Chapter 15
The Choice
The Dinner
Rachel picks the restaurant, which is how Marcus knows the conversation is serious. She does not spend money on a table to deliver good news; good news goes in an email.
"The numbers are a Series B," she says, once the ordering is done with. "Revenue, retention, the curve — the board has seen it, and they want to pour fuel on it. Which means, in their heads, the thing every deck means by scaling sales. A VP of Sales. Under him, SDRs to prospect, account executives to close, managers to manage the closers, an ops team to feed the machine. The whole apparatus. Eighteen months and a floor of thirty people." She watches him. "That's the plan they'll pitch you. I wanted to talk to you before they do."
Marcus turns his water glass a quarter turn and doesn't answer, because the honest answer is that the plan she just described makes him slightly ill and he can't yet say why.
"You're not selling it very hard," he says.
"No." Rachel leans back. "Because I've watched you spend a year proving a thing, and now the board wants to fund the opposite of it." She lets that sit. "You built a business where one person and some software outproduced a floor of reps. And their answer to it's working is to go build the floor of reps. In the year of our lord, when the SDR job is the first one getting automated out of existence at every company we're invested in. They want you to raise forty million dollars to hire, at great expense, the exact team the whole market is deleting."
"So why not just say that to them?"
"Because I'm your investor, and it's not my company, and the board is not wrong that you have to scale somehow — capping the whole thing at what you and Aisha can personally carry leaves most of the opportunity on the table, and the window doesn't stay open forever." She picks up her chopsticks, puts them down. "I learned something building my last company, Marcus. The choice is never really authentic versus successful. It's scale yourself versus scale the thing that made you work. Cloning yourself into thirty people fails, because what you do is uniquely you. The question is whether there's a way to multiply what you are without watering it down to nothing." A beat. "I think there is. I think you'll find it faster than I could describe it. So I'm not going to describe it. I'm going to buy you dinner and ask you what you actually want, and then go argue with the board."
Marcus looks out at the skyline. A year ago, Rachel's name on his phone meant a countdown. Tonight she is the only person in his life telling him not to take the obvious money.
"What I want," he says slowly, "is to never build the thing I spent this year escaping."
"Then don't," Rachel says. "But you'll have to tell them what you're building instead. And 'not that' is not a Series B plan."
The Void
He can't sleep.
At two in the morning he gives up and goes down to the office and opens the document he's kept all year, the one that started as a panic ledger and became, without his deciding it, a record of a person changing.
January. A man alone in a rented office, terrified of a telephone, who had just bolted an AI to his own worst instincts and fired twenty-four hundred emails into a void that answered with how did you get this email. A man who thought sales was a thing done to people, and hated himself a little for having to do it, and did it badly because you cannot be good at a thing you're ashamed of.
December. Whatever he is now.
He sits with the distance between those two men, and the thing he keeps circling is not pride. It's a small hard certainty, the surest thing he owns: I will never go back to cold calling into the void. Not because it was hard. Because it was empty — because it made him noise, and the world already has all the noise it can hold, more of it every day, most of it now written by machines that never get tired of being ignored.
So: not that. Never that again, and never that inflicted on thirty new hires in his name.
But Rachel is right that not that isn't a plan, and she's right that staying exactly as he is — two people, however good — leaves the thing half-built. He turns the binary over and over, solo or scale, and it stays a bad trade in both directions, and somewhere around three he realizes the reason it won't resolve is that it's the wrong question. Both answers are shaped like the pre-AI world. One assumes a person can only be a person. The other assumes a team can only be a floor of reps. He has spent a year living proof that neither of those is true anymore.
He stops trying to choose between the two doors and starts, for the first time, looking for the third.
The Third Thing
He calls David in the morning and they take it to coffee, the same place Marcus used to sit and dread his own pipeline.
"Rachel wants to know how I scale," Marcus says. "The board wants a sales floor. I'd rather close the company than build one. But two people isn't an answer either." He turns his cup. "I keep thinking there's a version that isn't either of those."
David, who has spent a year being wrong in an instructive direction, gets there before Marcus finishes.
"You already built the unit," he says. "It's Aisha."
"Say more."
"Aisha isn't a rep. A rep is a seat in a machine — you hire thirty of them because each one does a thirtieth of the work. Aisha's not that. She's a whole engine by herself: her own expertise, her own audience, her own relationships, and underneath all of it the same agent stack you run — the thing drafts her follow-ups, pulls her research, runs her cadence, remembers everything, so she spends her whole day on the part that's actually her." He's drawing on a napkin now, badly. "So you don't hire a sales floor. You hire more Aishas. Five people, each one a team of one, each one fascinated by a different corner of the market you can't see — one who lived servicing, one who came out of compliance, one who knows credit unions the way Aisha knows the front line. And under every one of them, agents doing the volume that used to need an SDR and a coordinator and an ops analyst. You're not scaling reps. You're scaling judgment, and letting the machines scale everything that isn't."
Marcus feels the click he's been chasing since the restaurant.
"A team of teams of one," he says.
"A team of teams of one," David agrees. "Which is the thing I got wrong all year, by the way. I kept telling you to pipe everything through a model. I had it exactly backwards. You don't automate the selling and keep the humans for overflow. You automate everything around the selling and hire humans for the only part a machine can't do — the part where somebody has to actually understand a stranger's business and care. I was trying to build a sales team out of software. Turns out the software's the easy half. The hard half is finding five more people worth pointing it at."
That evening he calls Rachel.
