Here at VVM when we start a new experiment/innovation, we start it small.
By starting small we ensure that if the experiment fails what we learn is more valuable than the resources we invested in the experiment. Knowing that the error cannot cause harm to VVM liberates us to run more experiments and builds our tolerance for prudent risk-taking.
If the experiment succeeds, we design a new, slightly bigger version of the experiment. We repeat this process until we determine this idea should be canned or have evolved it into something awesome.
You can find some more on this concept here: