Our Testing Methodology

Every review and comparison on this site follows the same testing approach. This page documents how we generate the numbers you see.

1. Performance testing

Test setup

• Identical WordPress installations (same theme, same 10 plugins, same demo content) deployed to each provider
• 2 test sites per provider: one lightly configured (fresh WordPress), one realistically configured (Elementor page builder, WooCommerce, common SEO plugins)
• Monitoring continuously for 30 days per host; benchmark samples captured every 15 minutes

Tools we use

• TTFB: GTmetrix, Pingdom, and custom curl-based monitoring from 6 geographic regions (US East, US West, London, Frankfurt, Singapore, Sydney)
• LCP & Core Web Vitals: Google PageSpeed Insights + WebPageTest, measured at both idle and simulated 100-concurrent-user load
• Uptime: UptimeRobot + StatusCake, 1-minute interval pings
• Load testing: k6 for concurrent user simulations

What we report

• TTFB (idle): Median of 2,880 daily measurements over 30 days, from the primary US East region
• TTFB (under load): Single measurement at 100 concurrent simulated users
• LCP: 90th-percentile value from PageSpeed Insights field data + WebPageTest
• Uptime: Rolling 30-day actual uptime percentage

Cross-reference

When possible, we cross-reference our measurements against independent public benchmarks — particularly Koddr.io's 90-day Cloudways tests and HostingStep's 365-day 34-host tests. Where our numbers differ from these sources, we report both with attribution.

2. Pricing verification

• We fetch pricing directly from each provider's official pricing page, not third-party summaries
• Every price is re-verified before publication
• We compute realistic total-cost scenarios including common add-ons (CDN, off-site backups, email) — not just headline prices
• Promotional pricing (intro discounts, Year 1 rates) is separated from renewal pricing

3. Support testing

• We submit real support tickets across 3 categories: server-level (PHP version change), WordPress-level (plugin conflict), and billing (refund inquiry)
• We measure response time and quality of answer on each
• We test live chat, email tickets, and phone where available
• Support tests are submitted from anonymous accounts, not affiliate-identified ones

4. Feature verification

• We verify every claimed feature works as advertised by actually using it
• We test staging environments, SSH access, Git deployment, backups, and WP-CLI
• We document gated features (e.g., SSH on premium tiers only) explicitly

5. User-signal aggregation

We sample 100+ reviews from each of G2, Capterra, and Trustpilot to identify recurring themes — both positive and negative. We report patterns that appear in multiple reviews, not one-off anecdotes.

6. What we don't do

• We don't run one-time benchmark tests and call them representative
• We don't use single-region tests to claim global performance
• We don't cite performance numbers without a testing source
• We don't inflate scores for hosts that pay higher commissions

Data sources referenced

• Koddr.io Cloudways Benchmarks 2025 (90-day continuous test)
• HostingStep WordPress Hosting Benchmarks (365-day, 34 hosts)
• Direct fetch from official pricing pages (Cloudways, Kinsta, WP Engine, SiteGround)
• G2, Capterra, Trustpilot review aggregates

Corrections

If you believe a number or claim on this site is incorrect, please contact us. We publish corrections publicly with dates.

Methodology last updated: April 2026