"I know how I want to scale. It's a sales team. It is emphatically not the board's sales team." He tells her the shape of it: no SDR floor, no funnel of interchangeable seats. Five hires over a year, each an expert in a corner he can't see, each running the same agent stack Aisha does — a team of five doing what the board thinks needs fifty, because the machines carry the volume and the people are there for the two things machines never will be, fascination and judgment.
Rachel is quiet a while.
"That's not a sentence most investors want to hear when they say 'sales team,'" she says finally. "Five people instead of fifty. Half your headcount is software."
"It's what works. Aisha's the proof — one hire, and she closes deals I never could, in a voice I don't have, with an agent stack doing everything underneath her that used to take three people. Give me five of her and a year, and we'll do what the floor of thirty would do, at a fraction of the burn, without building the exact thing the market is busy killing."
"The board is going to fight you on the headcount."
"Let them. The headcount is the point. Anyone can raise forty million and hire a hundred people to send email. In a year the email will all be machine-written and worth nothing, and they'll be laying those hundred people off. I'm not building the company that has to lay them off."
He can hear Rachel smiling, which she rarely lets him hear. "That," she says, "is the version I couldn't describe to you at dinner. I'll take it to the board. Come to the meeting hungry — you're going to have to win it. But I think you already know how, because you just did it to me."
New Year's Eve
December 31st, and the whole company is in the office, which takes about one round of drinks to accomplish, because the whole company is six people. Marcus. David. Aisha. The two engineers and the designer who spent the year building the product while Marcus built the way to sell it. Someone has brought decent champagne and no one has brought glasses, so they drink it out of the mismatched mugs from the kitchenette.
A year ago tonight, Marcus sat alone in this room and dreaded January — the investor call he knew was coming, the runway he could already see the end of, the reckoning he had no answer for, the calls he'd have to make and would rather have died than make. He'd been so sure the whole enterprise was going to end with him admitting, to Rachel and to himself, that he simply wasn't a person who could sell.
He looks around the room now. He was wrong about the smallest possible thing and the largest one at once. He was wrong that he couldn't sell. And he was wrong about what selling even was.
David lifts a mug. "To a year none of us saw coming."
"To the year we build next," Marcus says, and means the thing he and Rachel and David have started calling, only half-joking, the third thing.
Aisha, who has been quiet, raises hers last. "To not having to become somebody else to be good at this." She glances at Marcus. "You know that's the part that got me. Every other sales job wanted me to stop being a loan officer and start being a closer. This one wanted me to be more of what I already was."
"That's the only version that works," Marcus says. "It's the only part they can't automate."
On the far monitor, which nobody is watching, the work of the company goes on without them. The agent stack is drafting the follow-ups that go out the morning of the second, pulling the research for Monday's calls, keeping two dozen slow enterprise threads warm with the small useful touches Marcus decided on last week and will never have to send by hand. It is the busiest thing in the building tonight, and the only one not drinking. A year ago that machine would have frightened him — it was the thing that was supposed to make him unnecessary. Tonight it is the reason six people can do the work of sixty, and the reason the six of them get to stand around a kitchenette drinking mediocre champagne instead of grinding a pipeline no one could hold. The threat and the tool were the same thing the whole time. Which one it became came down to whether there was anyone in the room worth pointing it at.
Near midnight he steps to the window with his phone. The skyline is doing its year-end thing, the whole city lit up and briefly sentimental. He thinks about writing something careful and decides against it. He writes the true thing instead, the way he learned to this year, and lets it be shorter than it wants to be.
One year ago tonight I was dreading the sales calls I'd have to make in January. I was going to spend the new year cold-calling into a void — firing thousands of messages nobody asked for at strangers who'd never answer, and hating every second of it, and calling it a numbers game.
I'll never do that again. Not because it's hard. Because the void is full now — everybody's got a machine that can flood it for free — and shouting into a full room is worth exactly nothing.
Here's what I learned instead. The machine can write the email, pull the research, run the follow-up, remember everything, work all night. It cannot be fascinated by your customer's problem. It cannot care. It cannot know what actually matters, because it has never wanted anything. So let it have all the work that was never really yours — and become so good at the part that is that people come looking for you. That's the whole game now. Not one person doing everything. One person who knows what matters, with machines doing the rest — and, if you're lucky, a few more people like that, each one irreplaceable in their own corner.
AI came for the sales team this year. It's going to get it, too — the old one, the floor of people doing what software now does better. But it is never going to come for the fascinated expert who wields it. That seat is safer than it has ever been.
Become that. Become a sales team of one. Then go find the others.
He hits post a little before midnight, and by the time the city starts counting down he has stopped watching for likes, because for the first time in a year he isn't waiting for anyone out there to tell him it worked. He already knows.
Behind him the machine hums through its nightly work, tireless and content and utterly without a clue what any of it is for. In front of him is a room full of people who know exactly what it's for.
The clock turns over. The new year starts with the office full.
Endnotes
Marcus, VaultPath, Rachel, David, Aisha, and the year they had are invented. The choice at the center of the chapter is not: it is the real decision in front of every founder and operator right now, and the "third thing" is the argument of the whole book, stated plainly at the end. The revenue motion goes agent-first — machines carry the volume across the business — and the people you add are hired for the two things machines don't have, fascination and judgment, each of them running the same engine so that a very small team does what a large one used to. Not the lone operator, which was always a rare exception. Not the traditional sales floor, which is being automated out from under the people standing on it. A team of people who each operate as an expert team of one, with AI doing everything underneath.