Delegation: How Founders Escape the Firefighting Trap

There is a stage most founders reach where the day is full and nothing has moved. Every hour goes to a decision someone else could not make, a fire someone else could not put out, an approval only you can give. The business is busy and the founder is the reason it cannot grow.

The advice is always delegate more. The reason it does not work is that most founders have already tried, it went badly and they quietly concluded it is faster to do it themselves. They are right about the first time. They are wrong about what it means.

Why the first attempt fails

• The task was handed over, but the standard was never described, so the output was judged against a bar nobody could see.
• The authority was withheld. Someone was made responsible for an outcome and required to ask permission for every step towards it.
• It was given away at the worst moment, mid-crisis, to someone learning it for the first time.
• It was taken back at the first mistake, which teaches everyone watching that ownership here is temporary.

Delegate the decision, not the task

Handing over tasks produces someone who needs you constantly. Handing over decisions produces someone who does not. The unit of real delegation is a category of judgement with a boundary: all discounts under ten per cent, all refunds under five thousand, all hiring below a level, decided without asking.

That boundary is what makes it safe. It caps the size of any mistake, so the founder can genuinely stop checking, which is the entire point. Approval on everything is not delegation, it is supervision with extra steps.

A sequence that works

• Write down what good looks like before handing anything over. If you cannot describe the standard, you cannot fairly judge the result.
• Do it together once, then watch them do it once, then leave.
• Set the boundary explicitly: what they decide alone and what still comes to you.
• Agree how you will know it is going well, so checking is a number on a schedule rather than an interruption.
• Let the first small mistake stand and fix the system rather than the person.

Start with what is frequent, not what is easy

The instinct is to hand over the trivial things. Those are the ones costing you the least. Hand over the decisions that recur most often, because those are the ones filling the day and a decision made twenty times a month is a decision worth teaching properly.

The test of whether it has worked is uncomfortable and simple. Take ten working days off and change nothing else. What breaks is the map of everything still routed through you.