Decagon

Treat moats as discovered, not designed, and ship first. Decagon's founder Jesse said moats are discovered, not designed, and Anish Acharya of a16z agrees. His advice is not to wait for a business plan that survives MBA and VC scrutiny before shipping.

Cursor was criticized for lacking a moat, but being a high-end product let it capture reasoning traces and train its own Composer models. Granola is beloved and dominant despite early moat criticism. Classic moats such as network effects, scale, brand and proprietary data still work, because none depend on how hard software is to make. Quote: "Moats are most often discovered, not designed."

Lenny's Podcast▶ 54:55