Most founders automate the wrong things first. They point at the work they understand least — deployment pipelines, reporting, onboarding — and hand it to a machine before they’ve ever done it by hand. I believe the opposite: the work you understand least is the last thing you should automate.
I recently sat down with Data File Host to unpack this, and the conversation turned into one of the most honest pieces I’ve been part of. You can read the full breakdown on their site.
The bakery test
One of my favorite examples in the piece isn’t from my software ventures at all. It’s my mother’s bakery. Every night, someone spent thirty minutes translating the day’s orders into a production sheet for the next morning. Unglamorous, perfectly understood, endlessly repeated. I built software that does it in one click. That project will never make a pitch deck — but it’s the exact pattern I look for everywhere: high frequency, low ambiguity, and a human who can describe every step without hesitation.
Automation amplifies whatever it touches
A well-understood process becomes faster and more reliable when you automate it. A confused one becomes confused at scale — and now nobody remembers how it worked by hand. That’s why the founders who get this right do the manual version first, long enough to know exactly where the friction lives. It runs against the “delegate everything” advice, but it produces sharper decisions.
The expensive lesson
The article also covers a moment I don’t sugarcoat. My team once recommended multi-region redundancy to a client whose budget didn’t stretch that far. A regional cloud failure later took their application down for roughly eight hours. The relationship survived because we’d documented the recommendation and handled the incident transparently — but the lesson was expensive. Every automation you skip sends an invoice later.
Creativity needs systems
There’s a worry in creative businesses — including marketing operations — that automation flattens the work. I reject the premise. Creativity brings ideas to life, but systems make them sustainable. Automation should handle the repeatable substrate so people can focus on judgment and craft.
And with AI making it nearly free to build automation, the stakes on judgment just went up. Bad workflows can now multiply across a company in minutes. AI is leverage for founders who already have clarity — and an accelerant for those who don’t.
Read the full article on Data File Host →
Understand the swing completely. Then, and only then, build the machine that repeats it.
