#324 Using Behavioral Science to Hack Your Customers Minds with Richard Shotton, Founder at Astroten

Generative AI is transforming industries at an unprecedented pace. But as AI changes how you or your team work, one thing is clear. Your skills also need to evolve. At datacamp, we offer everything you or your team need to adapt and thrive with AI. Whether it's business users looking to get the Most out of ChatGPT and Copilot, or developers and data scientists looking to fine tune models. You can learn the entire AI skills spectrum on Datacamp. Power your AI transformation today. Start learning@datacamp.com.

So if you use language, people can visualize. It's very sticky. If you use language, they can't visualize, it's completely forgettable. Now you take that principle and then you look at how Apple advertise and you see them using it again and again. So the classic example would be ipod. Everyone else wanted to convey the benefit of storage for their MP3 players. So they talked about 100 gigabytes or 50 megabytes of memory. But the problem with that is it's completely abstract. You cannot picture a megabyte. You can't imagine 56. So people understood the message but they couldn't remember it. Whereas Apple, when they wanted to convey that exact same benefit of memory or storage, they use this brilliant phrase, a thousand songs in your pocket and you can picture a pocket, you can imagine a song. So because it's visualizable, it taps into the principle of concreteness. When you are trying to influence other people, when you're trying to persuade them to change their behavior. If you know about the bias, you work with human nature. If you don't know about the bias, you're trying to overturn human nature. And it is always more effective to work with human nature rather than against it. Behavioural science is basically just the study of what are those biases and how do you practically use them.

Welcome to Data Framed. This is Richie. A recurring theme on Data Framed is using data science for marketing. The goal is to understand your customers better so you can provide them with more value or sell them more products and services. Data alone will only help you so much. You'll see better results by combining data with psychology. That is, in order to go deep into your customers minds, you need to use behavioral science. I want to know about the benefits of behavioral science, what data practitioners need to know about it and how you use it responsibly. Our guest is one of the leading thinkers in behavioral science. Richard shorten is the founder of consultancy Astro 10 where he helps companies apply behavioral science to marketing. He's also the author of the Bestselling the Choice Factory and hacking the human mind. Let's learn how to hack our customers minds. Hi, Richard, welcome to the show.

Oh, thank you very much for having me.

Yeah, I'm looking forward to this. So I guess to begin with, I want to know what's the most impressive example you've seen of a company hacking their customers minds?

Oh. So lots and lots of companies use behavioral science once really brilliantly. But I think for someone that uses behavioral science again and again and again, I would turn towards Apple. So there is a principle known as concreteness. So the original studies were done back in the 1970s by Ian Berg at the University of Western Ontario, and he did a super simple study. He reads out this list of 22 word phrases, and half of the phrases are what he calls abstract phrases. So it's things like basic truth or subtle mistake. And half of the phrases are what he calls concrete phrases. So it's things like square door or flaming forest. So a concrete phrase is something you can visualize. An abstract phrase is an idea or a concept you can't visualize. And what Beg found is when he later on asked people how much they could remember. People remembered, on average, 9% of the abstractions, but 36% of the concrete words. So you get this massive, massive fourfold difference between the memorability of a concrete phrase and an abstract phrase. And Beg's explanation is super simple. He just says, vision's the most powerful of our senses. So if you use language, people can visualize. It's very sticky. If you use language they can't visualize, it's completely forgettable. Now, you take that principle and then you look at how Apple advertise and you see them using it again and again. So the classic example would be ipod. Everyone else wanted to convey the benefit of storage for their MP3 players, so they talked about 100 gigabytes or 50 megabytes of memory. But the problem with that is it's completely abstract. You cannot picture a megabyte. You can't imagine 56. So people understood the message, but they couldn't remember it. Whereas Apple, when they wanted to convey that exact same benefit of memory or storage, they use this brilliant phrase, a thousand songs in your pocket. And you can picture a pocket, you can imagine a song. So because it's visualizable, it taps into the principle of concreteness. And that phrase about the eye was much, much more memorable than anything Philips or Sony used. So I love Apple because you see them. Not just that principle, there are others as well, but you, you see them using these ideas again and again and again. It's that consistency that differentiates Apple from other people.

That's one guess. Certainly Apple's got amazing marketing campaigns, are very memorable. And I do like the idea. I mean, already, I mean, I can remember you said flaming forest. I can't remember any of the abstract phrases you used at all. So already that's working.

Absolutely. And it's interesting, you know, once we are aware of these ideas, they're not complex. Once you know visualizable language is powerful, that's quite an easy thing to repeat. But if you just knew a thousand songs in your pocket is a great strap line, how the hell do you as a communicator learn from that? But if you know it's the visualization part that's important, then it's very, very easy to mimic.

Yeah, I feel like there's a lot of lessons there for like a data scientist, data analysts, when you're trying to communicate the results of your analysis, often it can get very abstract, very technical, and just having something visual there in when you come to report things, that's probably going to help have an impact.

