Methodology
Review Methodology
Every antivirus product on this site is assessed against the same seven criteria. The criteria are weighted toward the things that actually protect a home user, not toward feature counts.
Last reviewed 2026-08-09
1. Protection evidence
We look at published results from independent laboratories over multiple consecutive test rounds rather than a single headline score, because a one-off perfect result tells you very little. Consistency matters more than a peak.
2. System impact
Slow protection gets disabled by its owner, which is the worst outcome of all. We weigh measured performance impact on ordinary hardware, and we note products that are noticeably heavy during scans or on older machines.
3. Usability and defaults
Most people never open the settings. A product is judged on whether its default configuration is safe, whether alerts are understandable, and how easily a non-expert can run a scan, quarantine a file or exclude a false positive.
4. Upsell and interruption pressure
Free tiers are funded somehow. We record how insistent a product's upgrade prompts are, whether it installs additional components by default, and whether it alters browser settings without a clear opt-in.
5. Privacy practices
Security software sees everything on a device. We read the developer's published privacy documentation for telemetry scope, data-sharing arrangements and browsing-data handling, and we note past incidents where they are documented and relevant.
6. Support and remediation
We consider what happens when something goes wrong: rescue and offline scanning tools, removal tools for a broken installation, and the availability of human support on the plan an ordinary buyer would purchase.
7. Pricing transparency and renewal
First-year discounts are normal in this industry; opaque renewals are the actual problem. We describe how a product's pricing is structured, what the renewal behaviour looks like, and how many devices a plan really covers.
What we do not do
We do not run our own malware detection benchmarks, and we do not present laboratory results as our own testing. Detection measurement done properly requires a controlled sample set and infrastructure that independent labs already operate; we cite them instead of imitating them.
Frequently asked questions
- Do you accept products supplied free by developers?
- Where a developer supplies a licence for hands-on assessment, that has no bearing on the outcome and it is noted on the page. Most assessments use publicly purchasable or free versions.
Trust and transparency
All product names, logos and brands are property of their respective owners. Use of these names does not imply endorsement. Questions about anything on this page can be sent through our contact page.