The Paradox of Data

January 12, 2020
paradox business

Data is the new oil.

— Clive Humby

“Data-driven” has been a massive trend in management, at least partially due to the incredibly visible successes of certain stories such as the Obama campaign, Moneyball, or Google’s ad ecosystem. At first blush, it seems common-sensical, if not tautological - a gut reaction to an executive saying “We use facts to justify the decisions we make about millions of dollars of investment” might be “I don’t understand why you are wasting time telling me obvious things”. Almost everyone, however, can come up with examples from their lived experience where this didn’t occur - often, a manager who vetoed a proposal based on uninformed intuition. Rightfully, this decision making modality is increasingly on the back foot, as the global appetite for data has skyrocketed.

Despite the sterling reputation that it has, and the mindshare that it has won, it’s hard to actually pin down a definition of a data-driven culture. Often, we say something generic like “we back our decisions up with data”, and then juxtapose it with a straw-man: “Managers won’t be flying by the seat of their intuition pants!”. This can be a dangerous level of understanding to have, as there isn’t a constructive definition - only a negation. Our past bad experiences with intuition-driven decision will tilt us towards the belief that data-driven decision making is the “generally superior” modality, but without a constructive understanding, we can find ourselves taking data-drivenness to an extreme, attempting to exorcise all forms of intuition, and attempting to divine corporate strategy from reams and reams of metrics.

It’s important to remember that the fundamental problem is not usage of data, but quality of decision making. It’s not about how much data you use, but how good your epistemology is. It’s not about how many observations you have, but whether or not you can determine the truth from your observations. It’s a common adage that there are many ways to lie with statistics, and the presence and usage of data does not guarantee that you have moved closer to the truth. If you forget that the purpose of data is to lead to effective decisions, you may find yourself using more data but achieving worse outcomes. Understanding this distinction and keeping it in mind is key towards becoming an effective data-driven organization.

2020/01/05 Winning The Paradox paradox