What should an entrepreneur delegate first?

Whatever is high-frequency, low-risk and easy to check. Routine purchasing, scheduling, standard customer responses, stock decisions within a limit. Frequency builds the other person’s competence quickly and gives you back real time.

How do I delegate when my staff are not skilled enough?

Usually the constraint is not raw ability but the absence of a written standard and a boundary. Write down how the work is done, set the limit within which they decide, then review. If capability is genuinely missing after that, the answer is training or a different hire — but check the system first, because it is the cheaper fix.

Is delegation different when the employee is a relative?

The mechanics are identical; the difficulty is that the boundary and the standard have to be explicit and written, because family relationships rely on implicit understanding that does not translate well to work. See managing family members in your business.

How Entrepreneurs Can Learn to Delegate (Without Losing Control)

Every founder has tried delegating and had it come back worse. The work was late, or wrong, or done in a way that had to be redone. The reasonable conclusion — faster to do it myself — is exactly how businesses end up permanently the size of one person.

The failures are usually structural rather than personal. Below is what tends to be missing.

What most delegation attempts leave out

A transfer of responsibility needs four things. Missing any one of them reliably produces the outcome founders complain about.

The decision, not just the task. "Handle the deliveries" is a task; it returns to you the first time something unusual happens. "You decide the delivery schedule and you can spend up to 50,000 shillings solving a problem without asking" is a decision. Only the second one reduces your load.
A boundary. The amount, the circumstance, the threshold beyond which they must come to you. Without a stated boundary, people either ask about everything or make a call you would never have made — and both outcomes get blamed on delegation.
A visible standard. What does a good job look like, in terms they can check themselves against before you see it? If the standard exists only in your head, the person is being asked to guess and then criticised for guessing wrong.
A review date. A specific day when the two of you look at how it went. Not "come to me if there are problems", which puts the entire burden of judgement on the least experienced person involved.

Accept that the first attempts cost you

The honest arithmetic: delegating something well takes roughly three times as long as doing it yourself the first time, twice as long the second, and then less time every time after that.

Founders abandon delegation in the first two rounds, when the cost is real and the return has not arrived. Knowing the shape of that curve in advance is most of what it takes to push through it.

You are not paying for the work to be done. You are paying for the capability to exist in someone other than you.

Start with frequency, not with importance

A common mistake is to delegate the big annual thing — the audit, the licence renewal, the major tender — because it is the biggest weight. Do the opposite. Delegate what happens often, because frequency is what builds competence and what saves you meaningful time.

Ten small decisions taken weekly by someone else will change your calendar more than one large one taken annually, and they will make that person genuinely more capable within a month.

Handing work back on purpose

When the person comes to you inside their boundary — and they will, out of habit — the temptation is to just answer. It takes nine seconds and moves the day along.

Answer once and you have quietly taken the decision back. The alternative is a sentence you will use often: "That is yours now. What do you think we should do?" Then wait. Then, unless it is genuinely damaging, back their answer even if yours would have been slightly better.

That last part is where most founders struggle, and it is the whole game. Being right is worth less than someone else becoming able.

The "slightly worse" rule

If a delegated decision comes out 80% as good as yours, that is a success, not a problem — as long as it is trending upward. A business run entirely at your 100% has a ceiling of one person. A business run at 80% across five people does not.

When delegation genuinely should not happen yet

Some work should stay with you. Anything where a single wrong call is existential — a bank guarantee, a major contract term, a decision about a person’s employment. Anything where you are the only relationship the counterparty trusts, until you have deliberately introduced someone else. And anything you cannot yet describe well enough to explain, which is a documentation problem rather than a delegation one.

That last case is common and worth naming: if you cannot write down how a thing is done, you cannot hand it over. Start with writing the process.

The signals that it is working

You hear about a decision after it was made, and you agree with it. Someone tells you they handled a problem you did not know existed. A customer is served well by a person you did not brief. Your week contains a two-hour block you did not have three months ago.

None of it feels dramatic. Delegation done properly is mostly the absence of things — the calls that did not come, the approvals that were not needed.