A B Tests Are a Great Way to Validate a Hypothesis

Experiments are complimentary to customer development, not a replacement. An additional way to validate a hypothe s is to run an experiment where you build something to test your hypothesis. For example, gogel had a hypothesis that the shade of blue used in advertising links affected click through rates. And a b and c, d, e, et cetera. Testing of different shades let gogel find a shade that generated an extra 200 million dollars in revenue.