He already knew the answer would be no. That was the reason he went.
He was three days into a project made entirely out of rejection, and this was the ask he had built to fail. A doughnut shop. A counter. A request for five doughnuts, linked and colored like the Olympic rings, made to order, right there. There was no way they could say yes. He was counting on it.
The worker took out a piece of paper. She wrote down the colors and the rings. She asked how she could make it. About fifteen minutes later she came back with a box, and the doughnuts were arranged like the Olympic symbol, and the man who had walked in hunting for a refusal stood at the counter and could not believe it.
That is the scene people remember. It is the one whose video passed five million views, the one he says turned his life upside down. And it is a little strange, once you notice it, that the most famous moment from a hundred-day hunt for rejection is the moment someone said yes.
The practice he was doing is trending again. It started as a card game invented by a Canadian IT worker in the months after his wife left him, when he sat alone in a one-bedroom apartment and worked out that the thing running his life was the fear of being turned down. His fix was a rule. Get rejected by someone, on purpose, at least once a day. A yes did not count. A yes meant you had not asked for enough. Years later, the man at the doughnut counter was turned down by an investor, took the no so hard he nearly quit working for himself, searched the internet for how to stop being afraid of rejection, found the game, and stretched it from thirty days to a hundred. He filmed all of it. Now the same idea is back as a social media challenge and, this year, as something an AI will coach you through.
The story we tell about this is a story about courage. You are afraid of no, so you go out and collect a hundred of them, and by the end the word has lost its sting. Exposure. Toughening up. Face the thing enough times and it shrinks. It is a good story. It is also mostly wrong about what is actually doing the work.
Start with how common the fear is, because it is not a quirk. In the United States, about one in fourteen adults has the clinical version, social anxiety disorder, in a given year, and nearly a third of them are seriously impaired by it. Across dozens of countries, the average person who has it loses close to twenty-five working days a year to it. About half of the people who have it never get treated at all, and the ones who do usually wait fifteen or twenty years before they start.
Notice what the fear actually is. Not a feeling that arrives after something bad happens. A prediction that it will. The clinical descriptions run in the future tense: what the person fears will happen. They will blush. They will be judged. They will be waved off. And the body signs on early. The sick stomach, the sweat, the pounding heart, all of it firing in advance of an event that has not happened yet. On the first day of his project, the ask was to borrow a hundred dollars from a stranger. He describes it as the longest walk of his life: down the stairs at work, toward a big man behind a desk, the hair on his neck up, the sweat, the heart going. He got his no, and he ran. Only later, watching the tape, did he see that the man had asked him why, had wanted a reason, had left the door open. He had fled a rejection that was not even trying to be one.
So if the fear is a prediction, the real question is what corrects it. The folk answer is repetition. Do the scary thing enough and you go numb. The clinical literature is blunt that this is not really how it works. The old fear does not get erased or sanded down. It stays intact. What changes is that you build a second, competing memory next to it, and you build that memory fastest not by piling up reps but by being wrong. The term for it is expectancy violation. You predict a specific bad thing, you watch the world fail to deliver it, and the gap between the two is the active ingredient. A good exposure works like a hypothesis test: you say out loud what you think will happen, and then you go find out. And it has nothing to do with manners: the research does not support the idea that anxious people lack social skills. They know how to behave. They just predict, wrongly, that behaving will not be enough.
This is where the doughnut counter stops being a cute story and becomes the actual argument. In these projects, again and again, the asks do not break the way the asker predicts. The braced-for no turns into fifteen minutes of careful work and a box of rings. A stranger refuses to let him plant a flower in his yard, then, asked why, explains that his dog would only dig it up, and sends him across the street to a neighbor who wants it; half an hour later the flower is in the ground. He knocks on professors' doors asking to teach a class, fails the first couple of times, comes back with a real lesson prepared, and one of them tells him to return in two months. He teaches the class. The forecast fails, over and over, in the direction of yes.
And when the answer really is no, it is not the catastrophe the body promised. The founder's first stranger would not drive him across town, but the words were, I'm not going that way, buddy. In record after record the refusals come back kind, or logistical: not going that way, the dog would dig it up, a sweet idea but no. The thing the body kept warning about, in these cases, does not arrive.
The fear is a bad prediction about other people, and what corrects it is other people, proving the forecast wrong while you watch.
