Transform Your Choices: Rory Sutherland’s Game-Changing Insights on Scientific Advertising

Thank you very much. At some level, I think every single advertising agency group in the world, uh, has got the strategy wrong for the business. And I can explain this in, well, one slide. Everybody's trying to make agencies more efficient at what they already do. In other words, you know, the kind of typical problems where a client comes along and says, "I have an advertising budget and I need to spend it, but where? Why?" I think we've completely failed is in actually expanding the scope of the kind of problems we're asked to fix. And I think there's an enormous amount of potential for taking psychology and the new learnings from psychology, behavioral science, and so forth, and simply looking at them and saying, as one very good psychologist made the point, he said, uh, "There's been more progress in psychology in the last 20 years than there was in the previous 100." And saying this is, if handled rightly, kind of another internet. It's another opportunity to change the way that businesses function, engage with customers, the way, uh, products and services are designed and experienced, and so forth. And so I'm a small part, a very small part of this, this aim. I mean, one of the things we've done is, um, found Ogilvy Change. We've hired, not, not that we will exclusively hire like this, but we've hired five psychology graduates, average age about 24, um, who've been, uh, studying psychology and behavioral science. And what we try and do, where we really sort of chalk it up as a success, is one, if we get money from a department which isn't a marketing department. That would be one. Because, you know, at the moment, if you look at Cannes, for example, it's a kind of weird agency marketing bubble. Um, but the problem with that bubble is we've lost the talent of actually talking to anybody who isn't a marketer. There, a very brilliant colleague of mine called Alistair Graham made the point. He said, "The language of marketing always strikes," he said, "a bit like the language of astrology. It's absolutely fine if you're talking to fellow believers, but when you talk to somebody else, you just sound mad." You know, you know, you can imagine this. If you've got, if you believe in astrology and so does your friend, you can have the kind of conversations which go, you know, "Oh, I know he's playing up a bit, isn't he?" "Yes, he's a Sagittarius." "Oh, they're like that, aren't they?" Okay. And that's a perfectly reasonable kind of chat. To somebody listening in who just doesn't buy this stuff, you sound borderline deranged. And as you know, I once said, you know, going to someone who's in, let's say, a Chief Financial Officer or Chief Operating Officer, and talking about brand iconography is a bit like going to their head of thoracic surgery at St. Mary's Hospital, Paddington, and saying, "We must trust to the healing power of the crystal." You know, it's that kind of weirdness.

So, just a few examples, uh, two things we've done in Ogilvy Change where we can chalk up a success that we're not necessarily talking just to the marketing function, and we're not solving the conventional problems of, "We've got some stuff we need making." One is going to a call center. Two of them traveled up on the train to a call center near Manchester. They spent the morning listening in on calls. They spent the lunch break suggesting just a few conversational gambits which behavioral science would suggest might help. And they just suggested to the call center people, "Try these sentences out. If you don't like saying them, don't say them. And, um, if they don't work, don't use them. But we think these might work." And then in the afternoon, they just spent a little more time listening in on the people trying the various sentences. And an example would be, during the course of the telephone call, which was about, um, uh, subscription to a publication, newspaper subscription, basically. In the course of the call, if you want to subscribe to this thing, you actually have three choices, A, B, and C. We'll call them. They're not called that, but you get my point. And they just said, "Actually, here, what do most people choose?" They said, "Most people choose B." Say that. "So, in future, the people said, 'Or tried saying, uh, most people choose option B. But if you like, you can have A, which is this, or C, which is more like that.'" Now, there are two reasons this is good. One, it's social proof. So, someone who's completely undecided can go with the default of what most people do. You know, "It's 27 people are looking at this airline seat right now" kind of stuff. The second thing is that three equal choices, as opposed to a default choice and two alternatives, are to the human brain a different kind of decision. And there are about four or five or six things, some of them that were suggested, that people said, "Oh, I don't really like saying that." So, he said, "Don't say it then. That's fine." Um, and then they went back on the train again. Two people, average age 24, called back a few weeks later. What's the change? The conversion rate of the call center has gone up by over 100%. It's more than doubled.

Now, the interesting thing there is that those kind of things don't get much attention in business precisely because they're not very expensive. Arguably, they don't get much attention from agencies because they're not very expensive. And therefore, it's harder to charge for things that aren't expensive. I mean, one of, you know, one of the great advantages of being an estate agent is it's a total rip-off, but compared to the rip-off price of the house, it seems relatively modest. And so, one of the things you want to do in the advertising business is attach yourself to extraordinarily expensive projects so that your own hours seem a relatively modest part of the endeavor, proportionately. But that, if you think about it, now, one of the reasons it's important, I actually, I take this lesson from direct marketing, which I still think, how many people, interestingly, have read, say, Claude Hopkins' "Scientific Advertising" or one of those early books from mail order gurus of the, you know, let's say the 1920s and '30s? Still read them because they were half right. And why they were half right is those people were mail order sellers, but they did test and they learned an enormous amount about behavioral science simply because, not because they were particularly clever, but because they did have results from A/B testing of advertisements containing coupons in the 1930s. And through testing those A/B splits, that the, the fact that newspapers had tended to have dual print runs so you could test different copy head-to-head, they found out a lot of interesting things, many of which actually later went on to, when dressed up as behavioral science, in a sense, went on to win people Nobel prizes and so forth. But a lot of that stuff was kind of known by those people early.

The second point they made, where I think they were also right, is they had a very strange philosophy. Dr. Ton Bird and various people told me this, uh, when I first started writing direct marketing advertisements or or off-the-page advertisements which hardly exist anymore. They always had the approach, "Start with the coupon." So, you design the coupon first and spend quite a lot of your time writing the copy that went in the bit that you tick the boxes in, and only then did you start writing the rest of the ad. Now, that seems completely weird. But they were right on two levels. One of which is that actually the design of a choice, you know, whether you frame it as B or A and C, or A, B, or C, for example, has an enormous effect on what people choose. So, paying a lot of attention to that is perfectly sensible for that reason alone. Um, the Scottish referendum result, by the way, had they had an option B, had they had nothing, Devo Max, and independence, there's not a chance in hell there'd be a majority in favor of independence. So, if you don't believe me in choice architecture, this stuff really, really matters. What you're given to choose from determines what you choose in a much more complex way than standard choice theory would suggest.

The second reason why they were right about this, "start with a coupon," is if you think about it, if you don't get the last mile or the last 100 yards of any purchase process right, everything else you do, let's say your coupon is only half as good as it could be, that means that everything else you do is only half as effective as it could be. So, you could produce the best advertising in the world, the most fantastic body copy, the most spectacular, uh, brilliant advertisement, but if the actual choice it leads to is crap, it looks to you like an ineffectual advertisement. And that's a really important point. By the way, "start with the transaction and move out" is a really, really good way to think. Because if you don't get that bit right, just as you could produce the best website in the world, you know, the best, you know, if Amazon had a really crap checkout process, okay, the quality of everything else in the store would barely matter. And I, I'll show you a few examples of of this kind of thing, uh, later on. But, but my fundamental point is that what we should be doing as an industry is looking for the wider application of the kind of talents we have within it, rather than trying to get, you know, marginally more efficient in a kind of trivial, tedious way at things we already do.

Now, I did have a worry, by the way, when I first started doing this, that actually, since universities produce a lot of very good behavioral scientists, what use is there for us? Like, that was a serious worry. And the very best of the behavioral scientists, Richard Thaler, would be probably the chief person there. They are also, uh, instinctively very good marketers. But most of the time, what I've found is that the talent, the mix of talents you get within an agency, is just particularly good at the very simple agency business of turning an insight into a really valuable action. And there are lots and lots of people who have interesting insights about human behavior, but turning those into interventions in a creative and interesting and oblique and original way, that's a different skill. And hopefully, as an agency, if you have both skills, you can do things which even the best academics mostly can't.

This is another sort of mantra we have. We have a few mantras at Ogilvy Change. One of them is "Dare to be trivial." "Little ideas from big thinkers" is another one. And our point is that, uh, because advertising is a human behavior is a complex system, you know, um, there is no reason to believe that you can't achieve big effects from small things. We tend to assume that if you want to achieve something big, you do something dramatic and expensive. And if you do something small, its effects will only be proportionately small. This isn't really true. In, once you involve humans in the mix, once human behavior is much closer to meteorology than it is to Newtonian physics, and once you have humans in the mix, you do have butterfly effects, really, really small things which actually create very significant effects.

Now, I've heard, whether this is true or not, I don't know. I've heard lots of anecdotal stuff, um, uh, that in a sense, the weird and almost accidental decision to put ice in cider was very largely responsible for the cider revival. I've heard various explanations as to why they did it. One of them was that the bottles were too big to fit in the fridge and they needed something else desperately to cool it down. I've also heard that they couldn't get fridge space from competing brands, so they had to find something dramatic. What seemed to happen, and this may be a total post-rationalization, so don't count this as science, but what some people have suggested happened was that it made drinking cider noticeable. Previously, you assumed the person was drinking lager if they had a kind of straw-colored drink. Suddenly, because people had ice in the thing, it just made the whole thing more salient. It's very, very difficult to order something in a pub which somebody isn't already drinking. Guinness know this very well, by the way. It's called the lighthouse effect. And if there's somebody in the pub already drinking Guinness, the proportion of other people who drink Guinness goes up dramatically. We're a herd species to a great extent, okay? Also, you can try this experiment for yourself. If you're the first person ever to order drinks and you order a Bloody Mary, someone else will have one as well. It's one of those strange things. You know, to a great extent, you know, they're all manner of kind of instinctive herd behaviors in humans. But there's no reason to believe that, um, you shouldn't do small things. We also believe test counterintuitive things on the borderline of stupid, actually. You know, test even stupid things.

Now, the reason is, if the things you're testing are kind of logical, rational, you know, it all makes sense, we all agree on this one thing, well, that's fine. And you might discover some valuable improvements from doing it, but so will your competitors. The really valuable discoveries come from testing some aspect of a product or service which no one's even thought might make a difference because it seems so weird. And those are the really valuable findings. I've always argued, I think one of the most extraordinary insights behind a business, uh, you know, I mean, you know, a weird one would be, nearly all really successful businesses arise from something which defies conventional economic thinking. You know, Starbucks. You know, the general assumption was you can't charge $3 for a coffee. The question I suppose Starbucks, and there were, there were precursors of Starbucks, but the question they asked is, "What if your place is all about coffee? How much can you charge for coffee?" Then everybody else sold coffee as an add-on, you know, an add-on to their main menu. "Oh, yeah, you can have coffee if you create a thing which is all about coffee, what can you charge then?" You could have, call it the Jack of all trades, turistic, you know, that if you only do one thing, people judge you in a different way to if you offer lots of things. Ray Kroc of McDonald's, I think, had one of the most fascinating comments when he said, "People don't want the best burger in the world. They want a burger that's just like the one they had last time." Now, he's actually, probably a good Darwinian psychologist would say he's absolutely right. We've evolved a very strong preference for food that's identical to food, food we've eaten before, because if we're still alive to wonder about eating it the next time, it was probably okay, wasn't it? You know? So, actually, that was an extraordinary, no one in the market research group was said, "What you, what sort of food do you like?" "Really consistent food." Okay. No one's ever going to answer that. Everybody would say, "No, make it better." Actually, in many cases, you know, the extraord, I don't know if any few of you probably worked at McDonald's, and you have these rules like, you toast the bun for 72 seconds, not 76, not 68. It's very, very precisely done to ensure complete continuity and consistency.

