The Shared Delusion

There’s something founders and employees never say to each other, even though both know it’s true.

The founder knows that no one will ever care about this company the way they do. They know that the engineer who says “I’m so excited about our mission” would take a better offer next month. They know that equity, at the levels they can afford to give, doesn’t actually change behavior. They know they’re asking people to act like owners while treating them like employees.

The employee knows that “we’re a family” means “until the runway gets short.” They know that working nights and weekends won’t protect them if the Series A falls through. They know their equity is probably worthless, and that even if it isn’t, the founder will make 50x what they make. They know they’re being asked to perform a passion they don’t fully feel.

Neither is lying, exactly. But neither is telling the whole truth.

And the gap between what’s said and what’s known is where trust goes to die.

---

I’ve been on both sides of this.

As a founder, I’ve hired people and told them we were building something meaningful together. I genuinely believed it. I also knew, in some back room of my mind, that I would care more than they ever could, and that this asymmetry would eventually cause problems.

As a consultant, I’ve been paid by the hour to solve problems as fast as I could. And I’ve felt the uncomfortable pull of a perverse incentive: if I solve this too quickly, I make myself unnecessary. The rational move is to slow down. Not dramatically. Just enough to stretch a week of work into two. Everyone does this. Almost no one admits it.

The first time I recognized this impulse in myself, I felt sick. It seemed like a small betrayal—of the client, of my own standards. I decided I would always work as fast as I could, deliver as much value as possible, and trust that it would work out.

Sometimes it did. Sometimes I finished fast, the client was delighted, and they hired me for the next thing. But sometimes I finished fast and they just... didn’t need me anymore. The reward for excellence was obsolescence.

I kept doing it anyway because sandbagging felt worse than economic irrationality, plus relationships and reputation matter. But I understood, for the first time, why people stretch work out. They’re not lazy. They’re not dishonest. They’re responding rationally to a structure that punishes speed.

---

The standard answer to this problem is incentive alignment. Give people equity. Pay bonuses for hitting milestones. Make their outcomes depend on the company’s outcomes.

It sounds right, but it never works.

I once asked a lead engineer—a guy who’d been with me for 4 years, who did great work and was genuinely invested—whether he thought his team would work harder if they had more equity. He thought about it for a moment and said, honestly, no. They wouldn’t.

This wasn’t because they were bad people or didn’t care. It was because a 1% stake in a startup that might be worth something someday is too abstract to compete with the concrete reality of a stable job. What motivates people day-to-day is not the expected value of their options. It’s not wanting to look bad in front of colleagues. It’s the satisfaction of solving a problem. It’s going home at a reasonable hour without guilt.

Founders have trouble understanding this because they’re not wired the same way. For a founder, the company is an extension of their identity. They work obsessively not because they own equity but because not working feels like disappearing. You can’t transfer this to an employee by giving them stock options. You’d have to transfer the whole psychological complex—the insecurity, the grandiosity, the terror of insignificance.

Most people don’t want that.

Most people are healthier than founders.

---

So if equity doesn’t work, maybe something else does? Bonuses tied to completion? Profit sharing? Some clever compensation structure that finally aligns everyone’s interests?

I’ve tried to design such systems. They always fail for the same reason.

Any structure that rewards speed creates a definition of “done” that becomes a site of conflict. Hit the milestone by March 1st and get 50% of your salary as a bonus. Sounds clear. But what counts as hitting it? What if you’re 90% there? What if you did heroic work and the problem turned out to be harder than anyone knew? What if you missed by a week through no fault of your own?

Now you’re arguing about definitions instead of doing the work. The founder feels like the team is gaming the criteria. The team feels like the founder is moving goalposts. The system that was supposed to create trust has created lawyers.

And even if you get the definitions right, there’s the cliff effect. Miss the milestone by a day and you get nothing. Everyone knows this is arbitrary. Everyone knows the person who finished on March 2nd worked just as hard as the person who finished on February 28th. The bonus structure broadcasts a message: we value the appearance of precision over actual judgment.

