Transcript for:
Rory Sutherland on Creativity in Marketing

So, without further ado, let me please ask you to welcome in a big way the living legend that is Rory Sutherland. Thank you. I've got to rush through this. The people in the front three rows, of course, being paying Mad Masters things, you get extra eye contact as well as part of the deal.

Okay, now, I call this selling fast and slow because it occurs to me... that I'm fascinated really with a behavioural science question which is a kind of meta question for everything we do as marketers or creative people which is why are people so unappreciative? of verging on hostile to creative solutions to problems?

That's the really, really interesting question, in fact. Because when you look at what you're doing when you're actually applying creativity to a problem, you're looking at it in different ways and expanding the solution space. You're giving people, almost for free, five different ways they could look at the problem and solve it differently.

And by the law of averages, some of those solutions... will either be more effective than, or cheaper than, or both, than the solutions they've arrived at through purely rational sequential logic. So what's going on with the hostility? Now about a week ago I discovered a book that more or less explains what's going on, because I've been puzzling over this for ages.

So why does creativity irritate people so much? What you've got to understand is there's a fundamental law of nature. And apologies for the people on the course because you've heard this before. Bees have a waggle dance. It tells the other bees where you find pollen, where you find nectar, where you find something else.

I think it's resin they collect as well. And they obey the waggle dance. And they follow instructions which say there's a good source of pollen a mile and a half to the northwest.

I should head off there. and off they go. And it's extremely efficient in the short term and in the long term. Generally, energy recovered is more than energy expended on collection. And the accountants love that bit of the hive, okay?

But bee scientists noticed pretty quickly that 20% of bees seem to ignore the waggle dance. They go off at random. I thought this is pretty weird because you've had 20 million years of bee evolution.

Evolution hates wastage, okay? I mean, drones in the hive, once their job is done... are actually kicked out of the hive and left to die because they have no further value.

They're that ruthless, okay? And it's like GE under Jack Welsh or something. Bees are that ruthless, okay, in disposing of waste in one way. So why do they indulge these random bees? And the answer is, this actually crops up in computer algorithm design.

It's called the explore-exploit trade-off. And there's a trade-off between optimising, narrow optimisation on what you already know for maximum... efficiency, which is actually a fairly simple mathematical formula.

You can tell when you're getting it right. And then in accompaniment to that, there always has to be an open-ended question, which is, what don't we know? Now, the reason there are no beehives where everybody obeys the waggle dance is that over time, they all starve to death. They get trapped in a local maximum.

They get over-optimized on the past. And two other things happen, of course. If you don't have the random bees, you can't get lucky. lucky. There's no exposure to positive upside luck, like some fantastic flowers only 500 yards away in the opposite direction coming into bloom.

You can't benefit from that. You almost certainly go extinct because once your current source of pollen and nectar becomes depleted, you don't know where else to go. And the third thing is you can't grow. You can be perfectly efficient at doing what you already do, but if you don't have exposure to discovery... and upside good fortune, it's impossible to grow.

Hence, this isn't an inefficiency. It's part of the whole process. But looked at through the lens of an obedient bee, the scout bees, as I think they're known, look like a cost. The payoff isn't immediate. The payoff isn't linear.

You need a fundamentally different way of actually accounting for the value of those bees. And it's very easy to look at them... as if they're kind of wasteful. So you can see two things happening.

You either cut them. There are two problems. Okay, it's not a trade-off, okay?

It's actually a feedback mechanism. Discover, exploit. It doesn't work unless you know how to exploit what you've discovered, but it also doesn't work if you purely exploit without discovering.

It's actually a benign feedback loop. It's not a trade-off at all. But there are two things that cause that feedback loop to break. One of which is that you refuse to invest in Explore because you're too busy doing exploit, and you don't understand the point of Explore because you have to meet your next quarterly target.

Okay? Or... The bees that make a discovery come back to the hive and no one fucking listens to them. Okay?

And I think we can see in modern business both of those things happening in parallel. Right? Which is, can we cut back on everything which isn't contributing? to our short-term financial measurable quantifiable goals.

And when these people come in and say there may be actually another way to get pollen, we fucking ignore them. And I think both happen. That's why I think it's so important that we understand why it is. that people in business now view what is a good news message, there may be another way of doing what you're trying to do, and they react with hostility and discomfort rather than gratitude.

