Should I fire them or coach them?

You've been carrying this decision for a while. Probably months. You've had the conversation in your head both ways, and you keep landing on wait a bit longer.

That delay has a cost and it isn't mainly the person's salary. It's that everybody else can see it too.

The test

Before you decide anything about the person, answer three things about the role.

Did they ever have a clear definition of good? Not a job title. A written answer to what success looks like in this seat, with numbers or examples. If it only ever existed in your head, they've been aiming at a target they couldn't see.

Did they get told, early and plainly, that this wasn't working? Not hinted at in a busy week. Told. If the first time they hear it clearly is the termination, you skipped the part that was yours.

Do they have the authority to fix what you're holding against them? Owners routinely hold people accountable for outcomes they can't control, then read the resulting paralysis as a performance issue.

Three yeses means you have real performance data and you can trust your read. Any no means you don't have data yet. You have an unspecified job and an inference.

Where the honest answer usually lands

Most of the time, in a company under a few million, at least one of those is a no. That isn't a defense of the employee. It's information about what will happen if you replace them.

If you fire into an undefined role, you'll hire someone else into the same undefined role, and in eight months you'll be here again with a different name. That's the loop that makes owners believe good people can't be found. See why can't I find good people.

So the sequence matters. Define the role, say the hard thing clearly, give it a real window with a date on it. Then you'll know. And you'll know fast, usually inside six weeks, because a capable person given a clear target moves immediately and a mismatched one doesn't.

When it's actually time

Some situations don't need the diagnostic. Stop optimizing and act.

Character, not capability. Dishonesty, expenses that don't reconcile, lying about what happened. Skill problems are coachable. This isn't, and tolerating it once teaches the whole company what the standard really is.

They're costing you other people. If good employees are managing around one person, you've already made a decision, you just made it in their favor. Everyone else is watching you choose the difficult person over them, repeatedly.

Clear target, real support, no movement. You defined it, you said it, you gave it ninety days, the number didn't move and the behavior didn't either. That is your answer. Give it any longer and you're protecting your own discomfort.

You've stopped giving them real work. Watch what you actually do, not what you say. If you've quietly routed the important things around someone, you fired them a while ago and haven't told them.

The thing owners get wrong about the cost

Owners weigh this as though keeping someone is neutral and firing is the risky move. It's the reverse, at your size.

In a ten-person company one wrong seat is ten percent of the company. It sets the pace everyone else calibrates to. It decides what your strongest performer believes about whether excellence matters here. And it consumes your attention in a way that's almost impossible to see from inside, because the time gets spent in fragments.

Add up the hours you spent on this person in the last quarter. Checking their work, redoing their work, discussing them with someone else, thinking about this decision while driving. That's the actual price, and you've already paid it several times over.

If you do it, do it in a way you can defend

Not legally. Personally. You're going to see this person at a gas station in four years.

Say it directly and in one sentence, near the start. Don't open with fifteen minutes of context that leaves them hoping. Don't relitigate the history, it helps nobody. Be specific about what happens next, money and dates and logistics. Let them keep their dignity, including the option to tell other people whatever they need to tell them.

And take the part that was yours. If the role was never defined, you can say that out loud. It costs you nothing and it's true.

What this has to do with the numbers

People is the constraint owners name most often in the Platform One Index, 46 percent, more than marketing and finance combined. It's also the lowest scoring of the five areas we measure, at 32.9 out of 100.

But when owners describe the problem in their own words, the majority aren't describing a bad employee. They're describing work that only they can do and nobody they trust to take it. That's a specification problem wearing a personnel problem's clothes, and firing doesn't touch it.

Which is why the test at the top of this page comes before the decision, not after it. See is it a people problem or a process problem.

Start with the measurement

Twenty questions, about ten minutes, free.

You'll get scored in five areas and see which one your answers point at. If you're weighing a decision about a person right now, it's worth ten minutes to know whether the seat or the company is what's broken.