Managed WordPress host TTFB benchmarks (HostingStep 2025, 34 hosts, 1,960+ days of 24/7 monitoring)

Source: HostingStep 2025 benchmark, hostingstep.com/wordpress-hosting-benchmarks/. 24/7 Pingdom monitoring of identical WordPress test installs, US-East probe.

TTFB (p75 mobile-equivalent, US):

• Hostinger Business: 351 ms
• WP Engine: 365-367 ms
• Rocket.net: 373 ms
• WPX: 411 ms
• Cloudways: 449 ms
• Kinsta: 459-469 ms
• Liquid Web (Cloudflare-backed): 497 ms US / 349 ms global via CDN

For comparison (edge platforms, from multiple 2026 third-party benchmarks — directional, vendor-influenced):

• Cloudflare Pages/Workers: 12-18 ms (BKK), 45-50 ms (EU); V8 isolate cold starts <5 ms
• Vercel: 32-70 ms (EU/SEA static); SSR adds RTT
• Netlify: 48-90 ms

Confidence: High for HostingStep (independent, methodology disclosed). Directional for edge platforms — the gap pattern (Cloudflare < Vercel < Netlify) is consistent across benchmarks but absolute numbers are sensitive to region.

Shared hosting WordPress (entry-tier): 900-1,400 ms pre-cache (CoreDash field data, 2025). This is the baseline most KW SMB sites are starting from.

Single most damning datapoint in the platform-comparison literature: ~32% of WordPress origins have good TTFB. A site whose TTFB is 800 ms cannot achieve a 2.5 s LCP on mobile networks without serving the LCP element from cache. A Cloudflare-hosted Astro site delivers its first byte ~20× faster than the typical WordPress origin.