Mastermind · May 4, 2026
The three-hour rule
Last week I picked up a client. A thousand dollars per session for three hours of vibe coding on Tuesday mornings, on a six-week deal. We sit at a coffee shop near Stanford, walk through the project, and I answer questions while showing them how to ship things they didn't know were possible. So every week, I'm paid to teach for three hours.
Xisen pointed something out on our call this morning that I hadn't named clearly. "You're basically having this strategy, right, where we're sustaining ourselves with some work. Something like three hours for the whole week, and then for the rest of the time, you're accumulating." He's right.
Three hours of paid work covers your week. The other one hundred and sixty-five compound.
Most young founders treat the choice as binary. Either you're a full-time founder raising enough capital to live, or you have a job and the startup is a hobby. There's a much better third option, and most people skip it because it feels like a half measure. You take one client. Charge them rent-coverage rates. In exchange, you spend the rest of your week building things that don't pay yet.
Three hours a week is one morning. Most founders waste more time than that scrolling X for vibes. The trade is absurd if you actually do the math.
Tim Ferriss had a version of this twenty years ago. The Four-Hour Workweek had the math right and the destination wrong. Most readers took it as a script for beach life and outsourced inboxes. The more useful read: compress paid work to the minimum, then redirect the rest into the work that doesn't pay yet. For a founder, those freed hours aren't for leisure. They're for the product, the network, and the events that will pay back six months from now. Same minimization, different hammock.
Here's why this is different from "have a side gig." A side gig is a job that drains you. Three hours with someone you like, teaching things you'd be doing anyway, isn't a job. The vibe coding I show them is the same vibe coding I'd do for MentorMates. The architecture decisions we walk through are the same ones I'd be making for DayDreamers Academy. They're not paying for my time. They're paying for the path I was already on.
The other thing about three hours: it forces you to charge more. If three hours is all you have to give, you can't afford to give them away cheap. You start at the rate that covers your week, and you find the person willing to pay it. That's a much stronger filter than "find ten clients at one hundred an hour." Ten clients is ten relationships. One client at the right number is one relationship, run well.
I almost didn't take this gig. The previous version of me would have said it's a distraction from the real work, which right now means the incubator, the demo day, and the hackathon. But the real work doesn't pay anything yet. Founders Bridge is in week three. We have three companies in the cohort, demo day on Thursday, and a Saturday angel pitch with a group of angel investors. We have a hundred forty-seven applicants in the pipeline. None of that is cash. It's all setup. So the three hours isn't a distraction from the real work. It's what makes the real work possible. Without the three hours, the real work has a deadline against my savings account. With it, the deadline disappears.
There's a temptation, when you find a structure like this, to scale it. Take a second client. Then a third. Pretty soon you're a consulting business and the startup is the hobby again. The discipline is to refuse. The whole point is three hours, not six, not ten. You picked the number to cover the week. Every additional hour after that is a tax on the compounding work.
I think about this a lot in the context of the founders I'm currently incubating. Most of them are doing it the hard way: full time, no income, founder-timeline pressure, raising before they have anything to raise on. A few of them could pick up one client tomorrow and instantly buy themselves another six months of optionality. They don't, because nobody told them they could. The whole startup ecosystem is built around the binary, and if you ask permission to do the third thing, nobody gives it to you.
The mistake people make is assuming the constraint is finding the client. It isn't. The constraint is being good enough at something specific that one person, somewhere, will pay rent-coverage rates for it. Once you can do that, the rest follows.
This week I'm hosting the Stanford Founders Demo Day, closing pilots for a partner integration, locking in a co-sale deal with another incubator, and running my own demo day Thursday with a Saturday angel pitch. None of that is paying me yet. It's all setup. But Tuesday morning I'll spend three hours at a coffee shop, and that pays for everything else.
Find your three hours. Refuse the fourth.