Many of the challenges in international development are complex in nature. They involve many actors in uncertain contexts and with unclear solutions. Our work has proposed an approach to addressing such challenges, called Problem Driven Iterative Adaptation (PDIA). This paper is the most recent in a series intended to show how one can do PDIA, building on the first paper, "Doing Problem Driven Work.” The current paper addresses a key part of the approach one moves to once a problem has been identified, performing real-time experimental iterations. This is intended as a practical paper that builds on experience and embeds exercises for readers who are actually involved in this kind of work.