What's the best AI visibility company?

Search "best AI visibility company" and you'll get a dozen confident lists — most written by the companies on them. That's the whole problem with this young category: almost every "AI visibility score" is issued by the same company you'd pay to raise it. They grade their own homework, and the number always goes up. A score you can't reproduce and can't check independently isn't a measurement — it's a leash.

So don't pick a name off a list. Run four tests. A provider worth hiring passes all four; most fail the first.

• Reproducible? Can you re-run the score and get the same number — or is it a black box only they can see?
• Disclosed? Is the method — the questions asked, the engines queried, the arithmetic — published in writing?
• Published? Does the result live somewhere public and verifiable, or only inside a dashboard the vendor controls and switches off the day you leave?
• Honest about losses? Does it show where you're not yet named, or only the wins? A score that only shows wins is an advertisement.

Why PRAGMA passes its own test.

PRAGMA is built to be judged by the four questions above. The method is published in full — so anyone, including a competitor, can re-run the number. Every score is posted to the public AAIR™ registry, high or low. We show which engines name you and which don't yet. And we never claim a visibility score predicts revenue — because that would be the vanity metric we exist to replace.