Somebody brings you a problem. You know the answer before he finishes explaining it, so you give it to him.
Ninety seconds. The job moves. He leaves with a decision and you go back to the other four things waiting.
It feels efficient because it works.
It also teaches him to bring you the next problem.
That's the trade nobody prices. The answer is the cheapest thing you can hand somebody. It solves the moment and leaves the person exactly as capable as he was before he walked in.
What it costs to answer everything
A company where every hard call routes to the owner has a ceiling, and the ceiling isn't the owner's ability. It's his hours.
Decisions wait, which turns a two-minute call into a two-day delay on a job with a schedule. Good people stop growing, because they can execute anything and decide nothing, and the ones with the most capacity have the most reason to leave. And eventually the owner can't step away at all, because everything requiring judgment requires him.
None of that's a delegation problem. It's a judgment-transfer problem, and they aren't the same thing.
Judgment doesn't transfer by being explained
You can teach a person to read a schedule. You can't teach him to feel that a schedule is lying.
You can teach someone to run a job-cost report. You can't teach him what happens in an experienced stomach when a number looks fine and something about it still isn't.
That feeling is judgment, and it got built by being wrong about a schedule before, being wrong about a number before, and remembering what each one cost.
Hand somebody a conclusion and he gets the conclusion. He doesn't get the machinery that produced it, and he can't build his own until he's made decisions, watched them meet the work, and lived with what followed.
A decision has to carry weight before anybody learns whether the reasoning holds.
Which means the transfer requires something that's hard to say out loud: letting somebody be wrong on purpose.
Why "delegate more" doesn't fix it
The standard advice is to delegate, hire better, or start a mentoring program. Each one misses.
Delegation moves tasks. Judgment isn't a task. Handing over the work while keeping every decision produces a faster executor, not a decision maker.
Hiring better buys judgment that was built somewhere else. That helps, and it doesn't transfer anything. It also costs more and leaves when someone offers more.
A mentoring program creates a calendar. Monthly meetings about development don't develop anybody. Judgment gets built inside real decisions with real consequences, not in a conference room on the third Thursday.
The thing that actually works is smaller and harder. Give somebody a decision with a real consequence, sized so a wrong answer is a lesson instead of a wound.
That's a judgment about somebody else's judgment, made in advance, with company money. It's the actual work, and there's no shortcut through it.
The tool
Four parts. Run them in order.
1. Size the decision before you hand it over
Score the decision, not the person. Five questions, yes or no.
| Question | Hand it over if the answer is |
|---|---|
| Can the company absorb the worst realistic cost without changing a plan? | Yes, and you can name the number |
| Can it be undone or contained? | Yes, or the damage stops where it starts |
| Who sees it if it fails? | Nobody outside, or one client with credit in the bank |
| How fast would you find out? | Days, not months |
| Does he have the information to decide well? | Yes, or one conversation closes the gap |
Four or five yeses. Hand it over whole.
Two or three. Hand over a piece. Let him do the analysis and the recommendation while you keep the final call.
Zero or one. Not yet, and be specific about what has to change first. "Not yet" without a reason is just no.
These thresholds are a starting point, not a formula. Adjust them to your own tolerance and your own company, and expect the first few to be miscalibrated.
The scoring matters more than the score. An owner who can't name the number in question one has found something more important than the delegation question.
2. Define the line before he needs it
Write down what comes back to you no matter what. Four or five items, specific.
- Anything above a dollar threshold you name
- Anything that changes a client's price or schedule
- Anything involving a safety decision
- Anything a lawyer, CPA, or broker should see
- Anything he isn't sure about
That last one has to be said out loud and said as a welcome, not a permission. Otherwise he'll guess rather than look uncertain, and a guess made to protect his standing is the most expensive kind.
A line defined in advance is what makes the rest of this safe enough to do.
3. Say the reasoning twice
Before. Ninety seconds, not a lecture. Here's what I'd weigh. Here's what I'd watch for. Here's the thing that would make me walk away.
After. Here's what actually happened. Here's what got misread. Here's what was surprising.
The second conversation matters more, and it's the one that gets skipped. A decision reviewed after the result is known is worth several decisions made and never examined.
