This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Find out more here

Wednesday, 16 April 2014

Questioning the nature of research

Ogilvy Group UK vice chairman Rory Sutherland is a leading proponent of applying behavioural economics theory to the practice of marketing and advertising. He’s interested in what people actually do, and how they actually make decisions about the products they buy, rather than what they say they do.

For him, survey research is an inherently unreliable means of getting to the truth of consumer behaviour and emotions. It’s still “better than ignorance in many cases”, but he says questions must be asked about how research can be “significantly improved” given all we now know about the human brain, thanks to behavioural economics and neuromarketing.

Research: Why is it important for researchers to try to measure consumer emotions?
Behavioural economics shows that affect has a strong effect on behaviour, and the mood we’re in and all manner of contextual variables affect the decisions we make. Since the majority of money spent on market research is spent on de-contextualised market research – which is essentially asking people to make a theoretical decision – how much predictive value does this really have?

We have to ask some questions about how market research can be significantly improved in light of what neuromarketing and behavioural economics teaches us. The value of research doesn’t diminish in the slightest – if anything it’s heightened – but the nature of the research that we do has to be questioned.

So the ad industry isn’t currently well-served?
For television, where it is automatically assumed that some sort of conscious memory of the commercial and its message has to be carried between the consumption of the commercial and the purchase of the product, there’s some value to using conventional research as a vague measure of likely effectiveness of TV advertising.

“As you get closer to the point of decision, there are factors at work which never really appear in conventional market research. They are contextual factors, social factors”

However, what we realise is that as you get closer to the point of decision, there are factors at work which never really appear in conventional market research. They are contextual factors, social factors. No-one in any research group would ever say, “If there are four brands of shampoo, I’ll buy the one that has most bottles on the shelf”, or “I’ll choose the one that’s on the third shelf up because it’s the one that doesn’t require much reaching down” or “I’ll look at the prices of three products and choose the one in the middle.”

In reality, we use heuristics and shortcuts and cognitively miserliness like this all the time. The mistake that quite a lot of advertising methodologies make is assuming that brand preference translates perfectly into purchase behaviour. It’s also making the assumption, of course, that preference is formed in advance of behaviour. Quite a lot of evidence from both behavioural sciences and from neuroscience suggests that we act first and form our opinions in light of our actions.

If there happens to be a Starbucks on the way to work you will end up going to that more than Costa Coffee, and as a result of that action and in order to make sense of that action, you’ll decide to like Starbucks. And when a market researcher comes up and asks what is your preferred coffee chain, you’ll say Starbucks. But in effect, your preference is a product of your behaviour and not the origin of it. I often say that if you hear a man complain that his wife doesn’t understand him, it doesn’t mean he’s going to have an affair – it means he’s already had one.

Quite a lot of the assumptions on which research is based are perfectly sensible in many ways and a pretty good proxy for what is needed – it’s better than ignorance in many cases. Although, if you spoke to Orlando Wood at BrainJuicer and a company called Real Eyes, they would say that our emotional reaction to a message has more predictive value in terms of behaviour than our stated rational take-out from a message.

So how do we take research past this point, so it’s no longer just a “pretty good proxy”?
I’m very favourable disposed to any research that takes place at or close to the moment of decision, because I think the way we think we decide and the way we actually decide don’t have that much in common. The conscious rational brain isn’t the Oval Office; it isn’t there making executive decisions in our minds. It’s actually the press office issuing explanations for actions we’ve already taken.

One of the reasons we chose the marketing model that we did – where we overtly state a proposition in a dramatised or comedic form that contains within it a persuasive argument for why you should buy a product… Advertisers are very comfortable with it because it absolutely defends you against charges of manipulation. You are dealing purely at a conscious rational level of awareness and preference. The only problem is it’s not really true. Some of the work by [marketing professor] Byron Sharp on mental and physical availability… his assertion would be that distinctiveness is more important a quality in advertising than persuasiveness and differentiation. The most important thing is that a person remembers your brand as having done a lot of advertising because in making purchase decisions we have an availability bias, which pushes us towards buying things we’ve heard of before.

“The conscious rational brain isn’t the Oval Office. It’s actually the press office issuing explanations for actions we’ve already taken”

This is not necessarily irrational. There are perfectly good evolutionary reasons why we’ve developed a lot of these heuristics. The fact that we trust advertised brands more than unfamiliar brands is very sensible. If you spend the money on advertising you are playing a long game and you have faith in your product and a reputation to lose. It’s not in my interest to make a quick buck if it destroys that brand. Our preference then is perfectly rational, but it may be unconscious or subconscious – it may be some Darwinian social instinct we’ve evolved around trust, which is, “Trust people with a reputation and something to lose; don’t trust people with nothing to lose.” It’s why, by and large, I wouldn’t lend money to crack addicts.

