Is anybody actually buying through AI assistants yet?

You will find plenty of confident percentages about AI-driven commerce. Most trace back to a single vendor survey, and we are not in a position to verify them, so we are not going to repeat them in order to sell you something. That would be the same failing this product exists to correct.

What can be said plainly

Assistant traffic is, for almost every brand, still smaller than search traffic. It also tends to arrive later in the buying process: the person has already asked which products to consider and has been given a shortlist, so they come to you having been pre-qualified by something you did not control. A smaller number of visits, each worth more, and each preceded by a conversation you never saw.

Why waiting is the expensive option

Two reasons, and neither depends on believing a growth forecast.

• You cannot measure backwards. If you start in a year, your first reading is your position in a year's time. You will have no way of knowing whether it went up or down to get there, or which of your competitors moved. A baseline is cheap now and impossible to reconstruct later.
• Absence is invisible in your analytics. If an assistant is telling buyers to consider three of your competitors and not you, nothing in your reporting will show it. There is no lost-traffic line for a conversation you were never part of.

The honest version of the pitch

We do not know exactly how large this channel is for your category. Neither does anybody else with real evidence. What a scan tells you is a fact about your own market rather than a forecast: here are the questions buyers in your category ask, here is what the assistants currently say, and here is who they name instead of you. Whether that matters enough to act on is then a decision you can make with the actual numbers in front of you.