By the way, when I say things to people as a suggestion, I'm totally happy for them to go, that's interesting, Rory, it's a different way of looking at things, but it's shit, and here's why. I don't mind. In fact, 90% of what I say should probably be met with exactly that reaction.

Because context... Textually, a lot of the things I suggest probably don't work or are stupid. But what I don't understand is you get genuine hostility. And I've got a few examples of this coming up. Most business questions, in fact, are actually really open-ended questions.

They are not questions with a single right answer. They're not like high school maths questions. Most business effort, I think, is spent pretending the opposite. It's spent taking open-ended questions which leave room for creative exploration and interpretation and pretending.

their high school maths questions. We've got these three metrics, and we've got to optimize them. In other words, you reduce them to something which has a single right answer. Now, let's take high speed, too. You give the brief to a load of engineers.

They define the problem in terms of a couple of variables which are capable of mathematical expression and optimization. Those things are basically speed and capacity, speed, time, and capacity. Now, what would have happened if you'd given that brief to a load of engineers? and the budget to Disney. I'm using them as an example of a genuinely creative organization.

They would have said, well, first of all, the question is wrong because speed isn't necessarily a benefit. I'll come up with examples of that already. Actually, train journeys can be too short.

There's evidence that people actually live further from work than they actually could because they prefer to have a kind of mental buffer between the two spaces. It's me time. You're commuting time. I'll come to that later. Disney would have said the right question to ask is how do you make the journey from London to Manchester by train?

So fucking enjoyable. They probably wouldn't have said that they're clean family organization How do you make the journey from London to Manchester by train? So enjoyable people feel stupid missing that taking the car people feel if they drive to Manchester.

They're actually missing out That's a much more interesting question, which you could solve actually much more economically than by fetishizing speed and capacity Why does that question never get asked? Because it's an open-ended question. It doesn't have a single right answer.

It requires some sort of imagination and subjective judgment to decide on the right solution. Is it champagne on the trains? Is it ball pits on the trains? Do you actually just have really slow trains from Harry Potter going to Manchester, which means that every single fucking tourist who visits London also visits Manchester, okay? There are loads and loads of solutions to that problem.

And I'm saying that's why people don't like it. like it. They're much more comfortable in an institutional decision-making setting making a model which has a single notionally right answer which is actually a bit crap because you don't get blamed for it. If you can practice what I call effectively performative rationality, you can give the illusion that you've arrived at your recommendation through a series of data-driven rational sequential logical steps.

Okay. It may be a a pretty crap answer to the question, but you won't get into any shit. OK? Now, this is basically what everybody tries to do. And we are doing it all the time, partly to avoid blame, I think, partly to win arguments.

We're pretending that complex questions, which submit to genuine creativity, redefinition, exploration, we're pretending they're mathematical optimization problems. And if you want a discipline that does this dishonestly to an absolutely hideous degree, it's mainstream neoliberal economics. It basically takes every problem, takes a couple of mathematical factors, usually something like price, Okay, and basically says here's your answer By definition the creative opportunity cost of pretending that questions are these kind of questions is absolutely immense This is the book I mentioned which you must all read because I think it has massive explanatory power in effectively say that people are effectively using rationality, processes, computerization, they'll soon be using AI, but they're using these things effectively to design processes which absolve them of blame.

Effectively, a large part of scaling up requires these things he calls unaccountability sinks. But they're effectively the kind of decision that doesn't look like a decision because you've pretended it's a completely rational, inevitable course of action, okay? And as a consequence of doing that, you've destroyed any possibility for more creative, more interesting modes of approach. The truth is, in real life, there's never a single right answer.

answer, there's more than one way to be right, okay? And one of the things is that, by the way, sometimes, as I say, the opposite of a good idea is another good idea. And someone said, I'm always quoting you saying that, and I go, actually, I have to confess, not me, it's actually Niels Bohr in a conversation with Einstein, was the first person.

Bohr said, in boring physics, the opposite of a good idea is wrong. In kind of high complexity physics, the opposite of a good idea can be another good idea. It's also true, not only in high complexity physics.

It's true in psychology. Sometimes we want to do things fast, okay? Now, virtually everything in our modern world is being optimised to save time.

But sometimes we like to do things slow, okay? The actual time consumed in the pursuit of this activity contributes to the meaning, enjoyment of the activity. Now, it is impossible to have a single optimisation model where people might respond in two completely different ways. different ways to the same thing. And so as a result, the opposite always gets completely frozen out.