The other point about testing things is, one of the things we've got to worry about is readiness to fail. Because marketing ROI is a pretty good idea if you apply it to the whole marketing function. What gets dangerous about ROI is when you apply it to every single activity. Because the problem there is that patently, if your marketing department doesn't create value, it's a bit right, I accept that. The problem is that in, in the process of creating value, it probably has to try things that don't work. So, the idea of of trying to measure and trying to justify everything you do, as though everything you do has to get some sort of return, is mistaken.

I'll give you an example. There's a trade-off, in a sense, between optimization and, um, and flexibility. And, and this is perfectly illustrated in nature by bees. Do a thing called a waggle dance, which, when they found pollen, they go back to the hive and they do a dance where their direction of dance in relation to the sun points to the direction, sorry, in relation to the vertical, I think it is, on the wall of the hive, um, points to the direction of the pollen source, and with some additional information about distance relative to the direction of the sun. And what surprised scientists was that whereas large numbers of bees follow the waggle dance and are indeed influenced by it, there's a significant number. It varies according to species and so forth, but let's say 20% of bees ignore it. I don't know whether it's the same bees or whether it's 20% of the time, you know, bees, it's kind of difficult to recognize them, isn't it? But I will try and find that out next time. Um, and they were a bit baffled because they said, "You had 50 million years of evolution. You'd think, wouldn't you, that, you know, higher levels of bee compliance would have been enforced over the 50 million years? You know, that little bee accountants would have come along and said, 'You know, we really need to get better at this because you're wasting your time going off on these random forays, uh, into the countryside when there's a known source of pollen at, you know, 37 degrees west or whatever it is.'" And then they modeled it again and realized that without these rogue bees, the hive would starve to death because it's the rogue bees who provide the new information as to how the foraging patterns, the collection patterns have to adapt, um, when faced with changing circumstances. One source of pollen gets exhausted, another one crops up somewhere else. And there's an, this is the R&D function, the rogue bee. And there's an optimal level of what you might call, you know, harvesting and hunting in any, in any organization. And the hunting thing has a higher failure rate, but when it succeeds, it has a very large upside. You know, that's the kind of Silicon Valley business where you're probably going to fail, but if you do succeed, you succeed big. The other 80% of the hive is basically doing the, you know, the efficient processing stuff. And getting the balance between the two right is really important. And a great business is not one where 100% of the bees obey the dance.

So, one of the vital things is continue to test things and actually test weird things. Because, as I said, if you find that patch of pollen that nobody else is going to find, it's a much more valuable discovery. And so, what partly obsesses me about, um, behavioral science is that it gives you a license and a vocabulary to justify trying eccentric things which previously everybody has assumed to be either obviously stupid or almost unimportant to the point of irrelevance. Really, really small things, you know, tiny little tweaks to the color of a button, etc., or interesting counterintuitive things.

I'll give you a nice example of the counterintuitive thing. The assumption would be, you have a website, it has your product on it, you make a lot more money when people buy your product, uh, direct from you than when you refer them to a partner retail site such as Amazon or Argos or something else. So, what every person does, because it seems perfectly logical, is they make the button that directs people to their own site to buy direct from the manufacturer, they make that green and really attractive. And the button that links you to Amazon or Argos or JohnLewis.com or whatever, that's kind of gray and shabby looking and probably even has little turds on top of it to make it disproportionately unattractive. Okay. And everybody goes, "What a sensible thing to do. I see you're using a nudge to drive people to shop direct with us rather than elsewhere." However, there's a contrary thing to test, which is, you make the button that goes to Amazon and JohnLewis.com green, and your own button gray. You're going, "What the hell is that? Why on earth would you do that? You're actually encouraging people to shop through an intermediary where you make a lot less money rather than coming direct to you." The answer is, actually, in some cases, you make more money overall. Because people instinctively want to shop from a third-party retailer. And if the design of your site coheres with their basic instinctive urge, you get a lot of people clicking through to buy, and actually, you sell more. A lot more. You sell a far, a higher proportion of people will go to the intermediary, but you still end up selling more from your own site. So, you, you know, proportionately lose out, but overall you gain.

Why is that? Beautifully described by a Romanian chap living in Holland in a book I've just downloaded, um, on Kindle. He describes this as "Paving the cow paths." There's a thing in architecture where you don't actually put paths in across a grass court. You wait to see where people naturally walk, and then you put concrete paths where the muddy bits have formed, because you now know where it is that people want to walk. If you just design paths, it's based on a whole load of assumptions about where people are going that you can't possibly know. He refers to this as "Paving the cow paths." In the sense that if you make choices feel instinctively comfortable, lots of people will make them. If you make choices feel effectively contrary to your natural instinct, people find it really, really off-putting. So, that's why, that's why if you think about that, that's a completely small and counterintuitive thing to try. But in the event, when you have the results, it proves to be brilliant.

Then the final thing is here. This is just a quote I always put in. I would argue that anybody with a creative background, with a psychology background, humanists, history scholars, postmodernists, or artists, as these people have said, have been frozen out of the decision-making process because economists and people who effectively pretend that human behavior can be mathematically modeled enjoy a disproportionate dominance and influence. Because by using numbers, you simply make yourself seem credible and scientific, even though the numbers you're punching into your model and your, the model itself, may actually be really, really bad rubbish and useless. An awful lot of decisions, business decisions, particularly true in civil service decisions, people love a mathematical model precisely because if they allow the model to make the decision, they're offloading responsibility for their own judgment onto somebody else. And it's a very clever ask covering exercise. Say, "Well, it wasn't me. My decision can't have been faulted because I chose the one which came out at 3.1, whereas the alternative only came out at 2.8." And you've therefore made your decision by making it seemingly rational, you've made it effectively immune to criticism.

Now, don't get me wrong. I don't suggest that you should put humanists, postmodernists, artists, creative people, planners, etc., in charge of everything. I think, you know, you need a counterbalancing rational judgment to curb their worst excesses and to check that what they're doing actually makes sense, is feasible, um, is affordable, um, you know, makes general economic sense. The only problem you've got is that so, I would argue that, you know, all good creative people should submit their work to appraisal by more rational types. My problem is that it doesn't happen the other way around. People who make seemingly rational judgments based on mathematical figures don't have to submit their work to imaginative people who, as these people said, "Until these insights are recovered, we'll be unable to effectively deal with the internal problems of our age." It's from "Complexity in the Art of Public Policy." And these people understand that in complex systems like human behavior, in some cases, it's actually the artists, as it were, not the economists, who understand sort of oblique interventions and psychology and so forth, who have the better answers. My problem is that the rational people, because when you have a rational idea, it's effectively allowed to go unpoliced. You go, "Look, the answer is 3.2, therefore we should do this." There's no need to actually expose this thinking to anybody else who might question it, least of all someone with an imaginative bend.

So, it's very interesting, you know, talking to interesting groups of people. So, one of the things I've also done at Ogilvy Change is talking to people who have, for example, um, uh, the, the opposition to High Speed 2. A far better idea is to build a high-speed railway across the north, connecting Liverpool to Newcastle via most of the northern cities. You then create a kind of counterbalance to London because you create, you know, effectively a well-connected megalopolis. If you went all the way from Newcastle, I, I guess you go Leeds, York, etc., Manchester, Liverpool, probably a bit of Sheffield on, you know, etc., okay? Now, the rail connections there are appalling at the moment. I mean, the journeys take ages. It's a trans-Pennine railway for part of it, it's rubbish. And someone made a very simple point, which is that the problem you have if you're a northern city is very interesting. It's actually, it's the fault of feminism largely, which is that women now work. And therefore, if you think about it, if you have a place where neither of you can feasibly commute to another nearby city because the journey time takes too long, you can never move. Because in order to move, you've got to find not only yourself but your partner a job somewhere else. That's one of the reasons why places like London become so enormous. Because when both of you work, the number of places in the world where you can both find optimal jobs becomes reduced to about 10. That's a complexity question. By contrast, building a high-speed railway line to Manchester and reducing the journey time by an irrelevant amount on the assumption, by the way, the economic justification of this is that everybody who is on a train is economically completely unproductive for the entire duration of the journey. Now, that might be a good idea, by the way, because I've often argued that the solution to long-distance travel isn't actually better jet aircraft or better trains, it's actually better drugs. Because if you could effectively just knock yourself out or go into a trance for the duration of a journey, you wouldn't care how long it was. But we'll park that idea for now. Okay? Um, yeah, I, I think that's GSK's problem. I don't think it's engineering firms at all, you know.

Um, did have a friend, actually, brief inter, who, uh, worked for a bank, and they do urine testing at random banks. So, the only time he could ever smoke pot was the day before he went on holiday, because then you have two weeks for it to clear the system. And therefore, if Goldman Sachs do a random test, you're okay. And the one thing he said, because, because he'd always get stoned, the table we went to holiday, occasionally I'd have these awful things where they'd say, "No, I'm afraid your flight from GFI's got to be cancelled. We're going to put you on a coach to Luton." And he said, "Under any other circumstances, I would have been livid." He said, "I was just there the morning, go, 'Yeah, fine. Coach to Luton. Whatever. Cool.'" Okay? So, maybe, maybe we're on to something here. But what I, my point about that is, if you look at this question in human terms of the journey from London to Manchester, okay? If I go to Manchester, which I do about four times a year, now, it involves a unit of time which we business people technically call a day out of the office, right? 40 billion pounds later, when you've built a high-speed line and the journey has been reduced from, uh, 2 hours and 10 minutes to 1 hour 26, okay, my trip to Manchester will involve a day out of the office, okay? Now, my point is that, you know, there are, there are huge gains to high-speed rail. If you, if you take Paris-Lyon, which was four hours and you get it down to two, as the TGV did, that's a big difference. You get France's third biggest city and France's biggest city which are now reachable within the day. That's important, okay? Frankly, 2 hours 10 to 1 hour something or other, 1 hour 20 something. My other argument is, look, the real problem of getting to Manchester is getting to Euston, which is a total hole, okay? It's connected to absolutely no tube lines at all. Uh, it's, if you're going to catch a taxi there, you've got to go into this stygian pit underneath the station where you're virtually gassed to death, okay? You know, getting to Euston, particularly with luggage, is just horrible. The bid on the Virgin train thing to Manchester, that's the good bid, right? So, why are you spending 40 billion pounds reducing duration, the only part of my journey that isn't, which is the bit when I'm on the train? And so this is my point, which is that the, the obsession, what you've, what you've got, Anthony Tasol, the planner, is very good on this. He says we've created a kind of thing called the Arithmocracy, which is a group of people who, because they have access to the numbers and the spreadsheets, have effectively seized the levers of power and decision-making to a completely disproportionate extent.

