Richard Shotton: burden of proof
"The Choice Factory" is the most popular new book for marketers so far this year. Well read, well discussed and well liked from everyone, it casts an eye on the the history of how customers come to given decisions, and the psychological background behind them doing so. We caught up with its author Richard Shotton on the book's themes, and their applicability to contemporary marketing.
What was your thinking behind writing a book?
RS: The initial reason for writing the book was laziness. I was running lots of psychology experiments for my agency, trying to show how behavioural science could be applied to advertising. Afterwards, planners would come up to me and say - you ran an experiment for such-and-such a brand a few months ago, can you share the results so that we can apply them elsewhere?
Unfortunately, my filling was so woeful that I could rarely find the results, or all that I had was a three-slide presentation of full-bleed images. That was frankly useless, so I ended up re-running many of the experiments.
That was when the laziness kicked in. I thought - whenever I do an experiment, I'm going to write it up, send it to Campaign or MediaTel, and I'll have a permanent record. That way, the next time people ask me for the data I can just send them the article. The plan worked well and saved me loads of time.
However, I began to think, if I've written all these articles, why not use them - not as a cut and paste job - but as source material for writing a book?
That was the honest reason for it. The post-rationalisation is that there is something slightly crazy about marketing today.
Everything that we're trying to do concerns behavioural change – whether that's getting consumers to pay more, switch brands, or buy more regularly. It's all about behaviour change. So why wouldn't you as an advertiser draw on the science of behaviour change?
There are lots of amazing books about popular psychology, but few of them apply it to advertising. I felt that there was a real gap for a practical, simple book that applied behavioural science to advertising..
How does the book help the reader to understand the theories and practices regarding social psychology?
RS: I was keen that the book did two things.
First, that it was practical. Each chapter covers a single bias – it looks at the academic evidence, my own experiments and most importantly, the advertising applications. The bulk of each chapter covers how you should adapt your marketing in the light of the bias – whether that's pricing or promotion, planning or creative.
Second, there has been a lot of coverage in the industry in the last few years of a handful of biases. Social proof, loss aversion and the like. I wanted to discuss some of the lesser known biases, like the pratfall effect and Goodhart's Law. These are hugely relevant to advertising but haven't had as much coverage.
What are your thoughts on the relationship between behavioural science and the advertising industry today?
RS: For many people outside of the industry, I think that they'd be surprised at how little psychology and behavioural science are used by marketers.
There's a historical reason for that. From the 1930s onwards you had famous, cutting-edge psychologists in advertising: people like Louis Cheskin and Ernest Dichter advising brands.
Cheskin, for example, was the guy who turned margarine yellow, realising that when you tasted this grey gloop, it wasn't the taste that people disliked, it was the aesthetics. So, a lot of these great psychologists were in agencies, or worked with them.
Let's now fast forward to 1957, when Vance Packard wrote The Hidden Persuaders. He made the false claim that advertisers used subliminal messaging to manipulate consumers. Because of this, psychology fell out of favour in agencies. They felt that its application was ethically unsound. I don't think that advertising has fully recovered from that. The gap has been filled with the more rational school of thinking in business, and psychology is left out.
But, on a positive note, that is changing. Over the past couple of years, I've noticed that clients are coming with more and more questions about behavioural science.
Why should advertisers trust behavioural science and psychology as a means to improve campaign effectiveness?
RS: Despite the £20bn spent on advertising in the UK per year, a worrying proportion of marketing decisions are based on flimsy evidence. Too often they're based on the opinion of the highest-paid person or most eloquent person in the room. Basing your decisions on peer-reviewed evidence from psychology and behavioural science is certainly an improvement on that.
However, there are some specific objections to behavioural science. For example, some critics point out many of the classic findings are based on lab experiments involving students. That may have been a fair criticism in the 1960s and 1970s, but many modern experiments are conducted outside of the lab.
For example, there's a wonderful study by Elliot Aronson, a psychology professor at Harvard in the 1960s, who came up with the idea of the Pratfall Effect. His original lab study involved recording an actor taking part in a quiz. The actor had been given the answers in advance so unsurprisingly he performs brilliantly and answers 92% of the questions correctly. However, after he finishes the quiz, the participant makes a blunder. He stands up and knocks a cup of coffee down himself.
Aronson takes that recording and plays one of two versions to people: either the entire recording with the final mistake, or just the quiz. What he finds is that the contestant is significantly more appealing when they have exhibited that flaw.
Aronson termed this the pratfall effect – essentially the idea that people or products who exhibit a flaw become more appealing. What's interesting about that finding is that it was initially based on a lab experiment, but it has recently been proved in the real-world.
In 2015, North Western University undertook analysis of more than 111,000 product reviews. They looked at the likelihood of purchasing a product after reading a review. They discovered a clear pattern whereby the likelihood of purchase increases as the review score improved. However, as Aronson predicts you then reach a stage - between 4.2 and 4.5 out of 5, dependent on the product category - in which likelihood to purchase peaks, and then declines afterwards.
