Why I Run My Business as an Operating System

I don't have a team. I have a system, and three businesses run on it. For years I thought growth meant more people, until I saw that most roles I'd hire for were a decision, a rule and a handoff. So I stopped hiring people to carry those rules and wrote the rules down instead.

Can I run a business without building a team?

Yes, if each recurring decision gets a written rule, a named owner and a place where the result is recorded. Work then moves without anyone being asked to move it. When the business does hire, the hire adds capacity the founder chose on purpose, instead of covering a gap nobody had defined.

The Personal Operating System I Built to Run Three Businesses Without a Team

I don't have a team. I have a system, and three businesses run on it. Most of the work that would normally need people is handled by software I built and still supervise. I know that sounds like a boast, so I'll say how it happened: I built this personal operating system slowly, mostly out of frustration, after fourteen years of running businesses the usual way.

The technology is the smaller part of the story. What changed was the thing I decided to treat as the asset, and the rest of this piece explains that change.

The Frame I Ran on for Fourteen Years

For most of my career, growth had one shape: more output meant more people. I ran operations inside businesses doing hundreds of crores, and the instinct was always the same. A new problem showed up, and the fix was a new hire, a new team or a new layer of managers. If the work grew and the team didn't, people assumed something was broken.

That frame built most of the companies people admire, so I don't think it's stupid. It does hide a cost, though, and I only saw the cost when I stepped out of the operator seat and into the builder seat. Every person a business hires is also a dependency that someone now has to hold together.

The Uncomfortable Thing I Noticed

When I started building instead of executing, I looked closely at the roles I would normally hire for. Most of them turned out to be a decision, a rule and a handoff. Someone receives a trigger, applies a judgment and passes the result to the next person. In many of those roles the human was wrapped around three lines of logic that nobody had ever written down.

Once I saw that, I couldn't stop seeing it. I had been hiring people to carry rules I had never bothered to write. So I stopped writing job descriptions and started writing the rules themselves.

How the System Actually Works

The method is boring, which is why it works. When a task shows up more than twice, I write the rule down as plainly as I can and hand it to a subsystem instead of a person. Then I watch it, because I don't trust any rule on day one. For a few weeks it runs in the open next to my own manual version of the same task. If the two agree, the rule earns a little autonomy, and if they don't, it gets an approval step with a human in the loop, who is usually me.

Some rules never graduate. They stay supervised for good, because the judgment inside them is too subtle to write down. I'm fine with that, since the point was never to remove myself from everything; the point was to remove myself from the tasks that only looked as if they needed me.

Over time the rules that hold start to pile up. The business runs on them whether I'm at my desk or on a flight with no signal. That pile of tested rules is the operating system: a set of decisions that keep working while I'm away, which is a different thing from an app or a dashboard.

Why the Company Stopped Being the Asset

This is the shift that reorganised my thinking. I used to treat the company as the asset: the name, the people, the clients. Now I treat the operating model as the asset and the company as what it produces.

A company is people plus a name plus momentum, and it's fragile in one specific way, because it depends on the founder holding the middle together. An operating model is the set of rules and decisions that keep producing the right result when no single person is watching. A firm that has one can lose almost anything else and rebuild quickly. A firm that lacks one can show healthy revenue and still own nothing that scales.

None of this is an argument against people. Good people make an operating model sharper, and the judgment calls a system can't make are exactly where the best people earn their place. My mistake, for years, was treating headcount and capability as the same thing.

What This Changed in Practice

The clearest change is what I protect. I used to protect the org chart, and now I protect the rules. When something breaks, I no longer start by asking who dropped it; I ask which rule was missing or wrong. That question moves the diagnosis from blame to structure, and a structural fix tends to stay fixed.

The second change is speed. Decisions that used to wait for me now wait for a rule, and a rule doesn't sleep or take leave. I still get pulled in, but only where my judgment is the real bottleneck rather than a habit.

The third change is quieter. I have more room to think, because when the routine runs itself, the scarce thing is my attention rather than my hours. Attention is what I actually wanted back.

The Rule That Failed First

The first rule I trusted too early was a payment approval. I told a subsystem to clear any vendor payment under a set limit if it matched an earlier invoice. It worked for a week, and then it approved a duplicate, because the second invoice carried a slightly different reference and my rule only checked the amount. Nobody lost much money.

What I lost was the belief that a rule is finished once it runs. A rule is finished when it has failed in a few ways I didn't predict and I've decided what it should do in each case. So every rule I write now has an explicit unsure branch: when it can't tell, it stops and asks me. That branch is what makes the autonomy safe.

What I Refuse to Systematise

Once this starts working, it's tempting to point it at everything, and I've learned not to. Anything that sets direction, prices a new kind of work or decides which clients we turn down stays with me. Those calls have no stable rule underneath them, and being slightly wrong on them compounds over months. The system earns my trust on the repeatable middle of the work, so my attention is free for the few decisions that actually move the business. When a founder automates the wrong layer, the risk doesn't go away; it moves to a place nobody checks anymore.

The Principle

Most founders build a company and hope a system grows out of it. I built the system first and let the company come out of it. The order matters more than it looks. When the company comes first, the system stays an afterthought that never gets built properly, and when the system comes first, the company inherits its discipline from the start. By operating model I mean neither the org chart nor the software, but the set of decisions that keep working when the founder isn't in the room.

I'm three businesses in and still not sure where the ceiling is. Some weeks the system surprises me by handling something I'd have sworn needed a person, and other weeks it fails in a way that reminds me supervision isn't optional. I've accepted both kinds of week, because that tension is the job.

If your business disappeared tomorrow but your way of running it stayed intact, how much would you actually lose?

Pick a time that suits you.

Book a call