Now, if you're looking for counterintuitive things, if you're looking for things that basically, you know, the Arithmocracy won't look for, the way to look at it is, there's a very, very big dark side of the moon kind of thing, which is nearly all assumptions about people and their behavior come from two, you know, lenses on a particularly dodgy pair of binoculars. One lens is neoclassical sort of economic theory about people being utility maximizers and so forth. And the other lens is market research, which is also like economics, useful at times, but dangerously spectacularly and, uh, actually, you know, frankly, 180 degrees wrong, uh, in many, many cases, I mean, spectacularly wrong. John Mayor Keynes, the economist, made this very good point. He said, arguing against this kind of precise mathematical approach, he said, "To be absolutely honest, I'd rather be vaguely right than precisely wrong." And there are cases where following the assumptions of either market research or economics will actually make your assumptions terribly, terribly wrong.

Here's something which interested me because it defies both. It completely defies economic logic and it completely defies market research. First, market research. When the people tested this, a large research company has huge experience testing taste testing carbonated drinks. It was the worst result they'd ever seen in their life. I mean, responses were not just "I'm not very keen on this." They were getting responses like, "I wouldn't drink this piss if you paid me to." Okay? In research, it was determined to be absolutely revolting. And yet, they went ahead and they priced it at about three times the price of a can of Coke, certainly double. How the hell do they manage to do that? Well, one way they did it, they made the can smaller. And the human brain instinctively, I don't think we went through this rational process, just instinctively thought, "That's weird. That can, it's like really small. That must mean the drink is like incredibly potent because if I had a full, um, 300, what is it, if I had a full 750 milliliters of the stuff, like a can, presumably I'd just go, 'Do!' or have a heart attack or something." So, it must be really hardcore. The fact that they have to put it in this small can, so because it's really hardcore, suddenly the price point is, well, whoa, this is a different kind of drink. Now, let's be kind about that. That might be how it worked unconsciously. No one in a market research group is ever going to say this. "There's no way I'm paying 1.50 for a for a soft drink. It's ridiculous." "Oh, wait, wait, hold on a second. I would if you gave me less of it." Okay? No one's ever going to tell you that. And this is why the successes that come from sort of weird, eccentric, um, counterintuitive things are so, so valuable because no one else has got there because they're using the mental equipment of the of the old dodgy pair of binoculars.

Economics should work like this. Physics sits on maths. Chemistry sits on physics. Biology sits on chemistry. That's where you get complex, particularly although you do have complexity physics and so forth, so let's not get too simple. Psychology sits on biology. And economics should kind of sit on, on, uh, psychology, on psychology. What they've done in order to make themselves look kind of mathematical and neat is they sort of pretend it sits somewhere there that you can model human behavior as if it's, you know, kind of maths. I object to maths teaching in schools, actually, because the typical maths question always goes, you know, "Two buses leave a town at the same time, traveling at a constant 47 kilometers an hour. How long will it be before they are, you know, 16 miles apart or whatever it is, 16 kilometers apart?" Okay? That's the kind of maths question you get. And there's a single demonstrably right, true answer. But most real-world questions, there isn't a rational, demonstrably right, absolute true answer. Most real-world questions are like, "Some buses are leaving around about this time, and they might get there about when. How, you know, if you're picking up your friend from the bus station, when should you turn up to pick them up?" Okay? I mean, their buses, when do they ever travel at a constant speed in a straight line? I mean, it's ridiculous, right? So, what the economists have done, they've assumed this kind of neat world for the purpose of mathematical neatness of buses traveling at a constant speed. And of course, it bears no relation to real-world behavior at all. And the best physicists, of course, and the best mathematicians have really spotted this. Wonderful guy, Murray Gell-Mann, who is the chap who discovered quarks, and Nobel prize-winning physicist, he makes the very simple point. He said, "Imagine how difficult physics would be if atoms could think." You know, it's all very well modeling the behavior of sort of inert single particles of one kind or another. But trying to understand the business of things which are influenced by the things around them, once you, you know, that are influenced by the things around them and have some degree of independence of movement, that's a completely different matter.

The other great thing about research is to understand, I think Jonathan Haidt's brilliant analogy. How many of you read Daniel Kahneman's "Thinking, Fast and Slow"? Any few of you? Actually, quite a quite a healthy proportion. More of you next time, but that's pretty good. He uses System One and System Two. System One is the automatic bit. System Two is the bit that consciously thinks. The example Kahneman gives is, if I ask you, "Five plus five?" or "One plus one?" you effectively know the answer without thinking about it. If I said, "What's 17 times 34?" you'd have to go, "Duh," and it would hurt a bit, and your eyes would go up in your head, and you'd start muttering to yourself, and eventually you'd come up with, after a bit of pain and annoyance, you'd probably come up with the right answer. But it's a completely different mental process. Haidt uses a bit more of a poetic metaphor, H-A-I-D-T, it's spelled. He refers to the brain as the rider on the elephant. The elephant is the System One. Um, the rider is the conscious part of the brain. The vital thing to remember is that if you do market research, there's a huge danger, well, a vital, vital point. This is poetic, is because the elephant is kind of wise, it's possibly cautious, it's a creature of habit, it's a herd animal, and it doesn't talk. It can trumpet a bit, but it can't actually talk. The rider, on the other hand, is capable of performing advanced mathematical equations and so on, but he's actually a bit of a bullshitter in that he's convinced he actually controls the elephant. And so what he does is, as Haidt's language, he thinks he's the office, when in reality he's the press office. Most of the time he's post-rationalizing explanations for things the elephant did under the pretense that they're what he wanted to do all along, okay? And the problem is, if you do conventional market research, I mean, I, I really, really qualify this with totally conventional stuff, the danger is you end up talking to the rider and listening to the rider exclusively. And you don't re, the problem is that fundamentally, the truth of the matter is the rider cannot, you know, he can go, the elephant, coax it, train it, nudge its attention, but it can't get the elephant to do anything it instinctively does not want to do. Just as you can't use the power of reason to get someone to fall in love, or or even buy a car they don't much like, okay? And the elephant ultimately executes, or does not execute, what the rider suggests. And understanding this is really, really important because a lot of the time, I think in advertising, we might end up writing software for the wrong operating system, obsessing about how to talk to the rider when actually the rider isn't really the obstacle to purchase here. There's something about, there's something about your product or your service or whatever that the elephant doesn't like. And if you can redesign things to make them elephant-friendly, you can often find you make a huge, huge difference through. And this is what design understands very well. Design generally talks to the elephant pretty well.

You know, I, I know one of the things I'd argue is the obsession with advertising on a proposition, the idea that in order for an advertisement to work, it must contain some, you know, overt stated phrase. I think is a dangerous assumption. We've always, we've always had this belief. But actually, if you look, a lot of successful advertising, it works on multiple levels. In fact, um, another model for the brain, the digital camera model. It's got an automatic mode, which, I mean, let's be candid, those of you with digital cameras, you use the automatic mode. When you first get the camera, you dick about a bit with manual override, but after a while, you actually, if I use the automatic mode, it's may not be perfect, but it's always pretty good. And it's never, okay, generally. You might, in special circumstances, you'd use the manual mode if there's a massive amount of backlight or you're taking pictures of fireworks or something, okay? But most of the time, you think, "Great thing about the automatic mode is it's always pretty good." Okay? You know, it's not perfect, but it's always pretty good. And so, most of the time, what we're looking for in life, for good evolutionary reasons, is solutions which are pretty good and definitely not shit. Because math leads us to believe that the definition of good is the very best something can be. But the actual algebra of decision-making in the human brain is not only concerned with how good a thing is, but how bad it is. It might be in a worst-case scenario. Satisficing is the business of actually looking for something that's definitely pretty good and very, very unlikely to be. Which why McDonald's is, in a sense, the best restaurant in the in the world. Now, that scandalizes people. No, the best restaurant in the world is that place in Denmark where they do foraging food or, you know, whatever. No, no, no. The reason McDonald's is so successful is not because it's a very good restaurant. There isn't a single town in the world where McDonald's is the best restaurant in town. In that sense, it's incredibly good at not being bad. Very interesting question. You know, I, all of us here, you're probably not yet old enough, but anybody who's about 35 in the advertising industry, you know, they know that about one time in eight, you go to a mid-range restaurant, you end up with the shits, right? Okay? It's a very good restaurant, but the downside is pretty. You never get ill going to McDonald's, do you? No. There's always one person. Is there somebody here who disputes? Occasionally, one person comes up to me afterward and says, "Actually, I've got it. I think they're just blaming it. I think they're hungover or something, and they're just, you know, they're basically just offloading the blame on Ronald McDonald because it's, but I mean, you know, there are cases, I'm sure." But actually, if you think about it, you don't have any anxiety eating there in terms of hygiene. You know, the absence of negatives is something that's that's probably processed differently in the brain from the presence of positives. And I think the effort that marketing makes in looking at how can we say this thing is great versus the effort it makes going, "How can we remove emotional hurdles to purchase?" I think it's a really, really interesting question.

I was talking to George Loewenstein, who's a very good behavioral economist at Carnegie Mellon. And the first thing we said is, very simple example of this, okay? People are trying to sell solar panels for people's homes. And they use the language of, "It should pay for itself in 10 years." Right? That sounds, doesn't it, right? Okay. So, I get this thing, I shell out a load of money, and 10 years later, I've just got the money back. Now, interestingly, if you frame that as an investment product and say, "10% annual return," that's a hell of a lot better than anything else you're getting. So, how you describe something, just a simple tweak to how you frame something. Do you frame it as an investment or as a kind of outlay which you have to grudgingly recoup? Okay? Nobody sells a 10% investment product on the, "Get your money back in 10 years." So, it's a very, very interesting point that, you know, that elephant-friendly stuff. A lot of the elephant is kind of cautious. It doesn't like new things. It does, you know, it's slightly resistant to change. It's a bit of a conservative. For good evolutionary reasons, I mean, you know, I mean, a lot of those elephant defaults, like herd, you know, herd instinct is really intelligent. You don't have perfect information, okay? It's not a bad policy to copy what everybody else does because you don't know. And what you do know is that if everybody does this thing, it may not be a very good thing to do, but it's unlikely to be really terrible. Do you see what I mean? If you're lost in the fog on a hillside and you were stumbling around, you know, in the Brecon Beacons or something, and you happen, you only had sort of, you know, three feet of, of vision, and you happen to stumble on a path, you'd follow the path, wouldn't you? And that's just your herd instinct, okay? Lots of people have walked this way. I don't know where it goes, but it's unlikely it's going to head straight off over a cliff, okay? Because other people in better light on other times have taken this route. And that, you know, it's unlikely that 35 people have all walked straight over a cliff. So, if there's a path, it probably leads somewhere, okay? So, and a lot, an awful lot of the time, in evolutionary terms, herd instinct, for all manner of different reasons, actually is essential, you know, is is part of our human nature.

