Nobody Warns You That Focus Is Also a Risk

Every founder gets the same advice within the first six months: pick one thing and go all in. It's good advice, and I mostly ignored it. Antialias has run for over a decade, Hacoco is newer and growing, Nikhaar Foundation has been quietly working on water and children's welfare for years, and I just put out a book. None of these are side projects pretending to be serious. They're all real, at the same time, on purpose.

I want to make the case for that, carefully, because the usual version of this argument is a humble brag dressed up as a philosophy. This isn't that. Running more than one thing has real costs. I've paid most of them.

What focus actually protects you from

The advice to focus exists because it's true for the failure mode it's designed to prevent: a founder who spreads six units of energy across six ventures and delivers one unit of output to each, badly. That founder is common, and the advice is correctly aimed at them.

But the advice assumes a specific shape of person, someone whose best work comes from narrowing everything down to a single obsession. I'm not that shape. My best work has always come from having somewhere else to put my attention when the first thing stalls. A hard week at Antialias doesn't spiral, because there's a Hacoco decision that needs me at the same time, and switching to it is closer to recovery than to distraction.

That's not a universal truth about entrepreneurship. It's a fact about how I'm specifically built, and I spent years assuming it was a flaw because it didn't match the advice.

The single-focus founder and the parallel founder aren't running the same race differently. They're running different races, and the coaching for one is often exactly wrong for the other.

The costs, honestly

None of this works without structure, and I don't think it works forever without a real cost somewhere.

Context switching has a tax, and I pay it daily. The version of me that walks into a Hacoco call twenty minutes after an Antialias client meeting is not as sharp as the version who'd been thinking about nothing else all morning. I've made peace with that tax. I haven't eliminated it.

Depth suffers in ways that compound slowly. There are things I would understand more deeply about any one of these ventures if it were the only one. That depth matters more in some businesses than others, and I've had to be honest about which of mine actually need it and which don't.

People assume you're not serious, and some of them are right to wonder. The market has a strong prior that a multi-venture founder is really just a single founder who hasn't committed yet. Sometimes that prior is correct about someone. Proving it wrong about yourself takes years, not a sentence in a bio.

What actually makes it work

I don't think parallel ventures work by sheer will. They work because of a few unglamorous mechanics.

Each venture has its own person running daily operations. I'm not the daily driver anywhere except in the earliest months of a new one. My job is direction and unblocking, not execution, and confusing those two roles is the fastest way to turn parallel ventures into four badly run ones.

I protect specific hours, not specific days. A calendar split by venture per day sounds clean and rarely survives contact with reality. What actually works is smaller and stranger: certain hours are always Antialias hours regardless of what day it is, because that's when the people I work with there are actually online and thinking clearly.

I let the ventures be different sizes of important. Nikhaar Foundation doesn't need me at the intensity Hacoco does right now, and pretending otherwise, giving every venture equal daily attention out of some idea of fairness, would just mean shortchanging the one that actually needs the hours.

The honest version

I don't think everyone should run four things. I think most people shouldn't, and the standard advice is right for most people. But there's a smaller group who've been told for years that their instinct to build in parallel is a discipline problem, when it's actually just a different operating system, and nobody wrote the manual for it.

This is me writing a page of that manual, for anyone who's been quietly doing this and wondering if they're the only one who never managed to want just one thing.