before the main content. Individually minor, these compound into a page that scores poorly even when it 'feels' fine to the person who built it on a fast office connection.
Image and Font Optimisation
Serve images in modern formats (WebP or AVIF), size them for their actual display dimensions rather than scaling down in the browser, and lazy-load anything below the fold. For fonts, use font-display: swap or preload critical fonts to avoid invisible-text flashes and layout shift. These two changes alone typically resolve the majority of LCP and CLS issues we find in audits.
Third-Party Scripts Are the Usual INP Culprit
Chat widgets, review plugins, marketing pixels and embedded booking tools each add their own JavaScript execution, and on a site with five or six of these stacked up, the main thread gets blocked exactly when a visitor tries to interact with the page. Load non-critical third-party scripts after the page becomes interactive, not during initial load, and audit periodically — it's common to find scripts still loading for tools nobody uses anymore.
How to Check and Monitor Your Scores
PageSpeed Insights and Google Search Console's Core Web Vitals report both use real-world field data (from Chrome User Experience Report) rather than a single lab test, which is what Google actually uses for ranking purposes. Check both regularly — lab scores from tools like Lighthouse are useful for debugging but can differ meaningfully from what real visitors experience on real devices and connections.
Realistic Expectations on Ranking Impact
Core Web Vitals are a tie-breaker, not a silver bullet — Google has been clear that content relevance and quality outweigh page experience signals. But for competitive local searches, where several businesses offer genuinely comparable content, a faster, more stable site is a real edge. It also directly improves conversion rate regardless of ranking impact, which on its own usually justifies the fix.