I've always wanted to build and sell my own software.

But I imagined nasty things like taxes, legal, support, and maintenance coming my way. I would be taken to jail for accidentally violating some esoteric tax regulation. My favorite part was always building, and the employee trade-off of having more constraints and no grown-up responsibilities seemed like a great deal.

Not only did it seem like a great deal, it WAS a great deal. Most of my years as a wage-slave have been quite enjoyable. Like Nassim Taleb says, "The three most harmful addictions are heroin, carbohydrates, and a monthly salary."

Running a niche software business in the cloud now requires so ridiculously little overhead that it is almost a rounding error. Most of the taxes, accounting, compliance, payments, servers, networking, databases, and marketing can be automated away if you are willing to pay for it. Take my money!

There is another shift, too: generative AI. It now writes most of my code, helps communicate with customers, and even paints pretty pictures. None of the output is of great quality, but the bang for the buck is fantastic. Pouring energy into high-quality artisan output is mostly not necessary. I find that a few things matter a great deal, while most things don’t matter at all.

First of all, organizations are absolutely vital. The greatest things in modern life are generated by large organizations. It’s the only way to make a dent in the universe. But a visit to an average-sized meeting of an average-sized company quickly reveals how little actual customer value is being generated on average. In my native language, there is a saying, “joukossa tyhmyys tiivistyy,” which roughly translates to “the stupidity of the crowd.”

I have news for you, though. If you are willing to be lonely, you will get away with murder! Flying solo is an unfair advantage.

A solo operator never attends weekly meetings, asks for a budget, interviews candidates, creates presentations, does code reviews, defines onboarding processes, pitches to investors, fights with co-founders, or books Christmas party venues. When I think about the competitors in my market, I imagine them in a brain-dead meeting, choosing the font size for their next slide, trying to make sense of someone else's code, or preparing for their 47th investor pitch. I have the unfair advantage of not doing any of those things.

A strong argument for creating value collaboratively is specialization. The marketing person focuses on marketing, the salesperson on sales, and the engineers on engineering. What people forget is that in an average company, the engineer doesn’t actually engineer that much. The bigger the organization, the more time the engineer spends on meetings, Slack, code reviews, interviews, emails, and Christmas parties. When they DO write code, it needs to be polished for collaboration: clean, structured, unit-tested, and according to best practices. Why? Because another engineer will eventually look at it.

Now, picture all your potential tasks sorted by priority. Optimally, employee #1 works on the highest priority, employee #2 on the next, and so on down the list. If the top marketing task ranks 263rd, you can bet employee #10, specialized in marketing, is still doing it, no matter what. In a solo operation, you focus only on the most important task. Period. This truly cuts through the bullshit. But how do you know what the most important task is? You don’t really, but that’s a different problem. And who is more equipped to figure it out? Is it the organization where everyone is in their specialized silo or the solo operator doing everything?

To be honest, I'm not really arguing for solo operations. I’m arguing for smaller operations. I believe software products should be built by the absolute minimum number of people possible.

Increasingly, as technology advances, that number is one.