I make this point time and time again. I had the other side, actually, the perfect example. I apologise to people who have to listen to this for the second time. I was going on stage, and like all people who go on stage, you go and have a piss five minutes beforehand. It's a nervous thing, OK?

And some bloody prankster had swizzled the nozzle on the wash basin so it was pointing upwards. And so as soon as I went to wash my hands, it deposited a vast quantity of water on the wash basin. on the front of my beige trousers, okay?

Now, to be honest, you could probably get away with that in your 20s, in your 50s, that's not a good look, okay? So I tried the hand dryer, absolutely no dice, I would have had to barbecue my own genitals in order to actually achieve the desired effect, and I went outside preparing to do a runner, basically, okay? And...

Then a miracle happened. It started to rain really, really heavily. The way it does in London very, very rarely. And I walked out into the street and just stood in the rain for five minutes until I was completely drenched, you see? And admittedly everybody thought that's strange.

She seems to have fallen into a swimming... pool, but it was better than the alternative interpretation. So what I mean is, sometimes in life in psychology, you need to dry the wet patch, but sometimes the answer is to wet the dry patch.

Okay? They both work. A lot of great advertisers. takes a weakness and turns it into a strength.

Good things come to those who wait. This is one of my favorite psychological ideas of all time. It's the Uber map. Don't change the time people spend waiting for a cab, okay?

Improve the quality of the waiting time by adding a high degree of certainty. The thing people really mind in waiting for things, really useful psychological insight, is it's not actually the duration. It's actually not knowing. People would rather basically wait 15 minutes for a train knowing there's a train coming in 15 minutes, than wait six minutes for a train in a state of total uncertainty.

This is the story I was telling you about, where a creative and useful insight was met with hostility. I'm going to be really careful not to name them in case I get them into trouble, but a brilliant person I've worked with who's kind of a psychology person at Transport for London. I'll keep it vague.

I'll keep their gender hidden as well, okay? But they came up with lots and lots of robust research. that showed that people actually quite enjoyed their commute. Now, it's quite interesting.

I think, by the way, it was more male than female, slightly. There was a slight gender bias in it, if I got that right. And I can explain that very easily. Men are like skyboxes. They've got a standby mode, OK?

They just switch off and stare, OK? Okay, what is the sport with the single greatest gender difference in the world? I think it's course fishing.

Course fishing is basically staring with equipment, okay? That's basically it, okay? So men particularly like staring.

staring into space thinking of absolutely nothing. He went and reported this finding and everybody involved in transport planning said, you can't say that. You mustn't say that. Do not reveal that research. And he said, well, why?

It's true. And I've mentioned the fact that people live further away from home than they really should. He said, all our models rely on the assumption that time spent in transit is a disutility.

Therefore, if you introduce evidence that suggests that time spent in transit may be pleasurable, it will mess up our. maths. So you literally have decisions being taken not for the purpose of human benefit, but for the purpose of the mathematical tractability of the model, hence, I think, avoiding accountability, if I go back to that book plug, okay?

It's basically, if we can have a mathematical model, we can escape any kind of intimation that we've used, subjective judgment or imagination, which are dangerous carbon-based subjective things, and if we can suggest ours is a purely silicon-based decision, we're not going to get... to get into any shit. So I work with a brilliant copywriter called Colin Nimick who told me something fascinating. In New York, they speak fast. In the American South, they speak slow.

They are both forms of politeness. In New York, you speak quickly. to recognise the fact that the person you're talking to, their time is valuable and they haven't got much time to spend on you.

And in the South, you speak slowly to respect the other person by intimating that they're worthy of a lot of your time. They're both ways to say the same thing, but in completely opposite ways. And I think these perceptual kind of, what you might call perceptual paradoxes, are all over the place. Now, I'm going to skip this, but... Everything we're doing in marketing at the moment is basically prioritizing speed.

It's fetishizing speed, okay? If you look at things like programmatic marketing, these things are massively weighted towards people who will make short-term purchases very quickly through low-cost media. to a point where that's starting to define who people's customer base is. Your customer base is your potential customers, some of whom may be cheap to convert, some of whom may be expensive to convert, some of whom may be quick to convert, some of whom may be slow to convert.

slow to convert. And what we're doing with programmatic is we've created a ridiculous feedback loop where your propensity to interact on impulse in low-cost channels is effectively what defines your value as a customer, not your actual lifetime value or anything like that. It's merely your readiness to actually produce what you might call high-speed attributable data, justification, very, very rapidly. I would suggest that the people who are easy to sell to are probably potentially in long-term Respects not your most valuable customers.