Here is what the courage story misses. What actually works in rejection therapy is not the no. It is the discovery, over and over in these projects, that the person you braced against turns out warmer and more willing than your forecast claimed. The fear is a bad prediction about other people, and what corrects it is other people, proving the forecast wrong while you watch. Which makes this the least private self-help there is. You cannot do it in a mirror or a journal. The whole thing lives inside someone else's response, given by a real person who has to stop what they are doing and answer you.
Say the limit plainly, though. That mechanism is the clinical one, borrowed from exposure therapy, and the clinical evidence is for treatment: an assessment first, a trained person watching, a pace, a way to stop. No one has shown that a run of filmed asks does the same job. The trend took the mechanism. It did not take the proof.
You can push back on this, and the pushback is strong, so let me make it as well as I can. The asks are tiny. A weird doughnut order. A request for a hug. A question about whether the oat milk could be free. Nobody is harmed by being asked something odd and getting to say no to it. And there is a deeper version of the objection. If the whole lesson is that people are kinder than you feared, then the small kindness a stranger spends on you is not a cost of the therapy. It is the therapy. The content and the cure are the same substance. Complaining about it is like complaining that a swimming lesson uses water. And the thing being learned is genuinely good, maybe the best thing a frightened person can learn: that the social world is not the hostile place the body keeps insisting on. If ten seconds of a cashier's patience buys someone a smaller fear for the rest of their life, most of us would make that trade for them without blinking.
I find that almost convincing. Here is what it leaves out.
The stranger's yes is not free, and the people the practice leans on hardest are the ones with the least room to decline. The worker at the counter gave fifteen minutes of real labor to a stranger's personal experiment. A cashier is paid, in part, to be pleasant, so her no is not the clean, free no of a stranger on a sidewalk; courtesy is a condition of the job. When the ask targets a person behind a counter, the experiment is quietly rigged. You have chosen someone whose freedom to refuse you is already shorter than yours. The refusal you collect from her is poor evidence about human rejection, gathered inside a place where being nice to you is part of the wage.
Then there is the camera. Somewhere between the card game and now, the private exercise turned into content; the tag has been reported in the tens of millions of views. In one filmed version of the money ask, a woman films herself while her hands shake, asks a man for a hundred dollars, and turns to the lens after he says no to tell it, I did it. The camera is on her, not on him. But the exchange is a public thing now, and the record does not say whether the other person knew, or agreed, or was ever asked. The question is not new. Under the original doughnut video, years back, one archived commenter asked whether the worker knew she was being filmed, and whether anyone had asked her before it went online.
And it does not land the same on every nervous system. Clinicians watching the trend keep repeating a version of one warning: done alone, without support, a steady diet of deliberate rejection can wear a person down and push the anxiety the wrong way. For some people the word rejection is not a stubbed toe. One therapist put it flatly: for a sensitive enough wiring, rejection can feel like dying. The clinic knows this, which is why real exposure comes wrapped in an assessment, a pace, a person keeping watch, and a way to stop. The trend keeps the mechanism and throws the guardrails away, then hands the whole thing to a phone and, lately, to a chatbot that will cheerfully draft your next ask.
Not every ask is the problem. The problem is the ask that only works because someone's freedom to refuse is already short: the worker on the clock, the stranger turned into footage, the person chosen because they cannot easily leave.
So look at what we actually have. A practice named for rejection, whose best moments are acceptances. A fix for a private fear that runs on other people's patience. A lesson that the world is generous, taught by drawing down exactly the generosity it promises you will find. That is not a scam. It might even be true and useful. But it is a strange kind of self-improvement, one that cannot be done by the self, and that sends part of the bill to a stranger who was never shown the total.
I have done it myself, just now. I used her as the hinge of an argument she never agreed to be part of.
Go back to the counter one more time. The talk on the big stage never gives her name. The record elsewhere does, and then it stops: it keeps her work and almost nothing else. Not what she thought. Not whether anyone asked. She took out a piece of paper. She wrote down the colors and the rings. About fifteen minutes later she carried out a box that looked like the Olympic symbol, for a man who had come in hoping she would turn him away. That is the whole of what we have of her. Someone's fear got a little smaller that afternoon. The cost went to her, and no one wrote down what it was.