If you want to survive as a company, you need to learn fast. If you want to learn, you need to measure. If you want to measure more and get more data, you need as much development speed as possible and ship fast.
This is the Build-Measure-Learn loop, as described by Eric Ries in The Lean Startup. The shorter the loop, the more you learn and the faster you'll be able to figure out what does and what doesn't work. He calls this smallest slice a Minimal Viable Product.
Thinking MVP is actually much harder than you think. It is a real exercise to break down the complete, polished thing to the smallest, tiniest slice you can ship as fast as possible. It requires lots of creativity.
These questions might help you in the process:
- Which features can be removed?
- Which features can be implemented by using a third party tool?
- Which things can be done by hand for the time being?
- Is it actually necessary to build the feature? Or can it just be announced to see how people react and gather data?
- It it possible to choose another form? Just a sketch? Just a design? A video? A non-functioning button?
- Can I learn from what others did?
Most people dramatically overestimate what is needed in a minimal version.