Well, absolutely. So if people are thinking about applying this to data analysis, I think there's a few things you could take from that Apple example. Firstly, use visualizable words. But the second one is keep it simple. And that's actually harder than it sounds because if you are an expert in an area and you're presenting to lay people what's simple is much simpler than you think it is. And I think sometimes experts in their area don't want to think they're dumbing things down. They think, well, if I explain something very simply, it reflects poorly on me as a communicator. But the evidence is the opposite. There's an amazing study. It's the best ever academic paper title. It's from Daniel Oppenheimer. I think it's 2010, maybe 2009. He's at Princeton and he does this paper and he says it's the name is consequences of erudite vernacular utilized irrespective of necessity. Colon the problems with using long words needlessly. And what he does in this study is he gets abstracts from academic papers. So these are the short precis at the beginning of the journal article that describe what's going to come ahead. And he takes some of these abstracts that have lots of unnecessary jargon. He gives them to readers and then he says to the reader, how intelligent do you think the author is? He gets that data he then gets another group, completely fresh group of people, shows them basically the same abstract, but he replaces the complex words with simple alternatives. And when that group rate intelligence, the author, they come back with scores that are 13% higher. Oppenheimer's argument is most of us are laboring under this misassumption, that if we talk in a complex way, the audience will admire us and think we're wise and intelligent. But actually if you confuse your audience, that irritation they feel, they're not going to blame themselves, they'll blame you. So yes, learn from Apple and it's use of visualizable language. It moves from the concrete world, the abstract world, to the concrete world, but also the simplicity. A thousand songs in your pocket is so much more simple than most technology describes itself. So simplicity and visualizability, I think those are both things that data analysts could apply.

Absolutely. Yeah, certainly simplicity is a really important thing actually. So recently I took all the transcripts from this podcast from Data Frame and threw them into like a word analyzer to see like how sort of complex the language was. I think normally I speak at. I think it was like a 10 or 11 year old reading age level or something in most of this. And then we had, we're trying to work out when guests talk technically, is that better for engagement or worse? No strong pattern. Apparently our audience is very smart, they can understand all the guests. But yeah, it's interesting that the simpler you go in general, it works well.

I think there's an element of accessibility. Yeah, an awful lot of people haven't got great reading skills. That's one reason to speak simply. But even if you're talking to a PhD audience or you're sending an email out with your findings, don't fool yourself into thinking that they have dedicated half an hour of thinking time for your proposal or your argument. They're probably got other things going on. You know, crap's going on at home, the phone's ringing. So they might have PhD level abilities, but they are stressed and time poor. And if you make something hard to understand, often people will not engage and leave it for another time, then never get around to doing it or they will get frustrated that you've added an extra complex task to the situation. So this is not just a tactic for dealing with people who haven't got great reading abilities. It's a tactic we should use with everyone.

Nice. Okay, so I'd actually like to take a step back because I think behavioral science is quite new to a lot of our audience. So talk us through the Basics, like, what does behavioral science involve? What are the goals?

Yeah, so if people have never heard of behavioral science, it's nothing complicated. It's essentially the study of what actually influences people rather than what people claim influences them. And crucially, those two things very, very different. And if there's a broad theme, it's summed up very well by a Princeton psychologist called Susan Fisk. So she came up with this lovely phrase, she says, when people are making decisions, one of the key factors is that they are cognitive misers. So she came up this brilliant phrase of cognitive misers. But one of her colleagues put it a bit more wittily. So Daniel Carneman said, think he is to humans, Swimmy is to cats. You know, we can do it, but we prefer not to. Neither of these academics is trying to be rude about people. You know, both of them were at Princeton. Both of them deal with very, very intelligent people. So they are completely aware that humans are capable of amazing flights of logic. You know, we can build iPhones, send people to the moon, build an electric car. But they're also aware that from an evolutionary perspective, energy is a very, very scarce resource and thinking is energy intensive and it's effortful. So what we have evolved to do as creatures is ration deep, considered thought. And most of the time, in most situations, we don't respond to them in a fully considered way. We often make snap, intuitive decisions. So we're faced with a situation, maybe we're faced with a proposal. We feel positively to it or cold towards it. You know, we feel positive or negative. Once we've had that initial reaction, then we use logic and rationality to justify to ourselves why we feel that way. But often it's that emotional quick, snap decision that comes first. Now, the way that we make those snap decisions is to use what psychologists slightly pompously call heuristics, what we would just call in layman's terms, rules of thumb. And everyone has hundreds of rules of thumb in their mind that allow them to come to these quick, snap, intuitive decisions. And what's interesting is that those rules of thumb are prone to biases. And essentially the main argument is if you know what these biases are when you are trying to influence other people, when you're trying to persuade them to change their behavior. If you know about the bias, you work with human nature. If you don't know about the bias, you're trying to overturn human nature. And it is always more effective to work with human nature rather than against it. Behavioral science is basically just the study of what are those biases and how do you practically use them?

That's fascinating. And certainly I work in education, so I'm really keen on the idea of people thinking. But also there are definitely days where, yeah, I don't really want to think. I also like not thinking.

Yeah. And also it's that kind of order effect of first of all, I have this emotional reaction or intuitive reaction to a situation and then often the thinking kicks in. And that's why even the most intelligent of students are prone to these biases. The more intelligent they are, the more able they are to generate almost endless reasons about why their initial reaction is right. So intelligence is often isn't an antidote to the biases.

Interesting. Okay, can you tell me more about the biases then? Like what are the different biases you need to be aware of?

