Challenging the conventional wisdom of data-driven decision-making, marketing professors and behavioral scientists Stefano Puntoni of the Wharton School and Bart De Langhe of KU Leuven and Vlerick Business School argue that many analytics efforts flounder because data analyses are disconnected from the decisions to be made. In this excerpt from their forthcoming book Decision-Driven Analytics: Leveraging Human Intelligence to Unlock the Power of Data, they argue that the key to making good decisions with data is to start by putting data in the background.

 

Let’s start with a simple problem. Joey went to the store and bought a pack of chips. In this store, a bottle of water costs $3, a pack of chips costs $1, and a pack of gum costs $2. So, how much did Joey spend?

If you answered that Joey spent $1 because he only bought a pack of chips, you’re correct. However, one in four individuals surprisingly answers this question with $6, incorrectly adding up all the prices listed. This error demonstrates a common problem: doing math just for the sake of it, without considering the context.

Now reimagine the scenario with more intricate figures: Joey went to the store and bought a pack of chips. A bottle of water costs $1.05, a pack of chips costs $0.75, and a pack of gum costs $1.70. How much did he spend in total?

This reveals a curious insight. One-third of individuals now incorrectly sum the costs to $3.50, even though it’s more difficult to calculate than $6. When the numbers involved are more precise, people are more likely to engage in unnecessary calculations.

Divers love crunching numbers. Runners feel the heartbeat of the business world. It’s crucial to blend the best of both.

Now, let’s think about how this relates to the world of business and data analytics. Through our work teaching data analytics in business schools and companies around the world, we’ve noticed something. When faced with a business problem and a dataset, people usually fall into one of two groups: divers and runners.

Divers are those who love crunching numbers and can’t wait to dive into the data. They see the dataset as a fun puzzle, and they’re eager to solve it. But just like those who mistakenly added all the prices in the scenarios above, divers sometimes rush into statistical analysis without thinking about the business side of things.

Runners, on the other hand, feel the heartbeat of the business world. They feel the race against time, the constant push to make an impact, and the challenge of managing limited resources. They might not be best friends with math and data, but they have a knack for keeping businesses going. If you found yourself more interested in the outcome of Joey’s shopping than the calculations, you might relate to the runners.

In business, data analysts are often like divers, while managers and decision-makers are often like runners. Sometimes, this can create an unintended divide, with each side failing to see the other’s value. This lack of mutual understanding can make it harder for everyone to work together effectively. It can lead to misguided decisions.

When data is bigger, more precise, and complex, it has a different effect on runners than on divers. Runners might want to rush forward and focus on the immediate business goals, while divers may want to dig deeper into the data, unintentionally increasing the divide between the two groups. This is like what happened in our example, where some people were more likely to do unnecessary math when the numbers were harder, while others, the runners, became more likely to skip the math. The reactions vary, but the cause is the same. When data analytics become more intricate, divers plunge deeper, and runners push harder.

For a business to do well, it’s crucial to close this gap and encourage runners and divers to understand and value each other’s skills and strengths. Effective decision-making in business requires a balance between decision-makers and analysts. To help with that challenge, we offer a new approach to using data for decision-making, one that we call decision-driven analytics. In this approach, decision-makers play a more active role in data analytics, blending the best of both runners and divers to drive better business results.

Thinking Without Data

Picture yourself overseeing your company’s car fleet. You’ve got two types of cars: SUVs that get 10 miles per gallon (MPG), and sedans that get 20. Each type of car makes up half of your fleet, and they all drive 10,000 miles a year. You’ve got enough money to replace one type of car with a more fuel-efficient model. So, should you replace the 10-MPG SUVs with 20-MPG ones, or the 20-MPG sedans with 50-MPG ones?

At first, it might seem like upgrading the sedans is a better deal. After all, 50 MPG is a lot more than 20, right? But actually, replacing the SUVs would save more fuel. Here’s why: Right now, each SUV uses 1,000 gallons of fuel a year (10,000 miles divided by 10 MPG), and each sedan uses 500 gallons a year (10,000 miles divided by 20 MPG).

Let’s look at what would happen if you replaced them: If you replace the SUVs, each new one would only use 500 gallons of fuel a year (10,000 miles divided by 20 MPG). That’s 500 gallons less than before. If you replace the sedans, each new one would use 200 gallons a year (10,000 miles divided by 50 MPG). That’s 300 gallons less than before. So replacing the SUVs saves 500 gallons per vehicle each year, but replacing the sedans saves only 300 gallons per vehicle. Therefore, it makes more sense to replace the SUVs, even though the MPG number doesn’t increase as much.

What this example seeks to highlight is the point that starting from the data you have isn’t always the magic solution to making great business decisions.

Let’s revisit the car-fleet scenario again. You’re in charge of a fleet of cars, and you’re trying to save on fuel costs. You’ve got two types of cars, and one type could be upgraded. The difference is this: Nobody’s telling you the miles per gallon for these cars. Instead, they’re asking you what info you’d want in order to make your decision.

Now, your first thought might be, “Well, tell me how much fuel each car uses, or give me the miles per gallon.” It seems obvious, right?

Sometimes the best answers come when we’re not buried in a bunch of numbers. It might seem strange, but taking a step back from the data can actually help us see things clearer. That’s the punch line of this whole thing: Thinking without data can, in a weird twist, improve how you use data to make decisions.

Moving From Data-Driven to Decision-Driven

Bart was trained as a psychologist and later moved toward statistics. Stefano was trained as a statistician and later moved toward psychology. Our paths crossed in the middle when we became business school professors.

Around 15 years ago, the emergence of Big Data began to dramatically transform marketing. As a result, universities around the world launched new business analytics programs. Bart, then a budding assistant professor, was assigned to instruct customer analytics. He noticed that most existing courses focused solely on technical aspects, overlooking the vital role of behavioral science in decision-making. Determined to offer a more holistic approach, Bart integrated psychology into his curriculum, confident that this would better prepare his students to tackle business problems with statistical models.

Stefano later established a research center focused on AI and psychology to help disseminate the findings of his research among businesspeople. During his interactions with managers, he discovered a recurring theme: Analytics projects seemed to fail often, but hardly ever due to technical issues. The issues were always people-related.

Most discussions on analytics tend to highlight data and tech prowess. This is understandable, given the rapid technological advancements, but the risk is forgetting about the business decisions that need to be made and the people making them.

Our approach to teaching analytics is counterintuitive because it pushes data to the background. But our experience shows that the real game changer in improving a company’s data usage isn’t acquiring more data or devising complex models. It’s better integrating data into the decision-making process.

With the exponential developments in artificial intelligence, we see an alarming gap growing between managers and decision-making on one side and data scientists and data analytics on the other. It’s time to put decisions squarely at the center of data analytics. It’s time for business to focus on the decision-maker and the core principles of decision-making. It’s time to embrace decision-driven data analytics, a framework for combining the best of human intelligence and judgment with the power of modern data analytics and technology.

 

Published as “Bringing Divers and Runners Together” in the Spring/Summer 2024 issue of Wharton Magazine.