Decisions Amid Growing Data: Think Deductively but Act Inductively

As CIOs look to mine their growing mountain of structured and unstructured data, they are hungry for faster ways to analyze that data against old assertions and for new, more effective solutions for using this data. For these leaders, asking why something happens may be a natural impulse, but it can take a lot of time and effort and comes with a hefty price tag. The key question for companies, therefore, may be instead to ask what first.

Asking what is the hallmark of inductive reasoning, which helps decision makers narrow their set of hypotheses for subsequent deductive—or why-based—testing.

Often, executives make decisions based on intuition, previous experience, or other intangibles. As industries continuously transform, new technologies emerge, and consumer play a greater role in sharing information on products and services, this type of approach is less of an asset and more of a hindrance. Yes, some executives may strike pay dirt using this approach today, but I would wager that fewer will succeed this way going forward, given the speed of change.

Let’s look at hypothesis testing, which is one example of deductive reasoning. You start with a general assertion (a belief statement, if you will) and test to understand its validity. For example, during tornado season, consumers in high-risk areas buy emergency supplies. Stores in these areas, therefore, should all stock up on these items during the tornado season. 

There may be a negative correlation between distance from an emergency zone and purchase behavior, however. For this reason, inductive analysis may be quite useful. It would dictate reviewing data, understanding trends, studying correlative analysis, and building a hypothesis from the ground up. This approach often leads to surprise findings—previously unknown and unforeseen data correlations that can help improve hypotheses. In the end, you may find that only stores within a certain radius of an affected area will need emergency supplies. 

Inductive reasoning allows organizations to take a look at the what in terms of what is occurring, before they start looking at the why. The why, after all, may not be as important as the what. The fact that something occurs over and over again may be enough. Starting there, and perhaps developing a framework or process based on data orientation, can turn a query based on deduction into a wisely strategic move, backed by induction.