I don’t recall how many projects I was involved in stalled because the team tried to put too many things in there – while thinking back, a simpler approach could have lead to a more positive outcome.

Since it’s not possible to predict whether a software we build will be accepted and used by many users, we need to ship as soon as we can our products so we can see if it sticks.

The only way to have a fast release cycle for new products therefore is keeping a very tight scope. Support just one account instead of allowing as many as the user wants. Support a simple search box instead of advanced search functionality. Postpone everything that’s not essential to the app. There’s time for that later, when you have users that demand those functionalities.

Remember that it’s always easier to add stuff later, than to remove things once they’re backed in.