Preparedness Over Prediction Model
In an unpredictable world, stop trying to predict the future and start preparing for multiple futures
Margaret Heffernan argues that our obsession with prediction and efficiency has made organizations fragile. When we optimize for efficiency, we eliminate the redundancy, slack, and experimentation that provide resilience when predictions inevitably fail. The most important human skills in an unpredictable world are not analytical but imaginative: the ability to envision multiple possible futures, build coalitions of diverse thinkers who can spot different threats and opportunities, and invest in preparedness rather than prediction. Heffernan draws on examples from supermarket chains that replaced human judgment with algorithms (and suffered when reality deviated from the model) to argue that human imagination, relationships, and judgment are not inefficiencies to be eliminated but the core capabilities needed for navigating uncertainty.
- Efficiency is the enemy of resilience in unpredictable environments
- Prediction fails precisely when you need it most—during unprecedented events
- Human imagination, not algorithmic prediction, enables response to genuinely novel situations
- Coalitions of diverse thinkers see more possibilities than any individual or algorithm
- Preparedness means investing in capabilities you may never need, which feels wasteful but saves you
- Stop Optimizing for Efficiency and Build Slack Into the SystemIdentify where your relentless pursuit of efficiency has created fragility. Efficiency eliminates redundancy, but redundancy is what provides resilience when things go wrong. Build slack into your systems, your schedules, and your budgets—not because it is comfortable but because it provides the adaptive capacity you need when predictions fail and plans break.Pro tipAsk: what happens to this system if any single component fails? If the answer is catastrophic, you have optimized for efficiency at the expense of resilience.WarningThis is not an argument against efficiency in all contexts. In stable, predictable environments, efficiency is valuable. The point is that most modern environments are less predictable than we pretend.
- Invest in Human Imagination and Diverse CoalitionsBuild teams that can imagine multiple possible futures rather than just optimizing for the most likely one. This requires cognitive diversity—people who think differently, come from different backgrounds, and see different possibilities. It also requires creating space for imaginative thinking that is not constrained by data and models. The most important human skill in an uncertain world is the ability to imagine what has never happened before.Pro tipRegularly run scenario planning exercises that explore wildly different futures, including ones that seem implausible. The exercise builds imaginative capacity even if the specific scenarios never materialize.
- Practice Preparedness Rather Than PredictionShift investment from trying to predict what will happen to preparing for multiple possible outcomes. This means building capabilities, relationships, and resources that are useful across many scenarios rather than optimizing for one predicted future. Military organizations call this maintaining readiness for a range of contingencies rather than planning for a specific battle.Pro tipFor every major plan, identify three ways it could fail and what your response would be. This is not pessimism—it is the practical form of preparedness.WarningPreparedness is expensive in calm times because it looks like waste. But the cost of unpreparedness during a crisis is always higher than the cost of maintained readiness.
Heffernan describes an American supermarket chain that replaced human supervisors for meat, produce, and bakery with an algorithmic task allocator in pursuit of efficiency. The algorithm could not respond to unexpected situations—a sudden weather change affecting supply, a local event changing demand patterns, or equipment failures requiring creative workarounds. Human managers, with their imperfect but adaptable judgment, would have handled these situations naturally. The algorithm, optimized for prediction, failed when reality deviated from its model.
Heffernan developed this framework after observing the repeated failure of prediction-based strategies in business, politics, and public health. She saw organizations invest billions in predictive analytics while cutting the human capabilities—imagination, relationships, experimentation—that actually enable adaptation when predictions fail. The COVID pandemic later proved her thesis dramatically.