‘Behavioural science is not a straitjacket for creatives and planners’ /

Richard Shotton became interested in behavioural economics after reading about the 1964 killing of Kitty Genovese and the bystander effect in The Tipping Point.

It was 2004 and Shotton was working on a ‘give blood’ campaign for the UK’s National Health Service with Charlie Snow (then head of planning at DLKW) when he read about Genovese in Malcolm Gladwell’s best-seller.

Genovese had been walking home when she was killed, and supposedly 38 people witnessed her attack at various different stages but did nothing to help her, giving rise to the theory that the more bystanders there are to an emergency, the less likely it is that anyone will intervene.

‘I remember reading that and thinking, well, this is exactly the problem we’re faced with,’ says Shotton, who at the time was a planning manager at media agency Zenith. ‘We go out and ask everyone to donate blood, and most people ignore us. So that got me thinking, could we apply this idea from psychology?’

It later turned out that the story about 38 people ignoring Genovese’s murder was most likely bunk. A 2007 article in peer-reviewed journal American Psychologist found no evidence for almost any of the claims about apathetic bystanders, which were first made in the New York Times two weeks after Genovese’s murder. And a research paper published in 2019 has since cast doubt on whether there exists at all such a thing as the bystander effect in human behaviour.

But whatever the truth about Genovese’s murder, reading about it gave Shotton and Snow a useful idea for the ‘give blood’ campaign. Instead of a message about blood stocks being low across the country, they would tailor the ads about shortages to individual towns and cities, to instil a sense of personal responsibility among communities.

Bystander effect or no, the idea worked and the campaign prompted a larger-than-usual response from people, and Shotton says that he has been hooked on behavioural economics ever since. Today, Shotton is one of the ad industry’s most committed and informed proponents of behavioural science. He published his first book on the subject, The Choice Factory, in 2018, and that same year he left his job as Manning Gottlieb OMD’s head of behavioural evidence to establish his own consultancy, Astroten, through which he has helped brands such as Google and BrewDog apply behavioural science to their own marketing.

We spoke to Shotton shortly after the release of his second book, The Illusion of Choice: 16½ Psychological Biases That Influence What we Buy, about what experts underestimate when it comes to changing behaviour, what it means to apply a bias laterally, and whether ad agencies make life more difficult for themselves by trying to look clever.

In the book, you talk about removing friction or barriersto a desired behaviour. Do you think that removing friction is more important than actually encouraging or motivating people to engage in a behaviour?

It’s interesting that [Nobel prize-winning psychologist] Daniel Kahneman, when he’s asked what’s the single biggest thing that he’s learned from his 60 years of running behavioural science experiments, says, ‘make it easy’.

It’s often far easier to change behaviour by removing friction than it is to change behaviour by motivating people to want to change. And one of the nice things about behavioural science is that no one is ever taken on their word alone. Everything has to be proved experimentally. But there is an awful lot of experimental evidence behind that idea.

In the book, I talk about the Bergman and Rogers study. They worked with the Department of Education in America in 2017 on a new service where parents could sign up and be texted advice on how to encourage their children to work harder. If the signup mechanic was slightly complicated, and you had to fill in a form to enrol, 1% of people did so. If they removed that bit of friction, and people could just enrol by texting the word ‘start’, enrolment rates went to 8%. And if people were told they’re already enrolled, they just have to text the word ‘stop’ to get out of the programme, 97% enrol.

What’s really clever – and I think this is the important part – is they then do a follow-up where they recruit experts in education. They tell them about the three different scenarios and they ask those experts to predict what they think will happen. They know that removing friction will boost uptake, they get the direction of change right, but they’re wildly wrong in terms of the scale. Those experts think, if you remove friction, you boost enrolment rates by 27 percentage points. But in reality, it was a 96 percentage point boost.

That’s a big theme, this idea that experts, including marketers, repeatedly underestimate the impact of friction. And what that means is most companies will have a misalignment of budgets; they will put too much money, time and effort into motivating people to want to change, and a smaller proportion into making the desired behaviour as easy as possible.

It’s a very fundamental human thing, isn’t it? Aren’t there studies where people are asked to improve a design and they all try to add something to it when the most efficient thing to do is to take something away?

