Not building. In the first year the job is finding out what to build, getting the first customers, and not running out of money — in that order.
10 min read
The founder's job is the work you are worst at, which is precisely why it does not get done. Building is comfortable, measurable and immediately rewarding; talking to strangers who might say no is none of those. Six months can pass in which every day felt productive and nothing was learned. The discipline is not working harder — it is noticing which of the three jobs your calendar is quietly avoiding.
In the earliest stage a founder has exactly three jobs:
Everything else is either in service of those three or is a distraction wearing a convincing costume.
Because technical founders in particular will default to the part they are good at, and building is the most comfortable, most measurable and most avoidable of the three.
You can spend six months building and feel productive every single day while making no progress on whether anyone wants it. Progress is not the same as motion.
A technical founder tracks their week honestly and finds: 34 hours coding, 3 hours on the website, 1 hour on a customer call, 0 hours on finances.
They feel like they had a productive week. But of the three jobs, they made real progress on none: they built features nobody had asked for, spoke to one person, and did not know their own runway.
The next week they invert it — 6 customer conversations, 2 hours on the numbers, and only the code those conversations justified. It feels less productive and moves the company far more.
First-time founders often believe that once the product is good enough, customers will follow. Products do not sell themselves at the start — the founder does, personally and awkwardly, for the first several dozen customers.
The other one: treating finance as something to sort out later. Later is when you find out you had six weeks of runway and did not notice.
Not where you meant them to go. Write it down honestly. Most founders are shocked by the split.
Three a week, minimum, at the early stage. Book them before the week fills up, because it always fills up.
Cash in the bank, monthly burn, months remaining. Fifteen minutes. Never be surprised by this number.
Some things genuinely require the founder — talking to customers, deciding direction, hiring. Some do not. Guard the first category ruthlessly.
It drifts back towards comfortable work every single time. Expect that and correct it rather than being disappointed by it.
Every week at the early stage. The balance shifts as you grow — later the job becomes hiring, direction and capital — but the first-year version is these three.
Do not apply this rigidly once you have a team. A CEO of thirty people who spends all week on customer calls has stopped doing their actual job, which by then is people and direction.
In Airbnb's early days the founders travelled to New York to meet hosts in person and photograph their listings themselves, because the listing photos were poor and bookings were low. It did not scale, it was not engineering, and it is one of the most cited early-stage decisions in the industry.
The unscalable manual thing is often the fastest route to learning what is actually broken. Doing it yourself is the point, not a stopgap.
The shape almost every early technical founder finds when they track a real week: thirty-plus hours building, one or two hours with customers, zero on the numbers. Nobody plans this. It happens because building is the only one of the three that never requires you to ask anyone for anything.
Track one week honestly before deciding you are busy. The split is almost always worse than the founder expects, and it is fixable with three calendar entries.
Create a free account to read the rest and apply it to your own startup.