Experimental Strategy With Entangled Ideas

Prior work in experimental strategy has largely assumed that actors make decisions about when and how to experiment on one business idea at a time. Even when prior work allows for the possibility that an actor may possess multiple business ideas, this work has yet to consider how decisions about when and how to experiment on one business idea might affect decisions about when and how to experiment on another idea. Using a formal model, this paper introduces the possibility of entangled ideas - ideas that are connected such that information about the value of one idea informs an actor about the value of another idea - and considers their implications for experimental strategy. Analysis of the model shows that the possibility of entangled ideas overturns many of the recommendations in prior work - changing whether an actor experiments or not, how the actor crafts a program of experimentation, and what the actor does after reviewing the results of an experiment.