When should I hire my first employee — hiring the first assistant
Hire against a defined outcome, never against a task list. Write down what "done well" means, what the reporting rhythm is, and which exceptions come back to you — before you post anything. That decision is worth making before you spend another month or another hire on it.
Why this happens
Hiring starts before the outcome is defined. No written definition of the outcome, so no way to judge a candidate or their work. Until that layer is fixed, every extra task, tool or person amplifies the same problem instead of solving it.
The signals that confirm it
- Quality swings wildly between similar tasks — ask yourself: is one deliverable excellent and the next one unusable?
- You are redoing work that should have been right the first time — ask yourself: do tasks take three rounds of revisions before you give up and do them yourself?
- You cannot measure your VA’s actual impact — ask yourself: could you prove with data that your va is worth what you pay?
- Feedback feels awkward, so it gets avoided completely — ask yourself: do you notice problems and say nothing for weeks?
What to change first, this week
- Build one quality checklist for your most common task type and use it on the next three jobs. (Checklists are not restrictive, they are clarifying. Guessing stops, consistency starts.)
- For the next complex task, record a three-minute video walkthrough and attach an example that got it right. (Visual instructions remove roughly 80% of revision rounds. The first submission lands close to done.)
- Choose two metrics today — tasks completed, hours removed from your calendar, tickets resolved — and review them every Monday. (Delegation stops being a cost and becomes an investment you can actually judge.)
- Calendar a recurring 15-minute weekly review. Use numbers, not feelings: "45 tasks completed, target is 60." (Small weekly corrections replace big quarterly confrontations — and the work actually improves.)
The rule that decides it
The BBW method. Score each task on founder value, repeatability and automatability. High founder value stays. High repeatability with low founder value gets automated. Judgment-light but human-shaped work gets delegated. Low value with low impact gets eliminated.
What good looks like in four weeks
A role brief with owned outcomes, a quality bar and a paid test assignment. Hiring before diagnosis moves chaos onto someone else's desk and calls it help. First find out which work should leave your plate at all — then decide who or what executes it.
The objection worth answering first
"My business is too specific to delegate" The judgment is specific. The steps around it rarely are. Split every task into decision and execution — the decision stays, the execution moves.
Common follow-up questions
Where should I start?
With the task that has the most recurring hours and the lowest cost of error. Finish it end to end — documented, handed over, running without you — before you start a second one.
Do I need to hire someone before fixing this?
No. Hiring into an undefined outcome is the most expensive way to discover the outcome was undefined.
How long does this take?
One focused week for the first system. The compounding starts when the second one takes half as long because the first is written down.
The chain behind this
- SymptomQuality swings wildly between similar tasks
- CauseYou do not have an inconsistent VA. You have undefined standards — they are guessing what "good" means.
- BlockageNo written definition of the outcome, so no way to judge a candidate or their work.
- What good looks likeA role brief with owned outcomes, a quality bar and a paid test assignment.
Which of the four systems is missing in your case?
Four questions, no email. You get the constraint, how sure we are, and the one thing to do next.