Chapter 6
Building the Audience
Fourteen Days
The email comes in at 1:52, eight minutes before the call.
Marcus — so sorry, something blew up here overnight and my whole day just went. Can we push? I still want to have this conversation. — K
He reads it four times looking for the part that means no. It isn't there. It also isn't a date.
He writes back something short and gracious, offers three windows, and then sits in a chair in an empty office at two in the afternoon with thirty minutes he'd cleared and nothing to put in them.
By the end of the day she hasn't picked one. Not a rejection. Just a person with a job.
Fourteen days left.
The numbers, meanwhile, are excellent.
Followers past seven hundred. Engagement up every week since he stopped sanding. Comments from people whose titles he'd have paid to get on the phone in January. He can watch the graph go up, and it has become the only part of the day that reliably feels like progress.
By Monday night he's been on LinkedIn for three hours. Not scrolling — working. Answering every comment. Leaving comments on other people's posts. Reading the feed for what's moving, so tomorrow's post lands into a conversation instead of beside one.
The writing isn't what's eating him. That got cheap two weeks ago and stayed cheap. It's everything around the writing.
David catches him on the way out.
"The API integration spec."
"I know."
"You've been getting to it for two weeks. I'm not nagging, I'm blocked." He stops in the doorway, and then says the other thing, which is worse because it isn't a complaint. "Also you look terrible."
Rachel's check-in that Monday afternoon had been the same conversation in a different key. She'd looked at the engagement numbers for a while.
"These are interesting," she'd said. "When does interesting turn into a number I can put in a spreadsheet?"
He'd told her the Karen Walsh call was Friday. That was before Friday.
His phone buzzes at 10:10. Tom.
"Coffee. I can hear you burning out from here."
Marcus looks at it for a while. He is not going to ask how Tom knows.
What Are You Optimizing For
Same coffee shop, Wednesday. Tom looks like a man who slept. Marcus does not.
"So," Tom says, before the coffee arrives. "What are you optimizing for?"
"Audience."
"Are you? Or are you performing for one?"
Marcus doesn't answer.
"Here's what I can see from outside it," Tom says. "You're working harder than you were in January. In January you were cold calling, which is the worst job in the world. You look worse now."
"The numbers are up."
"Which numbers?"
"Followers. Engagement. Comments—"
"Right." Tom nods. "Now name one person who'd notice if you stopped."
Marcus opens his mouth.
He runs the list. Seven hundred and some followers. He can name the servicing manager in Boise who left the one comment worth having. He can name Karen Walsh, who has not written back. Two people in DMs who are pleasant and going nowhere.
He gets to four. He is fairly sure about three of them.
"That's the distinction," Tom says. "A following is people who see you. An audience is people who'd notice the absence. Only one of those has ever bought anything."
"So the follower count is worthless."
"It's downstream. It's the smoke." Tom turns his cup. "And I'd let you keep optimizing the smoke another month if this were 2019, because back then a big following was at least evidence you'd done something hard."
"And now everyone's posting."
"Everyone's posting daily, in complete sentences, on schedule. Most of it costs nothing to make now." He shrugs. "You know what a feed looks like when the cost of posting goes to zero. You're in it every night."
"So reach stopped being proof of anything."
"That's the whole thing, yes." Tom sets the cup down. "Which is why the play isn't wider. It's deeper. In a flood the only thing that gets found is signal, and the only thing that keeps getting found is a person somebody decided to trust."
Marcus writes that down.
"And notice where the trust comes from," Tom says. "Not from the posting. From the thing underneath it — the part you spent February building in a file nobody's read. Audience without that is noise with a schedule."
"I have fourteen days."
"Twelve. And you're spending three hours a night manufacturing smoke."
The Engagement Hierarchy
Most people who set out to build an audience build a following instead, and then can't figure out why it doesn't do anything.
The distinction is testable. If you went quiet for a month, who would ask where you went? Those people are your audience. The rest are a number that goes up.
Almost everyone stalls on the first of four rungs.