Consumers know from bitter experience that there is always a trade-off in life. No product is perfect. If you don't reveal where your flaw is, you leave it to consumers to guess. And they may assume that flaw lies in a crucial area.
Many of the best advertising campaigns have used the Pratfall Effect. VW: "ugly is only skin deep". Guinness - "good things come to those who wait". Listerine - "the taste you hate twice a day".
The best ads have harnessed this bias - partly to gain trust. Consumers assume that advertisers don't tell the truth so one of the ways to solve that trust gap is to admit a weakness, and tangibly prove your honesty. Your other claims then become more believable.
The best advertisers also pick a weakness that has a mirror strength. VW - we're ugly because we don't care about aesthetic fripperies; we care about engineering excellence. Listerine - we taste horrible, but so do all potent medicines.
When brands make mistakes, they are often slow to act, which causes negativity within their customer base. How can this situation be improved?
RS: The interesting thing about the brands that have used it best - Stella, cream cakes (Naughty but Nice) and so on - they have not blundered and then apologised for it. They have actively publicised their flaw, which is the most powerful way to do it.
A brand's reaction to a mistake is fascinating though. Advertisers too often believe in an input model, when they should be focussed on the output model. The input model, according to Jeremy Bullmore, works as follows is: I say that I am humble, which conveys humility. The output model is such that it doesn't matter what I say, it's what is received by the consumer that is key. If I go out and say that I am humble, then the consumer in the output model will take a very different reading from the situation.
That applies to the lack of apology; the brand sticks its head in the sand and assumes that people will think nothing, but actually if there is a void, a vacuum, consumers will fill it with possibly the worst possible interpretation available.
How can clients and agencies be persuaded to change?
RS: I think that we should have a more nuanced approach to risk than we may currently have. If I said that you should completely change your brand approach based on the Pratfall Effect, you would probably want a huge amount of evidence [to change].
However, if you were thinking of changing your banner or search copy, and I told you about the idea of social proof - a well-known bias that when a course of behaviour is seen to be popular it becomes more popular still - then the risk of testing that idea is minimal. You should only require a low burden of proof.
The fact that there's a well-proven theory of social proof should give you the confidence to run a small A/B test. That will then give you the proof as to whether the bias applies in your specific circumstances. Then you can consider using it in on a larger scale.
But, don't go and ask consumers whether a social proof message is powerful. Consumers will invariably deny it, they will claim that they act in a rational, logical manner. But, a whole body of evidence shows that this is not the case.
Robert Cialdini, a brilliant psychologist at Arizona State University, devised an ingenious experiment at a hotel in the US. He randomly rotated the messages that encouraged people to reuse their towels. One third of the rooms had a notice which read: "Please reuse your towel as it's good for the environment". That's the control and in that scenario 35% of people re-used their towel.
His test, using social proof, was "Please reuse your towel because most guests do so". Even though he jettisons all rational messaging, the reuse rate went up to 44%. The final third of rooms had a notice: "Please reuse your towel because most guests who have stayed in this room have done so". Reuse increased to 49%.
This is genuine behaviour change in a real-world experiment. What is fascinating here is that there is an amazing amount of nuance to it. The first point is that social proof is powerful, more effective than the control message. The second point is that scale does not matter; it's about making a relevant link with the audience.
The third point, the twist in the experiment, is that after the real-world test Cialdini recruits another group of participants and showed them the three messages. Cialdini asks them which one would motivate them to change their behaviour.
Overwhelmingly, people claimed that it was the environmental message that would affect them. That's very different to their actual behaviour - and it's different in a predictable way. People try to project an image of behaving in a socially responsible, rational manner. No-one wants to admit that they are swayed by the herd. Yet, the evidence suggests that we are.
What are the behavioural changes that you would like to see in advertising?
RS: A willingness to test ideas from social psychology. A greater reluctance to believe that consumers have fundamentally changed in the last few years. There is a worrying fixation in marketing with fads and fashions. The logic is the warped version of: "Technology has changed loads, therefore how can the things that once worked still work now?"
That's a big mistake, because human nature is remarkably static. There's a wonderful Bill Bernbach quote: "It took millions of years for man's instincts to develop. It will take millions more for them to even vary. It is fashionable to talk about changing man. A communicator must be concerned with unchanging man, with his obsessive drive to survive, to be admired, to succeed, to love, to take care of his own."
Bernbach said that 50 years ago, but it's ignored too much in advertising now.
I recently did a test with over 200 people in agencies, where I invented a fictitious study from the Harvard Business School. I told them about the Pratfall Effect. I said that Harvard Business School found that if you admit a weakness, you become more appealing. I then asked how credible that was.
The twist was that half of the group was told that it was from 1973, and the other half was told that it was from 2013. People were 10% more likely to find the claim credible if it was positioned as coming from 2013. So, there's still this emphasis, this prioritisation, towards modern, new, novel things rather than the proven bedrock of insights into human nature which we have in social psychology.
Richard Shotton is Deputy Head of Evidence at Manning Gottlieb OMD. The Choice Factory is published by Harriman House and is available to buy now.