The thing you were good at stops being your job. Founders who do not notice become the bottleneck they built.
10 min read · Level 4
The thing you were best at stops being your job, and the transition is uncomfortable in a specific way: you move from work you are demonstrably good at to work you are new at, where progress is slow and mostly invisible. That discomfort is why founders unconsciously retreat to the comfortable work — and the retreat is what turns them into the bottleneck they built.
The founder's job changes substantially at each stage:
1–3 people — do the work. Build, sell, talk to customers. 4–10 — set direction and make sure people are unblocked. Still hands-on. 10–30 — hire, manage managers, communicate context, and decide what not to do. 30+ — direction, capital, key people, culture. Almost none of the original work.
Because the skills that got you here actively work against you later. A founder who keeps writing all the code at thirty people is a bottleneck, and the company slows to their throughput.
And because the transition is uncomfortable in a specific way: you go from doing work you are demonstrably good at to work you are new at, where progress is slow and less visible. Many founders unconsciously retreat to the comfortable work.
A technical founder at three people writes most of the code. Correct.
At twelve people they still review every pull request. Now every engineering decision waits for them, they are the constraint on all four engineers, and they have no time for the things only they can do — hiring, direction, key customers.
The fix feels like a demotion and is a promotion: hand code review to the strongest engineer, spend the recovered time on hiring and direction.
What makes this hard is that code review is measurable, immediate and satisfying, while hiring is slow, ambiguous and mostly rejection. The comfortable work is comfortable precisely because you are good at it.
The test: list what only you can do. If most of your week is not on that list, your job has changed and you have not.
First-time founders often hold on to their original work too long, usually justified as 'I'm faster at it'. You probably are — and that is not the point. Your being faster caps the company at your speed.
The second mistake: swinging too far the other way after reading about delegation, and disappearing from the product entirely. Founders who lose contact with the product and customers make worse decisions, not better ones.
The third: not telling the team the job has changed. People keep bringing you decisions because nobody told them not to.
Direction, key hires, culture, major customers, capital. Short list.
Compare against that list. The gap is the problem, and it is usually large.
Counter-intuitive and usually correct. It is the highest-value handover because it is the biggest bottleneck.
Explicitly. 'Take these decisions to X now, not me.' Otherwise the old routing continues by default.
A few customer calls a month, using your own product weekly. Losing this makes your decisions worse.
The job changes at roughly 5, 15, 30 and 100 people. Each transition needs a deliberate reset.
At every significant growth step, and whenever you feel like the bottleneck.
Do not delegate direction, culture or key hires. Those remain yours at every size, and handing them over is how companies lose their identity.
At three people, correct. At twelve, every engineering decision waits on one person, four engineers are throttled by their availability, and the things only the founder can do — hiring, direction, key customers — get no time at all. The handover feels like a demotion and is a promotion.
Hand over the thing you are best at. It is counter-intuitive and it is usually the biggest bottleneck precisely because you are good at it.
Founders who read about delegation sometimes disappear from the product and customers entirely, and their strategic decisions get worse rather than better. Losing contact with what customers actually experience removes the input that made their judgement good in the first place.
Keep a deliberate, small line open — a few customer calls a month, using your own product weekly. Delegate the work, not the contact.
Create a free account to read the rest and apply it to your own startup.