Delegation

Definition
Handing ownership of a task or an outcome to someone, or something, else, so it gets done well without you doing it yourself.

Why it matters

Without delegation, the founder is the bottleneck on everything. Growth stalls at the number of hours one person can work, and the best use of their time, such as strategy and key relationships, gets squeezed by tasks others could own.

The usual failure is delegating the task but not the trust: hovering, re-checking and taking the work back the moment it looks different from how you would have done it. That costs more than doing it yourself, and the other person learns nothing.

Done well, it also builds the team. People who own an outcome learn to make decisions, and the founder gets back time for the work only they can do.

How to apply it

  • Write a clear definition of done before handing anything off, so success is not a judgement you make afterwards.
  • Give the person, or the system, the access and authority needed to reach the outcome.
  • Agree how the result will be checked. Review the outcome, not every step taken to get there.
  • Let the first attempt run before stepping in.
  • Start with something reversible and low in stakes, then widen what you hand off as trust builds.

What it is

Delegation is more than handing over a task. Real delegation transfers responsibility for the result: the other party owns the outcome, decides the reversible details themselves and comes back only for the calls that cannot be undone.

It helps to think in levels. At the lowest, the person researches and reports back. Next, they recommend an action and wait for approval. Then they act and report afterwards. At the top, they act and only tell you about exceptions. A good delegator picks the level on purpose and says which one applies.

Common mistakes

  • Handing over a task without the authority to complete it.
  • Accepting "reverse delegation", where the problem keeps bouncing back to you.

Delegating to software

The same rules apply to automation and AI agents. Describe the outcome, grant only the access it needs, and decide in advance which decisions stay with a person. A founder who answers every refund query herself might write the refund rules once as an SOP, so a new hire can follow them. For a repetitive step such as routing incoming leads, she could build a workflow in n8n and review only the exceptions it flags.

Worked example

Suppose a founder of a twelve-person firm wants new enquiries answered without her touching each one. She writes the outcome first: every enquiry gets a reply within an hour, the CRM is updated, and anything about pricing comes to her. She then sets up a Zapier Agents agent with access to the shared inbox and the CRM, and nothing else. She tells it to handle routine replies and report back on anything else.

In the first week she reviews results rather than every draft, and she steps in only when an answer is wrong in a way that matters. After two rounds of changes to the instructions, the agent handles most routine enquiries and she is only pulled in for pricing questions. Handing over the outcome, with clear limits on what the agent can touch, is what made it workable.

Tools in the example

Some links are affiliate links: we may earn a commission at no cost to you. It never decides a ranking. How we work with partners

  1. Article

    Handoff

    The moment work passes from one person or seat to another.

  2. Article

    Standard Operating Procedure (SOP)

    The written steps that make a task safe to delegate.

  3. Article

    Accountability Chart

    The map that shows who a task should go to.

  4. Article

    RACI

    A simple grid for who does, approves and is told about a task.

  5. Article

    Human-in-the-loop

    Keeping a person in charge of key decisions when work is delegated to AI.