Methodology
How WhatVPN scores a VPN
Scores are editorial judgements, but they are built from evidence that anyone can check, weighted the same way for every provider. Here is exactly how a score is produced.
The weighting
Every provider gets a 0–100 score in five categories. The overall figure is a weighted average — weighted the same way for all of them:
What feeds each score
- Ownership and independence — who ultimately controls the provider, and whether they also own VPN 'review' sites (a conflict we flag).
- Jurisdiction — the legal environment the provider answers to, and any mandatory data-retention obligations.
- Audit history — the auditor, scope and recency of independent no-logs and security assessments. Reports older than 24 months count for less.
- Logging and architecture — the exact wording of the logging policy, and whether infrastructure is RAM-only.
- Published third-party benchmarks — speed and streaming results from named testers (e.g. TechRadar, Tom's Guide), noted where they disagree.
- Feature verification — protocols, kill switch, split tunnelling, port forwarding and dedicated IP, checked against current apps.
What we don't do
We don't publish invented test figures, we don't suppress a provider's weaknesses or a rival's strengths, and no provider can pay to change a score or a position. A provider without verifiable test and audit data is held out of the scored ranking until the evidence exists, rather than given a placeholder. This is both our editorial line and, under the UK's Digital Markets, Competition and Consumers Act 2024, the standard the law expects of published rankings.