Every founder eventually runs the same experiment. Someone on the team is clearly running on empty, so they are told to take a proper break. Two weeks later they come back tanned, and within four days they look exactly as they did before they left.
The obvious conclusion is that the person did not rest properly. The more useful conclusion is that rest was never the missing ingredient.
Table of Contents
Tiredness and burnout are not the same problem
Tiredness responds to sleep. Burnout does not, because burnout is not primarily a deficit of energy it is a change in the relationship between effort and meaning.
The recognisable version in a small company looks like this. The work still gets done, often to the same standard, but the person has stopped expecting it to matter. They no longer argue in planning meetings. They stop proposing things. They answer messages promptly and say almost nothing in them. Managers frequently read this as improved professionalism, right up until the resignation.
A holiday removes the load for a fortnight. It does nothing about the mechanism that produced it, which is why the effect wears off almost immediately on return.
What actually produces it in a team of ten
In large organisations, burnout is usually discussed in terms of workload. In small teams the workload is rarely the deciding factor, because people who join small teams generally expect to work hard. Three other things do more damage.
Unbounded responsibility. In a company of ten, everyone owns everything a little. That feels like trust at first and like a permanent low-grade alarm later, because nothing is ever definitively not your problem.
Invisible switching. A developer who is also handling support, and occasionally sales, is not doing three jobs badly. They are paying a cognitive toll every time they change context, and that toll is entirely invisible in any measure of output.
No completion. Startups rarely finish things; they move on from them. Without the moment where something is done, effort stops producing the small closure that makes it sustainable.
None of these are solved by time off. All of them are structural.
What helps instead
The interventions that work are unglamorous and mostly involve subtraction.
Give people a domain that is definitively theirs, and harder domains that are definitively not. The relief comes less from ownership than from the permission to stop thinking about everything else.
Protect one uninterrupted block per day per person. Not a policy about focus time; an actual agreement that between specified hours, nobody is expected to answer anything. In small teams this needs to be stated out loud, because the default assumption is instant availability.
Mark completions deliberately. A shipped feature, a closed quarter, a finished migration. Five minutes of explicit acknowledgement does more for sustainability than a company offsite, because it restores the link between effort and outcome that burnout erodes.
And watch for withdrawal rather than exhaustion. The person to worry about is not the one visibly struggling it is the one who has gone quiet and agreeable.
A fuller description of why rest alone does not restore capacity, and what does, is set out in this piece on burnout recovery.
The founder problem
Founders are unusually bad at spotting this in themselves, for a structural reason: they cannot step outside the work to see it, because the work does not have edges. There is no point at which a founder is definitively off duty, so there is no contrast against which tiredness becomes visible.
The practical test is not how tired you feel. It is whether you can still imagine the company in two years and feel anything about it. If the picture is clear but flat, the problem is not sleep.
Small teams rarely lose people to overwork alone. They lose them to a long stretch in which effort stopped connecting to anything, and nobody noticed because the output never dropped.