Mars is absolutely right. Met Daniel Kahneman. He was at the Hay Festival and someone asked him, not me, they said, "Are there any System One businesses?" And he said, "Yes, advertising." And then he paused for a bit and said, "And politics." Interestingly, this is where you have to talk to the elephant. Now, another way of looking at it is the elephant instincts have evolved for all manner of reasons. And if you're a real kind of, um, uh, hardcore rationalist, you say, "Yes, but of course, it evolved to survive on the Serengeti. And the environment in which we find ourselves like is nothing like that. So, we have totally inappropriate responses to things based on the fact that we spent a million years evolving in circumstances which are nothing like modern-day life." And an interesting one is the, the hardcore, you know, there are people that sort of weird dating, what the pickup artists, you know, there's basically a, a dating strategy for men, okay, which basically says, "Proposition every girl you like who you meet. You'll get rejected 11 times, but you get successful on number 12." Okay? Now, most men are incredibly reluctant to practice this because the humiliation and just the embarrassment is absolutely monumental. But the pickup artists tell themselves the evolutionary story, "Ah, we evolved to live in communities of 100 people where this kind of behavior would get you beaten up or ostracized." Just going and chatting up and propositioning every girl you meet in a community of 150 would get you clubbed over the head or ostracized. But in London, hey, okay? So, those people, those are the kind of people who go, "Ah, yes, but, you know, a lot of our evolutionary instincts are completely stupid and irrelevant." I think, I think that's true, some of the time. I think most of the time, actually, our instinctive judgments, it, once you understand satisficing, are instinctive judgments like habit, "Do what I did last time," "Do what everybody else does," um,

I think they're actually pretty good, and in any case, you can't escape them anyway. I mean, those pickup artists manage to do it. Um, uh, but, but I, I, I, I think most people, you can't escape them. If you go into a restaurant, if you, if you've ever been sort of beckoned into a restaurant by the owner when you're in a place like Seville or something, and you finally go, "Okay, much cheapness, I'll come in here," and you go into the restaurant, you're the only person there. Okay, you just feel like, don't you? Okay, that's the herd instinct at work. You might argue it's totally irrational, but in a sense, there's kind of nothing you can do about it. You just feel bad.

And a very complex question I asked the other day, which is that, look, um, I know rationally, okay, that I'm just to say flying Malaysian Airlines now than, you know, I was three years ago, and that I am flying any other airline really within re, you know, you know, to at least, you know, the sort of seventh decimal place, okay? And my problem is therefore, is it irrational for me to be reluctant to fly Malaysian Airlines? To which I answered, well, it's very complicated because even though I know statistically that the whole thing is completely irrelevant, if I get on a Malaysian Airlines plane, I'm going to be three times as frightened as I am on British Airways. Is it worth 300 quid to avoid to avoid feeling really frightened for nine hours? Well, probably yes. So the extent to which you can say, well, it's rational is really difficult since you have no control over how the elephant feels.

Now, the other way to look at it is, look, when you design a steering wheel, it's designed for the way hands evolved. Now, human hands did not evolve to steer Mercedes. Okay? Um, well, Mercedes only been around for 100 years. They may have been a small amount of, I mean, we should have evolved a really long finger for doing the indicating, I suppose. But, um, the fact is, we, nobody says we're exploiting your hands there because, you know, you're designing something. All good design works around our evolved physique. You know, door handles, kettles, etc., work well with the bodies we happen to have, regardless of what those bodies is evolved for. And in the same way, I'd argue good design, good design of experiences, good design of advertising, good design of strategy, just has to work with the evolutionary, uh, psychology we have as well. You know, we don't regard it as immoral to exploit people's hands for the purpose of car steering. So I think designing choices, um, in line with the way we've evolved to make choices, is simply good design. It can be manipulative and horrible, by the way, and I'll show you a few examples soon. But the fact is that nonetheless, on many occasions, it's the best you can do.

Now, I'll skip this because it's too long. It's a terrible, ugly slide, but there are quite a lot of things that are outside the view of the binoculars, outside its field of vision. Things like path dependence, it doesn't understand framing, it doesn't understand complexity, it doesn't really understand AC, that's a thing I'll explain later, psychophysics, it doesn't understand loss aversion, it doesn't understand, doesn't understand satisficing, doesn't understand signaling, doesn't understand ethics and trust. Okay? So if you take those two, these things tend to be instinctive elephant things. Uh, let me give an example, U, of, of to what extent actually you can up the elephant and the elephant doesn't really know why it's being screwed over.

This is an example of an extraordinary thing. Does anybody here familiar with driving down the M20 and needing to turn right or left onto the M25 to go south or north? Have you ever messed it up and missed the turning? Yes? Right. Okay. Anybody else to use that junction? Have you, have you messed it up? Yeah, you've missed the turning. Right. There was someone in my previous talk today who, someone said, "Do you know that junction?" Yes. "Have you ever missed the turning?" Yes. Okay. Now, anyway, I, I go down to Deal quite a lot and come back to Sevenoaks down this route, and I have to turn left on the M25 to go south. And then the sort of last 12 times I've done it, I've messed it up four times. Okay? Now, if I just messed it up once, I wouldn't have thought, you know, if, if I only used the junction once and cocked it up, I would have just said, "You're unfamiliar with it, you're an idiot." Right? If I only cocked it up one time in 12, I'd just go, "You're an idiot." When I actually messed it up like four times out of 12, or five times out of 12, and have to drive into Addington and come back again, I thought, "Actually, this isn't me, is it? It's, there's something wrong with this junction." Okay?

Now, the interesting thing was, because the elephant does most of the driving, nearly all the driving you do, and sometimes the elephant does all the driving, you know that moment where you've been driving for 15 minutes, you go, "This is weird. I can't remember the last eight miles." You know, "I hope I didn't hit anyone." You know? Um, but because the elephant does a lot of driving, I had no idea what it was, and I just said, "Okay, so it's not, it wasn't as if I was getting that junction's really stupid because I had no idea. I just knew I'd cocked it up." So I went back to Google Street View. Now, this is the first clue you get. There's a junction. There's, I'm sorry, you, I'm a bit of a nerd on this. There's, there's no distance marker to explain how far away this division of the lanes comes. It mentions London on the right and on the left, and also the M25 is mentioned is next to a stupid yellow block. Um, and there's no primary route destination. It mentions Swanley and Dartford Crossing. Now, Swanley, I mean, the, that intersection on the M25 is known as the Swanley intersection. So whether Swanley is left or right of, is it's stupid information. And actually, the sign people admitted they said there should always be primary route destinations like Gatwick or Stansted. So they give you obvious directional clues because, I mean, other than the people of Swanley, I don't think anybody in Kent knows whether Swanley is north or south of the M20, right? And it's chiefly famous because, of course, it's the, the Swan interchange is where Kenneth Noy had that road rage thing. So it's the interchange, the most famous thing about Swanley is the interchange. Anyway.

Now, you know, okay, so there's no clue about what you're supposed to do there. There's just that thing. Right. Now, we move on a bit. You can see in the distance another gantry approaching. So if you don't understand that gantry, which you don't, because it's okay, you go, "Oh, well, never mind, there's another gantry coming up, so I will be able to get adequate and reliable information from that gantry." This is your elephant. This is the way the elephant works. And I will make a decision then. Before you can even read what's on the second gantry, this happens. They put a divider in the road at a point where it's completely illegible. And if you're in the fast lane, you basically go, "Okay." And then as you get closer, you can see what's on the thing, and you realize you've got 100 yards to do an amazingly dangerous and illegal swerve over there, or you can plow into those trees, or you can carry on eight miles out of your way to Addington.

Now, interestingly, it took me 10 minutes with Google Street View to work out, you know, why this is there. My elephant was just following cues, using unconscious inference effectively, you know, basically going, "Oh, well, look, there's a gantry, so I'll look at the gantry and I'll make a choice." And then suddenly you're forced to make a choice. Or you, Bob Dylan, any of Bob Dylan fans, "Brownville Girl," you know, "I didn't mean to trespass, but sometimes you just find yourself the wrong side of the property line." Okay? And this is the case where, before you're given the information necessary to make the decision, you're forced to be one side of the line or the other. And actually, I did, I sent it into the Highways Agency. They, nothing. I got nothing back for two days, and I thought, "This is useless." You know? And funny enough, I got a phone call saying, "We had a meeting about it yesterday. We all had a look at this, and we agree it's absolutely terrible, and they're going to change it." So all of the three of you who've messed up that turnoff, okay, I, but that is interesting because I knew there was something wrong. I thought, "Okay, the, the frequency I'm making a mistake now suggests there's some piece of poor road design." But I had no idea what it was. And then it was only when I ran over it.

Now, let me give you an example of this where all of you have made, I think, this mistake. Contentious view, okay? Which is that I don't think wine is all that good as a drink. Does anybody else agree with me that gin and tonic's nicer? Spirits are nicer. Beer's nicer. Cider's nicer. You know, white wine's not really a bloke's drink anyway, in my old-fashioned view. And, you know, red wine can be quite nice with certain sorts of food, but it's not that great. It's incredibly inconsistent. It's often really rubbish. Okay? Come on, somebody agrees with me here. It's not that nice, is it? Let's be absolutely honest about that.

Now, cocktails are bound to be better than wine mathematically because when you make a cocktail, you can take any liquid of any kind from anywhere in the world, alcoholic or not, and mix it in any ratio, at any temperature, right? Okay? Now, that gives you quite a lot of freedom to make an interesting drink. With wine, the rules are, you see those grapes in the field over there? Use them and just them, not any other grapes, to make an alcoholic drink. Well, that's just self-limiting, isn't it? It's rubbish, right? Whereas with cocktails, you can mix anything, like beer and ginger beer and wine and whiskey and anything, okay?

Now, what happens when you go to a restaurant? Restaurants want you. They do the same thing as the, the M20 signage. They want you to order wine, right? Because it's much more profitable because you can mark up wine insanely. People know what whiskey costs, and they know what beer costs, so you can't, like, charge £8 for a pint of beer, and you can't charge sort of, you know, £9.50 for a single shot of Johnny Walker Red, because people go, "Look, I know what a bottle of Johnny Walker Red costs, and if you're charging nine shots, you're basically getting, you know, £300 quid a bottle out of me. That's a scam. Piss off." Right? But with wine, you get some sort of wine, French wine, with a plausible-looking label. It cost you £6, and you charge people £30 for it. They go, "Oh, it's lovely, this, isn't it? So sophisticated." Right?

So what happens when you go to a restaurant? How do they actually force you to drink wine? Well, first thing happens is you sit down at a table which already contains wine glasses. It's a pretty clear prompt, isn't there? Four, there's a wine glass every single, sometimes there are two, okay? You haven't said anything about wanting to drink wine, they put the bloody glasses down already. Then what happens is you sit down, okay, and they bring you the menus, and they bring you, do they bring you a thing called a drinks list? No, it's called the wine list, right? They've already bloody nudged you in that direction with the glasses. Now they're bringing you a wine list. Now, how is the wine list laid out? The first eight pages are bloody wine, right? And then the final back page, for like perverts and deviants, they might deign to list like whiskey, cider, beer, and other drinks, okay? So the choice architecture.

