"I don't trust surveys. We shouldn't be asking people for their opinion. We should be trying to understand how they actually behave when faced with a real decision"

In recent conversations, the above refrain has become increasingly common. It hints at a desire to really delve deep into the mind of a customer, to understand what interests them, what truly motivates them, what causes them to act upon their interest and ultimately convert to a sale.

Interestingly, a similar mindset has been driving the growth of one of the most fashionable areas of academic research, behavioural economics. Frustrated with classical economic theories which seem disconnected from reality and concerned with the use of hypotheticals in psychology, behavioural economists seek to understand the quirks of human behaviour in a truly robust and objective fashion.

Their tool of choice for this analysis is highly controlled experiments, with real decisions to be made, with real money at stake. Take the concept you are interested in, apply it to a treatment group, and observe how their behaviour differs from a control group. This approach obeys the fundamental principles of the scientific method. Of course, for a business seeking to understand how its customers behave, there exists an alternative approach. Driven by the decade of data capture since the early 2000s, data analytics can provide unique insight into a wide range of key business questions.

At its core, the insight provided by data analytics relies on two foundational stones which allow you to draw meaningful conclusions about customer behaviour. First, you have a data point which describes how your customers behaved and, second, you have a data point which describes a key factor which will have influenced their behaviour. For organisations with sufficiently rich data the ability to investigate these key factors through analytical models and techniques can be incredibly powerful.

However, the lesson which can be learnt from behavioural economics is that not everything which influences behaviour is easily represented as a data point. Framing, emotions and memory, for example, are well known to cause irrational behaviour but their impact is unlikely to reveal itself using conventional data analytics.

If we choose to limit ourselves to studying only the data we have what lessons will go unlearned? The academic approach, highlighted above, suggests we should be willing to go beyond the standard ideas and standard theories, and look deeper.

When considering a key business question, it can be valuable to look through the lens of behavioural economics in addition to the insight found through data analytics. Once the behavioural theory has been established, quick experiments can be devised to test the theory allowing for the true impact of the hidden factors to be revealed to the organisation. The knowledge that this approach brings can be simultaneously surprising, simple to understand and cheap to implement.

It is fair to say that behavioural economics is not a traditional or typical tool that a firm might employ to understand more about its customers. But in a competitive world, where truly unique insight is incredibly valuable, it remains to be seen for how much longer that will remain true.

Stephen Lovelady
Stephen is an analytics and behavioural economics specialist in the Marketing & Insight team within Deloitte's Customer practice. He has worked on analytics projects for major FSI clients and is currently assisting with agency management transformation for a large CPG firm. He joined Deloitte from the University of Warwick having previously lectured in econometrics and behavioural economics.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.