The management system an industry standardizes on
There’s a pattern I’ve now lived three times, in three industries that have nothing to do with each other. It starts the same way every time: somebody’s operation is drowning in its own coordination, and the software that’s supposed to exist for them doesn’t fit. It ends the same way too. The system that fits becomes the system everybody in that industry wants.
Let me tell you about the first time, because it set the template.
The floor covered in spreadsheets
In 2017 I walked into a rotational-molding plastics factory. Scheduling the operation looked like this: the owner and five of his factory managers crowded over printed Excel sheets. Not on a desk. On the floor. The schedule was literally spread across the floor, because that was the only surface big enough to hold the operation.
Here’s the thing most software people get wrong about a scene like that: those spreadsheets weren’t dumb. They encoded years of real operational knowledge (which molds fit which machines, which jobs couldn’t follow which, who could run what). The floor was the database. The problem wasn’t the knowledge. It was that the knowledge lived on paper and in six heads, and the operation had outgrown both.
So I didn’t start by writing software. I walked the floor. I learned the process from the people running it, mapped how tools, machines, people, and jobs actually connected, and only then architected and built the complete initial SaaS product: Django backend, frontend, the works. Scheduling, molding logs, reporting, BOM and machine workflows, all of it automated end to end. The factories ran on it.
That foundation became RotoEdge Pro, which today is positioned as industry-standard software for rotomolding plastics factories worldwide. I built the beginning; later versions are others’ work. But the lesson stuck with me permanently: the software that wins an industry isn’t the cleverest one. It’s the one shaped around how the work actually moves.
Twice more, same pattern
The second and third times were painting contractors. New-residential painting is another operationally messy, physical business where generic software fits like a borrowed suit, so I built the system of record for it: MyPaintBuckets, the company I co-founded and still run. Project details, material ordering, extra-paint-order capture, scheduling, invoicing. Same method: learn the operation, then automate the operation you actually found (not the tidier one you wished you’d found). Rebel Paint, the AI-leveraged side of the same vertical, is next up.
And the third industry is my own. Software teams that mix humans and AI agents are, operationally speaking, a factory floor: work has to be scheduled, claimed, logged, checked, and shipped, by workers with very different capabilities and very different failure modes. The FleetHarbor suite is the management system for that industry, and I’m building it the same way I built the first one. Walk the floor (mine, daily). Learn the process. Automate what’s really there.
Why the pattern repeats
People ask what rotomolding, painting crews, and AI agents could possibly have in common. Honestly? Everything that matters:
The real process is never the documented process. The floor knows things the office doesn’t. Coordination cost, not labor cost, is what’s eating the business. And the moment a system actually fits, adoption stops being a sales problem, because the operation pulls the software in rather than having it pushed.
You can’t shortcut this from a desk. Every failed vertical-software product I’ve ever seen failed the same way: built from what the industry says about itself, not from what its floor does on a Tuesday.
The honest version of “industry standard”
I want to be careful with that phrase, because it gets abused. I don’t mean I set out to build The Standard and the world complied (nobody gets to plan that). I mean the humbler, more repeatable thing: when you automate a real operation faithfully, the result generalizes, because the operation you studied is shaped like every other operation in that industry. The spreadsheets on the floor in one factory look a lot like the spreadsheets on the floor everywhere.
Walk the floor. Respect what the paper knows. Build the system around the truth of the work. Do it right once and an industry has a habit of noticing.
If your workflow is messy, physical, and too important for shallow software thinking, that’s my favorite kind of problem. Let me know what’s on your floor.
I'm Joseph Dattilo — engineer-founder in Lansing, Michigan, author of the FleetHarbor suite, and founder of Date Palm Media. More about me · More writing · Get in touch