Strategic thinkers and planners know that failures large and small are part of long-term planning and execution. Training a workforce to make micro-models, prototypes, and create pathways for high-quality feedback helps our strategic actions exceed our expectations.

Expansive planning models, carried out over large spans of time, need to be built for failure. What was true (or assumed true) on Day 1 is unlikely to remain valid over time. By Day 93 or Day 500, a variance, large or small, between the model’s reality and actual reality will be present and, eventually, quite noticeable. Long- and short-term planning need to account for this reality variance. Part of building better plans beings with thinking about necessary failure, micro-models, and feedback loops.

Necessary failure. Reality variance contributes to model failure, but eliminating or reducing reality variance only enhances the model’s changes of meeting expectations. At Day 1, the model will have been tested, prodded and broken as it was created. These are safe failures. When the wheels fall off on paper, that’s OK (inconsequential failure); when they fall of at 120 mph, that’s less OK (failure with consequences). A model’s robustness reflects its ability to handle failures with consequences that are both predictable and unpredictable.

When a model does what we want it to do over time, we consider that a success. What we really want are models that outperform themselves; we want beautiful failures. Part of beautiful failure involves micro-modeling and feedback loops.

Micro-modeling. The success or failure of any large-scale model depends on the people executing it and, specifically, their ability to create the small, nimble micro-models for day-to-day actions. The quick pen and paper sketch of known information and available resources, the problem to be solved, and how to make that solution real means that people are keeping the inconsequential failure rate high. A higher inconsequential failure rate results in models that have eliminated some potential for consequential, in-the-wild failures. Some failures with consequences are unpredictable in both nature and magnitude.

How can strategic minds incorporate necessary failure and micro-modeling into large-sale plans? They can make sure that when systems speak, people are listening and prepared to respond with a better system.

Feedback & feedback loops. Feedback is information generated by our actions. For feedback to be useful, the information returned must be timely (if your parachute doesn’t open, you need to know now), clear (parachute tangled or not deployed?), contextualized (my parachute or a parachute?), and concise (you don’t want to wade through information on location, temperature, wind speed, and altitude to find your parachute status).

Having feedback that is timely, clear, contextualized and concise is worthless if it isn’t connected to micro-models and large-scale models. Micro-models, given how often they are created and adapted regularly incorporate feedback since they are based in part on the information available when they are created or modified. These feedback loops occur between models, at model inception, and not within them. Inception feedback loops help us make better models. Micro- and large-scale models designed to incorporate internal feedback loops make themselves better as they function in the wild.