Behavioural economics has become an important discipline for brand owners eager to understand why their consumers sometimes make irrational decisions that run counter to economic model predictions. Yet all too often, time and money is wasted by the challenges involved in turning behavioural insights into effective marketing practise.
Consider British Airways which set out to discover what to put into new self-service fridges it planned to introduce in First Class. Following focus group research, BA filled the fridges with salad and fruit. But when an air stewardess added cake and chocolate bars it was these indulgent snacks, rather than the healthier options, that were eaten. The reason? First Class passengers want indulgent snacks when they wake at night.
BA’s experience neatly illustrates one of the problems brand owners face when gathering and acting on behavioural insights. In a laboratory in the day time, focus group participants like the idea of fruit and salad, but they are considering their preference in a very different environment to the one experienced by an air passenger mid-flight when he or she wakes in an oxygen-poor, low-pressure climate in the middle of the night.
In short, artificial results come from conducting research in an artificial environment. And at a time when digital has fundamentally changed how marketers understand modern consumers and how those same consumers rationally and irrationally participate in dialogue with brands, this is a pressing and serious concern. However, there is a solution.
Three decades ago, when champions of behavioural economics first started arguing for a seat at the market research table, behavioural research was done in limited contexts – typically, in full-blown laboratories or with students on university campuses. The idea was to apply scientific principles with the use of a controlled, well-observable research environment.
Thirty years on, many marketers still trust lab-based research – especially for testing the usability of web sites or new product development. But as BA found to its cost, lab-based studies risk emphasising behaviour that just would not happen in the real world. So for this reason, additional experiments must play an important role in validating if and how far lab-based insights, including those from behavioural economics, might actually work – an approach that comes into its own with A/B testing of websites or emails.
Another challenge when turning behavioural insights into effective marketing practise also relates to the way in which behavioural researchers have long generated behavioural insights.
For decades, behavioural economics revolved around finding irrational influences – or ‘biases’ – which can affect human decision-making. Hundreds were identified, but they were a vast collection of unrelated things rather than part of a coherent, overall model or framework. As a result, there is no clear path through the seemingly unlimited and unrelated variables that might make a consumer behave in an irrational way.
One solution for marketers is to build their own framework of biases which impact on their consumers’ decision-making in their own particular marketing context.
At House of Kaizen, we have developed a framework for bias which accounts for four key sources of online shopping bias: network (social influences); identity (self-concepts); cognition (limitations in the way prospects process rational information); and emotions (feelings). Each, in turn, features five or six individual variables ensuring the framework is both focused and relevant.
So, when applied to an a leading antivirus software brand’s online shop, our framework reveals that contrary to expectation, emotions are highly relevant for purchase decision-making amongst antivirus shoppers. Furthermore, 75.9% of these consumers do not just believe they know ‘little’ or ‘moderately little’ about such software, they also don’t want to learn more about it. Rather than making informed decisions, they tend to buy if the product removes conscious and subconscious anxieties.
All of this is in sharp contrast to perceived rational economic behaviour. And to test our findings, we developed new content for the web site that up-played anti-anxiety facts.
Three different designs for the presentation of this content were then trialled – delivering exactly the same content against the original page, one looked ‘scientific’, one ‘tech-y’ and one ‘casual’. And the results were conclusive. When compared to the control page, only the scientific one produced a sales uplift – of 21.1%, so not insignificant. Why? Because a scientific approach is perceived as the most reliable and trusted way to counter virus threats.
In our experience, turning behavioural insights into effective marketing practise depends on two things. First, experimentation within the actual context of how a consumer will engage with a brand’s product or service. Second, frameworks built and tailored for an individual marketer’s specific context. For real and credible insights capable of being effectively turned into marketing practise can only be achieved by combining the two.
Used in tandem, then, these two approaches eliminate the fundamental problems associated with acting on behavioural insight that otherwise weaken the important role in marketing that behavioural economics can – and should – play.