And this is a bit of final genius which only occurred to me about five days ago. It's even cleverer than that because how many wine lists do they bring you? One. Now, by bringing you only one wine list to the table, you presume that one person is going to order for the whole table, which effectively forces you to order wine, right? Because you can't say, "Well, I'm going to decide for all of you, right? And we're going to have a 2-liter bottle of Vimto this evening." Right? So basically, the second you've made the assumption that one person is ordering drinks, you can't say, "Okay, it's Harvey Wallbangers all around, folks."

So what happens, and this is where the real doom sits in. Someone asks the question, "Red or white?" And it's game over for real human choice, free will, freedom of action. It's all, once someone said red or white, you're all forced to play along with this ludicrous charade of pretending you enjoy a stupid French alcoholic drink which doesn't get you drunk in a nice way and doesn't taste very good either, right? Now, admittedly, I've just been presenting to Diageo, so I might be a bit biased on this one, but it is, it is, when you think about it, that single wine list is the craftiest bloody nudge I've ever come across because you basically, someone has to go, "Right, we've got to reach consensus here," and the lowest common denominator is to say, "Red or white?" And then once you've said red or white, other things they do in restaurants still or sparkling. It's basically a choice framing designed to prevent you ordering tap water, right? Because if they say, "Would you like anything to drink?" or "Would you like water?" people can say, "Just tap." The second you said still or sparkling, you've basically nudged people into this choice.

And what you could call that, red or white, still or sparkling, to some extent, tea or coffee as well, it's placebo choice. It gives you the illusion that you're in control, where in reality, almost all your options have been frozen off, a bit like the M20. Okay? If you don't believe in placebo choice, by the way, the vast majority of door close buttons in lifts are placebo buttons. They're not connected to anything. They're just there so that irritable people feel they're in control and can jab it in an impatient fashion. It has no effect on the behavior of the lift at all, okay? So watch this stuff because the way in which choice is designed.

I mean, another interesting example, I, I often give is airline websites. Always fascinate me because they're all designed, if you look at the path dependence of choosing a flight, it's designed for business travelers, okay? First question you ask: "Where do you want to go and when?" Well, that's pretty good for business travel because it's unlikely that my boss will come to me one day at 9 AM and say, "Hey Rory, sometime next week, could be Tuesday, Wednesday, or Thursday, I'd quite like you to go somewhere sunny." Generally, my business trips are, you have to be in Frankfurt on Wednesday, okay? So for a business trip, the way that that airline websites are arranged makes perfect sense. The, the path dependency of choice is where are you going, when, what class of traveler you allowed to fly, okay? The problem there is that it's completely useless for leisure travelers. Leisure travelers might just go, "I want a bargain. Here are the dates I'm on holiday. Just give me a bargain." Okay? That might be a perfectly sensible way of doing it. You might say, "I, these are my exact dates. I just want to go somewhere interesting, sunny, historic, etc." Okay? I know, because you've got your Facebook generation, aren't you? So you've got to go somewhere really weird and crap on holiday to show off to your friends. How many people have been to Machu Picchu? Come on. Okay, we got, okay, six, five. How many people have been to Lincoln Cathedral? Actually, that's about equal. Okay. Let me just tell you, Lincoln Cathedral, okay, costs about 3% as much to get to it. You can do it in a day, and it makes Machu Picchu look, okay? Just, just giving you that one. You don't have to go miles away. There's no correlation between how elaborate a holiday is and how much pleasure you get out of it at all, okay? It's also good to go on holidays that are cheaper than the ones you can actually afford because then when you're on holiday, you feel justified in being extravagant. But anyway, that's a separate thing.

But the second thing is, they ask you the stupid question upfront: "What class do you want to travel?" Well, obviously first class, but it depends what it costs, doesn't it? And we all want to go first class, but not bloody idiots. But if it costs £20 more, we'll all go first class. And if it cost £20,000 more, none of us will. So they're asking a totally stupid question before you've had the information that made that decision possible. Exactly like the M20, they were asking you to choose.

So one of the interesting things that airlines now do is when you present people, even if the person has just said economy class, you present them with the premium economy and business class flight, and it makes them over £10 million in extra revenue, just changing the choice architecture like this. And if you think about it, you probably look at that and you, you know, you know what I mean? You know, you know, if it were £500 and £1,000, you'd probably go, "Nah, it's twice the price." If it were £800 and £1,000, I think, you know, you go. So it's an impossible decision to make unless you have these pieces of information. I think I just think I just find it very interesting because what's so fascinating is that like that, like that junction, we know something's wrong, maybe, but we don't really know what. And that's because our elephant is uncomfortable when being asked to make choices that it doesn't like.

Um, I'll come on to another thing which I think is one of the most important concepts. I've talked a little bit about not about removal of negatives rather than just the addition of positives as being a really important part of marketing. How many people know about the, the "Save More Tomorrow" mortgage? Shlomo Benartzi and Richard Thaler. They realize that mortgages, um, basically, I'll go even further than this. I don't, I think if you got the world's most brilliant psychologists in a room and asked them to design the worst possible financial product for the human brain, they'd come up with a pension. Actually, the current pension, if you're 25 years old, someone comes to you and says, "Here's a good idea. You pay me £300 now and £300, uh, every other month, every next month, and then I might give you some money when you're unimaginably old." Right? And the way you can't see that money, the only reason you need the government to subsidize pensions to the tune they do at a cost of something like, what is it, £50 billion a year, is because the pension is designed so badly in the first place that without the government tax relief on pensions, no one will get them at all, right? So all you're doing is using billions of pounds of taxpayers' money to overcome the fact that the product has been designed by morons.

Now, in Canada, they've apparently designed a pension which works quite interestingly, which is you put the money in, and first of all, it's an instant access savings account, and then when that reaches a certain level, it tips over into a medium-term savings account, and then it tips over into a kind of pension thing. But you can get a certain amount of the money back at any time. And then when it goes into your pension thing, it gets all the tax benefits that it would have done had it been in a pension all along. Do you see what I mean? Now, what these people did is they did a very clever thing. Instead of asking, they said, "People are more high-minded about their future earnings than they are about their present earnings and expenditure." Secondly, "We don't view a loss the same way as we view a reduced gain. We hate losses, whereas we don't mind reduced gains." So the way they designed a pension which got twice as many people signing up and on average saving twice as much, each of those people saving twice as much, um, the system was, you chose 20%. Someone signed you up for a pension, you chose, let's say, 20%. I'll explain what that means in a second. Um, nothing happened. You, you signed up to the thing. Nothing happened. It, nothing happened the next month, the months after that, probably nothing happened. But every time you got a pay rise, 20% of the increment went into your pension rather than into your pocket.

Now, the vital thing to understand that as a blow is, there is a huge difference in the, to an economist, there is no difference between getting less richer and getting poorer. To the human brain, there is a monumental, for good evolutionary reasons, a monumental difference. Why? We don't mind getting less richer, but we really mind getting actually poorer. And so that was just an example of how really good psychological design makes a thing attractive to twice as many people, and for each of those people, it's twice as good as getting them to save. Thaler partly explains this. This is one of his brilliant experiments that he says, "I think everybody in advertising should know this." So, so you're lucky to know this.

Um, um, he asked, he asked his students an experiment. You're sitting on a beach and you're looking at the waves, and you've been there for a few hours, and you're pretty damn thirsty. And your friend says, "I'm just off to get you a beer from X. Tell me how much you're prepared to pay for a beer from X." And if the price, it's a bottle of beer from the fridge, okay? If the price is lower than the price you quote, I'll buy it. And if it's higher, I won't. What Thaler found fascinating was the price people, I've doubled these because the experiment was done in the 1980s, so this is not really scientific, but I just wanted to make it clear. If, if the place X was a boutique hotel for a bottle of chilled Heineken, people were prepared to pay $5.30. If it's like a beach shack, they're prepared to pay $3. Now, that doesn't make any sense in economic terms. A bottle of chilled Heineken is a bottle of chilled Heineken. Why should we care where it's bought from? And why should we be prepared to pay, boutique hotel? The reason is that we have an innate sense of fairness, and we accept that a boutique hotel has higher overheads generally, and therefore we regard a fair price from the boutique hotel as significantly higher than a fair price from this place, even though our utility value of consuming our acquisition value, sorry, our acquisition utility or acquisition value of consuming the beer is identical in both cases. The amount we, it's a hugely significant experiment that. And what Thaler surmises is that actually, any purchase involves two mental things. There's acquisition value, which is the value you get from the thing, having a bottle of cold beer on the beach, okay? But there's also transaction utility, which is how it feels to make that particular transaction.

Now, once you understand those two things are different, an awful lot of marketing activity makes a huge amount of sense. There are really, really good products that are difficult that we just don't like buying, okay? And there are really products which feel great to buy. So you, you've all got a pair of shoes that you bought in the sales because they were reduced from some huge amount to a much lower amount. You've never worn them, you don't like them very much, but it felt great buying them, right? Equally, there are things which are the other way around, which actually we really, really like, but we hate buying. Now, why you know this, by the way, is most of the things people buy people for Christmas are those things that they have a high, um, acquisition utility, but a low transaction utility. So if you think about it, people like owning really expensive perfume, but they feel guilty buying it. Blokes, I think, feel a big, you know, things like iPads. Blokes all want them, but they feel a bit guilty spending the money on them. And so why those things crop up at Christmas is because they're things that everybody wants to own, but which people find painful to buy. And, you know, if you think of free gifts, it's often the same thing. It's something that you really, really like, but you feel was difficult to justify to your wife or your husband or whatever. You know, it might be embarrassing to buy, it might be guilt.

