Your org chart was built by accident

Nobody designed your org chart. I've asked a lot of founders to show me one, and what comes back is almost always a picture drawn after the fact — a tidy diagram of a structure that actually accumulated, one urgent hire at a time, over three or four years.

That's not a criticism. It's how every early-stage company works. Something breaks, you hire the person who can fix it, and their job quietly becomes "the things that were breaking in the month we hired them". Do that fifteen times and you have an organisation shaped entirely by the sequence of your historical problems.

The problem is that your historical problems are not your current business.

How to tell it's happened to you

You don't need an audit. Four symptoms, and if three are true it's happened.

• Decisions climb to you that shouldn't. Not the hard ones — the ordinary ones, arriving in your inbox because two people each half-own the outcome and neither will call it.
• No single name sits against your most important number. Ask who owns revenue. If the answer is "well, sales and marketing and me", nobody owns it.
• Your best person is doing three jobs. One of which they're brilliant at and two of which they inherited during a crisis eighteen months ago and never put down.
• New hires take a quarter to work out who to ask. If the structure can't be explained to a competent adult in five minutes, it isn't a structure.

Rebuild it in the right order

The mistake is opening the existing chart and moving boxes. You can't fix it from inside it, because every box has a person's face on it and you will unconsciously protect them.

Do it in three passes, in this order, and don't let yourself skip ahead.

Pass one — write the outcomes. No names, no titles.

On a blank page, list the eight to twelve things that must be true in twelve months for this to have been a good year. Not activities — outcomes, with numbers. "Revenue at $4.2M." "Churn under 4%." "Series A closed." "Gross margin up six points."

This list is the actual shape of your company. Everything downstream serves it.

Pass two — one owner per outcome. Still no names.

Beside each outcome, write the role that owns it. One role. Not a committee, not "sales & marketing" — one seat that carries the number and reports on it.

You'll discover two things here and both are useful. Some outcomes have no plausible owner, which means you have a gap you've been absorbing personally. And some roles end up carrying six outcomes, which means that seat is really two seats and you've been asking one person to be two people.

Pass three — now put the names in

Only now. And be honest, because the whole exercise is worthless if you flinch at this step.

Most names land somewhere reasonable. A few land in a seat that's smaller than the one they hold today, which is a hard conversation you now have to have. One or two seats have no name at all, and that's your hiring plan — written from what the business needs rather than from what hurt most recently.

Hire people who make you the least impressive person in the room

This is the belief I've held longest, and it's the one founders push back on most: when I walk into a room and feel like the dumbest person there, I know I've hired right.

It isn't modesty. It's risk management. People who are better than you at their function will tell you when you're wrong, and they'll tell you early, while it's still cheap. Ego protects you from feedback, and feedback is the only thing that stops small failures from compounding into expensive ones.

An org chart built by accident is usually an org chart of people who were available and agreeable. An org chart built on purpose is uncomfortable to sit in, and worth considerably more.

What an investor sees

I've sat on the other side of this table — chairing an ASX-listed board, running a CEO search, and taking companies through diligence. Nobody asks to see the org chart because they're curious about your reporting lines. They're checking one thing: does this business function if the founder is unavailable for a month?

An accidental chart answers that question badly, and it answers it before you've said a word. Every outcome that traces back to you personally is a line in someone's risk memo, and it comes off your valuation.

The rebuild takes about two weeks of honest work. It is the highest-return fortnight available to most founders I meet, and it costs nothing but the willingness to look at the thing properly.