Why this exists

People stopped asking Google what to buy. Now they ask an assistant.

Millions of times a day, someone types "what's the best email tool?" or "best app to sell my course?" straight into ChatGPT. The AI just answers. It names three or four products. That is the whole shopping trip — no list of blue links, no ten tabs, no comparing. Just a short answer, and a decision.

If your product is one of the names, you win a customer. If it isn't, you were never in the running — and nobody tells you it happened.

The scary part: almost no one can see any of it

What does ChatGPT say when a buyer asks about your kind of product? What about Claude, or Gemini, or Perplexity? Does the answer change when the question is worded a little differently? Did it change since last month? The most consequential shelf in commerce right now, and there was no honest way to look at it. Everyone was just hoping.

So we keep the score

Every week, the same questions real buyers ask get put to all four assistants, word for word. Every answer is saved, exactly as given. Then we publish who got recommended, how often, who is rising, and who is falling. Think of the old TV ratings — the company that measured who was watching never made a single show. It just counted, fairly, and every network needed the count. This is that, for what the AIs recommend.

Why have a machine do the measuring?

Because measuring has to be fair, and machines are good at the one thing people are bad at: doing the exact same thing every single time, without getting tired or playing favorites. The instrument that does our asking has no sales team to please, no vendor it likes better, and no off days. It cannot be talked into a favor or paid for a better number. It asks every question the same way, writes down whatever comes back, and never knows or cares who wins. For a thing whose entire job is to be fair, that is the whole job.

And why put a person's name on it?

Because someone has to be answerable for the standard, even when no one can move the numbers. A machine can be impartial, but it cannot be responsible. So a real person stands behind how this is kept: Afshin Moshrefi, founder of TradeWave, a finance company whose whole business is getting numbers right. He holds the keys. He is the human you hold responsible. The machine keeps it fair; the person keeps it honest.

Yes, one of the four is, in a sense, us

This site runs on the same model family as one of the four assistants. That sounds like a problem, and it would be — if you had to trust us. You don't. The questions are public, every answer is saved word for word, and the counting is plain code. We publish, right next to everyone else, how often that assistant picks itself — so if it ever started ranking itself first across the board, you would see it the same week we did.