So there are literally thousands of experiments. I mean, the first study in kind of modern psychology or behavioral sciences back in 1880. So you've got thousands of these studies, but there are some that occur more than others, There are some that are more powerful than others. So one that lots of people might be familiar with but is a very influential bias would be something like social proof. So this is the argument that humans are a social animal, we're a herd species, that if we believe a course of behavior is commonplace or popular, that behavior will become more appealing and people are more likely to adopt it. So often, if you want to change someone's behaviour, if you want to get them to adopt a proposal that you're putting in front of them, rather than just default to giving people a long list of rational reasons about why to obey, what you want to do is create the impression that lots of others have already adopted this behavior. Now, interestingly, this is a study because it's such a well known one, there are hundreds of experiments into it. And there are studies that show, yes, this affects people when they're buying cans of Coke or bags of crisps in the supermarket. But there are also studies that show professionals are deeply influenced at work by social proof. Now there's a study from 2018 by the Behavioral Insights team and they work with the Australian government and they wanted to get doctors to give out fewer antibiotics. So send out letters to doctors. Over 6,000 letters are a nice big sample. And sometimes the letters try and just rationally explain to doctors, educate them about the downsides of over prescription. And when those letters go out, there is a 3.2% reduction prescription rates. So some reduction, but pretty small. Other doctors get exactly the Same letter, all the same reasons, but at the top of the letter there is a social proof argument. Now the exact line would vary according to the situation of the doctor, their individual prescribing habits. But let's say it's a very over prescribing doctor in Sydney. It might say something along lines of, you're giving out more antibiotics than 90% of other doctors in Sydney. And when that message went out, there was a 12.3% reduction in prescription rates. Now that's interesting because doctors, you know, they define themselves as being logical, rational, evidence based thinkers. But even they, even this group were very influenced by what they thought their peers were doing. So yeah, if you're at work and you've put a argument together about a particular approach, don't think that logic alone is going to be your most successful tactic. Yes, give people rational reasons to adopt your suggestion, but also try and create this impression that lots of others are behaving the way that you're suggesting.

That's absolutely fascinating. We've had a few episodes on the show around doing transformation programs, like trying to do data transformation, get more people using data and things like that. And one of the consistent pieces of advice is show off what you're like, create champions. So if someone does something cool, you make sure everyone else knows about it. And so you're getting that social proof.

Yes. So if internal behavior change is of interest to people, I would say that one of the biggest mistakes I see again and again with internal change programs is emphasizing the scale of wrongdoing. So let's say, you know, take a mundane example, you want to get people to fill in their timesheets. What many companies do, and I remember this from experience, is I used to work at an agency, a marketing agency, and every couple of months we would get emails saying, you know, none of you idiots are filling in your timesheets and it's causing us all sorts of financial problems. That is completely misusing social proof. That argument of lots of people are misbehaving uses social proof in the wrong direction. You're emphasizing the scale of wrongdoing. That means you remove a sense of transgression and the message likely backfires because most people will think to themselves, well, everyone else isn't adopting this policy, they must have a good reason. And I'd be a bit of a mug to do it myself. So with internal behaviour change, the first thing to do is think, how can I honestly emphasize the scale of positive behavior, the uptake of the behavior I want? That might be the absolute Scale. Or it could just be the fact that more and more people are doing it. Emphasize that, not the large volume of people who aren't behaving the way that you want. Wonderful.

And I'm sure if you didn't know about that kind of, that bias that idea then is a very easy mistake to make. Like you think, okay, if I shame people publicly, they'll stop doing that. But it doesn't work as a tactic.

Yeah, you see it all the time. So Cialdini has run some studies around stealing, which essentially if you put signs up telling people how regularly something's been stolen, it gets stolen more often. He said this is the government's big mistake because it feels such a logical thing to do. You know, he said too many government campaigns will go out and say loads of youths carry knives, loads of students turn up late, loads of people don't turn up the doc's appointment. These are all very, very well meaning messages, but the problem is they use social proof in the wrong direction. They emphasize the scale of the thing you don't want, not the scale of thing you do want.

Okay, so it seems like a very simple messaging switch can, can help there. Now, you mentioned before that there's kind of two sorts of thinking. There's this quick, heuristic thinking, then there's the sort of deeper sort of more considered thinking. Are there any cases where you want people to use one or the other and how do you make them switch?

So most decisions will be made in this kind of quick, intuitive way. And my argument would be, rather than thinking, how do humans tend to make decisions? How can we overturn that? I would say work with human nature. If you know about social proof, if you know that we copy others, use that to your advantage. Now that doesn't have to mean lying. You know, there is an ethical use of this, but I used to work with the blood service. You know, they were trying to get people to donate blood and they used to come out and say only 4% of people donate blood. That is a misuse of social. You make it feel like a weird, aberrant behavior. And people think, well, if no one else is bothering, why should I? What you can completely honestly say is two and a half million people give blood every year. Now, 4% and 2 and a half million is the same number. In a, you know, roughly, math might not be perfect, but in a, in a country like Britain, now, if you say two and a half million, you are emphasizing how lots of people are doing this behavior. If you emphasize 4%, you're emphasizing how lots of people aren't doing the behavior. Both of them are true. There's no ethical head start for either of them. But using the one that works with human nature is the one that's going to be more effective.

Okay, I love that. It's just same statistic, prevented, presented in different ways.

So, yeah, and the underlying argument is deal with the world as it is, not the world as you want it to be. Maybe the world would be better if people were thoughtful, logical, rational decision makers, but we don't work in that world. So tailor your communications and your attempts to change behavior to reflect what actually influences people.

Okay, nice. And I think in the sort of data AI world, technology world, a lot of the things you have to do are persuade people to adopt new ideas or adopt new technologies. How do you go about getting people to adopt new things, for example?

