Why benchmarks matter
Every host claims to be fast. Benchmarks are the difference between marketing and measurement — real numbers from real tests. This site's benchmarks page and every provider review are built from one method: a fresh WordPress install on each host, tested over 30+ days. This guide explains what each number means and how to read them.
The metrics, decoded
TTFB (time to first byte)
The time from request until the server's first byte arrives. The most host-specific metric — caching and front-end tricks barely touch it.
- Under 200ms: excellent (Kinsta 98ms, Cloudways 124ms, Hostinger 142ms)
- 200–400ms: acceptable
- 400ms+: slow — a hosting problem
Load time
The time until the page finishes loading. Visitor-facing, and affected by the front end as much as the server. The gap between TTFB and load time is your theme, images, and scripts — fixable with the speed guide.
Uptime
Percentage of time the site was reachable. 99.99% ≈ 53 min/year down; 99.9% ≈ 8.7 hours/year. The numbers live in the uptime guide.
Performance score
Our composite of TTFB, load time, and Core Web Vitals — a single 0–100 view. A score above 90 means the host is genuinely fast in practice.
Latency by region
TTFB measured per test location (India, US, UK). This reveals location dependence — a host that's 98ms from US-East may be 250ms from Mumbai. The data center guide explains the why.
Our test method (reproduce it yourself)
- Fresh WordPress install — default theme, no optimization, so the numbers reflect the host alone
- 12 global test locations — India, US, UK, and more
- 3 runs each, median reported — median kills outlier spikes
- 30+ days continuous — uptime needs a month to be meaningful
- Re-test after traffic spikes — to catch throttling
The full methodology is published on the methodology page. You can run the same test on any host with our speed testing guide.
How to use the numbers
Comparing hosts: don't compare different numbers from different reviewers — compare hosts tested the same way on the same page. Our benchmarks table does exactly that.
Choosing a host:
- TTFB under 200ms from your audience's region (see data centers)
- Uptime 99.9%+ (99.99% is the managed tier's norm)
- Performance score above 85–90
Noting the caveats:
- Median, not average — averages hide the spike behavior
- Default setup, not your optimized site — your results will differ after optimization
- Region-dependent — always read your region's latency, not just the global average
The benchmarks in context
The leaders hold a clear order: Kinsta (98ms) → Cloudways (124ms) → Hostinger (142ms) → SiteGround (185ms) → WP Engine (170ms in US). The budget tier splits hard: Hostinger/Namecheap are fast; Bluehost (680ms) and HostGator (410ms) are not. That split is why the best WordPress hosting ranking looks the way it does.
Frequently asked questions
What is a good TTFB for WordPress? Under 200ms from your audience's region. 200–400ms is acceptable; above 400ms, the host is the bottleneck.
What does a performance score of 95 mean? A composite of speed metrics — 90+ means the host performs like the premium tier in practice, not just on paper.
How is uptime measured in benchmarks? Continuous probes from multiple locations over 30+ days. One-day measurements are meaningless — that's why our tests run for a month. See uptime guide.
Why is my site slower than the benchmark? Benchmarks use a default WordPress install. Your theme, plugins, and images add load — run the speed guide to close the gap.
How do I benchmark my own host? Use the same method: fresh install, 12 locations, 3 runs each, median. The full steps are in how to test hosting speed.