Now, I'll give you a wonderful example of this. I was talking to the person who ran the Boots Advantage program. And the interesting thing about those points is every single Boots Advantage point is worth a penny. In fact, it's worth slightly less than a penny because you can only spend it in Boots, so you can't use it to buy a car or a speedboat or something like that, okay? But they're worth a penny. However, to the people who've collected them, they're worth more than a penny. You're going to go, "What the hell is this guy talking about? How can a Boots Advantage point, which you can redeem for a penny in Boots, be worth more than one penny to the person who owns it?" Let me explain what happens. I was talking to the woman who ran the program for 10 years. Um, typically, the Boots Advantage card holder is female because Boots, I mean, is largely baffling if you're a bloke, to be absolutely honest. There are lots of words like, you know, hair care and stuff which we don't really use. You know, I've always argued that Boots would have better choice architecture if you just had a shelf that just said, "Stuff blokes might conceivably buy," and you put all the toothpaste and shaving stuff there, and then the rest of the store, we could totally ignore, okay? But they don't do that. So bloke blokes find Boots slightly alarming, frankly. And, um, but the one, and but the interesting thing is, what, so what happens is, person, wife X, all year goes off to Boots, buys nappies, essentials, and things. And towards the end of the year, she's accumulated, let's say, um, let's say, um, 2,000 Advantage points, worth £20, £20 maybe, it's 2,000, maybe it's a bit more, 2,500, okay? £25 worth of Advantage points. And then one evening, she needs a bit of extra stuff like some more loo roll and some other things, okay? And she sends her husband off to the store. This happens all the time, apparently, okay? And she says, "Just as her husband's leaving, she says, 'Oh, by the way, take my advantage card so you can collect the points. Look, here's the money.'" And takes her advantage card. Bloke goes and buys all these things, loo roll, nappies, etc., gets to the till, and the woman behind the till says, "You know, you can buy all this on your points." And the bloke goes, "Brilliant!" Okay? So he buys it all on points and gets home, and wife is furious. Right? What is going on there? The value of points is that you can spend points guiltlessly on extravagances. If you spend money on an extravagant thing, you think, "This is money I could use to actually, like, pay the gas bill, and this is money I could use to, like, feed my children and, you know, pay off the mortgage." Okay? And if I spend that money on a bottle of Chanel No. 19, I feel a bit guilty. Whereas if I buy that same bottle of Chanel No. 19 on points, it's okay because it's points, okay? And that genuinely, you know, so weirdly, one of the interesting psychological things is, in, in the early days when I worked in advertising, you had really expensive company cars. It's a brilliant idea because they gave you the kind of car you'd feel too guilty to buy with your own money. People always say, "No, it's economically irrational. We should offer people the cash alternative." No, don't do that. It ruins it. Now, I want you to give me an Aston Martin and insist that I drive it. You see, because I can never buy one because, you know, my wife would get angry and I'd feel bad, okay? But if my company insists I have one, I'm actually much happier. So the benefits of free will are grossly overrated, in my opinion.

You know, but I mean, the point is that, you know, an Aston Martin for me has an enormous acquisition value, but the transaction, uh, utility, you know, the pain of actually buying it would be too great to go through. So the only way, so there's some extraordinary cases where, if you think about it, people can be very, very seduced if you buy them something they've always wanted but that they've never felt able to buy. And one that distinction is really, really important because I think there are a lot of really, really great products out there. I think, by the way, they're terrible products that people buy, precisely the opposite. I mean, Bernie Madoff would have been the perfect person, you know, if you think about absolutely appalling financial products. But Bernie, his marketing was sheer genius, you know? People will go to him and say, "Please, can I invest with you?" Bernie says, "No, I'm full up this year." All that kind of stuff he did, which was really, really clever. But the products were awful. But I think there are also great products, um, that exist in the marketplace which people don't buy. And it's nothing to do with the product or the value they get from owning it. It's just that they haven't found a way to feel okay about buying it. And there's, you know, there's a lot of weird stuff that goes on. I mean, people, I mean, the weird stuff I've bought that I regret buying, I bought at airports, and it was really just, you know, I feel like a treat, I deserve a treat to cheer myself up after this grueling trip. Oh, look, there's one of those. And, you know, the stuff I really regret buying is often bought at particular locations where you're just desperate for a kind of, you know, a bit of transaction value.

The other thing I can't understand is, um, in, basically, unconscious inference. The elephant is going around all the time making deductions from things that you're not even aware of. The elephant, the workings of the elephant are opaque to introspection. Has anybody been to Dum, really good Indian restaurant? Fantastic, right? Okay. Now, one of the interesting things is, I, I went there and we had this chat, and I said, "How have you managed to find the Holy Grail of Indian restaurants, Dum, which is that you're actually really busy at lunchtime?" Because lots of people go to Indian restaurants in the evening. Lunchtime, they're basically dead, aren't they? Okay? And we said, "I think the answer lies with the elephant in part." And we had an interesting discussion. And the first one is, they open for breakfast. Now, what that means is at sort of 11, 11:30, 12, 12:30, there are people milling around in the restaurant, so you feel it's okay to go in because there are already people there. People are terrified of being the first customer in a store because you place yourself in a kind of state of social obligation by going in that you can't walk out again. And I had to advise a jewelry store. I said, "Very simple. When your store's empty, you've got four staff there, make three of them pretend to be customers because then other people will come in. Whereas I don't want to be the only person in a jewelry store because I'll get oversold." So the first thing is, they open for breakfast. The second thing is, if you go on Tripadvisor and look for Indian restaurant with outdoor dining, you find almost nothing. They've got tables and chairs on the pavement. The elephant just says, "Oh, look, tables and chairs. It's a suitable place to go for lunch." That's the kind of inference the elephant makes. Goes, "Oh, look, furniture on the pavement, kind of like a cafe. I'll go there to eat." Now, understanding, I think there are businesses that do things which are insanely off-putting because they get the elephant wrong. They're not aware they're doing it. I have one business near me which, um, they put so much in the windows that you couldn't see that it was open. It was a shop that was open till 10:00 at night, but it was on a busy road, and they put so many posters in the window saying "Much Cheapness, Pencil £1.99" that no one knew it was open, okay? Because the elephant doesn't look for a sign that says "Open." It just goes, "Lights on, place is open."

My local coffee shop went bust under three different owners. It's now done very well. It's on the same busy road. It largely succeeded because they put tables and chairs out on the pavement. Someone driving in a car, the elephant just goes, "Tables and chairs. It's open." Because if you were closed, you'd lock your tables and chairs away to stop drunk people throwing them about. And we make those inferences without being conscious of the intelligence behind them. I think that's a really interesting thing. We're really, really clever at spotting things like this, but we're not aware of the mental processes that go on. I think you can genuinely redesign the world, uh, to make it elephant-friendly. If you want people to finish the antibiotics, don't give them 24 white pills, give them 18 white pills and six blue ones and simply say, "When you finish the white pills, take the blue ones." People will finish the antibiotics. That's just how it works. I don't know why the elephant works that way. If you put little chunks and little, uh, little things to achieve halfway through a task, okay, you're more likely to finish the whole task.

Um, anybody here use Google Mail, Gmail, for work? It does that really annoying thing where the number of mails isn't the number of mails you've got, it's the number of unread mails. And you go, "Brilliant, I've only got 12 to do, and then I can go to bed." So you plow through all 12, and then suddenly it says, "And you've got 23 red ones." Okay? Now, that's a kind of, you know, that's a kind of chunking where if you give people a task to do that is broken up into different things, the likelihood they'll complete it will be much, much higher.

The other final point, you may have seen this. B, I better go now because otherwise I'm going to keep you people hideously late. Um, we don't, the vital thing is we don't, it's not in our evolutionary interests to be objective. We don't perceive the world objectively at all. The way we perceive the world is the way we perceive those, um, tables and chairs. The way we've evolved to look at the world is to make smart bets on what is likely and probable given the limited information we have available. So we have a kind of betting gambler's mentality, not a kind of, "I will now find perfect information and I will act upon it." That's the way our senses work. The reason being, bluntly speaking, evolution doesn't give a damn about truth. It cares about fitness, and it is not to your advantage in terms of relative fitness to wait for perfect information before you decide. The best thing to do is to actually take a bit of a bet on what information you have available right now. And our whole sensory mechanism actually works this way. It takes bets.

Here we go. At any one moment, we are being bombarded by sensory information. Our brains do a remarkable job of making sense of it all. It seems easy enough to separate the sounds we hear from the sights we see, but there is one illusion that reveals this isn't always the case. Ba ba ba. Have a look at this. What do you hear? But look what happens when we change the picture. Ba ba ba ba ba ba ba. And yet the sound hasn't changed. In every clip, you are only ever hearing "bar" with a "b." It's an illusion known as the McGurk effect. Take another look. Concentrate first on the right of the screen. Now to the left of the screen. The illusion occurs because what you are seeing clashes with what you are hearing. In the illusion, um, what we see overrides what we hear. So, um, the mouth movements we see as we look at a face can. What you're doing is making a bet. The brain goes, "This is a bit weird. I'm hearing 'bar' but I'm seeing 'far'." Let's take a punt on this. It's more likely that I've misheard than I've misseen. So we'll just overwrite, kind of error correction, a "b" with an "f." What's even weirder, I've noticed, is if I look away when the guy's saying "bar," I continue to hear him say "bar." If I look away when he's saying "far," I continue to hear "far" because the brain is just making a bet there, which is, it's probably likely that he's saying the same thing he was saying just now when I could see him. And most of our sensory information, we basically take punts on.

This is the brain making its bet. You know those funny things where, um, according to the shading, something appears to be sticking out of the page or bent in concave or convex? The way the brain makes a bet on whether it's concave or convex is it just goes, "That one's dark at the top, and that one's dark at the bottom." Okay? Dark at the top probably means it's concave because it's more likely that any light source comes from above than from below. And so that's why I'd assume some of them are outy, sticky-outy, and some of them are sticky-inny, because it basically goes, "On average, light generally comes from above." But we don't know whether this is, by the way, evolutionary, whether it's innate or whether it's learned, because in order to find out, you'd have to bring babies up in a locked room with a brightly lit floor, which would be a bit cruel. Um, okay, okay. But, but the, the brain basically goes, "Look, I'll make a punt on this. The reason for that shading, given that the light's probably coming from above, is that it's sticking out versus, you know, concave and, and sticking in." And that's a bet. That's all we do. We, we basically. Now, if you take this one, if you hold your fingers up to cover this area between the two things, you think it's a gray button and a white button kind of joined together. If you cover up the middle, what you'll actually see is the top and the bottom are exactly the same color again. We're just, it's just the brain taking a punt. Pareidolia, where you see faces in things, is the brain just taking.

Now, the interesting thing with pareidolia, by the way, is, uh, facial recognition software also suffers from pareidolia because you've got to calibrate it at a certain level of sensitivity. And if you calibrate it so it never sees a face where there isn't one, it will also fail to see lots of faces where there is one. So you have to get a balance in the calibration between oversensitivity and undersensitivity. So computer software is no different from the human brain in that the optimal level is to make some false positives to avoid making too many false negatives. And so to actually to call this an optical illusion is a bit dubious. It's actually your brain being really flaming clever. So to call it an illusion as if you're fooled is a bit dubious. So that's the first point. The way we perceive things, the way we perceive price, color, taste, is not objective.

A massive effect on your perception of wine or alcoholic drink will be the quality of the glass. If you pour wine from a heavier bottle, it tastes better than if you pour it from a light bottle. Analgesics and painkillers are more effective if they are branded, and they're also more effective if you tell people they're expensive. What the brain does is it reacts in accordance with a kind of expectation. It goes, "Well, there's a likelihood that a heavy bottle of wine contains high-quality wine, so I will then have heightened awareness for the quality." Now, whether you can call that actually a hack is a bit dubious because it actually makes the wine more, more enjoyable, and it makes the painkiller more effective.