There are. It's a tough task. So we are creatures of habit. And there's some really nice work that suggests that there are particular moments when people are more open to changing their behavior. So some of the work comes from Catherine Miltman at Wharton, and she came up with this idea called the Fresh Start Effect. So her argument is a big driver of behavior is the desire to be consistent with our past selves. So we look down on people who are inconsistent. Maybe in America they might call them flip floppers. In Britain we might call them hypocrites or inauthentic. So most of the time people aren't considering alternative ways of behaving, they're just doing the same thing again and again and again. But behavioural scientists don't just list the problem of changing behavior and how hard it is. They also come up with some solutions. And Milkman's argument is when we enter new time periods, our link with our past self is weakened and we become more open to change. So she does some lovely data analysis. She looks at volumes of Google search terms over time for things like quit smoking, start dieting, stop drinking. She looks at gym registration and attendance data. She looks at visitors to a website called Stick, where people make public pledges to change their behavior. And for each of these data sets, she sees the same pattern. So there is this pronounced spike in openness to change as people enter new time periods. So that could be the beginning of the year, beginning of the month, after a birthday, after a public holiday. Any moment in which this is almost this kind of fracture in the progression of time, people are more open to change. So if you want to get them to adopt a new behavior, if you send out your message at a random time, it's likely to fail. But if you, let's say, wait until the first week of September when loads of people come back from the holidays, it's more likely to work. If you wait till the 1st of January, it's more likely to work. If you wait till after an office move, it's more likely to work. If you put your efforts into people who just changed jobs, it's more likely to work. Now, I know that sounds like such a tiny little change, which loads of people are skeptical about the genuine influence, but here's a brilliant campaign by the West Midlands police. So the Behavioral Insights team, which is the government's nudge unit, they work the West Midlands Police. 2017 might be 2018, they launched a new service. It was essentially this confidential hotline. They let known criminals know about the hotline and they offered them the choice. Now, either we're going to hunt you down, they probably weren't that drastic about it, but either, you know, we know who you are or we will help you find a new job or help you move to a new area and start a life afresh. Now, they sent out 2077 messages to the criminals and sometimes they just went out at random moments. Other times they sent the message just after people birthdays and the people who received the message just after their birthday at a fresh start moment, I think it was over 56% more likely to phone the hotline than those who just got at a random moment. So even a group like hard and criminals people, you know, this is where the most difficult audience to change their behavior. You can imagine, even they are affected by a tiny thing like the fresh start effect. So that might be one consideration I would have about introducing new schemes. Don't just think about making the products appealing, think about the moment in which you, which you launch it.

That's really fascinating, I think. I mean, at the moment it's September, so we've got a load of like back to school advertisements for everything. In January you always get like, okay, go to the gym, change your life. But I do like the idea of like find if you know when your customer's birthday is, that's. Or your prospect's birthdays, that's going to be a really good time to message them.

Oh, absolutely. If you've got a lot of data, then it opens up even bigger moments. So, so when I read about Milkman's study, I thought, okay, well if a small thing like a new month opens up people to change, what about more Meaningful life changes. And what I found was there was an even bigger effect with what I called life events. So if someone had moved house, if they got divorced, if they'd start a new job, all sorts of unrelated behaviors were much more likely to churn. So people got divorced, wanted to try new lagers, people move house, might try new TV programs. It seems like so much of our habitual behavior is driven by the context. If you have one of these big changes, lots and lots of other seemingly unrelated behaviors in life become destabilized and you have this small window of opportunity to more effectively influence people. So yeah, if you've got that level of data, think about can I identify people who've just moved jobs? Can I identify people who've just started university or just moved house? And then they are much, much more malleable.

That's interesting. I suppose, like maybe the thing that just popped into my mind there is like there's a slightly sketchy way of doing that. Just you go around scraping social media to see what's going on with people lives. Are there more ethical ways of doing that?

Let's. So LinkedIn for example would know whether someone has just changed jobs. Mobile data providers can tell you if someone's moved house because what they'll infer is, let's say someone used to be in Basildon between 6pm and 8am and now they're in Thurrock constantly. Well, they've just, they would assume that's moved house. So there's so many data providers out there that you can get this kind of information. Or if you know, if you're something like Giant supermarket and suddenly people are used to be shopping at the Basildon Superstore and now they're at the Ipswich Superstore. Well again you could infer that they have, they have moved house. Your own data set could be quite useful as well.

Okay, so it just seemed like all this information on people's life events, it is available to find if you want to. Another thing I think behavioral science ought to be useful for is getting people's attention. I mean there's so much going on, like so many different distractions. It can quite often be hard to just make sure people are aware of your products or services. What are you, what are you? What can behavioral science help us with here?

