Show Me the Data

For more than 50 years, Americans were drinking the same beers—mostly just four brands of pale lager. Then suddenly around 2005, craft beer started taking off, rising from 3 percent to 20 percent of the market in less than two decades. Overwhelmingly, the phenomenon was being driven by millennials, who started turning 21 around that time, and the popular press churned out hot takes about how the generation was upending traditional product selection with a new preference for artisanal goods.

“They implied that millennials were somehow wired differently and changing all the rules about branding that we’d learned from boomers and Gen Xers, so we’d have to throw it all out,” says Jean-Pierre Dubé, the James M. Kilts Distinguished Service Professor of Marketing and a Charles E. Merrill Faculty Scholar. Something about that explanation fell flat for Dubé, who is also faculty director of the Kilts Center for Marketing. He suspected that the real reason for the rise of craft beer had more to do with greater variety in the marketplace.

Fortunately, he had a way to test the hypothesis—using unique datasets of consumer buying habits from the global measurement and data analytics company, NielsenIQ, which distributes their data to academics through Chicago Booth’s Kilts Center for Marketing. The dataset is a gold mine for understanding nationwide buying habits dating back to 2004, gathered from more than 100,000 consumers who scan all of their weekly purchases—amounting to more than 65 million transactions a year.

Dubé was able to cross-reference the data with specific information on when craft-beer brands became widely available in different states and cities. By comparing buying habits of millennials in different areas at the same time, he determined that almost the entire generational differences in craft-beer consumption could be explained by the assortment of beer available in their city when they turned 21. Millennials had many more beer brands, craft and national, from which to choose when they turned 21 than older generations of Gen Xers and Baby Boomers.

“The generation into which you were born had virtually no predictive power,” Dubé says. Rather than being a story about a generation’s radically different buying habits, it turns out craft beer’s rise is actually about the power of capturing brand loyalty while someone is young, he concluded.

Expanding the Ecosystem for Marketing Researchers

The study is just one example of how the Kilts Center is revolutionizing consumer marketing research through the power of its data relationships. In addition to two datasets from NielsenIQ and one from Nielsen, the center has relationships with other data providers including TransUnion, Numerator, Syndigo, and Visa. By providing separate access to each of these unique datasets, Kilts offers an unprecedented level of information for analyzing consumer habits and preferences.

“These datasets allow researchers to study how people make decisions, and observe outcomes with economic stakes,” says Art Middlebrooks, clinical professor of marketing and executive director of the Kilts Center. At the same time, he says, it helps the center “promote The Chicago Approach to marketing, which is a data-driven, analytical approach—starting with a theory, gathering evidence to test that theory, and evaluating the results to draw conclusions.”

By offering open datasets with granular information down to the household level, the center enables researchers to test and expand upon others’ research using the same data. “It creates a centralized database where the whole marketing community can use these data to build upon each other’s work,” Dubé says. “Someone writes a paper, and then hopefully over the next few years, several hundred people write papers building on it.”

The unique collaboration with NielsenIQ and Nielsen grew from the school’s relationship with Booth alumnus James M. Kilts, ’74, who led the founding of the Kilts Center in 1999. In 2009, he became supervisory board chairman of Nielsen (a single company at the time; Nielsen later split off its data analytics business to form NielsenIQ in 2021). Shortly thereafter, the center began talks about housing Nielsen’s US datasets containing consumer household purchases, retail sales, and advertising occurrences. It took several years to prepare these datasets before making them available to academic researchers worldwide in 2013. Since the launch of this program, more than 1,600 researchers from 200 academic institutions have made use of these datasets housed at Kilts, publishing hundreds of papers in top-level marketing, economics, and finance journals and presenting at major conferences.

The TransUnion relationship followed in 2017. “We have stakeholders across the ecosystem—lenders, consumers, government entities—who mention the important role academic research plays in understanding industry trends at a deeper level,” says Michael Umlauf, AB ’90, senior vice president, data science and analytics at TransUnion. “The Kilts Center provides a hub of access so we can get the right data into the hands of researchers more efficiently.”

For all of the companies, Booth serves as an intermediary, receiving and organizing data, and vetting proposals from researchers. “What’s unique is the skin that Kilts puts into the game to manage the dataset, maintain control over it, and help people understand what’s there,” says Umlauf.

Implications for Policy and Beyond

Not surprisingly, many of the most-cited papers using Kilts datasets are by Booth faculty themselves. In one of the most downloaded papers using Kilts data, Günter J. Hitsch, the Kilts Family Professor of Marketing; Bradley Shapiro, associate professor of marketing and True North Faculty Scholar; and Northwestern’s Anna Tuchchman compared data on television advertising to sales of products over five years using Nielsen data. They found that, on average, television ad campaigns increased sales by only one percent, a fraction of the amount estimated by previous research. Despite that disappointing number, however, they found that television advertising can still increase profits substantially for certain brands and products.

Kilts datasets have also allowed for examination of deeper policy questions. In another study, Dubé and others looked at the concept of food deserts and the prevailing wisdom that people in poorer neighborhoods are unhealthier as a result of the lack of access to grocery stores with nutritious foods. The researchers find that when a grocery store moved into a neighborhood, people did indeed shop there, “but the healthfulness of the stuff they put in their shopping baskets didn’t change,” says Dubé.

He and his coauthors were able to use UPC data to examine the nutritional value of purchases, finding that access only explained 1.5 percent of the nutrition gap. An analysis of the vast panel of consumer transactions, on the other hand, reveals that most of the nutrition gap is due to differences in food preferences. Further, consumer surveys reveal these differences in food preferences are heavily correlated with nutrition knowledge. “Instead of putting in supermarkets at a very high price tag, we should be spending this money on nutritional education,” Dubé concludes.

Sarah Moshary, assistant professor of marketing and a Robert King Steel Faculty Fellow, along with Northwestern’s Anna Tuchman and Cornerstone Research’s Natasha Bhatia, similarly used NielsenIQ data to debunk the concept of the “pink tax,” which holds that products marketed to women are more expensive. Former Booth professor Neale Mahoney (now at Stanford) along with coauthors from Harvard, UCLA, and UC Berkeley used TransUnion data to examine the geographic concentration of medical debt and its effects on consumer credit.

Those are just some of the many questions academic researchers have shed light upon in recent years, thanks to the relationships the Kilts Center has forged with leading data providers. In the future, the center hopes to grow its offerings, potentially adding new datasets that might include information on durable goods and restaurants, online purchases as well as in-store purchases, and consumers from other countries.

The ever-expanding collection of data serves a dual purpose, says Middlebrooks, of allowing new insights into thorny marketing and public policy questions and spreading Booth’s reputation far and wide. “By housing these datasets at the Kilts Center, Chicago Booth becomes known for cutting-edge research in marketing,” he says, “and what we call The Chicago Approach to marketing gets broader visibility in academia.”