Mind Money Notes

Can robotics startups actually become profitable companies beyond the demos and the LOIs? That's the argument everyone in this space is having, and I think the answer lives in some very unsexy details. First, figure out who actually pays. In construction it's rarely the person using the robot. Most of the time a subcontractor or an equipment leasing company buys the machines, but the general contractor is the one who asks them to carry it. Somebody high up who cares about the project's bottom line and bidding lower than competitors makes the decision, and the foreman and superintendent care about work done on time with fewer injuries on the crew. You have to sell to all three. Second, offer lease and purchase. Different customers want different things, and forcing one model on everyone is just leaving deals on the table. Third, the service ratio is the business model. If every robot needs its own babysitter, you built an expensive tool. If one technician runs 10 to 20 robots, you built a margin. Fourth, reuse your platform. Our next robots share the same chassis wherever we can, so every new task costs a fraction of the first one. A demo gets you a round, and unit economics get you a company. I spent years in factories watching that difference up close.

I read Elon's biography on repeat. The moment I finish, I start again, because whenever I'm stuck on a hardware problem, there's a page where he was stuck on something worse and kept going. In Elon we trust.

Small design detail I love: when a robot runs out of wire it blinks yellow, when the battery is low it blinks red, and whoever manages it gets a phone notification either way. Jobsites are loud and dusty, so a robot has to talk like a traffic light.

Robotics rounds keep getting bigger, and I'm happy for every team that lands one. It's good for all of us. Our pre-seed was a $4.3M party round, and the lead came in last with a $1M check. Nobody tells you it can work that way. Smaller rounds force a different kind of engineering. You can't buy your way out of a bad design, so you fix it on the bench at 6am. I think that makes for tougher robots.