So you're absolutely right, there is. Well it might be one of the oldest studies around really. There's a 1933 study from Hedvig von Restorf, this brilliantly named German psychologist and she did a super simple study. It bastardized it slightly just to make it a bit clearer. But she gives people these list of information and let's say it's got 10 items on and nine of them are all animals. Cat, dog, weasel. But one of the items is a bit of furniture, like a chair. She gives people lists of information. They might have a couple of minutes to look at the list. She then takes the list away and then she asks people what they can remember. And what she finds is you see a very clear pattern and we are hardwired to notice what's distinctive. If you give people long lists of animals with one or two bits of furniture in, people remember the furniture or vice versa. If you give your long lists of furniture with one or two animals sprinkled in, they'll remember the animals. So that became known as the Von Restor effect or the isolation effect. And it's essentially this idea that our attention is drawn to what's distinctive. So if you're thinking about getting some detention, what you want to think about is what are the norms of behavior in my category or what are the normal ways people will communicate who are surrounding my message. And then once you've, you can think of that it's worth, you know, putting those norms of behavior into a list and then being quite candid about categorizing them. Some of those things you think you might believe are crucial, but others you might think are just there for tradition's sake. If you break those more traditional areas, then you are onto something very powerful. You tap into one of the most well proven and longest standing principles of von Restorff effect. So simple one to work for me on occasion is something called the pratfall effect. And it's this idea that if you would admit a flaw, you become more appealing. Now that is a brilliant way of applying the Von Restel effects in marketing because virtually every advertiser, 99.9% of advertisers will brag. So if you come out with an admission of a weakness in your product, then you are going to stand out and you will get noticed. And if you're noticed, that is 9/10 of the battle. Most information is completely ignored. And if people think this sounds like a completely ludicrous approach, or just think about some of the products in your home, think about Guinness. Good things come to those who wait. That is admitting the flaw of slowness. Maybe a bit older one, but stellar. Reassuringly expensive. That's admitting the floor of being expensive. Marmite. Good. Sorry. Love it or hate it, me half the Population hate your product. There are all sorts of amazing campaigns that have applied that principle and I think not all of the success, but an awful lot of the success is it breaks what people assume is the cardinal of advertising, which is to, is to brag.

Interesting. I do like the idea of doing things that are a bit different to stand out and also admitting your flaws. I mean it's a tricky thing to do because I guess it could backfire spectacularly if you do it badly.

Oh. So absolutely this is not a suggestion you should just randomly admit problems. But what you really need to do is be very honest yourself and think what is your core strength? And maybe you are a service provider and you are very, very high quality. Well if you're high quality you're probably going to be quite pricey and often people will be a bit embarrassed about mentioning that they're more expensive than their competitors. But actually it's very hard to prove that you're high quality. One of the best ways to do it is actually draw attention to your price and if you admit that you're a bit more expensive than other people. Generally people assume, well if they're pricey they're probably high quality. Generally people believe in, in trade offs in life. So it's about admitting the right flaw, what's your core strength? And then think is there a mirror weakness that if I admit it will emphasize my, my core strength. And that I think is something that if you, yeah, don't do it randomly, don't do it willy nilly. But it's something that can be very, very powerful.

Absolutely. Yeah, certainly it seems like if you do it right then it can be an incredibly powerful thing. I guess we're new data frame slogan. Sometimes we're slightly intellectual. It seems that you mentioned the pratfall effect is kind of like it sounds like a comedy thing. I get pratfall comes from comedy. Right, so and you use comedy in order to get people's attention in this.

Way that's really interesting. But often some of the best brands that have used a prattful effect, it's quite self deprecating and it could often lead to amusing situations. So I think you're, I think you're absolutely spot on with humor. It actually links back to your original question about noticeability. So there's a psychologist at the University of Amsterdam called Fred Bronner and he did this very simple study. He gets huge sample, 1287 people, gets them to flick through a newspaper and then he asks them what mood they're in, good or bad. Happy or sad, relaxed or stressed. And once they've answered those questions, he then asked them to try and recall as many ads as they can. And what he finds is when he cuts the recall data by mood, people are much more likely, I think it's 50, 51% more likely to remember the ads if they are in that positive mood, happy or relaxed versus unhappy or stressed. So there's this argument that if you either reach people in a good mood, maybe you reach for the weekend rather than when they're on a Monday morning, or you use humor to put them in a good mood, they are much more open to what you're saying. And then on top of that work that I've done suggests not only are people more like to notice your advertising, if they're in a good mood, they are more likely to believe what you say and they are more likely to be open to spending. They're less price sensitive when they're in a good mood. And none of this stuff is speculation. Now all these insights are based on really, really simple experiments. So for the price sensitivity point, all I did was with a couple of other people, Ben Davies and Ben Sachs and Chris Davis. I think it was the three of us that did this one. We got 821 people. We showed them a series of promotional ads. So it might be a case of beer for 20 quid or a meal deal at a supermarket. We got people to say whether it was a good deal or a bad deal on the five point scale and then we asked them about their mood. And what we found is that the proportion of people who thought the deals were good value was 60%. If people were in a bad mood, it went up to 76% if they're in a good mood. So exactly the same offer. People are 26% more likely to think it's a great deal if they're in a positive mood. So there's lots of evidence this, it's not speculation. And I think it's not just a quirk of a one off study either because I think there's a very sensible rationale behind it, which is if you're in a bad mood, you are skeptical, you focused on the opportunity cost of spending money. If you're in a good mood, you focus on the upsides, all the benefits you're going to get for the product and minimize slightly the cost to yourself. So yeah, there is noticeability, believability and even price sensitivity relations to upside of putting people in a good mood.

Okay. But yeah, I can certainly see how if you try and put your customers in a good mood before you ask them for money. That's gonna work really well for, I guess, for sales teams in particular. You know, you have a conversation, get your customer happy, and then that's the point where you start talking about pricing.

