01We measure. We do not operate what we grade.
RadRank is a measurement tool. We run prompts against AI engines, count the brand mentions in the answers, and report the result. We are not in the business of running the brands we score, and our scoring does not change based on any service relationship.
02No pay-to-rank. The rule above all others.
You cannot buy a higher RadRank score. There is no path to. No advertiser, no client, no operator, and no amount of money moves a number. The scoring code takes no "client," "sponsored," or "operator" flag as input. Run the same brand twice — once tagged a paying customer, once a stranger — and you get the byte-identical score and the identical rival list. There is no boost branch because there is no branch to boost.
03Identical methodology for everyone.
Same prompts. Same name-matcher. Same engines. Same cadence. For every brand — including operators we work with, their competitors, and brands that have never heard of us. No operator-supplied data influences a score. The method described on our Methodology page is the method applied to all.
04We refer to a set, never a single hidden door.
When RadRank suggests where you might get help improving your visibility, we point to a set of operators — never one secret, preferred vendor wired in behind the scenes. There is no hidden door. The measurement and any referral are separate things, and the measurement never bends toward the referral.
05We grade our own sibling on the same public board.
RadVeo is our sibling company. It appears on the same public board, scored by the same prompts and the same matcher as everyone else — no boost, no performative penalty. If RadVeo is winning a category, the receipts show why. If it is losing, that shows too. We hold ourselves to the rule we hold everyone to.
06Never "guaranteed #1."
We will never promise a ranking, sell a placement, or stamp anything "RadRank-certified." There is no such thing. What RadRank offers is honest: see where you stand, and improve your odds — measured. Every suggested move to climb is labeled odds-improving, never a guarantee. Anyone promising you guaranteed AI placement is selling something we refuse to sell.
Why you can trust this
Every claim above is checkable, not just stated:
- The score is reproducible — same inputs, same number, every time.
- The evidence is visible — every mention links to the answer that earned it.
- The method is published in full on the Methodology page.
- We never fabricate a brand name — only names that appear verbatim in an engine's answer.
- When an engine does not answer, we say so; we never imply more engines ran than actually did.
Transparency is the neutrality proof. If you can audit how the number is made, you do not have to take our word for it — and that is the point.
Read the full Methodology → · or check your RadRank →