So the interesting question is, is that if you, the way Ludwig von Mises described, he said there's no sensible distinction to be made between the value created in a restaurant by the man who cooks the food and the value created by the man who sweeps the floor. And by that he meant by the man who sweeps the floor, marketing and advertising. And his point is that to produce great food, it's not nearly necessary to produce great food, you also have to produce an environment, setting, and context in which it's possible to enjoy it. If you cook the best food in the world, and you serve it in a restaurant that smells a bit of sewage and has flaking paint, okay, then you haven't created any value. And part of the job of manufacturing is to manufacture the context in which the thing can actually achieve subjective value. I, I don't, because we like to believe there's an objective part of a product, which is the bit that's good or bad, and then there's the fluffy magic dust. But the brain doesn't work like that. You know, the brand of something affects our enjoyment of that thing regardless of the innate qualities of the thing because we don't perceive things objectively. We will perceive exactly the same thing differently. The sound of a door closing, basically, is a different thing if it's on a BMW versus if it's on a Hyundai, even if the noise is objectively the same in many cases. So the fact that our brain is effectively taking what information is available and effectively trying to weave it into a plausible narrative explains why marketing actually is hugely important.

Now, I'll give you an example. You, you go, "Well, why would you serve fantastic food in a restaurant that smelled of?" Okay, well, actually, I'll give you an example of exactly this. Royal Mail, about 10 years ago, said, "At the moment, 98% of our, or 97% of our first-class mail is arriving the next day, and think it should be higher. We're going to get it up to 99. We're going to get, we're going to get this up from 97% to 99." And the effort virtually killed the organization. Now, it was absolutely destructive. Now, the questions to ask from a human psychological perspective are: One, how is anybody going to bloody notice? Because most of us don't post 100 letters to ourselves every day from different points in the country to count how many arrive, right? But the second, more important question, if you think about it, was that was a case of a restaurant serving Michelin food. It's already very good, the Royal Mail, okay? But the restaurant smelled of poo. Why did the restaurant smell of poo? Because if you ask people, "What percentage of first-class mail do you think arrives the next day?" the answers varied from 40 to 70%. The reality was 98%. Well, if your reality is much better than your bloody perception, what the hell are you doing trying to improve the reality? Bring the perception in line. And they actually did this with one very good campaign. A boring campaign would have said, in System 2 language, "97% of first-class mail arrives the next day." The System 1 campaign said, "The Royal Mail has a better rate of of reliability than the German Post Office." Now, if your System 1, if you're British, you got to understand this, basically better than the Germans, okay, is good enough, okay? Because we see them as ally-retentive, obsessive people who are obsessed with distribution and logistics. So if we're doing better than them, then that's fine. But also make the distinction. The value of a stamp does not depend on the statistical probability that it arrives the next day or the letter it's attached to arrives the next day. The reason I pay more for a stamp is because of my expectation that it will arrive the next day. So if my expectation is 40 to 70%, it's not a valuable stamp. And that's, that's that. Von Mises' point that all central points of Austrian school economics, all value is subjective, and therefore changing the subjective valuation of a thing by reframing it or whatever is perfectly valid as a form of value creation, puts marketing in a really elevated position. By contrast, neoclassical economics thinks that things have an absolute value, and that all human interchange, exchange, takes place in an atmosphere of perfect information and perfect trust. Well, of course, if you had perfect information and perfect trust, you wouldn't need marketing at all. But you, the fortunate thing for us is that it never happens.

Now, other things which the elephant cares about, but which the rider doesn't talk about very much, okay? Status, certainty, autonomy, relatedness, fairness. This is a brilliant thing. You could always phone for a taxi.

Mean in objective terms, in terms of the time you have to wait, all the other stuff, there's not much difference between Halo and Uber and phoning for Radio cabs, right? You would always phone for one. The real difference is that when you previously phoned for a cab for the next 5 to 10 minutes, you occupied the Twilight Zone of fear and uncertainty. Is he here yet? Maybe he hasn't turned up. They said 10 minutes, why isn't he outside? What's going on? Has he got lost? Can you go in the street and see if he's parked around the corner? Okay, that was a state of mind, mentally. Now, the objective. The food in the restaurant hasn't changed, but the restaurant has had the, the floor has had a really good sweeping because now I watch him approach on a map in real time, and my state of mind is totally transformed because I go, "Oh, look, he's stuck at those traffic lights. Never mind, I'll make myself a cup of tea."

The best thing London Underground did, uh, to improve passenger satisfaction per pound spent, uh, was not faster, more frequent, more comfortable trains, even. It was to put dot matrix signs on the platform. That's sweeping the floor, not changing the food. The food was the trains that were already running. The floor sweeping was the simple psychological insight that we'd rather wait 8 minutes for a train knowing it's 8 minutes, than wait 5 minutes for a train in a state of uncertainty. And that's, you know, information has a huge effect on behavior and, you know, a huge effect on your, on your mental state.

One of the things I said to British Airways is actually, I said, "When you've got a delayed flight, I said, put 'delayed 9 minutes'." Right? Okay. I said, "You just make it up." I said, "But whatever you do, never put the word 'delayed' on its own." Right? "Delayed 80 minutes." You go, "Oh, I'll go shopping. Whatever." You know, "I'll go to the lounge. I'll have a drink." Right? "I'll get my laptop out." "Delayed" on its own drives you practically insane because it's information that something bad and annoying is going to happen, but without enough information to enable you to react to that bad and annoying thing in any certain or sensible way.

Um, fairness. We have an innate sense of fairness and justice, which actually predates, uh, Homo sapiens. So, the final experiment that I want to mention to you is our fairness study. Uh, and so this, this became a very famous study, and there's now many more because after we did this about 10 years ago, uh, it became very well known. And we did that originally with capuchin monkeys. And I'm going to show you the first experiment that we did. It has now been done with dogs and with birds and with chimpanzees, um, but we brought, we started out with capuchin monkeys.

So, what we did is we put two capuchin monkeys side by side. Again, these animals, they live in a group, they know each other. We take them out of the group, put them in a test chamber, and there's a very simple task that they need to do. And if you give both of them cucumber for the task, the two monkeys side by side, they're perfectly willing to do this 25 times in a row. So, cucumber, even though it's really only water in my opinion, but cucumber is perfectly fine for them. Now, if you give the partner grapes, the food preferences of my capuchin monkeys correspond exactly with the prices in the supermarket. And so if you give them grapes, that's a far better food, uh, then you create inequity between them. So that's the experiment we did.

Recently, we videotaped it with new monkeys who had never done the task, thinking that maybe they would have a stronger reaction, and that turned out to be right. The one on the left is the monkey who gets cucumber. The one on the right is the one who gets grapes. The one who gets cucumber, note that the first piece of cucumber is perfectly fine. The first piece, she eats. And then she sees the other one getting grapes, and you will see what happens. So, she gives a rock to us. That's the task, and we give her a piece of cucumber, and she eats it. The other one needs to give a rock to us, and that's what she does, and she gets a grape, and she eats it. The other one sees that. She gives a rock to us. Now gets again cucumber. [Applause] [Laughter] She tests the rock now against the wall. She needs to give it to us, and she gets cucumber again. [Music] So, this is basically the Wall Street protest that you see here now.

One final thing, which is particularly valuable if I'm talking to anybody in the advertising industry, is something we noticed which I didn't realize there was a name for it. We coined our own sort of silly phrase, okay? And then we realized there was a serious name for it. Um, and it explains why it's difficult to sell brave creative work. And let me explain. The central mentality of the human brain when making a decision unconsciously, instinctively, is not, "What is the best decision I can make for all concerned in my role as marketing director, etc., etc.?" The central System 1 elephant question is, "In the worst-case scenario, what is the extent to which I will look like an idiot?" Okay? So, it's actually in game theory, any game, it's "minimax." In other words, words, I want to choose instinctively the route where the worst possible scenario is least negative for me personally.

And the way we discovered this is British Airways runs flights from London City to JFK on an Airbus A317. Any plane geeks here? I think it's an Airbus A317. It has 32 business class seats and no other seats. All business class service. Um, you board at London City, and basically, boarding time, you can check in about 10 minutes before the plane leaves, and boarding time is, "Oh, look, let's go to the plane. Oh, we're all on. Close the doors. Let's go." Right? You don't have to wait for like 300 backpackers to start putting luggage up in. And it's flaring fantastic. I've been on it myself once, and, um, it's brilliant, okay? And our question was, given this exists and is so unbelievably cool, how come there isn't a waiting list? How are we having trouble filling it?

Well, you wouldn't, if all people were interested in was the acquisition utility of going on this service, it would be insane popular. With, apart from anything else, anybody lives in Kent, anybody lives in Essex, anybody who works in the city, anybody works in Canary Wharf, etc., okay? The problem is interesting because it's a business class service only. All business class flights, pretty much, are booked by an intermediary. You have a corporate travel agent, and you have a PA or assistant. Their interests are slightly different, which is, "I want to do something that won't get me shouted at." Right?

Now, think about it from the point of view of a travel agent, or think about it from the point of view of your PA who's booking you a flight to New York. Right? If he or she does something really boring like books you out of Heathrow, all right, and something goes wrong, the flight's delayed by two hours, it's British Airways' fault because I've just done the boring default thing, and now British Airways have cocked up. If I put my head above the parapet as a PA or travel agent and recommend you fly out of London City, which is the eccentric option, and something goes wrong, it's my fault. If I go with the default, it's like, "No one ever got fired for buying IBM." Do you see what I mean? If I go with the default and something goes wrong, I'm unlucky. If I go with a brave decision and something goes wrong, I'm an idiot. Do you see how that bias affects all decision-making?

Now, just to prove that it's not just on airline bias, um, penalty taking at the World Cup. The best thing statistically for quite a few penalty takers to do, because in about 90% of all occasions, the goalkeeper dives right or dives left, generally at random, okay, very often. And this would be true for at least two people out of five taking a penalty kick. The best thing to do is to, well, the ball right down the middle with also a lower chance of missing, usually, okay? Very, very few people do that. Similarly, in many cases, the best thing for a goalkeeper to do might be occasionally to stay in the middle, but they never do. They always dive left or they dive right. What's going on?

The reason is that very simple. If you shoot the ball left and the goalkeeper dives left and saves it, you're unlucky. Shoot the ball right and the goalkeeper dives right and saves it, you're unlucky. Shoot the ball down the middle, you're more likely to score. But if the goalkeeper saves it, you look like a total fool because you couldn't even be bothered to go to one side or the other for the goalkeeper. Similarly, if he dives and the ball goes somewhere else, he's unlucky. If he stands still and the ball goes either side of him, he looks lazy and useless. And so that's called defensive decision-making, okay? And it's an innate thing. It affects everything from medical, massive, massive inefficiencies in the health service and in medicine where people refer people for unnecessary treatments or diagnoses or tests because they go, "Well, if I pass you on to a consultant, I can't get sued." Okay? "If I do nothing, you might sue me." And in, in American medicine, it's terrible. A massive problem in, in the public sector, which is, "If I do a boring thing, I will never lose my job." And it causes people to go with the default far more often.