Yeah, and there's an overlap here. So it's firstly, the evidence about if someone's in a good mood, they're more likely to be risk taking, to believe things, to spend money. But there is another part of this which is if you repeatedly put Peter in a good mood, they're going to like you. And there is an argument called the halo effect. So the halo effect is essentially the argument that when we evaluate a person or a product, what we probably should do is weigh up each of their characteristics individually. So I should make a decision about someone's good looks completely independent from their intelligence. But that would be time consuming. And we are machines that are designed to make quick, fast, snap decisions. So what tends to happen is people latch onto one standout characteristic about a person and then they use that as a guide to all the other characteristics. There's a. There's a studies going back to the 1930s, but there's a lovely study from 1977 by, it was Timothy Wilson and Richard Nisbet, and they get a colleague of theirs at university, a Belgian colleague who's working in America, to record a lecture. And this guy has quite a strong accent and he records a lecture about a particular topic and he does it twice. First time he comes across as very friendly and warm. Second time, bit rude and cold. The academics take that video, they play one of the videos to people and then they ask them to rate the academic. Now, just as you'd imagine, completely logically, the people that see the friendly lecturer, they rate him as warmer and more likable than the people who see the cold and rude lecturer. The interesting bit, though, is on completely unrelated metrics, you see the same impact. So when people rate how good looking the lecturer is, his mannerisms, his accent, again, it's that when he comes across as warm and likable, all those factors are scored higher than when he came across as cold and miserable. So we often take one factor about a person and then overextend it across all their other attributes. So if as a salesperson, you come across as likable, what will also happen is your intelligence will improve, your trustworthiness, improve your looks will improve. You push up one of these metrics and others will follow to a degree. So I think there's A double benefit of putting people into good mood, which is on that one to one basis is you're then going to benefit from this halo effect.

That's fascinating because you wouldn't necessarily think that likableness and good lookingness or intelligence ought to be related. But I guess people blur these things if they like you.

Well, absolutely. I mean, you are right. Logically they are not related. Absolutely they're not related. But there'll be a mantra, I suppose, for behavioral scientists, which is, we're not interested in what people ought to do, we're interested in what they actually do. And we see these metrics kind of moving in, in lockstep. And I shouldn't exaggerate. It's not like, you know, if you're seen as very good looking, you'll be seen as very intelligent, but your intelligence will move up compared to someone who, who didn't have that, that very strong standout characteristic.

Okay, fascinating. So I want to go back to something you mentioned earlier. So you were talking about how with the pratfall effect, it's about being honest with your customers. But also at the start you mentioned that Apple is like one of the kind of the masters of behavioral science and they're a very secretive organization. So I'm trying to work out, is it a good idea to be more transparent with your customers or more secretive? When is one or the other a good thing?

I think one of the interesting things about human nature is it isn't that there is one answer. And imagine, you know, you and I were given a task and we have to cross a river. You know, human nature's a bit like that. You could swim across the river, you could build a bridge, you could hire a ferry. There are lots of ways to cross that river. And just because there is the existence of a ferry, it doesn't negate the fact that swimming will also get you across. So there's this, I think that's the same with persuasion. There are lots of these different techniques and it's about matching the right one to the right circumstances. But it probably sets behavioral science, you know, and the practical use of it as much, I think as an artist, as a science.

Okay, so choose the right tool for the right situation. Actually, do you have any concrete examples of like one or the other?

Yeah. So often you'll be wanting to think about where, where your strengths lie. So Dyson, you talked about transparency. I think Dyson are transparent in a very, very interesting way. So, you know, one of the truthful stories about Dyson is that it took the inventor James Dyson years and years to go from the initial idea of the banks vacuum to the final product. And that is something that Dyson emphasize all the time. You look at their advertising, look at their website, you listen to their pr. In fact, if you pick up James Dyson's autobiography, the very first sentence is, I went through 5,127 prototypes to get to the bagler. Thank you. Now that emphasis of being open and transparent about the efforts you've gone to, that is a phenomenal tactic because exactly the same product will be rated as higher quality and preferable if people can see the effort that's gone into it. There's a famous study by Andrea Morales at the University of Southern California, maybe 2005ish. And what she does is she goes out and recruits a group of people in market for a house and every person is given 10 apartments that meet their requirements. So everyone gets the same list. But some people are told the estate agent took one hour to generate the list and they used a computer. That's the low effort group. Other people are told the estate agent took nine hours to generate the list and that list was made manually. Now when Morales asked people to score the estate agent, how good was their service? 0 to 100. She sees this clear pattern. Low effort group, they rate the estate agent at 50 out of 100. High effort group rate the estate agent at 68 out of 100. Exactly the same service, but you get this 36% difference in ratings. Now Morales argument is it is a complicated question to work out whether someone provided a good service. And when we face complex questions, we we often substitute them for simpler questions that give us an almost as good answer. And the simple question here is how much effort do I think the estate agent went to? So her argument is you as a communicator or as a business, you can't assume people will know what efforts you've been to. You have to be much more transparent about them. Now that's not about a low ethics approach of inventing stories of hard work. Most businesses are already doing that, they just keep those stories themselves. The argument here is you need to let people know about them and if you do, exactly the same, products will be rated higher. So sorry, that was a slightly long winded way of saying why for Dyson is it useful to really double down on that story? Because if you go through their history, there is a truthful story of EFF and therefore being transparent about it is both honest to who you are and working with you. Know this, this hardwired bias. If you're a company that doesn't have that effort story, well, then you need to look at other biases to tap into. So I think you're absolutely right. It's often matching it to the. The genuine insights about the. The organization. Okay.

No, there's a really interesting story about Dyson, like more than 5,000 failures. I think that requires a ridiculous amount of persistence. Like, I think a lot of people would have given up long before then.

And crucially, he normally says virtually ever, all the time. He doesn't say more than 5,000. He always says 5,127. And there are studies that suggest that precision has a benefit. Now, if you say more than 5,000, it fails. A bit ambiguous, you know, have they made it up? Is it plucked out the air? 5,127. Giving it that precision tends to be seen as more accurate and more credible. Okay, interesting. All right.

