Data-Driven Leadership: Why Intuition Still Holds Value

Making well-informed, data-driven decisions is the hallmark of a modern business leader. With oceans of data at our fingertips it’s easy to dismiss the dinosaurs of yesteryear, who made decisions using quaint and archaic methods such as “seat of the pants” and “gut instinct.” I jest, of course. However, it certainly is comforting to convince ourselves that decision-making is now a science, rather than an art. I’d argue that context has always been the critical factor in decision-making, even more so now as leaders and their enterprises grapple with data overload that can provide contradictory and often misleading inputs. Back when I was a management consultant we favoured a hypothesis-driven approach to decision-making. Early in the process we identified one or more theories around a possible solution, and then used data to prove or disprove those ideas. To the data purist, this might seem a little like “leading the witness” (because the data surely speaks for itself). However, our approach nicely illustrates the role that intuition and experience play in leadership, along with data’s necessarily supporting function. It is incumbent on any leader to develop a deep and fundamental understanding of their business, without waiting for individual data points to “inform” them. So how then can data be used most effectively to support business acumen, creativity, and experience? The most obvious consideration is whether you are using appropriate data at all. Is the data relevant to the objectives at hand, and can you be sure that it is accurate, complete, and free from biases or errors? If not, you run the risk of making data-driven mistakes. At the very least, you’ll waste your time. Finally, beyond the data itself and the decisions being made, leaders need to consider the broader business environment where relatively vague issues such as ethics and privacy, risk, and commercial imperatives come into play. These factors require judgement and prudence from leaders, and often lie in very grey areas from a data standpoint.