Something you thought was a great idea at the time, but looking back you'd do completely differently.
Could be: Launching too early Launching too late Building features nobody asked for Not talking to users enough Focusing on the wrong metrics Having zero launch plan