I think the reason for that misunderstanding is that we want to believe that humans are sensible, rational decision-makers; we want to believe that the reason we would decide to enrol in an education programme is because of the merits of that programme. It’s uncomfortable to recognise that a big driver of signup rates is little barriers. I think it doesn’t reflect well on us as humans, necessarily, and so we pretend it’s not true.

What does it mean to apply a bias laterally? That crops up in the book quite a lot.

I try to distinguish between the literal application of these principles and the lateral application. Maybe an example from social proof is the easiest place to show it. Social proof is the idea that we are deeply influenced by what others do. So, if we know that a product is popular, it will become more appealing.

Most brands and most academic behavioural science consultancies apply that in a very literal way. So, the Tax Office goes out, for example, and says nine out of 10 people pay their tax on time, and they see an improvement in repayment rates. Or Weetabix will say they’re the number one breakfast cereal in the UK. That is a literal application of social proof.

Photo by Viktor Forgacs on Unsplash

However, there are plenty of studies which show you don’t need to even state your  popularity, just implying it is enough. That is where we get into the area of lateral applications. So, a classic example, Red Bull. When they launched, they wanted to create the impression that they’re popular. They’re only a launch brand, so they can’t honestly go out and say they have millions of units sold per week. But what they do instead is they pick a handful of nightclubs, and as they’re closing, they go down and they fill up the surrounding bins with loads of crushed red bull cans. So anyone leaving the nightclub, they see all these Red Bull cans in the bins and they assume that everyone else in the nightclub must have been powered by Red Bull. That’s a lateral application of social proof.

It’s interesting to us as marketers because it means that behavioural science is not a straitjacket that restricts what planners, strategists or creatives can do. It doesn’t turn their job into some paint-by-numbers approach. All behavioural science does is provide robust insights into human nature. But those insights should be seen as just the starting point. What the strategist or creative has to do is then take that insight and think about the best way to creatively apply the idea. As much of the power in behavioural science is the execution as the underlying insight.

Is it generally the case that biases wear out or lose potency the more that you’re exposed to them? In the book, you advise against one specific bias – the generation effect – too frequently.

That’s not quite what I meant. Going back to this point about literal and lateral, the generation effect is this idea that if you involve the audience and they have to generate the answer themselves, it becomes more memorable. If you apply that literally and leave blank spaces in your ad [copy], that, I think, feels like it has limited creative life. If you applied the idea laterally, not leaving blank spaces in the ads, but creating a puzzle for the audience, I think it gives you more space to iterate on the idea.

The Economist applied the idea of the generation effect in that more lateral manner, right? It’s not that they leave visual blanks in their posters, but when they say things like, ‘I don’t read The Economist – management trainee, age 42’ that is harnessing the generation effect. They don’t tell you that you’re a failure, unless you read The Economist, what they let you do is work that out for yourself by solving the puzzle. And that slightly more oblique route, I think can be done again and again and again. So it was more of a reiteration of the point that if you apply these ideas literally, it pins you in a little bit. If you are a bit more free-spirited, you’ve got far more opportunities.

The book also raises the point that, in psychology, the opposite of a good idea can be another good idea. For example, it can be beneficial to remove friction from a decision-making process and it can also be beneficial to add friction. How does one choose between such seemingly contradictory approaches?

If you want to change someone’s behaviour, generally you remove friction. That’s a pretty solid insight. However, there are a series of experiments around this thing called the Ikea effect, which show the more effort you make people put into getting your product, the more they’ll appreciate it. That seems contradictory, but I think if you dig into those different sets of experiments, you see they’re all focused on slightly different things.

‘Making it easy’ tends to work when you want to change behaviour; adding friction tends to work if you want to change people’s perception of quality. So if your brand is struggling in terms of quality perception, the Ikea effect might be something that you want to prioritise. If you’re interested in converting people to action, then you want to be applying that principle of ‘make it easy’.

There are thousands of these experiments; the job of a planner or strategist or a marketer is matching the unique challenges of the brands to a specific study. And I think when you are clear about what the blockage or barrier is that you as a brand are facing, then it becomes much clearer about which psychological study is the most relevant for you.

So there are no true contradictions in behavioural science, it just comes down to understanding your challenge or your goal a little bit better?

