Yellow Kid Weil was an American con man in the early twentieth century, and he had a problem that had nothing to do with the quality of his schemes.
He could find the perfect wealthy mark. He could design something plausible. And then he had to walk up to a rich man he had never met and start talking — at which point, Greene says, "the sucker might become suspicious."1
Not because the scheme was bad. Because a stranger with an opportunity is a recognisable species, and wealthy men recognise it early and often.
So Weil stopped approaching marks. He went looking for somebody the mark already knew.
The person Weil recruited was specific: "someone lower on the totem pole who was himself an unlikely target, and would therefore be less suspicious."1
Two properties there, and the second one is the clever part.
Lower on the totem pole means the man has access to the mark — he works for him, or with him, or near him — without being the mark himself.
Himself an unlikely target means the recruitment approach is not suspicious to him, because a man with modest money does not expect to be worked. His whole defensive apparatus, such as it is, is calibrated for a world in which nobody bothers with him. Weil is exploiting the subordinate's correct belief that he is not worth conning.
That belief is what makes him cheap to convince. And once convinced, he becomes something no amount of Weil's own skill could produce: a trusted source inside the mark's own circle.
The mechanism turns on one detail: "Convinced the scheme was for real, the cat's-paw would often suggest, without prompting, that his boss or wealthy friend should get involved."1
Without prompting. Weil does not ask him to bring the boss in. He does not need to.
The subordinate has been shown a scheme promising incredible wealth, believes it, and does the natural thing — he thinks about how to make his own position in it bigger. Greene names his reasoning exactly: "Having more cash to invest, this man would increase the size of the pot, making bigger bucks for all concerned."1
So the introduction is generated by the intermediary's own greed, working on his own behalf. He is not being used as a messenger. He is solving his own problem, and the solution happens to be the thing Weil wanted.
That is why the approach never reads as an approach. Nobody is pushing anything toward the mark. A man he trusts has had an idea.
Put yourself in the mark's chair for a second.
A scheme is described to you. You cannot evaluate it — you do not know the industry, the counterparties, the mechanics, and the twenty minutes you would need to check it properly do not exist in your week.
So you check something you can evaluate: the man in front of you. And the man in front of you is your own trusted subordinate, who has known you for years, who has nothing to gain from lying to you, and who is putting his own money in.
The verdict comes back positive with full confidence, and it is a verdict about a completely different question.
Greene's summary of what the intermediary supplies is exact: "The cat's-paw establishes your credibility and shields you from the unsavory appearance of being too pushy in your courtship."2
Notice something about the structure that Greene does not stop on.
The intermediary is not lying, and he is not a plant. He genuinely believes the scheme. He genuinely thinks a bigger pot is better for everyone including himself. When he brings his boss in, he is doing his boss a favour as he understands it.
That means there is no point in the chain at which anyone is knowingly deceiving the mark.
Weil deceived the subordinate. The subordinate, sincerely, brought the boss. The boss checked the subordinate, correctly, and found him honest — which he was. Every link is sound and the chain still delivers a wealthy man to a con.
The deception is transitive and the sincerity is not. Weil's single lie is laundered by passing through one person who believes it, and it comes out the other side wearing that person's reputation.
Which explains why this is a cat's-paw rather than an accomplice. An accomplice knows and can be turned, bought, or frightened into confessing. The intermediary has nothing to confess.
You want to get in front of someone who does not take meetings with people like you.
The move you are being tempted toward is the direct one — a good email, a warm subject line, a specific insight about their business. It will not work, and the reason is not the email. It is that the category stranger who wants something has a filter in front of it and yours is one of three hundred.
So stop working on the message and start working on who is already inside the room.
Find the person adjacent to them who is not themselves a target. The chief of staff. The junior partner. The supplier who has been there eleven years. Somebody whose ordinary week does not include being pitched, and who will therefore evaluate you as a person rather than as an approach.
Now the part that separates this from ordinary networking, and it is the part to be careful about. Do you need them to make the introduction, or do you need them to want it made? If you have to ask, you have built a favour, and favours arrive marked as favours and get discounted accordingly. If they raise it themselves, unprompted, because bringing you in serves something of theirs, it arrives as their own judgement.
Which means the honest version of this and the Weil version differ in exactly one place, and it is not the technique. It is whether the thing you showed the intermediary is true. The mechanism is identical either way: their belief carries you across a threshold your own words could not. If the belief is warranted, you have used a relationship the way relationships work. If it is not, you have spent someone else's credibility without telling them, and they will pay for it when it collapses.
Greene does not report a failure case, but the structure names two.
The intermediary bears the whole cost of exposure. When the scheme collapses, the mark's grievance runs to the man he trusted, not to the stranger who has left town. The relationship destroyed is not the con man's; the intermediary's standing with his boss, built over years, is the actual thing that was spent.
And it can only be run once per intermediary. Not because he learns the technique — because he learns that his judgement produced a disaster, and so does everyone watching. His usefulness as a door was entirely a function of his credibility, and the operation consumes it.
