Understanding consumer behaviour


In this blog by one of our newer researchers Carlos Calmasini outlines his PhD passion, a PhD in Consumer Behaviour (Marketing/Economic). Being interested in all aspects of Analytics and Consumer Behaviour, Carlo likes to work in the boundary between academic and market research and is part of Direction First Analytics team.

The understanding of the main drivers behind individuals’ behaviour and the subtleties and nuances involved in consumption choices always raises interesting and challenging problems. At least, this is what I thought when I decided to dig into this field of knowledge and started my PhD. The field being very wide, first step is to decide what direction to take among multiple approaches, from psychology to economics/behavioural economics to marketing/consumer behaviour just to mention the main ones. The latter is perhaps the most applied one, and generally takes a lot of ideas from the former two. In this short writing I’ll summarize in very broad terms some aspects behind the recent evolution of consumer behaviour. This summary clearly reflects my education and research interests, which explore online services in both consumer behaviour and behavioural economics (i.e., marketing and economics fields respectively) with experimental/choice methodologies.

Behavioral models from both marketing and economics have traditionally relied on a consumer rational approach that has been increasingly criticized in the last two decades, starting with the work of Kahneman & Tversky and of Thaler (who were awarded the Nobel Memorial Prize in Economics in 2002 and 2017 for their work on behavioural economics). These contributions were initially mostly theoretical and aimed at demonstrating how consumers’ behavior deviated from the “prescriptions” of the rational models because of cognitive biases. Basically, the initial perspective behind the bias argument (you can just google “wiki list of cognitive biases” to see a list of a couple of hundred) was that the consumer was deviating from the “right” way to think, the rational one.

Since then many steps forward have been made and these days the approach is pretty much reversed, with the actual behaviour seen as the norm while the research effort aimed at explaining both the reasons for consumer behaviour and at modelling, or better mimicking, such behaviour. Several streams of research are trying to develop such perspective to understand and model bounded rationality in decision making.

From an understanding perspective among the most important contributions, just to mention two, are the work of G. Gigerenzer on heuristic decision making and the hype of marketing academic research toward the role of emotions in consumer behaviour. Along with this shift from the initial what and how biases happen to the current why they happen, applied research methodologies has been developed, from the most technological and complex (fMRI, EEG, Eye tracking) to the most approachable ones (e.g., implicit measures), each one with its pros and cons.

If actual behaviour is not easy to measure, modelling it can be even more challenging. However, recent academic research started to provide some interesting and applicable contributions. Following the heuristic decision-making field, essentially based on consumers taking shortcuts in order to deal with the cognitive load of decisions, in recent years researchers have been able to produce predictive models that could be used in applied market research. For instance, let’s take the recent concern of many marketers due to the new GDPR legislation recently mentioned also on this blog. Usually the business models of big providers of online services (e.g., Facebook, Google, Twitter, Uber) rely on consumers’ data being linked together and used for several purposes. We know that people do not read privacy policies and that they tend not only to forget information released in the past but also to underestimate the potential negative consequences of such release (e.g., dynamic pricing from Uber and Amazon). In more general terms this is the same kind of problem (i.e., problem of immediate gratification in intertemporal choice) that consumers face when for example buying financial products that cover medium/long periods. So, how to model this discounting behaviour for predictive purposes? An alternative explanation to the classic approach to rational discounting behaviour (i.e., an hyperbolic discounting function, rather complex) has been proposed in terms of simple heuristic-based model. The heuristic model, based on a simple average and on consumer’s reference point, has been designed to mimic the cognitive algorithms used by individuals and their resulting “biases”. In line with experimental evidence, researchers were able to conclude that this heuristic model explain intertemporal choices better than traditional “rational” discounting models.

So, research methods for both understanding and modelling current “non-rational” behaviour, if correctly applied, are proving capable of producing remarkably results and allows us to better explain the relationship between consumers and products.


Leave a comment