Yeah, I think that’s fair. And even ones that feel contradictory at first, if you dig into them, you can see there is a kind of rationale. There’s famously the foot-in-the-door technique, which is, if you want [someone] to make a big change in their behaviour, you ask them to do a little thing first, and then you follow up with a bigger ask later on. That’s based on the work of [Jonathan] Freedman and [Scott] Fraser back in 1966.

There’s a separate body of studies by [psychologist Robert] Cialdini, in fact, called the door-in-the-face technique, which is where you ask someone originally a massive request, you expect them to say no, and then you follow up with a much smaller ask. Now, both of those tactics can be effective. So it feels like you’ve got two wildly different ways of persuading people. But there’s an underlying thing that unites both of those, I would say the principle of ‘make it easy’. With the foot-in-the-door, you make the first step in the journey feel very hard for people to say no to. With the door-in-the-face technique, the value of that giant ask is it makes the genuine request look much smaller in comparison.

There’s no situational preference for either one of those, you can use them to achieve the same thing [but] you’re right that there are different times and places for different biases. If you go to something like social proof, which is far more widely studied, there are examples where if someone is uncertain, social proof tends to be more powerful. Social proof tends to be much more powerful among younger people. Social proof tends to be more powerful if people are scared. So there are studies into how the situation affects the power of the bias. But that doesn’t mean there is that level of research on every single bias. It’s still an evolving field.

In your book you include a quote from The Black Swan author Nassim Taleb about how people with no skin in the game have more interest in appearing smart to their peers than providing solutions. That seems like a provocative thing to throw into a book that will be read by a lot of agency people. Do agencies make their lives harder than they need to by trying to appear clever?

What Taleb argues is that if you are not the owner of a business, there is a danger that you become rewarded for perception, not results. Now, he’s talking as much about academics as any other category, but this book is mainly designed for marketers and agencies, so I wanted to think about how we can sometimes be guilty of this issue.

I talk about it specifically in relation to a set of studies by [Matthew] McGlone. What he shows is, if you give people a phrase, let’s say ‘woes unite foes’, [they] are more likely to believe it is true than people who see ‘woes unite enemies’. And we’re talking about a 17% swing in believability. That’s known as the Keats heuristic, and what it argues is people conflate truthfulness and ease of processing. And one way to generate fluent processing or easy processing is by using a rhyme. So rhymes have been shown to boost believability. I’ve also done some tests that show that rhyme also boosts memorability.

But what’s interesting is, if you then look at the prevalence of rhyme in advertising, it’s in long-term decline. You’ve got this body of evidence that says it’s successful, yet you have an industry that rarely uses it. The only way I can explain it is thinking that maybe it’s more about how using rhyme reflects on us as professionals – it doesn’t look very sophisticated. That, I think, is the explanation for its dwindling usage, rather than any data on its declining effectiveness.

When we last spoke you mentioned that the pratfall effect was one of the most effective and underused biases in marketing. Is there any indication that it’s enjoyed a resurgence?

I think it is a very, very useful tactic, but it’s one that’s still applied reasonably rarely. Paradoxically, I think some of that rarity explains why it’s so powerful. We know that one of the main things that drives memorability is distinctiveness. So, yes, the pratfall effect will boost appeal, it will boost trustworthiness because you’ve admitted a flaw. But also, when most ads brag, admitting a flaw will make you distinctive and therefore more likely to be noticed.

Is it still your favourite bias?

Some of the experiments on pricing are fascinating. There are so many of them out there, and it’s so easy for brands to apply them on their websites quickly to see if they work – there’s no cost of doing it. I often find that’s a brilliant area if I’m working with brands for the first time, to focus on. The problem is if you say, change your branding, change your positioning and admit a flaw, firstly, whether or not that is successful will be as much about the creative execution as the insight and, secondly, it will take years to see the impact. So it’s quite hard for someone to make their first leap into behavioural science using that idea. Whereas, if you say to someone, give a precise price or add on a third super-expensive option, or talk about your unit pricing, not your total pricing, there’s no risk. It is completely free, or almost free to test, and you’ll see very, very quickly whether or not it works. I really love talking about those ideas to brands because there’s nowhere to hide. We will see very quickly whether it’s a genuine insight.