So the technique's supply is not marks. It is people whose trustworthiness has not yet been spent, and each use burns one.
The general instruction he draws is broader than the con: "Devices like this are often the best way to approach a person of power: Use an associate or subordinate to hook you up with your primary target."2
And immediately after, the cheaper variant: "The easiest and most effective way to use a cat's-paw is often to plant information with him that he will then spread to your primary target. False or planted information is a powerful tool, especially if spread by a dupe whom no one suspects. You will find it very easy to play innocent and disguise yourself as the source."3
That second version drops the introduction entirely — you never meet the mark at all, and the intermediary carries only a fact. It is built out separately at Planted Information Through a Dupe.
🚩 SINGLE SOURCE · 🚩 SECONDARY WITHOUT PRIMARY. Weil's method is described in general terms with no specific con, no date, no mark named, and no citation. Weil is in Greene's Preface roll-call of practitioners and appears here as a generic exemplar rather than as a documented case.1
🚩 The account is a technique description, not evidence. Every element — "would often suggest, without prompting" — is presented as habitual behaviour. Nothing in the passage is falsifiable.
Tension with the chapter's premise. Law 26 is about keeping your hands clean of unpleasant work. Weil's intermediary does nothing unpleasant; he makes an introduction and invests his own money. The chapter's frame does not fit the case, and what actually unites them is narrower: the operator is not the visible cause of what happens.
Open question. The technique's supply is unspent credibility belonging to other people. Does that make it self-limiting in any closed network — and is that why con men moved from town to town?
Greene's own Law 13 says: when asking for help, appeal to people's self-interest. Weil is the demonstration, taken further than Law 13 goes.
Law 13's advice is to frame your request in terms of the other party's interest. Weil does not frame anything. He positions a person so that their genuine self-interest generates the request on his behalf — the subordinate is not persuaded to help Weil; he is helped to want a bigger pot.
The difference is between rhetoric and arrangement, and the arrangement version leaves nothing behind. There is no persuasive appeal for anyone to later reconsider. See Gratitude as a Burden, Not an Obligation for Law 13's own claim.
Psychology — The Question You Actually Answered
That page describes standing in front of a used car with no knowledge of engines. The question is this a good car requires expertise you do not have and cannot acquire in twenty minutes. So the mind quietly answers a different question — do I trust the person selling it to me — and hands you the answer with the confidence the hard question would have earned, without your ever noticing the swap.
Weil's whole method is a machine for controlling which question gets substituted, and that is the insight the pair produces.
The mark cannot evaluate the scheme. That is fixed and Weil cannot change it. What he can change is which person is standing in front of the mark when the substitution happens — because the substituted question is always do I trust the source, and the answer to that question is determined entirely by who the source is.
Approach directly and the swapped question is do I trust this stranger, which returns a fast no. Route through the subordinate and the swapped question becomes do I trust the man who has worked for me for eleven years, which returns a fast, confident, and correct yes.
So the technique does not defeat the mark's judgement. It selects the input to a substitution that was going to happen anyway, and the substitution then performs flawlessly on the wrong object. Which gives the defence, and it is not scepticism: a wealthy man cannot become an expert in every scheme. What he can do is notice when he is answering a question about a person and the decision is about a thing — and that noticing is available in the moment, unlike the expertise.
Business — Sales as Transference of Belief Over a Bridge of Trust
The reframe on that page removes the adversarial picture of selling. Sales is "passing conviction from one person to another… an education process over a bridge of trust." The claim underneath it is ethical as well as mechanical: if a prospect knew everything you know about the product, they should buy — and if they knew everything and still wouldn't, getting them to buy is the manipulation.
Weil is that mechanism running with the ethical clause removed, and setting them side by side isolates exactly where the removal happens.
The transfer is identical. Conviction moves from Weil to the subordinate, and from the subordinate to the mark, across bridges of trust that are entirely real. The subordinate is not performing belief; he has it. The bridge between him and his boss is not manufactured; it was built over years of honest work.
The difference is a single test, and it sits at the very start of the chain. Would the first person in it have bought if he knew everything Weil knew? No. Everything downstream — the sincerity, the unprompted introduction, the mark's correct assessment of a trustworthy man — is structurally sound and is transmitting a corrupted input.
Which delivers something the sales page's own frame implies and does not state: conviction transfer is indifferent to the truth of what is transferred, and its fidelity is highest exactly where the relationships are best. The strongest bridge carries the falsehood most efficiently. That is not an argument against the frame; it is the reason the knew-everything test has to sit at the origin rather than at the close, because by the time the belief reaches the second person it is indistinguishable from an honest one — including to the person carrying it.
Sharpest implication. Only one lie is told in this entire operation, and it is told to somebody who is not the target. Everything after it is sincere, competent and correct — the subordinate's belief, his self-interested reasoning, the mark's assessment of a man he has trusted for years. A single falsehood, passed once through a person who believes it, acquires that person's reputation and becomes untraceable, which means the operator's real product is not deception but laundering.
Generative questions