Website speed guides
Step-by-step fixes for every part of the loading timeline our 8-region test measures — DNS and connections, server response, page weight and caching. Written by the engineers who run the tool.

How to reduce TTFB (time to first byte): causes, targets and fixes that work
Reduce TTFB by caching HTML at the edge, fixing slow database queries, removing redirects and moving content closer to users. Targets, causes and configs.

What a CDN is and how it makes your website fast worldwide
A CDN serves files and cached HTML from edge servers near each visitor. Learn cache hits, keys, TTLs, purging and headers, and how a CDN cuts far-region TTFB.

Image optimization for web performance: AVIF, WebP, responsive images and lazy loading
Image optimization guide: pick AVIF or WebP, set quality, serve responsive images with srcset, prevent CLS, lazy-load correctly and prioritize the LCP image.

Cache-Control and browser caching: how to make repeat visits instant
Browser caching: set Cache-Control headers by file type, use immutable and stale-while-revalidate, bust caches with content hashes and speed up repeat visits.

Core Web Vitals explained: how to improve LCP, INP and CLS
Core Web Vitals explained: what LCP, INP and CLS measure, the good thresholds at the 75th percentile, and the practical fixes that move each metric to green.

Why your website is slow in other countries: latency, distance and fixes
Your website is slow in other countries because every round trip to a distant server adds latency, and a page load needs many. Learn the physics and the fixes.

How to eliminate render-blocking CSS and JavaScript
Render-blocking CSS and JavaScript delay first paint. Learn defer vs async, critical CSS, preload, font-display and third-party fixes, with copy-paste HTML.

HTTP/2, HTTP/3 (QUIC) and TLS 1.3: protocol-level speedups explained
HTTP/3 over QUIC and TLS 1.3 cut connection setup from three or four round trips to one or two. Learn how they work, how to enable them and how to verify.
Measure before and after
Run a free test from 8 regions, apply a fix, and test again to see the difference in each region.