Ulf Böckenholt

Photo: Claudio Calligaris

Enhancing the faculty of management's social science perspective, Ulf Böckenholt, the new Bell Professor of E-Marketing, studies choice behavior and develops methods of predicting it. Editor of the journal Psychometrika, Böckenholt's research interests centre around the quantitative aspects of marketing and psychological models of consumer behaviour.

"The most influential frameworks in the early days of the consumer research field were comprehensive models of buyer behaviour" he explained. "The implicit assumption was that buyer behaviour can be captured in one comprehensive model or 'grand theory.' This view has changed drastically over the past two decades and current research places much more weight on [isolated] real world phenomena and the study of consumer behaviour that occurs within the milieu of an actual marketplace."

To sketch a more realistic image of the consumer, Böckenholt's research draws from the fields of psychology, mathematics, statistics, sociology, anthropology, political science, economics, biology, behavioural ecology and neuroscience.

It doesn't hurt that his own educational experience spans several disciplinary boundaries. First studying economics at the University of Karlsruhe (Germany), Böckenholt graduated in 1982 with a bachelor's degree in psychology and computer science. His work in research methodology and quantitative psychology earned him a PhD from the University of Chicago in 1985. He taught in the department of psychology at the University of Illinois from 1986 to 2000, then in the department of economics at the University of Groningen (Netherlands) before arriving at McGill this past fall. He's teaching graduate-level courses in buyer behaviour and multi-level modelling, both with direct links to his current research streams.

One project seeks to better synthesize the vast amounts of data on consumer choice behaviour that has become measurable by several technological advancements over the past 20 years. "The availability of data and the level of access is unprecedented today," Böckenholt said. "First, scanner technology gave this kind of research a new way of looking at large samples over extended periods of time. Then the Internet gave us a window to learn how long a person spent at a site, what they looked at, what they bought, what they almost bought. Add to this the rise of computer power to analyze this data on a large scale and it's a good time to develop new methods, models and ideas."

The ultimate promise of this synthesis is targeted, one-to-one marketing, a vision of a personalized shopping experience fashioned from information gathered and analyzed with every click of the mouse. But it's not here quite yet.

"The fact that the data is so rich does not mean that we have a complete grasp," Böckenholt admitted. "Much work remains to be done before the information can be used effectively."

Böckenholt is hoping to create an e-commerce centre that will do just that, as well as establish McGill as an innovator in the rigorous scientific study of this burgeoning field. The range of topics to be investigated is as endless as the Web itself. For his part, Böckenholt has recently developed new ways of enhancing the predictive performance of choice models by incorporating information extracted from online shoppers' click-stream data.

"Typically, it is only considered how people differ in their preferences for products and their attributes, but not in terms of their actual decision processes," he said. "By taking a more procedural view of the decision-making, the goal of this research is to provide a coherent and unified account of the information searches and decisions."

Another current project "is concerned with both assessing and optimizing the return of marketing program activities on measures such as customer satisfaction, brand equity, market orientation, and market share," Böckenholt added. "Typically, these marketing activities are allocated at multiple levels ranging from geographical districts to stores and customers, which complicates determining both their effect on these variables as well as optimal allocation decisions."

Ed Kromer