Alan Mitchell: Reinventing marketing - Behavioural economics has yet to deliver on its promise
A view from Alan Mitchell

Alan Mitchell: Reinventing marketing - Behavioural economics has yet to deliver on its promise

The discipline may have undermined the assumption that consumers make rational decisions, but it has not come up with a credible alternative explanation for their motivations, writes Alan Mitchell.

Is behavioural economics (BE) really the marketing saviour that many believe it to be? IPA president Rory Sutherland, for example, suggests it's a chance for marketers and agencies to regain clout in the boardroom and to 'reboot' the advertising industry. Is it really?

BE is certainly generating intellectual shockwaves, as it challenges long-standing assumptions about how and why people behave as they do. It is probably the start of something big. Right now, however, the emphasis should be on 'start', not 'big'. Here are some reasons to remain cautious.

BE is not yet a science, and may never become one. To be sure, it borrows from a long list of scientific disciplines, such as social, cognitive and evolutionary psychology, as well as sociology, anthropology and economics. Yet, right now, all it does is jumble a selection of their findings under a new umbrella term without adding anything new.

There is a problem with this cross-disciplinary approach: it relies on modes of explanation from different disciplines that are often incompatible and fail to 'talk' to each other. Areas of interest to BE include themes such as 'anchoring', 'availability', 'salience' and 'framing' (where, in different ways, we consistently give irrelevant or unimportant information undue weight in our decision-making); the workings of heuristics such as 'a higher price means better quality' or 'if in doubt, do what everyone else is doing'; the influences of emotional drivers for things like safety, security and status; and powerful behavioural patterns such as inertia.Some of these themes relate to the perceptual or emotional hard-wiring of the brain, and some are learned in social contexts, while others relate to social instincts.

So far, however, BE has failed to offer any integrated theory to explain these phenomena in a unified way or any means of translating one mode of explanation into another. It has done nothing to fit the pieces of the jigsaw together. All it has done is create a list of interesting phenomena - a list that gets more confusing the longer it gets.

No alternative

In fact, so far, BE's main contribution has been negative. It has taught us that human beings are not 'rational', as once claimed by economists. However, it has not replaced rationality with any coherent, robust, rounded alternative.

That is important because, without this alternative, attempts to 'apply' BE to marketing strategies and initiatives are on shaky ground. So far, for example, BE experiments have uncovered a wide range of decision-making tendencies and propensities. Fascinating stuff, but it has not calibrated just how powerful or big these tendencies and propensities are, when they do or do not kick in, or whether they apply evenly across populations or differently to different people.

We know, for example, that peer pressure can influence how people behave, but, as yet, no BE theorist has explained why some people are more susceptible to peer pressure than others. Nor has BE, quantified or graded the size of different population segments according to their 'peer pressure propensity', accurately predicted the likely size of behaviour shifts driven by peer pressure, or demonstrated whether these differences are fixed or change over time. Without robust understandings of this sort, BE-based initiatives might not work as mark-eters hope or expect.

BE is also strangely quiet on the subject of how people change and learn, preferring to focus (no doubt for reasons of experimental simplicity) on isolated snapshots. Here's an example. One BE experiment revealed the power of priming and suggestion by giving two sets of students different CVs for a new lecturer. One CV said he was 'warm', the other said he was 'cold and distant'. When asked for their impressions of him after his first lecture, the students tended (note: a tendency, not an absolute) to echo the assessments that had been fed to them.

What the research did not do, however, is repeat this question for the second lecture, or the eighth, to see how long the effect lasted. Nor did it measure how the scores changed once students from the two groups started chatting together.

In other words, the experiment revealed a 'tendency' but did not show how strong it is in different circumstances, how long it lasts in the face of real-life experience or how robust it is in the face of competing tendencies and influences. It offered only a first, tantalising glimpse, not a conclusion.

Competing influences

Enter another problem with BE. Doing experiments designed to isolate what is being studied from other influences is the classic scientific approach. Great.

However, what it ignores is con- text, which just happens to be all-important for marketers. In the real world no BE effects or influences ever occur in isolation. They all happen together, in a jumble.

At any point, for example, an individual may be being influenced by 'framing' and 'priming' and 'status quo bias'. Is one of these influences always stronger than the others? If so, how much stronger? Do their relative strengths change according to context, or who the individual is (segments)?

Does the addition or subtraction of another influence (say, peer pressure) radically change the outcome, like adding a new chemical element to a molecule to transform its qualities? Do some influences always work in 'synergy' to create the same behaviour change, or do they sometimes cancel each other out to leave zero noticeable effect? If inertia and status quo bias tell me to stick with my current brand, but ad-driven priming and a retailer's framing tell me to pick another, which one wins out?

BE has no answers to such questions yet - and there are plenty more where that came from. For example, BE tends to focus on some aspects of human behaviour while ignoring others. Researchers have zeroed in on one mode of mental operation (unconscious, automatic decision-making processes) while ignoring or downplaying others, such as our conscious, reflective instincts to 'stop and think'.

Yet, stopping to think often overrides unconsciously generated impulses, and humans evolved this facility for a very good reason: it aids survival by leading to better decisions.

Priming, framing and inertia are all very interesting, for example, but exactly how powerful are they when they come up against a good, old-fashioned 'rational' incentive such as a price discount? So far, BE hasn't answered any of the fundamental questions about this relationship between 'automatic' and 'reflective' systems.

There are other instinctual aspects of human behaviour that BE largely ignores. Take reciprocity, where we trade favours with people we trust and enmity with those we do not.

Reciprocity requires a type of 'mental accounting' to keep tabs of who has returned favours and who hasn't. With this comes apparently hard-wired instincts to sniff out people who attempt to cheat, manipulate or lie - and to punish them. One instinctive way of doing this is via gossip: undermining the cheat's social reputation via word of mouth. Marketers attempting to use BE insights to manipulate consumers into doing things that are against their own interests might discover these instincts for themselves.

That is why Rory Sutherland's BE agenda - to 'use ideas to turn human understanding into business advantage for our clients' - is so dangerous. Where is the consumer advantage in BE? Is the client's advantage always the same as the consumer's? What happens if they are opposed?

BE is opening new doors of understanding that could transform how we think about, and do, marketing. For your own safety, though, proceed with caution and humility.

Alan Mitchell is a respected author and a founder of Ctrl-Shift and Mydex.