Broadcast. Publish, and hope. Necessary, insufficient, and where most people stop forever.
Respond. Somebody comments; you answer like a person. Not thanks! — a reply that moves the thought forward. This is where a reader stops being a metric and becomes a name you recognize.
Initiate. Go to them. Comment on other people's work with something worth reading. Send a message that isn't a pitch and isn't a preamble to one. Most people never do this, because it produces no number that goes up today, and because it is uncomfortable to be interesting at a stranger for no immediate return.
Connect. Introduce two people who should know each other. Be the person through whom things happen. This compounds without you pushing it.
Every one of those rungs is being automated right now, and the automation gets more obvious the higher you climb. A machine can broadcast and you cannot tell. It can reply and you can half tell. It can send you a message that isn't technically a pitch and you can always tell — because there is nothing behind it. No reputation on the line, no cost to being wrong, nothing it could lose by having been interesting at you and getting nothing back. That is the argument for climbing: the higher the rung, the more of you it takes, and the less of it can be faked by something with nothing at stake.
Go deep on one platform. Pick the room your buyers are in and live there. Three platforms at a third of the attention isn't diversification, it's being forgettable in three places. And the case for concentration got stronger, not weaker, when publishing got free: a thin presence everywhere now looks exactly like the automated presence everywhere, because that is what automated presence does. Depth in one room is the thing that doesn't scale, which is why it still reads as a person.
Then get them off the platform. A following on someone else's network is rented, and the rent changes without notice. That was always true. What's new is that the flood raises the rent on everyone at once — when everybody can post daily forever, organic reach compresses for everybody, and the compression falls hardest on whoever was relying on it. An email address is the one channel where your reach doesn't depend on how much noise other people are producing this quarter. It is also being flooded, so the same rule applies once you're there: the list is worth something only if the mail is worth opening.
Survival
The change that matters isn't the routine. It's what the routine surfaces.
Four days out, Marcus is back in the beta lenders' usage data for the first time since February, looking for something else entirely, and he notices a shape he hadn't gone looking for. The borrowers who leave don't scatter. They land in a small number of the same places, and they land there fast — within weeks of a rate move, not months.
He takes it to the servicing manager in Boise, the one who left the comment worth having, in a DM that isn't about anything.
The reply is four lines and Marcus reads it three times.
They're not waiting for those borrowers to shop. They buy the trigger the day the credit gets pulled. Your customer is being called by somebody who knows their loan better than the person who wrote it, and knows it before you do.
Marcus sits with that most of a day. Then he writes four sentences.
Retention gets sold as lifetime value. That's the wrong frame, and I think the wrong frame is why the money isn't moving.
A borrower you don't keep doesn't just stop being revenue. They become somebody else's proof that it's easier over there. And they say so, to people who trust them.
Meanwhile the people taking them aren't waiting for the shopping to start. They're buying the signal the day the credit is pulled.
For a traditional lender this isn't a growth strategy. It's whether you get disintermediated by a company that has never met your borrower and doesn't need to.
It goes up on a Tuesday.
Two hundred and forty-seven likes. Thirty-one comments. People tagging colleagues, which he has never seen happen to anything of his. The word doing the work, as far as he can tell, is disintermediated, because it names a fear people already had.
The graph moves more in two days than in the two weeks before it. Marcus notices, distantly, that it no longer feels like the point.
What does feel like the point: three of the comment threads turn into people asking whether he'd written this up anywhere, and he sends the plain retention PDF three times, and by Thursday the list has gone from fifty-two to eighty-one. Twenty-nine people who cannot be taken away from him by a ranking change.
He's reading the thread that evening when one comment stops him.
Mike Donovan. SVP, Sales Operations. First National Mortgage.
First National is top ten. The kind of company that does not answer cold email, has a procurement process, and could by itself be the reason VaultPath exists in two years.
"The trigger-lead point is the part nobody upstairs wants to hear. We've modeled it twice and shelved it twice. What's your read on why?"