But that would mean traditional measures of brand awareness are quite useful then?
Yes, awareness and fame and distinctiveness, and also the awareness that you advertise a lot, not just the “Have I heard of them?” unprompted awareness score. Which of course we don’t always we measure. We don’t ask people, “Which of these have you heard of most? Which is the most famous?” These are degrees of ubiquity but those things can be very valuable.

There’s some very good work done by the Institute of Practitioners in Advertising which shows that your excess share of voice – the extent to which your share of voice exceeds your share of market – is a very reliable predictor of whether or not you’ll grow. So I’m emphatically not downplaying the importance of fame, awareness, mental availability – whatever you want to call it. What I would downplay is detailed dissections of consumers stated reasons for adopting or planning to adopt a particular course of action.

Can researches ever get to understand what goes on in the Oval Office, as you put it, if every time you try to get in there the press office gets in the way?
Yes, very easily – through experimentation. Control-based tests and continual experimentation. That’s how Tesco and Walmart became successful: they tried something different every day. If you talk to academics, they are slightly baffled by the ratio of money spent on research versus the money spent on experimentation within marketing. Direct marketers would be frustrated with this because they – as are some digital marketers – very great believers in split tests.

“If you talk to academics, they are slightly baffled by the ratio of money spent on research versus the money spent on experimentation within marketing”

Quite often, people within a group will pretend they are a maximiser, when most of our decisions are taken as satisficers. We always claim in the presence of others that we are great connoisseurs looking for the best value for money we can find, but most of the time we simply don’t have the mental energy for that. It would be an insane use of our mental resources in any case. So what we do is look for something that is pretty much guaranteed not to be crap. That’s why in some ways you can never quite make sense of the popularity of McDonalds. Everybody, whenever you talk about food, they’ll always talk bullshit about health and Italy and olive oil. But actually, when it comes to eating, what we want is a place that won’t rip us off, won’t give us food poisoning, the toilets will be clean, the service will be OK, and everything will be pretty good.

Paul Dolan, the government’s wellbeing adviser, says: “Nothing is as important as we think it is when we’re thinking about it”.

What explains this under-reliance on the experimental method?
I have no idea.

Is it to do with budget constraints?
It might be idleness, sheer idleness in the sense that constructing experiments takes a bit of work, but it’s relatively inexpensive.

What do you make of efforts to directly measure emotional response? FMRI, EEG, biometrics – that sort of thing?
Very simply, I like it to be done and I’m glad it’s going ahead and it reveals some interesting and revealing things. But its immediate applicability seems to be… I mean the most important revolution in marketing is really just like the most important revolution in science 150 years ago, you know: the scientific principle of experimentation, hypothesis and verification. That’s the most important thing. What I think neuromarketing does is shines a useful light on some of this stuff, because it occasionally provides the why to the what. But as pragmatic marketers, the most important thing to us is the what.

Where do you want research to end up?
More behavioural economics. Behavioural economics is starting to provide aggregate experimental results, which show tendencies, common biases and heuristics. The second generation, the next generation, will be databases of susceptibility, that will shows us which consumers overall are affected by which biases and which heuristics.

  • This interview was conducted in May, for our June edition Special Report on Measuring Emotion. You can read the full print edition, plus web exclusive content, here.

Follow us on
Follow us on Twitter

Readers' comments (11)

  • @david - in reply to your question; no, the brain works in the same way irrespective of the category. The implicit system (System 1 in Prof Kahneman speak) governs c.90-95% of our decision-making (Prof Zaltman) and works with patterns/heuristics which have built since birth. This system pre-selects, and filters, what is passed to the explicit system (System 2) for what effectively amounts to post-rationalisation (Rory's press office analogy) ratification & checking. The explicit system acts like a gate-keeper but its policing is lax, particularly under any one of 4 conditions - time pressure, complexity (too many variables), overload (too many tasks) and low involvement. Think about a category and whether one of those conditions apply. A car manufacturer once challenged me on this, saying 'but we KNOW car purchase is purely rational. People go down a purchase funnel and make rational trade-offs eg budget vs model / engine size / specification'. When I asked why they made such trade-offs he replied 'so that they can get a BMW on their drive'. People don't 'think-feel-do', they 'feel-do-think'.

    Unsuitable or offensive? Report this comment

View results 10 per page | 20 per page

Have your say

Please add your comment. You can include links, but HTML is not permitted.
Your email address will not be displayed on the site. All comments are moderated.