This issue is about design principles for good, viably sneaky experiments. Good experiments are an appropriate way to act when faced with both uncertainty and an uncertainty-averse organization. At a minimum, a good experiment must have a hypothesis about a mechanism that the experiment will investigate, an actionable insight if the hypothesis is supported, and an actionable insight even if the hypothesis is not supported. A good experiment can be made easier for an uncertainty-averse organization to swallow by redesigning it to be as small, cheap, and fast to execute as possible, while providing all three of the abovementioned characteristics.