How We Rank Hosting Providers
Every score on this site is computed from a documented formula — not gut feeling, not pay-to-rank, not affiliate revenue. Here is exactly how it works.
The Composite Score (0 – 10)
Each provider receives a composite score per hosting type (shared, VPS, managed, etc.). The score is computed from five components, then normalised within the type's cohort so the best provider reads close to 9.8 and the weakest reads no lower than 6.0 — preserving relative differences without making everyone look mediocre.
| Component | Weight | What it measures |
|---|---|---|
| Editorial rating | 30% | Our team's assessment of support quality, reliability, UI/UX, and overall experience. Set manually after testing and research. |
| Price value | 20% | The provider's cheapest plan for that type, normalised against all active providers in the same category. Cheapest in the cohort = 100, most expensive = floor of 15. |
| Account limits | 25% | Disclosed quantitative and boolean limits (inodes, CPU, RAM, SSH, Node.js, etc.) normalised against the cohort. Higher limits = higher score. |
| Features & freebies | 15% | Bundled extras relevant to the hosting type: free SSL, free domain, daily backups, free migration, money-back guarantee. |
| Transparency | 10% | Percentage of relevant quantitative limits the provider publicly discloses. Providers that hide limits score lower here, and receive an additional −0.1 direct penalty per undisclosed key. |
Type-Specific Metrics
Scoring metrics are tailored to each hosting type. Shared hosting limits (inodes, entry processes, sendmail) are irrelevant to a dedicated server. SSH access is always available on VPS — scoring it there adds no signal.
Score Normalisation
Raw composite scores rarely reach 9+ because providers always have some weaknesses. To give scores meaning relative to user expectations, we apply min-max normalisation within each type's cohort:
The best provider in each type scores up to 9.8. The weakest scores no lower than 6.0. Relative differences between providers are preserved exactly. If Provider A outscores Provider B by 15% in raw terms, the same 15% gap appears in the displayed scores.
Scores are computed per type independently — a provider's shared hosting score is not influenced by their VPS offering.
Comparison Pages
The interactive score visualisation on comparison pages uses a relative score — computed only between the providers being compared, not against the full cohort. This means a provider's score on a comparison page will differ from their global score.
The same five components and weights apply on comparison pages. The only difference is normalisation against the two (or three) compared providers rather than the full type cohort.
Data Sources & Update Frequency
Pricing — checked regularly against provider websites. Promo pricing is used where available; regular/renewal pricing is shown separately.
Account limits — sourced directly from provider documentation, knowledge bases, and official terms. Each limit is tagged with its source URL. Limits marked as "undisclosed" mean the provider does not publish this information.
Editorial ratings — set by our team after reviewing support responsiveness, control panel quality, and uptime claims. Updated when providers make significant changes.
Composite scores — recomputed whenever pricing or limits are updated. The last computation date is not currently shown per-provider but will be added in a future update.
What We Don't Measure (Yet)
We publish gaps openly rather than inventing proxy metrics:
- ○ Real uptime — requires persistent monitoring infrastructure we don't yet have.
- ○ Actual page speed / TTFB — server-side performance varies by location, plan, and load.
- ○ Support response time — would require ongoing blind testing.
- ○ Renewal pricing impact — currently treated the same as promo pricing in the score.
Questions about our methodology or spotted an error in a provider's data? Contact us — we correct mistakes publicly.
Last updated: October 2026