If they can't trust it, they won't use it

Julia proposed the feature every founder proposes eventually: connect the restaurant menu to inventory, so a dish with no tomatoes cancels itself before a customer can order it. The frustration it solves is real. The debate that followed is the teaching.

The pushback wasn’t “bad idea”, it was “that’s an ERP”. Inventory means registration discipline, expired stock, debit notes, double entries, and the eternal question of who records that the tomatoes ran out and what happens when they forget ▸ 18:22. The menu is the easy square of the map; inventory, for a two-person team right now, is a bullet in the foot ▸ 18:02.

Behind the scope call sits a harder law. Coupa, the software from the corporate years, works because its people spent 13 years testing it, with a whole “solutions” team whose job was to fire thousands of inputs at it and find every bug ▸ 25:49. Robust systems win not because they are clever but because they survived that. And the waiter is the strictest judge of all: if they can’t trust it 100%, they won’t use it. And if they won’t use it, why did we build it? ▸ 26:47 A workflow tool that fails once during a rush doesn’t get a second chance; the old system, ask the kitchen, was slower but never lied.

el sistema viejo era lento pero nunca mentía →

The debate closed with a seed collecting interest: “I wasn’t attacking you, I was attacking the idea” ▸ 33:14, seed 3-1 learned at the Hult Prize three weeks ago, now doing live conflict-resolution work between co-founders. Ideas as children versus ideas as proposals. The feature stays parked; the partnership doesn’t…