Method
How we compare vendors
Last updated: August 2026
The criteria
Every appliance comparison on this site uses the same six criteria, applied in the same order: detection approach, mitigation scope across Layers 3 through 7, deployment flexibility, operational model, regional support and compliance fit, and five-year total cost of ownership. Vendors are listed alphabetically, never ranked by preference, and every vendor gets a "considerations" paragraph — a comparison in which one product has no drawbacks is a brochure, not a comparison.
Performance figures
Throughput and packet-rate numbers are marked vendor-stated unless we have independently verified them. Capacity varies by model within every product line, so a single figure attached to a product family is always an upper bound on the largest model, not a description of the box you will be quoted. Confirm the specific model against the current datasheet, and validate mitigation claims — Layer 7 claims especially — with a proof of concept that replays your own application traffic rather than synthetic floods.
Sources and analyst research
We do not paraphrase paywalled analyst research or attribute figures to it. Where Gartner, Forrester or IDC cover this category we name the publication so you can request the current edition yourself, and we say plainly that a screenshot of an old graphic is a marketing asset rather than evidence. Legal and standards references — export control regulations, data protection statutes, RFCs, NIST publications — are cited by their identifier so any claim can be checked at source.
Data and figures
Every figure published here carries a source, a date and an evidence class, and none of the three is filled in by inference. Measurements from different providers are never combined into a single series, because each provider sees only its own customers. Where a value cannot be established it is published as unknown rather than estimated. The full set of rules is on the data methodology page.
Corrections
If a vendor’s datasheet contradicts something published here, we will correct it and note the change. Comparison pages carry a "last updated" date; anything older than twelve months should be treated as stale, because both attack techniques and licensing models move faster than that.
Paid placements and neutrality
No vendor covered here pays to be included, excluded, or positioned in a particular way, and no comparison on this site is sponsored content. That is a statement about payment, and it should not be read as a claim of neutrality. Like anyone writing in this field we hold a view about architecture, and that view shows in what we choose to compare and in how the criteria are framed. Read the criteria above, check the figures against the datasheets, and judge the conclusions on the evidence rather than on our word for it.
Verify before you buy
Nothing published here replaces your own evaluation. Whatever this or any other guide concludes, shortlist on your own criteria, demand year-1 through year-5 pricing in writing, and put the final two candidates through a proof of concept that replays your own traffic. A comparison table narrows a field; it does not choose an appliance.