5127. Get that right, actually. So you talk about demonstrating effort being a good thing for customers. So at the moment, there's a big push with making use of AI to automate things. And so this is like a big trend in let's see how much we can automate and do less effort for the same task. So it feels like there's some big implications here.

AI is a more recent technological invention. So it's not like there are hundreds of psychological studies into its impact, but there are some. There's a 2023 study from Kobe Millet VR University in Amsterdam, I think, and he shows people products. So some people are shown this poster, and it's a drawing of a skull, and they're told that it was created by an AI robot. Others are shown the same poster, but they are told it was hand drawn. Millet then asks both groups to rate that poster on its creativity, its artistic merit, even on their purchase intent. And for every one of the metrics, he sees, you know, stark pattern. If people think it is AI generated, they will rate it lower than if they think it's hand drawn. I think with purchase intent, it is a 61% swing. So it's not a small change. And Millet's argument for this goes back to the illusion of effort. He says, look, most people's experience with ChatGPT is you put in a prompt and it spits out a blog post or a detailed answer in a second. They think of it as being low effort. So if you emphasize that you've used AI, you are telling people that little effort's gone into it, exactly the same product will be rated worse. So you have a bit of a dilemma here. Or we have a few different options to put it more positively. In my world of marketing, what you don't want to do is what Coke did at Christmas and emphasize your ads being made by AI. Why would you tell people that? You just got to reduce the impact. So that's one option. Don't positively tell people necessarily. But the other area is if you have to tell people that something's AI generated, maybe your client's pushing you to introduce AI and speed up the service. What you've got to be aware of is that unless you introduce counterbalancing measures, they will rate that product as worse. So what you might want to start thinking about is can we switch attention from speed of delivery to the length of time it took to set up the processes and get the programming right? And I think that's a subtle shift that you can do.

Absolutely, yeah. A lot of these things like, okay, you're automating stuff, but it's actually, it's a subtle art in getting all the prompts right and all the workflows right. So it is a bit of effort to get it set up in the first place.

Yeah. And don't shy away from telling people that. Don't assume that they will know. Make sure you're transparent about wasn't to do with AI But I think the implications are spot on for AI. There's a really famous story about the design of a logo back in the 1990s, might be early 2000s. So two big banks merged. They became Citibank, I think it was Travelers and Citigroup maybe. And they hired Paula Share to design the logo. So she's world famous designer, top of a field in the briefing meeting literally as she is being told what she needs to do, she is sketching out a logo. And you can see this online. She sketches out on a napkin. You can still, you can find the images of this online. A quick Google. And it is essentially to all intents and purposes maybe 95% the same as the finished logo. She delivers a few months later. And legend has it that Citibank are like, well, how can you in all good Faith ask for $1.5 million? That was the fee, $1.5 million for something that took you 30 seconds. And her legendary answer was, well, it may have taken me 30 seconds, but it took me 34 years to learn how to draw it in 30 seconds. And that's the same shift I'm Talking, talking about 30 seconds is the speed of delivery. 34 years is the amount of effort it took to get into that position. And you could think about that as an analogy for setting up these AI systems now.

Nice. That's a good comeback. And when someone's trying to cheat you out of your $1.5 billion, you need, you really do need a good comeback. All right, so suppose your boss is like, okay, I really believe in behavioral science. We're going to go all in on this. Like how do you get started?

I would pick something that has a, is a costless intervention. Because one of the things I love about behavioral science is these ideas don't necessarily need a million pound ad campaign to make them come to life. They can be tiny, tiny, tiny tweaks. So I would think about what is it we can change that won't cost us anything but will have a nice positive effect. Start small and it's easier to I think persuade people to adopt it. So one of the studies that I come across quite recently that I love is by Peterson I think at the University of Texas. And it's basically about social proof but it's really clever application. He gets a large group of people to go through an E commerce journey and sometimes an item isn't there and it's labeled as sold out. Other times it's labeled as unavailable or out of stock. After people have, you know, got done their, you know, purchasing journey, all the other tasks he's given them, he asked them how irritated are they with the website? And he sees this really clear variation. If the unavailable product was labeled out of stock or unavailable, people are very irritated if it's labeled sold out, but only a little bit irritated. And his argument is if you say it's out of stock or unavailable, you are drawing attention to your owners logistical ineptitude. But if you say something sold out, what you're subtly doing is, is using social proof. You're emphasizing this is a super popular product. You know, it's high demand that has meant it's not there and that minimizes that irritation level. Now that is a lovely study to put into practice if you own an E commerce site because, because you have to label your products that aren't there in some respect. So why not label in the way that has the biggest impact, has the biggest positive impact. And that's what behavioral science will tell you. It's this giant repository, hundreds upon hundreds upon hundreds of studies and just take those simple studies and apply them to your challenges.

That's amazing because it's such a simple change going from oh, the label says after stock to it's sold out. And that's just going to improve customer satisfaction. I love that. Low effort, big impact. So I guess if you want to do more of these things, you're going to have to run some experiments. What some experiments might you want to run yourself?

