I review developments in theory and methodology that may allow us to begin creating innovative forms of organizing, rather than rest content with studying them after they have emerged. We now have the conceptual and technical apparatus to prototype organization designs at small scale, cheaply and fast. The process of organization re-design can be seen in terms of multiple stages. It begins with careful observation of phenomena. Qualitative or indeed quantitative induction (i.e. data mining) can play a critical role here. Once we have some understanding or at least conjectures about underlying mechanisms, we can use the behavioral lab or an agent-based model to run cheap experiments to adjust the design. Once we have formulated a new design, we may want to run a field experiment with randomization. If the results look satisfactory, we can scale up and implement.