Hold it whether the outcome was good or bad. It matters most when the result was good for the wrong reason, because success hides weak reasoning better than failure does.
4. Move him up the ladder
Four rungs. The signal to advance is the shape of what he brings you, not time served.
| Rung | What he brings | You do |
|---|---|---|
| 1 | A problem | Decide, and explain the weighing |
| 2 | A problem with a recommendation | Decide, and say where you agreed and where you didn't |
| 3 | A decision, before he acts | Confirm or redirect, then review after |
| 4 | A decision you learn about in the normal course of work | Review only when the result is instructive |
The move from rung 2 to rung 3 is the one owners stall on, because it's the first rung where being wrong costs real money.
When somebody stalls. If a person sits on the same rung for months, check these three causes before concluding he's reached his limit. They call for different responses.
| Cause | What you'll see | What to do |
|---|---|---|
| The line isn't clear | He asks about things well inside his authority | Rewrite the line, more specific |
| He's missing information | His recommendations are reasonable but built on the wrong facts | Fix the access, not the person |
| He's avoiding the risk | He recommends and then waits to be told | Hand him one decision anyway, sized small |
Time on a rung isn't the diagnosis. What he brings you is.
When to move somebody down. It happens, and doing it badly costs more than the original mistake.
Move down for a pattern, not an outcome. One wrong call at rung 3 is the system working as designed. Three wrong calls with the same root cause is a signal he needs rung 2 again for a while.
Say it as a sizing decision rather than a verdict. The decision was bigger than it should have been, so the next few come back through you. That's a statement about your judgment, which is where the error actually was.
The arrival signal. It's unmistakable when it comes. He disagrees with you and has a case. Not an objection. A case, with the facts, the assumptions, and the tradeoffs named.
At that point the machinery is his, not borrowed.
Composite example
A contractor doing roughly seven million had a project manager with six years in who called about nearly everything, including a nine-hundred-dollar material substitution.
The owner read it as a confidence problem. It was a definitions problem. Nobody had ever told the PM what he was allowed to decide, so he asked about all of it, which was the rational response to an undefined line.
They wrote the line down. Under two thousand dollars and no schedule impact, decide and log it. Anything touching the client's price or date, bring it. Anything you're unsure about, bring it and that's fine.
Inside two months the call volume dropped by more than half. The PM made one call the owner would have made differently, on a sequencing question, and it cost about four thousand dollars in rework.
That four thousand bought a PM who now decides sequencing without calling.
It also bought something the owner hadn't expected. The mistake wasn't the price of the lesson. It was the price of finding out the line had been the problem all along.
What this doesn't do
It doesn't lower the standard. Sizing a mistake isn't tolerating sloppy work, and the review conversation is where that distinction gets enforced.
It doesn't apply to everything. Some decisions are too large, too regulated, or too permanent to be anybody's practice, and the sizing test exists to identify them.
It doesn't produce copies of the owner. Some instincts are principles and some belong to conditions that have changed, and a person who reasons well will eventually reach conclusions the owner wouldn't. That's the system working. What transfers isn't the answers. It's the habit of looking one step further down the chain than the problem in front of you.
And it isn't free. A mistake allowed will show up somewhere, and construction mistakes get poured, framed, installed, billed, photographed, and discussed in front of a client. Somebody pays for them, removes them, explains them, or lives with them.
There's no version where an owner keeps every decision, prevents every mistake, and still ends up with leaders who can carry the business.
Start with one decision
Pick one decision in the next thirty days that would normally come to you.
- Run the sizing test. Five questions. Write the answers down rather than deciding by feel.
- Write the line. What comes back regardless, including anything he's unsure about.
- Say the before. Ninety seconds on what you'd weigh and what would make you walk away.
- Let him decide. Then leave it alone.
- Hold the after. What happened, what got misread, what surprised him. Include where he was right and you'd have been wrong.
- Name the next rung. Say out loud what he'd have to show you to get the next decision without asking.
One decision. One line. Two conversations. Then again next month with something slightly larger.
Everyone in the company is already learning from what gets tolerated, corrected, praised, and ignored. The only question is whether they're learning to bring problems or learning to solve them.
Practical insight. Real context. Zero fluff.