The Bank That Knocks First

A bank reacts because it has to. It cannot see your project, so it waits for you to bring it and then judges what you brought. The initiative belongs to whoever holds the knowledge, and in ordinary lending that is always the borrower.

A signature house learns differently. Every promise it makes (this workflow, installed in this kind of business, will deliver this result) leaves a record. It did the work, it guaranteed the outcome, and it watched what happened. After a few hundred engagements, it knows things no borrower knows. It knows that a particular reservation system reliably pays off for restaurants of a certain size, that a particular billing workflow works in dental practices and fails in veterinary clinics, and that a certain kind of owner gets the most out of it.

At that point the knowledge has changed hands. And when the knowledge changes hands, so does the first move.

In the last piece I predicted that banks would eventually buy signature houses to get the one thing they cannot build: the ability to see inside a project. Follow that one step further and look at what happens to the local branch.

The branch already has what the house lacks: local presence, local relationships, and the trust of the people on Main Street. The house has what the branch lacks: a catalog of proven promises, and the knowledge of which ones work where. Put them together and the branch stops being a place where people come to ask for money. It becomes a local signature house with a retail bank attached.

And the job of the people who work there changes completely. The loan officer who used to sit and wait for applications becomes something closer to a relationship manager for promises. Their job is to know the businesses in their town, know the catalog, and match them. When the house learns that a certain promise works well and profitably, those employees go looking for the businesses that fit the profile, and they go to them first.

Here is the second principle. When you borrow from a bank, you carry the risk: if the project fails, you still owe the loan. That is why the bank only gets interest. It took little risk, so it earns a fixed, limited return.

The house has flipped that. It did the work, guaranteed the outcome, put up the money, and agreed that if the result does not come, it does not get paid. It is carrying the risk. And whoever carries the risk has always had a claim on the upside. Economists call that party the residual claimant: the one who bears what is left of the risk and keeps what is left of the gain. It is the basic definition of an owner.

So the natural contract is not a loan with interest. It is a share of the benefit. The house takes a portion of what the work actually produces, the recovered revenue or the saved costs, until it has been repaid with a return, and it earns more when the result is better than expected. The customer gives up some of the upside in exchange for giving up all of the downside.

This is not new either. Energy service companies already offer two kinds of contract. In one, the customer finances the project and the company guarantees the savings. In the other, called shared savings, the company puts up the money, carries both the technical and the credit risk, and the savings are split between the two for the life of the contract. The more risk the company carries, the more of the gain it shares. The branch that knocks first is the same trade, applied to everything else.

There is a version of this that ends badly, and we have already seen it. In the years before 2008, mortgage brokers went looking for customers who fit a profile and pitched them “no money down” deals. It was proactive, it was bundled, and it was a catastrophe, for one reason: the people making the pitch did not keep the risk. They sold the loans on to someone else and moved to the next door. After the crash, regulators had to require the firms that package loans to keep a slice of the risk themselves.

Everything I like about the house that knocks first depends on the opposite. It is safe, and it is good for the customer, only because the one making the offer is the one who gets hurt if it is wrong. The rule from the last piece applies with even more force here:

Keep the risk you sell.

A house that pitches proactively and holds the risk is an honest partner that has earned the right to go first. A house that pitches proactively and sells the risk on is subprime with better software.

The other thing that has to be right is measurement. A share of the benefit only works if both sides agree on what the benefit was. That used to be the weak point of shared-savings deals: endless arguments over what would have happened anyway. What is new is that AI systems record everything they do. The same machinery that does the work produces the evidence of what it earned, which is what makes a share of the result practical for ordinary businesses for the first time.

1. The loan officer becomes a promise manager.

   The branch job shifts from judging applications to knowing the local businesses and matching them to proven promises. The best people in the role will spend their days out of the branch, not behind a desk.

2. “Nothing up front, a share of the gain” replaces interest for proven promises.

   For work the house has done many times and knows will pay off, a benefit share becomes the default offer. Ordinary loans remain for everything the house cannot see into.

3. The shares are capped and they end.

   To stay fair, and to stay distinct from taking permanent ownership of someone’s business, benefit shares will run for a set period or until a set return is reached, then stop. The customer keeps everything after that.

4. The catalog becomes the bank’s most valuable asset.

   Banks will stop competing mainly on interest rates and start competing on their catalog: the record of which promises work, in which businesses, and how well. It is built only by doing the work and watching what happens, so it cannot be bought or copied.

5. The first scandal is a house that sold the risk on.

   Somebody will package benefit shares, sell them to investors, and keep knocking on doors with nothing at stake. It will end the way 2008 ended, and it will produce a rule that houses must keep what they sell.

6. The local house becomes a silent partner in half of Main Street.

   A branch holding benefit shares in hundreds of local businesses is a new kind of local power, closer to an old merchant banker than a modern bank. Who governs it, and whose side it takes when times are hard, will become a real local question.

Charleston AI’s pay-after model is the first step on this road. We do the work, the customer receives it, inspects it, uses it, and only then pays, so the risk that the work is wrong stays with us until the customer is satisfied. That is the risk-carrying half of the story.

What this piece describes is the other half, the half that only arrives with a record. Once a promise has been kept enough times, in enough businesses of the same kind, the one who kept it knows something the next customer does not. That knowledge is what earns the right to knock first. And carrying the risk is what earns the right to share the gain.

The set so far: The Signature Is the Product, The Signature House, and The Best Lender Is the One Who Does the Work.

Already true

Banks lend reactively because borrowers hold the knowledge. Shared-savings contracts already let a provider finance a project, carry the risk and split the gain. Revenue-based financing already ties repayment to what a business earns. And AI systems already log what they do in enough detail to measure a result.

What has to happen

Signature houses have to build real catalogs, records of which promises work where, and banks have to buy or partner with them to put that knowledge in the branch. Then the offer can go out before the customer asks.

Where I am probably wrong

Bank regulators may not let a bank take an ownership-like share in the businesses it serves, which would force the benefit share into a separate company beside the bank rather than inside the branch. And customers may simply prefer a plain loan: a confident owner who expects a big gain would rather pay interest than share it. If that happens, benefit shares stay a niche for cautious owners, and the proactive branch sells mostly guarantees instead.