In senior care, a dropped handoff is not a lost lead. It is a person not getting attended to. I ran that kind of operation for years before I ever saw a startup dashboard, and it permanently set my standard for what "the system caught it" has to mean.
Before fintech, I founded a preventative health startup in China and scaled senior care nationally. We launched more than 30 community service points and 5 senior care homes. Every home reached over 80% occupancy. From the outside that reads as a growth story. From the inside, almost none of the work was growth. It was procedure.
Care operations punish improvisation. A tech company can absorb a sloppy week. A missed follow-up costs a deal, someone apologizes, the pipeline recovers. A care home cannot absorb a sloppy shift. Medication schedules, night rounds, meal restrictions, a family's call about their mother: every one of those is a task that has to happen on time, done by whoever is on shift, whether that person started three years ago or three days ago. The building runs 24 hours a day and the residents do not pause while you find the one staff member who knows how things are done.
Where my first approach broke
My early instinct was the founder instinct: hire good people and trust them to handle it. It worked while the operation was small enough that I could see everything myself, and it stopped working the moment it wasn't. Good people carry the process in their heads, and heads go home at the end of a shift. The handoff between one shift and the next was where things slipped, because a handoff between two people is only as good as what got said out loud in the two minutes they overlapped. In our business, what slipped in a handoff was not revenue. It was attention owed to a specific person in a specific room.
So the rule became: everything repeatable gets written down, as a procedure a brand-new hire could run without asking anyone. Not a philosophy document. A checklist with times, owners, and what to do when something is off. Shift handoffs moved from conversation to a written record, so the incoming shift inherited facts instead of memories. When something went wrong, the fix was never "remind everyone to be careful." The fix was a change to the written procedure, so the same failure could not repeat quietly.
That is the discipline that got 5 homes past 80% occupancy. Families toured other facilities before they chose ours. What they were buying, whether they said it this way or not, was the confidence that the standard of care did not depend on which staff member happened to be on duty. Occupancy is a trust metric. Written systems are what made the trust rational.
What this looks like at a tech company
I apply the same discipline now at early-stage tech companies, and the transfer is direct. The stakes are different. The failure pattern is identical: the process lives in the founder's head or in the heads of the first three hires, everything depends on the right person being available, and handoffs happen in Slack threads that nobody reads twice. It holds until the first busy month, and then things start slipping in ways nobody can see, because there was never a written definition of what "done" looks like.
The care home years gave me a test I still use. A system works when the worst handoff, on the worst day, with the newest person on shift, still gets caught. Not when the average day goes fine. Average days are not what systems are for.
The takeaway a founder can apply this week: pick the one process in your company that currently depends on a specific person's memory, and write it down as a procedure a stranger could run tomorrow. Times, owner, and what to do when it goes wrong. You will find gaps the moment you try to write it, and every gap you find on paper is a failure that will not happen live. That was true with 5 buildings full of people depending on it. It is just as true with a sales pipeline.