How statistics can shed light on consumer behaviour
AUTHOR: Lachlan Colquhoun DATE: 30.11.04 ISSUE 3, 2004
Terry Elrod combines a love of statistics and an interest in psychology to create sophisticated models for consumer behaviour.
Professor Elrod, from the University of Alberta in Canada, is visiting the AGSM for six months, both to collaborate with the school’s researchers and also to introduce students to his Bayesian approach to modelling human behaviour.
“Understanding consumer behaviour and ‘knowing customers’ are never simple,” he says.
“Customers may say one thing, but behave in a different manner, and since each consumer is unique in their behaviour, modelling and predicting human behaviour are some of the most challenging tasks for marketers today.”
 |
{ | “These kinds of tools were not available 20 years ago, but now there is software out there that everyone can use to gain these insights.” |
PHOTO: Greg Newington (Terry Elrod)
Professor Elrod says the development of new software applying Bayesian theories allows researchers to use better and more realistic experiments on consumer behaviour.
“These kinds of tools were not available 20 years ago, but now there is software out there that everyone can use to gain these insights,” he says.
Modern consumer experiments are often administered using web browsers, so consumers can participate from home in real time.
“Consumers are not passive participants in experiments,” he says.
“I have been developing new experimental designs that take into account the complex behaviours we are studying and take advantage of the experiment being computer-based.
“For example, every consumer participant can receive a unique version of an experiment, and the more complex experimental designs allow us to develop more realistic experiments and more valid understandings of why people do what they do.”
Describing himself as ’the modelling guy’, Professor Elrod says most of his work involves collaboration with researchers, usually in the discipline of marketing, who have particular areas of study which benefit from advanced modelling or rigorously designed experiments.
Recently, for example, he has been involved in designing better surveys to help retailers understand how consumers respond to characteristics such as store appearance, convenience, service, product assortment and price.
“Each of these store characteristics needs to be measured using more than just a single question,” says Professor Elrod.
“For example, store appearance has to do with whether the store is visually attractive, whether the employees appear neat and tidy, and whether the promotional materials are visually appealing.”
Another area has been to design experiments on the effectiveness of releasing product information.
“Some current research indicates that companies should sometimes delay telling consumers about some of their product’s best features,” says Professor Elrod.
“It appears that companies should tell consumers enough to ensure that their product passes the ‘first cut,’ but then have an extra surprise advantage to their product that makes it stand out from the other ‘finalists.’
“I have been working closely with other researchers to develop advanced experimental designs which ensure this finding applies generally and is not true only for certain product features.”
Professor Elrod describes his approach as ‘harnessing statistics to gain insight’ and sees himself as a ‘conduit’ between the world of statistical theory and modelling and the real world of marketing products to consumers.
“When you have a statistics person like me with a psychologist it is quite neat, because we are both also marketers and we are looking at a phenomenon, and the challenge is to get a conceptual clarity that neither of us could develop on our own,” he says.
“I play the role of the listener, and I really want to understand what it is that they understand, so I can draw it out and design experiments which deliver insight.”
Professor Elrod said that, in addition to the Sydney lifestyle, he came to the AGSM to experience a world-class academic research environment.
“I had a choice of business schools to visit but I’m delighted I chose the AGSM,” he says.
“I have good relationships with several people from the school I’ve met and worked with before, and the school really is world class.”