The only way you could overcome this problem is actually for the manager to say, "Player number two and player number five, I want you to wellie the ball right into the middle." And then they have the defense. If the ball is saved, "I was obeying the instructions of my manager who told two of us to do this." But if you leave it to them to make the decision, none of them will wellie it in the middle. They'll go either side. And so that's just a really, really interesting example of a behavioral bias. It comes up in this book, which I say is one of the best business books of the year, Risk Savvy by Gerd Gigerenzer. Really recommend reading it. It's a fantastic and fascinating book.

Um, I also, by the way, when talking to Diageo, made this point about defensive decision-making. It's a very simple lesson. You will never sell men mixed drinks unless you have a menu which contains a photograph of the drink prepared in the glass in which it will be served. No man will ever order a cocktail, not because he doesn't like them, but because there's a 3% chance that this arrives. And the possibility of that total humiliation, however remote it is statistically, is basically the dominant factor in him saying, "I'll have a pint of Stella." Okay? Because the sheer terror of that possibility. It's very interesting. Hipsters are serving cocktails in mason jars. It's quite an interesting thing because then it makes it all about the drink, not about the terror of how you'll look drinking it, which I think is a very, very interesting psychological experiment.

Few more things. Anybody who wants to go, by the way, I'm very conscious that I've overrun. So, if anybody either needs to go to the loo or is running late, please don't feel you're doing anything naughty by leaving, okay? Um, the, the final thing I'll talk about is we don't really understand trust, even though it's the bedrock on which all economic activity rests. And that's because trust is an elephant process. It's not a rational process. Most of the things that determine whether we trust someone or not, or whether we trust a brand or not, or a company or not, are actually instinctive.

Why do banks always build marble buildings with big oak halls? Instinctively, we don't actually do this calculation. You see, we don't actually think of it consciously. We go, "Well, I'm just depositing £10,000 here, and my slight fear if I lend £10,000 to a random bloke in the pub is that he'll disappear and leave town." Okay? Someone who spends a fortune on an immovable asset with a high upfront cost automatically conveys to us trust because I feel that that guy's not planning to move anywhere in the short term. If he were, he wouldn't have invested all this upfront money on something that doesn't move.

Um, it fascinated me the extent to which our trust instincts are very, very clever, even though they're entirely unconscious. Um, an example would be four years after I left university, about three of us, all contemporaries, could just about afford our first crappy secondhand car. And secondhand cars are cheaper in London than anywhere else, actually. None of us bought a secondhand car in London. We all instinctively did the same thing without knowing why. We went back to the town where we'd grown up and bought a secondhand car from someone vaguely known to our dad. Now, that's quite clever because the guy we bought it from might sell dodgy cars, but he's not going to sell a dodgy car to the son of someone who drinks in the same pub that he does because my father, who's staying around in Molesey, has the prospect of doing him reputational damage. So instinctively, we understand these reputational currencies, even though we never talk about them, we never measure them.

A reason why people prefer and pay a premium for brands is that someone who invests a lot of money in their reputation has more to lose from selling a bad product than someone who hasn't got much of a reputation to begin with. They've got an upfront commitment. You know, if you think about it, for Phillips or Sony to sell 50,000 shitty TVs, um, is actually quite a worry because they will erode a huge amount of the billions of brand equity that enables them to charge a premium for TVs if their reputation falls. If you buy a TV cook from Dasu, who you've never heard of, the same reputational, you know, we like to, we, we basically trust people we can punish a bit, or who stand to lose from ripping us off. If you don't have a brand and you don't have a reputation, um, then you've got nothing to lose by ripping your customers off. We trust people who have something to lose, who have skin in the game.

The other thing is advertising is a reason to buy products because the fact that someone spends money on advertising is proof that the manufacturer confidently believes that their product will be widely and repeatedly popular because if they didn't believe it was going to be widely and repeatedly popular, they wouldn't spend money advertising it. So the act of advertising and the fact that it's expensive effectively says to us, "Well, I don't know if this product is any good or not, but the person who knows it far better than I do, namely the manufacturer, certainly believes that it's good." In the same way, if you met a racehorse owner and he said, "My horse is definitely going to win on Saturday," you go, "Everybody says that." If you saw him betting £50,000 on his horse to win on Saturday, which is equivalent to the act of advertising, it's an expensive, costly mark of confidence. Any serious gambler would regard that as very significant information indeed. So the fact that people prefer to buy advertised products is not irrational if you understand the fact that the brain is an instinctive game theorist who looks for signs of confidence not only in the product itself but in the person selling it.

So, six ways to trust a taxi driver, just to give an example, okay? Um, the knowledge. A lot of people think we don't need the knowledge anymore because we've got satnav. But maybe the knowledge isn't about memorizing London streets at all. Maybe it's the fact that because it takes you two years of upfront effort, or three or four years, to qualify as a cab driver and get a badge, right? Having sunk all that effort into becoming a cab driver, you're not going to risk your badge for the sake of ripping off an American tourist for 12 quid. Now, at the moment, I agree, black cabs are really expensive in London, but I've got two 13-year-old daughters, and I will happily put them in a black cab with a complete stranger and say to the guy, "Take them here." Okay? Would I be willing to do that if you could qualify for a London black cab driver's badge by buying a TomTom and spending three nights at evening classes? Right? The fact that it's so difficult, it's a barrier to entry, is a reason to trust.

Other ways: Addison Lee, they've got a reputation to lose. The knowledge, sunk cost. If you live in a small town, you probably all use the same cab firm all the time, and the reason is that you want them to know that you're quite valuable, so they won't want to rip you off because you've got a credible threat of defection. If you spent £200 quid or £400 quid with a cab firm in the last year, okay? They don't want to piss you off or cheat you because you might have a hissy fit and never use them again. If I spread my taxi fares across all six, seven Oaks taxi firms, none of them really care about me that much because I'm worth £40 quid a year to six of them, which is a totally, I'm a totally boring customer, and therefore my expectation of being able to trust them is reduced. So that's why we're loyal to butchers, you know, the guy isn't going to slip me any horse.

Now, interesting that did happen. There was a taxi firm I used a lot for all my airport jobs, and they said, "We mainly do airport jobs." I said, "That's fine." One day my wife broke down, I couldn't get hold of the usual firm. I rang this firm and said, "I know you normally do airport jobs, but could you just fix, you know, rescue my wife from the side of the road?" They said, "No, we only do airport jobs." You know, given that I use them sort of once or twice a month every single year, I never use them again for anything ever. And that's partly how capitalism is kept honest. It's the threat of retaliation. Longevity, company that's been in business. You know, there's American firms that say "Established 2003." You can't really do that in Britain. It's got to be like 1790 for you to put that up. But Americans will do that, you know, "Hey, we've been going for a decade." Wow, we're like traditional. Okay? But it's not an irrational message. A business that spent time and invested time and effort in securing a reputation is more trustworthy than one that's just started.

And also, religious. Religion is a mechanism for trust. Um, Jesse Bering, who's an atheist anthropologist, says that when he's in a strange country, he always chooses the taxis which have religious icons dangling down from the rearview mirror. He said, "I don't believe in a bloke in, in the sky, but I really like the fact that they do because if they think someone's watching them, they're probably less likely to club me over the head." Okay? You, you don't want Richard Dawkins as a taxi driver and trust the bastard as far as you can throw him.

So, the final point is, there are lots and lots of things in which we can take the advances in, in social science and psychology and meld it with the talent you find within an ad agency and create a much bigger business than we currently have. And I'll just end with this one example. This is my last slide, but one I promise you, okay? I went and said, I said, "Okay, why are you building High Speed 2?" Because it's, I've explained, I think it's silly. They said, "Well, it's partly greater speed, but it's also greater capacity, carrying capacity of the trains to Manchester. We need the capacity." They don't actually, most, you know, Euston is actually quite underused, etc. Anyway, but never mind that. They will claim its capacity. I said, "Well, you can solve that problem for £50,000 quid. You don't need to spend £40 billion." I said, "That's ridiculous." "How can you solve it for £50,000?" I said, "Very easy. Every time I go to Manchester, I book an advance first." Um, which is, you book about four weeks in advance, and you get a first-class ticket for about £30 quid instead of £50 quid instead of three million quid, which is what you pay if you book on the day. All right? So I book an advance first up, and I book an advance first back, and you get a breakfast as well on the, on the thing, okay?

Now, so the problem is, I come in from Kent, and I'm so paranoid about missing my designated train because then I'll lose the ticket I've already bought, that becomes worthless, and I have to buy another ticket, which costs £7,000 to go first from Euston to Manchester. So I can't afford to miss my train. So I leave home massively early and leave a huge margin of error, which means I arrive at Euston about 45 minutes before my train leaves, and I have to hang around at Burger King for the next 30 minutes. I said, "You could solve that problem, and you could reduce my journey time by 40 minutes." All you have to do is give me a mobile phone app where I can say, "I'm at Euston right now," and if there are empty seats in first class on the next train leaving in 5 minutes, they say, "Why not pay us £5 quid and you can jump on an earlier train?" That also increases carrying capacity. Sorry, I don't want to go into the details of yield management, but as EasyJet and Ryanair know, it's always okay to allow someone to occupy an empty seat on an earlier journey because it gives you the possibility then to sell the later seat to somebody else. You should never let people take a later journey than the one they've booked, but letting them pay £5 quid to travel on an earlier journey if there are available seats is really good yield management. So if you allowed people to do that, pay £5 quid to jump on a train one train or two trains earlier, you would have reduced their journey time to Manchester by 40 minutes, and you would have increased carrying capacity because railway seats are a perishable good. If you want to get into the technicality of this, right, and I've got that presented to people now, we still don't succeed. And the reason you don't succeed is nobody believes, because of the butterfly problem, no one believes that small things can have big effects. I think that's a perfectly fact. Does anybody disagree with that as a basic premise that, you know, okay, it doesn't save 40 minutes for everybody, etc., etc., but it will be a huge benefit in terms of reducing wasted time in Burger King, and it would also effectively mean fewer trains left partly empty. The problem you face is nobody believes that small things can have big effects.

And I was talking to Dan Ariely about this, and he said, uh, there's a fundamental human bias in this. And the way he explained it is that quite often you find that drugs that are used for relatively common conditions like asthma also work against rare cancers, but they have a weird problem. If you prescribe the drug, which is an anti-asthma drug, for a rare cancer, people look at the side effects and they go, "No, this can't be a serious cancer drug. It just says 'may cause drowsiness'." That's not a cancer drug. If it says "may cause drowsiness," that's like crap. So when they sell it as an anti-cancer drug, they exaggerate the side effects and say, you know, "May cause cardiac arrest and anal fissures or something," and people get, "Yeah, that's really unpleasant. So that's really going to be a good cancer drug." So the problem there is a fundamental problem that when you solve big problems in very, Zen, simple ways, no one believes you because they genuinely believe that the scale of the downside has to be commensurate with the scale of the achievement. And in a complex system, that simply isn't true. That's the reading list. Thank you very much. Please feel free to leave because I've kept you far too long. If anybody has questions, I'm very, very happy to answer them either on Twitter, where I'm at Rory Sutherland, or just by milling around here afterwards. Thanks ever so much.