In 1983, an academic journal called Marketing Science published an article about ground coffee. Co-authored by eminent MIT researcher John D. C. Little ’48, PhD ’55, it contained a statistical model showing how shoppers adjusted their coffee-buying preferences based on predictable factors such as brand loyalty, sale pricing, and in-store discounts. Little had been able to achieve this level of mathematical precision thanks to the emergence of the big-data revolution of that era, laser-scannable Universal Product Code (UPC) markings on food packaging, which for the first time made it practical to capture consumer behavior in enough detail to model it using scientific methods.
Thirty-five years later, MIT Sloan School of Management faculty member Juanjuan Zhang—recipient of a new MIT professorship created by Little’s family in his honor—remembers the ground-coffee paper vividly. “It was the first paper I ever read about marketing,” she says. Zhang, who studied game theory and econometrics at the University of California, Berkeley, cites Little as her inspiration for entering the field of marketing science—a field that Little, now an Institute Professor Emeritus and her longtime MIT colleague, is often credited with inventing.
An expanded concept of “marketing”
To Zhang, marketing is “a way of communicating—of bridging the gap between production and adoption, between intention and acceptance.” Her research shows how that gap can affect almost anything. For instance, in 2010 Zhang studied how kidney-transplant patients tend to reject viable organs when they are farther down on the waiting list. The reasons for someone higher on the list “passing” on a vital organ transplant may have nothing to do with the organ’s medical quality, but the rational tendency of subsequent patients toward rejecting what others have rejected leads to 10% of donated kidneys going unused.
To Zhang, this is a marketing problem: patients are making sense of incomplete information in a way that harms their health. But trying to close this “information gap between experts and laypeople,” Zhang says, can also have unintended consequences—as she discovered while studying the effects of mandatory labeling policies for genetically modified organisms (GMOs) in foods. In an ongoing project, Zhang combined field data with a game-theoretical model to show that transparency about GMOs actually led people to avoid the labeled food more often, even though the current scientific consensus holds that genetically modified foods are safe.
“It shows that governments can use marketing too,” Zhang says. “The popular view is that a transparency-motivated policy of GMO labeling can’t go wrong: ‘We tell consumers what is in the food, so that they can make better informed decisions for themselves.’ But, as marketing researchers, we know that consumers are not scientists—they read news, they try to guess what science has to say, and that’s when even government policies of the best intention can backfire.”
Marketers aren’t traditionally seen as scientists, either. But in the current age of hyper-targeted advertising and data-driven product management, being scientific has become part of every marketer’s toolkit. This systematic reliance on data largely began with John D. C. Little.
Little earned an undergraduate degree in physics from MIT in 1948 before shifting his graduate studies to operations research. Also referred to as “management science” or “decision science,” operations research focuses on methods for optimizing solutions to complex problems using rigorous mathematical techniques. Little rose to eminence within the field with his proof of what would become known as “Little’s Law”: L=λW.
That compact equation might as well be the e=mc2 of operations research. It describes so-called queueing behavior: how long it takes for items to arrive at and pass through any system, whether it’s cars through a traffic stop or patients entering a hospital ER. Little’s successful application of generalizable mathematical rigor to an entire category of seemingly unrelated practical problems revolutionized his field—and fired his curiosity about bringing that same rigor to another practical domain: marketing.
“One of the things people have been reluctant to do was to call [this field] a science,” Little told an interviewer in 2012. “But everything has fundamentals. And they remain to be discovered and utilized.”
It’s a perspective Zhang shares. “Our ‘laws’ may not be as clean-cut as Newton’s First Law of Motion,” she says. Still, “they’re based on theories that we refine using data.” And just as technological advances have allowed scientists to gather higher-fidelity information about the natural world, marketing science, too, has been driven forward by the availability of more and better data. The supermarket UPC codes of the 1980s, which inspired Little’s coffee paper, led to the e-commerce boom of the late 1990s and early 2000s, which then seeded the current marketing landscape of smartphones and social media that Zhang inhabits. In 2017, for example, Zhang and colleagues from Tsinghua University conducted a randomized study on Weibo, China’s largest microblogging site (similar to Twitter). They found that just 42 social media influencers could increase a TV show’s viewings by up to 57% simply by retweeting promotional content about the show to their millions of followers.
People, not particles
But just as Little changed the paradigm of marketing by incorporating scientific methods, Zhang’s research is now questioning the fundamental assumptions behind some of those methods. Her own study on the effect of tweeting, for example, assumes that a randomized field experiment provides a gold standard framework for establishing cause and effect. “The field experiment methodology is analogous to what has been broadly used in natural sciences. We measure the effect of a treatment in a ‘lab’ that we create in the field, where human subjects behave without knowing they are in an experiment,” she says. “But my current work tries to rethink if that’s actually the right method.”
Why wouldn’t it be? Zhang’s answer is simple: “People are not particles. Particles react—they don’t think. But humans do. Humans make inferences from every marketing message, every policy change, every natural stimulus they experience.” To Zhang, this basic fact complicates the supposedly “clean” data from field experiments—a problem Zhang refers to in her research as “belief endogeneity.”
If, for example, a smartphone company wants to see how demand for a new phone varies with price, the company could conduct a field experiment offering the same phone to different groups of buyers at $99 and $999. That buyers don’t know they’re being observed is the whole point: by comparing the two groups’ buying behaviors in this field lab, the marketers can “know” that one factor—price—is influencing demand.
Or can they? Buyers who don’t know they’re in an experiment may draw their own conclusions about the circumstances—assuming that price reflects quality, for example. “That layer of thinking contaminates the experiment,” says Zhang. “Our next step is to find out if there’s a systematic way to debias the data. We’re trying to provide a marketing scientific angle to this whole problem of big data, by saying that every data point is alive.”
Zhang’s work on that front is still in progress, but she says its paradigm-questioning purpose is inspired by Little’s example. “He’s only interested in the boldest ideas,” she says. “That’s my goal and the goal of many of the faculty here, to say: ‘let’s innovate, not just refine.’”
Zhang acknowledges that being named the John D. C. Little Professor was “a very personal and tearful moment, and an honor of a lifetime.” But even at age 90, Little is still combining big ideas with a practical bent, and he gave a parting gift in the latter spirit to Zhang. “When he retired last year, he traded offices with me—so I’m now sitting in his office, in his chair,” Zhang says. “These are some giant shoes to fill. But this is absolutely invigorating.”