I have been on the road most of this month. Almost every meeting turned into a conversation about the Trust Transfer Index, and those hours have been the best part of my summer.
In one of them, someone asked a question I haven’t been able to put down. Could we reverse-engineer the whole thing and figure out which kind of person each model connects to each card?
It is a good question because the answer is a mirror. The shopper a model hands you is the product line you built, and the pages you published about it.
So, I did it. I sent my agents as 12 different shoppers, from a student with no credit history to a chief executive who wants lounge access. Then I read the card lists an AI reads before it answers. Forty lists. 266 card mentions.
Capital One appeared in all twelve shopper lists. It is the only bank in the study that did. Chase appeared for eleven. Then it drops off quickly, and most banks appear for only a handful of the twelve.
Appearing is the first hurdle, and it is a real one. A bank present for three shoppers competes in three markets while watching the other nine, no matter how good its cards are.
This is my favorite interpretation of the study. Review one row at a time, and you are reading that issuer’s strategy.
Every bank reaches broadly from a range of cards, each built for a different person.
Capital One reaches all twelve from a range of seventeen cards. Discover reaches people just starting out with a range of seven cards.
The shopper a model hands you is the product line you built, and the pages you published about it.
The model reads what already exists. Product decisions from years ago, plus every page written about them since.
Look at your own row, and you are looking at the customer your published record describes. When that customer differs from the one you want in 2027, the work is to start publishing for that person now.
Each kind of customer is its own contest, and most banks are in a few of them. Winning one carries almost no weight on the next.
Want to walk through the study with your team? Reply, and we will book a review together.
