Responses to AV articles
I agree that using VirusTotal isn’t the best approach – far from it. But I have also heard AV-Test doesn’t use good criteria. I like the NSS Labs methodology myself, which shows higher numbers than Imperva, but much lower than most other tests. Their consumer report is free. and they also offer a companion report. But consumer products are often more different from enterprise versions than you might expect, and the tests weren’t against 0-day like the Imperva ones. These reports by NSS tested effectiveness against exploits using known vulnerabilities, rather than Imperva’s test of signature recognition of new virus variants.
Apples and oranges, but I am generally more interested in exploit prevention than signature recognition.