Okay, so the biggest advice there, the single biggest thing, is do not ask people directly, you know, if you have found any of these studies we're talking about. Interesting. If you go and say to someone, look, would you be influenced if you knew lots of other people would be behaving in a particular way? Do not expect them to answer honestly. They will say, no, no, no, I make my own decisions. I think everything through very clearly and then I make a very considered decision. But a theme of behavioral science is what people say and what they do are completely different things. So there's a brilliant psychologist at the University of Virginia called Timothy Wilson, and he wrote this book which sums up his ideas called Strangers to Ourselves. And the whole book is a series of studies showing people do not have full introspective insight into their own motivations. So if you stop someone and say, what would encourage you to exercise more or to buy this lager? They'll give you loads of reasons, but they will be a very small subset of the genuine reasons. They will give you reasons that reflect well on themselves, that make them look like sensible, logical decision makers. They will, you won't admit these tiny little tweaks have any effect on them at all. So the first thing to do is put far less emphasis on surveying people, getting focus groups with them, and instead far more emphasis on real world live experiments. And the basis of virtually every psychology experiment is you set up naturalistic as possible situation and you do it in two variants. Everything apart from one metric is kept the same between the two situations. And then any difference in responses you attribute to that single change variable. So you know, with the Peterson study, you get a thousand people to go through an e commerce journey, you make sure it's exactly the same, you know, it's the same type of people, the same time of day, the same offers. The only thing that changes is the wording of the label. And then you can feel reasonably confident that it was that wording change that affected behavior. So avoid listening to what people tell you influences them. Run simple test and control experiments and then ideally, don't take one experiment as gospel. It could be a statistical fluke, really. What you want to see is either other people have shown this insight in lots of different situations or you should be looking at repeating it more than once. Okay.

Yeah, so that seems pretty good. Don't trust what people say. And I like it.

That's the biggest one. Yeah.

And it seems like these basically these like a B tests that you typically want to run.

Yeah, yeah, they're very, very. So if someone works in digital marketing, you know that. That. That would be the language they would use. The only thing that I would say slightly different is often a B tests are about certainly my word of marketing. They're around the very final bit of creative. So you may be changing a color or some images you used. What I would say is take that same theory of research and same methodology, but do it a bit earlier in the process. What you really want to be doing is finding out which of the principles, like social proof or the Prattful effect, which work for your brand and your context and your situation. Find out what biases are powerful. And once you know that, well, that will be useful forever. Whereas if you're only testing that final. The kind of aesthetics, as it were, in a digital marketing situation, well, that information is really valuable for that campaign or a month or two. So apply what you do with final creative all the way back to understanding people's motivations.

Okay, wonderful. Yeah. So start early rather than realizing that your entire marketing campaign is using the wrong language. Okay. So it seems like when you're trying to hack your customer's mind to scope for some sort of ethical disasters here, are there anything you should avoid doing? Any red flags to watch out for?

That's a brilliant question because I think you're absolutely right on that. Behavioral science and knowing about these principles isn't a kind of moral get out of jail free card. You know, it is just a tool. Behavioral science is a neutral tool. And unfortunately, you can use this to sell cigarettes, or you can use it to get people to do immoral things, or you could use it to get people to donate to charity and to recycle more. You know, it's just a. It's just a tool. I always think a helpful analogy, something like rhetoric. Rhetoric might be the, you know, the skill of putting a persuasive speech together. Now, if we think about learning some of those rhetorical skills, we would recognize very, very quickly that those skills can be used by dictators to get neighbors to hate other neighbors, or it can be used by Martin Luther King figures to get people to love each other. It's the same skills often, but they're put to very different uses. So I think what people need to think about is, is the intervention I'm doing in the long term interests of my colleagues or customers. Or think about it as there's this lovely idea that Thaler covers with Sunstein in his book Nudge. He talks about the American philosopher John Rawls, who had the, I think calls the publicity principle. It's essentially, if people were to know what I was doing, would I be ashamed? Like if my local newspaper were to blast it on the front page, would I be embarrassed to look my neighbor in the eye? And if you would be embarrassed, then that is a very good alarm bell. So you've got to be careful about how you use these principles. They're just the principles of what effectively influences people. Obviously influence can be used for, for good or bad.

I do like the idea that if you're doing something and it's going to be blasted of a newspaper over social media and you think that's a bad idea, then maybe don't do that thing.

It's, it's a nice simple, simple rule.

Okay, wonderful. All right, so just to wrap up for people who are interested in learning more about behavioral science, where should they start?

Well, the great thing is this is a very popular field now. Lots and lots of good books out there. Slightly biased. I have three. So I've written a new one with a friend and colleague called Michael Aaron Flickr. We've wrote a new one called Hacking the Human Mind. But I've also got some earlier books called the Choice Factor and the Illusion of Choice. So, yeah, lots of material that people can find out more.

All right, Hacking the Human Mind completely on the, on the wall behind you.

Yes, yes.

Just happened to be there.

Yeah, nice. Unfortunately I've been waving my head in front of it, so.

Wonderful. And I always want people to follow on social media. So is there anyone's research you're particularly interested at the moment? Like who should I be looking out for?

So if you look at other people. I'm a big fan of Rory Sutherland, so works in similar industry to me, works in marketing, advertising. And he wrote a brilliant book called Alchemy. And I think what he's so good at is he has this very fertile mind and he takes a lot of the same experiments that other people might discuss, but the lateral implications, you know, completely left field. So maybe check out a couple of his TED talks if you like the style and you know, that, you know, might divide people. But I think, I think he's very, very funny. Phenomenally insightful, if you like. That style. Strong recommendation. Get his book, Alchemy.

All right. Wonderful. Definitely worth looking out for. All right, super. Thank you so much for your time, Richard.

Oh, thank you, Richie. Really nice to chat. Thank you very much. Sa.