Don’t waste time digging the entire hill.

Given the pace of change and the volatility in the world today, it’s not feasible to collect and analyse all the data needed to produce a predictable outcome (deterministic problem-solving). By the time you’ve gathered and reviewed everything, the situation may already have moved on. Sometimes the best way to find answers quickly is to use random sampling.

An archaeologist friend shared an analogy. When archaeologists approach a mound, they don’t excavate the entire site at once. Instead, they dig a trench which provides a snapshot of what lies underneath. This random sampling approach is more efficient and less resource intensive.

Similarly, when faced with a big decision or complex problem, instead of trying to gather all the information, start by taking several small samples of the situation to evaluate it. Broadening the search space while limiting the depth initially increases the chance of finding valuable insights. Starting sooner also means you learn faster.

Here’s a business example. When online shopping was evolving down different paths, Alibaba didn’t try to predict which model would dominate. Instead, they rapidly developed solutions for all three scenarios: search-based models like Google Shopping, multi-store platforms like Chewy or Farfetch, and general-store models like Amazon. Their early solutions weren’t perfect, but this broader approach gave them an advantage over competitors who bet on just one model. Today all these models coexist.

Problem-solving that starts with a small cross-section of observations can be more efficient. Look for patterns and pay attention to anything that stands out. What’s different about these anomalies and how can they help us reach a decision?