On building a new business

A friend is working on something new. We talked yesterday, and these ideas came up.

As with every single thing I ever do, this is not specific to business. This stuff is just a reasonable way to exist peaceably in the world. Building something new doesn't have to mean what it has meant.

Take it as far as you can before you involve someone else. Once you delegate its creation, it becomes someone else's interpretation of your idea.

This lets you stay close to the idea as it evolves — because an idea always does, the closer it gets to reality. No one is better able to carry it through that process than the one who picked it up first.

Also, this keeps you close to the practical realities of it. Working with folks who are detached from how it works is hard, and tedious. Better to work with folks who are close to the beating heart of a thing. (This is from the founder and CEO of a 14-year-old company with staff and whatnot who is still very much leading and implementing most engineering things.)

Only two things matter: (1) that it works today, and (2) that you want to return to the workspace tomorrow.

As long as that stays true, you are using your time extraordinarily well.

Be aware of which future-thoughts are fears about what could happen and which feel like being on the edge of your seat because it just might work.

Taking one step towards what just might work is much, much, much more reliably productive than trying to dodge what might not work. It's not always a choice between the two, but if you have to assess your options, see if this binary applies.

You're either creating or arresting your momentum, every time this choice comes up. Taking consistent steps toward what just might work is a good way to build the momentum of creation. Every time you switch strategies and take a step of fearful avoidance, you lose some of that positive momentum. The best stuff happens when you're at speed. Acknowledge your fears, but protect your creative momentum with the choices you make.

There is a track for doing something correctly or traditionally, and there is a track for doing something that works. Be clear with yourself about which one you're on.

It's not a moral decision. It's just preference. They're both valid. But they're super super different priorities, and your motivation for the work is gonna resemble more of one than the other.

You won't always have to choose between "correct" and "it works", but when you have to choose, stay awake to your priorities. It's not a sin to choose one over the other, but choose such that you'll want to return to the results of your choices tomorrow.

Sustainable from day one. That includes the money.

I'm autistic; I can't afford to create seasonal exceptions for myself in my patterns. I can't decide that for six months I'm gonna work unsustainably and then launch the thing I made at the sixth-month mark. Every day I work unsustainably, I get worse at making the thing.

So I don't do launches, or anything that requires unsustainable buildup for a hopeful payoff.


  1. I build super small, super simple things that inarguably make my life better in some way.

  2. I put the solution somewhere that others can reach it too, making sure that giving them access is sustainable for me.

    • This means my commercial stuff is subscription-only from the first day it's on the market. PWFG, but still, there are no "launch specials" to try to gain traction or something like that. Sustainable from day 1 means low stakes, low cost of operation, and honestly a low initial return. If it is healthy from day 1, the odds are much better of it making it to day 2, and eventually day 1000.

  3. I make the thing a little bit better the next day, and I share the results. And I do that the day after that, and the day after that.

  4. And maybe we have a birthday party for it after it's been running for a year. I prefer a birthday party when dealing with new living things, rather than holding a high-stakes press conference in the maternity ward at the moment of delivery.

Don't add something until the lack of it is causing trouble for you.

(... or causing trouble for your customers, or causing trouble for the system, either of which also means that it's causing trouble for you.)

This came up while talking about writing automated tests for software. Having approached testing from a ton of angles over the years, my preference is to move fast and loose and without tests until the shape of the working idea becomes clear. Only when it becomes important to hold certain specific pieces together do I start writing tests. Another way to say it: only when I start losing time to pieces rattling apart do I start writing tests.

This approach ensures that every piece of the system is a working part of the system — involved in the system breathing, if you will. When that's true, each part is naturally accountable to every other part, and you're going to be much more quick to notice (and notice accurately) when some specific part needs a tuneup.

Last updated