Here’s a video from the people that made the iPhone app Convert, showing all of the different versions that they tried:
There are a couple of things worth noting in this. First, they experimented a lot. They generated a ton of variety, all of which would have been pretty cheap. When I keep talking about failure, people often seem to think that it means that we need to launch products that don’t work, when in fact I mean almost exactly the opposite. We need to do what tap tap tap did, and figure out what doesn’t work before we launch. The second big point is that the big change that makes the whole thing actually work well doesn’t come until about 75% of the way through all of their tinkering (at about 1:10 in the video). This shows again that the big changes often don’t come until you start using things.
(Hat tip to Endless Innovation…)