Core Web Vitals — LCP, INP, CLS
Google's page-experience ranking factor boils down to three metrics measured in real browsers:
- LCP (Largest Contentful Paint) — time until the biggest content element (hero image, H1, video poster) is rendered. Good: < 2.5s. Improve by preloading hero image, server-side rendering, reducing JavaScript before first paint, using a CDN.
- INP (Interaction to Next Paint) — replaced FID in March 2024. Measures responsiveness of tap/click/typing. Good: < 200ms. Improve by minimising JavaScript main-thread work, breaking up long tasks, using
requestIdleCallbackfor non-urgent work. - CLS (Cumulative Layout Shift) — how much visible content jumps around after initial paint. Good: < 0.1. Improve by setting explicit
width+heighton images and videos, reserving space for ads, avoiding font-swap layout shifts (usefont-display: optionalor preload fonts).
Field data (real users, from Chrome User Experience Report) counts more than lab data (synthetic tests) for ranking. Lighthouse scores reflect lab data only.
The 80/20 of page speed optimisations
- Next-gen image formats — WebP saves 25-35% vs JPEG, AVIF saves 50%. Biggest single LCP win.
- Resource hints —
<link rel="preconnect">for third-party origins,<link rel="preload" as="image" fetchpriority="high">for hero. - Enable HTTP/2 or HTTP/3 — multiplexed connections, no head-of-line blocking. Cloudflare/Fastly/any modern CDN does this by default.
- Text compression — brotli saves 15% more than gzip on typical HTML/CSS/JS. Enable at the CDN or nginx.
- Defer non-critical JS —
<script defer>orasync. Analytics can wait; hero interaction can't. - Lazy load below-the-fold images —
<img loading="lazy">. Browser-native, no library needed. - Long cache-control on static assets —
Cache-Control: public, max-age=31536000, immutablefor hashed filenames. - Inline critical CSS — first paint doesn't wait for CSS download.
Frequently Asked Questions
How do I test my website's page speed online?
Enter the URL. The analyzer measures Time to First Byte, Largest Contentful Paint, First Contentful Paint, DOM size, transfer size, HTTP/2 status, compression (gzip/brotli), caching headers — everything that affects Core Web Vitals and user experience.
What's a good page speed score?
Google PageSpeed Insights score: 90+ is excellent, 50–89 needs work, below 50 is poor. Our analyzer's LCP target: under 2.5s (Core Web Vitals good). TTFB under 500ms. Total page weight under 2MB. Most sites fail these — they're real bars, not vanity numbers.
How do I improve Core Web Vitals (LCP, FID/INP, CLS)?
LCP: optimise hero images (WebP, proper sizing), preload critical resources, server-side render. INP: minimise JS execution, defer non-critical scripts. CLS: set explicit width/height on images, reserve space for ads, avoid font swaps. Our analyzer points at the specific offender on your page.
Is this like Google PageSpeed Insights or Lighthouse?
Similar metrics, lighter weight. PageSpeed Insights runs full Lighthouse (slower, deeper). Our tool runs a live fetch with targeted metrics — great for quick iteration. Use both: ours for fast feedback, PageSpeed Insights for definitive scoring.
Does page speed affect SEO ranking?
Yes. Core Web Vitals are a confirmed ranking signal (Page Experience update, 2021). Slow sites lose rankings and conversion. Amazon found every 100ms of latency costs 1% in sales. Speed = SEO + UX + revenue.
Smarter quality checks. Faster deployments.
Better apps.
