SEO Term Definition
Page Speed
Page speed describes how quickly a page responds, renders useful content and becomes responsive for real users.
Reviewed by Alexander Yarovenko · Updated: 2026-09-18
What is Page Speed?
Page speed refers to how quickly the content on a web page loads and becomes usable to a visitor. It encompasses multiple technical metrics: time to first byte (TTFB), first contentful paint (FCP), largest contentful paint (LCP), total blocking time (TBT), and cumulative layout shift (CLS).
Why Page Speed Matters for SEO
- Search context — Core Web Vitals are part of page experience, while relevance remains essential
- User experience — slow loading can interrupt reading and task completion
- Crawl efficiency — faster pages allow Googlebot to crawl more pages within the same crawl budget
- Conversion — measure whether speed changes affect completion of this site’s real user journeys
Key Page Speed Metrics
| Metric | What it measures | Good threshold |
|---|---|---|
| TTFB | Server response time | < 800ms |
| FCP | Time to first content visible | < 1.8s |
| LCP | Largest content element loaded | < 2.5s |
| TBT | Main thread blocking time | < 200ms |
| CLS | Layout shift score | < 0.1 |
How to Improve Page Speed
- Optimise images — compress, resize, use WebP/AVIF format, lazy load below-fold images
- Minify CSS, JavaScript, and HTML — remove whitespace, comments, and unused code
- Enable browser caching — set long cache headers for static assets
- Use a CDN — serve assets from servers close to the user
- Defer non-critical JavaScript — prevent JS from blocking the render path
- Eliminate render-blocking resources — inline critical CSS; defer the rest
- Improve hosting — better servers mean faster TTFB
Tools to Measure Page Speed
- PageSpeed Insights — real user data (CrUX) + lab data with specific recommendations
- Google Search Console — Core Web Vitals report with field data segmented by device
- WebPageTest — detailed waterfall analysis
- Lighthouse — built into Chrome DevTools
Decision guide
| Use when | Optimize the slow template or resource, not an isolated score. Field data describes real visits, while laboratory tests help reproduce the bottleneck and test a proposed fix. |
|---|---|
| How to verify | Segment by page type and device, record LCP, INP, CLS and server timing, then change one bottleneck at a time. Compare the same template after deployment and wait for field data to mature. |