Product roadmaps in a startup

The hardest thing in deciding the product roadmap or the next milestone is not deciding what to put on it, its what to take out. For every idea we decide to execute on at Livemocha, we have to defer 5x others.

Coming up with ideas is not the hard part but deciding which 3 things you can build this month and which ones you won't build, will over the long run decide whether you will be successful. It's not easy especially since you never have more than 60% of the information to make that decision. But if you don't focus and try to do too much, you won't do anything well.

Clearly there is a lot of art in this, but we could take some steps to minimize wrong choices

• Ask why should we build this often and try to get as much data to help guide hypothesis. Its hard for 4 people in Seattle to predict the right answer for 5 million users around the world.
  
• Try to test inexpensively. If it yields good results, then push on the pedal. Unless that feature is super strategic and a must-have, usually you can break it into phases and test while you build.
  
• Keep a relative sheet of prioritized features based on various criteria like risk, effort, etc. In my experience, every idea in isolation seems great, when you put it on the list and see how it stacks up, you can make a better decision.
• If you make a mistake and build stuff that no one uses, Admit, use the learnings and move on quickly.