People aren’t stupid. They see what you’re doing. You’re trying to substitute a formula for the discomfort of having to evaluate their work subjectively. And they resent it, because they know a formula can’t capture what they actually contributed.

---

The uncomfortable conclusion is that there is no system. There’s only judgment, trust, and the willingness to have hard conversations.

The founder has to make subjective calls about who did good work. The team has to trust that those calls will be fair. Every attempt to avoid this through structure is an attempt to avoid the vulnerability of saying: “I’ll evaluate you based on my honest assessment of your contribution.”

But trust doesn’t scale. In a team of five, everyone has seen everyone else tested. You know who delivers and who doesn’t. You know whether the founder keeps their word. Trust is based on observed behavior.

In a team of fifty, you’re relying on reputation and hearsay. You haven’t seen the founder handle hard decisions. You don’t know if “we take care of our people” is real or just something they say. So you protect yourself. You sandbag a little. You don’t invest more than you can afford to lose.

This is rational. This is healthy, even. But it’s also corrosive to the thing the founder is trying to build.

---

The startup ecosystem runs on a shared delusion because it has to.

Founders tell employees they’re building something important together. Employees pretend to believe it. Everyone knows that “together” obscures a massive asymmetry—in risk, in reward, in how much this actually matters to them. But the pretense is load-bearing. Without it, you can’t get people to stay late. You can’t get them to push through hard problems. You can’t create the intensity that startups require.

Strip away the fiction and you’re left with naked transaction. I pay you to do work. You do the work. Neither of us pretends it’s anything more. Maybe that’s more honest. But it also doesn’t build anything. Nobody ever changed the world through naked transaction.

So we maintain the fiction, knowing it’s a fiction, hoping we can succeed fast enough to outrun the contradictions.

---

What would radical honesty look like?

Maybe something like this: “I’m going to care about this company more than you do. That’s okay. I’m not asking you to be a co-founder. I’m asking you to do good work for fair pay. If we succeed, you’ll benefit. If we fail, I’ll help you land somewhere good. Let’s not pretend this is something it isn’t.”

I don’t know if this works. Maybe people need the myth. Maybe “we’re changing the world” is a necessary lie that enables coordination. Maybe stripping it away just leaves everyone feeling cynical and mercenary.

But I suspect there are people who would prefer this deal. People who are tired of performing passion they don’t feel. People who would rather have a clear transaction than a fake family. People who would actually trust a founder more for admitting the asymmetry rather than papering over it.

The current system selects for people who are either true believers (rare) or good actors (common). A more honest system might select for people who are simply professional—who do good work because it’s what they do, not because they’ve bought into a vision they secretly doubt.

---

I don’t have a solution. I’m not sure there is one.

What I’ve learned, being on both sides, is that the only thing that actually works is smallness. Keep the team small enough that trust can function. Small enough that everyone sees everyone else’s contribution. Small enough that the founder can make judgment calls and the team can trust those calls because they’ve watched the founder be fair.

This limits what you can build. It rules out certain kinds of ambition. But it also rules out the worst of the dysfunction—the quiet sandbagging, the performed passion, the slow erosion of trust that happens when the fiction becomes too obvious to maintain.

Maybe the real lesson is that startups work despite their structure, not because of it. The ones that succeed are the ones where the people involved happen to genuinely like and trust each other, where the work is interesting enough to override rational calculation, where everyone gets a little lucky.

And the ones that fail—which is most of them—fail for many reasons. But underneath the stated reasons, there’s often this: the shared delusion couldn’t hold. The gap between what was promised and what was real became too large. And once people see that gap, they can’t unsee it.

---

The founder lies awake at night, alone with their anxiety, wondering why no one else seems to care as much.

The employee updates their resume, just in case, wondering if they’re a sucker for giving so much to something that isn’t theirs.

Both are right. Both are trapped. And the startup churns on, powered by a fiction that everyone half-believes and no one can afford to question out loud.