Marcus's first instinct is a paragraph about the platform. He gets three sentences in before he hears Tom and deletes it.
What he sends instead: "My read is that it prices a problem nobody owns. Retention sits between marketing and servicing and gets funded by neither. Is that what happened at First National, or was it something else?"
He hits send before he can improve it into something safe.
Mike answers in under two hours, and it is four paragraphs long, and none of it is about software.
Ninety Days
The call with Rachel is on a Friday, and it is the ninetieth day, and both of them know it.
"I don't have a signed contract," he says. "I want to say that before you ask, so we don't spend ten minutes walking toward it."
"All right." Her face gives away nothing.
"Here's what I do have." He goes through it plainly. Karen Walsh, VP of Sales at a two-hundred-officer lender, who found him and rescheduled and is still in the thread. Mike Donovan, SVP at First National, who wrote four paragraphs about his own problem to a man he has never met. Two mid-size conversations that are real and slow. Eighty-one people on a list, up from fifty-two two weeks ago. And seven hundred-odd followers become a thousand-odd, which he mentions last and briefly.
"Ninety days ago you asked for pipeline and qualified opportunities. I have the beginning of both, and I got them by being findable instead of by chasing. I couldn't have bought that in January."
Rachel is quiet for a while.
"You know what I noticed? You didn't lead with the followers. In January you led with the roadmap."
"That's fair."
"It isn't a compliment. It's an observation about whether you've learned anything." She takes off her glasses, which in January was the worst thing that could happen and today just means she's thinking. "First National writing to you unprompted is the only item on that list I couldn't have gotten from a competent consultant."
"So."
"So I'm not having the other conversation today." She lets it land, and then does the thing he has come to expect, which is to give with one hand and set the price with the other. "But I'm changing the ask. Ninety more days, and at the end I don't want pipeline. Pipeline is a story about the future and you're good at those. I want a signature. One. Any size."
"That's harder."
"It's clearer. And Marcus—" She puts the glasses back on. "What you built in the last three months is the only real asset this company has added since I invested. Don't celebrate it. Convert it."
The call ends.
He opens the thread with Karen Walsh, unanswered for fourteen days, and writes the shortest thing he's written in a month.
"No pressure on timing. But I've learned something since we last talked that's actually about your problem, not mine. Twenty minutes whenever the fire's out."
She writes back in forty minutes with three times on Tuesday.
Later, walking to the car, he remembers that he is owed an answer.
Ask yourself again in seventy days, Tom had said, in February, with a book he hadn't finished sliding across the table. If you're still turning it over after Rachel's decision goes whichever way it goes, you'll have your answer. Panic stops when the emergency does.
The decision went. The emergency, such as it was, is over for ninety days.
He checks, the way you press on a bruise.
He is thinking about trigger leads. Specifically, he is thinking that if the online lenders are buying the signal on the day the credit is pulled, then somebody is selling it, and he does not actually know who, or on what terms, or whether a servicer can see it happening to its own book in time to do anything.
Nobody has asked him to know that. It will not close Karen Walsh. It is a Friday evening and the clock just reset.
He gets in the car and does not start it, and looks up how trigger leads are priced.
Endnotes
The first draft of this chapter rested on a widely repeated claim that social sellers outperform their peers by 78%. I could not find research that supports the number — it traces back through a chain of secondary citations to nothing I could stand behind, which is a different problem from a figure being contested. It is gone rather than hedged, because the argument does not need it, and because a book that spends Chapter 4 refusing to launder one unsourced figure has no business laundering another two chapters later.
What replaces it is an argument rather than a statistic: the ladder in the teaching section, which claims only that the higher rungs are harder to fake convincingly, and gives the reason — nothing is at stake for the thing doing the faking.
The trigger-lead mechanism the servicing manager describes is real and is standard practice in mortgage lending; the specifics of what Marcus's beta lenders can see in their own data are invented, along with Marcus, his followers, and the timing of Rachel's decision.
Draft completed: December 2025 | AI-era rewrite: August 2026