I'll give you a nice story everything important takes fucking ages We assume when we look back in history. You know Mary what the Montague came up with variolation It was a massive marketing the basic effort there was not the discovery of the thing It was the marketing of the idea She ultimately won because she persuaded the royal family to variolate her their children against smallpox and once the royal family did it, everybody else followed on. It's actually a marketing challenge.

Edward Jenner came up with a much better thing, the cowpox vaccination for smallpox. He basically had to spend 15, 20 years of his life changing people's minds. Changing people's minds is inherently slow. You cannot rush it.

Because of all kinds of factors, the nerds among you will know about the Bass Diffusion Curve. People are driven by habit and social proof, and in the early stages of a new behavior, it neither... It neither has the strength of habit behind it, nor the strength of social proof, and so it's very, very slow in the early days to be adopted.

Electricity. We completely forget the fact that the electrification of the British, and in this case the Irish home, actually required a marketing budget, and actually took about 30 to 40 years. That's an ad basically going, it's a really good idea to have electricity in your home.

about in more detail later. This is one of the most fascinating things because it's a slow, it's a Dixie piece of communication, it's a southern piece of communication in that they wrote handwritten letters to everybody in the street because they wanted, they had triplets, they wanted to move back. back to, what's that place in Derbyshire with the wonky church? I think it was there, anyway, Chesterfield, isn't it?

And they couldn't find a house they wanted. So they found the street they liked and chose the houses they could afford to buy on it, which weren't on the market, and they wrote handwritten letters to all the owners of those houses explaining their circumstances and saying, would you like to sell? About five people, I think, six people got back to them.

They had about five house tours. I think three people made an offer and they bought a house. Now, do you just realize how weird this is? None of those people had their house on the market, okay?

Now, on the strength of a piece of... direct mail, which is a high investment, low speed, slow form of communication. Okay.

Handwritten direct mail is about as slow. It's about as kind of Southern as you might get. Okay.

People who thought their house wasn't for sale. sale decided that it was. That's a massive decision, okay?

You'd think, well, there's obviously no point in trying to persuade people to sell their houses because if they wanted to sell their houses, the house would be on the market. Turns out it's not the case at all. This is a brilliant business called Inkpact, which actually sort of semi-automates the process of producing handwritten letters. The success rate they have is astronomical.

You know, literally response rates which are 10, 15, 20 times higher than you get in any other medium. maybe 100 times higher than you get in digital channels. I would argue that the effort, the expenditure of time, in southern terms, is fundamentally perceived as commensurate with the scale of a decision. Effectively, when we as humans process a message, we sort of process how much effort, time, and love has gone into the creation of this message, and we pay attention to the message accordingly. It's costly signalling of a kind, and this is the ultimate costly signalling.

Actually, shut up. The lady from Inkpac, Charlotte, is it somewhere floating around? Over there, waving. There we go.

It's an extraordinary business. It makes no sense until you realise that the opposite of a good idea is another good idea. I bought one of these.

Get them. They're fantastic. I did have to contend with my wife, who used to work in procurement, and has tedious, benowsic questions like, why are you spending £800 on a tap when you can buy a kettle for £50?

But they said, don't worry, we'll give you a demonstration for your whole family. whole family at our showroom in Manchester. I said, look, I like Manchester and I like cookers, but I'm not going to get...

They said, no, no, it's an online Zoom sales pitch. It lasts about 20 minutes. It's an expensive way of selling a tap.

The point is, I wouldn't have bought a tap otherwise. I wouldn't have bought one of them otherwise. That's the time it takes to make that kind of sale. It's expensive offering people a test drive, but that's the price of selling a car. And we're actually optimising our targeting of messages around those people who happen to be...

to be very susceptible to making impulse decisions in low margin categories or low value categories in inexpensive media. And we're saying, those are our target audience. And then we wonder why the economy isn't growing. Okay? It's like closing down the exploratory bees because you think they're a cost.

It's fundamentally, it looks logical, it looks efficient, but it's catastrophic. I better skip all this because I've got no time left at all. Stop trying to sell as efficiently as possible. possible and let's try selling as effectively as possible and accept that there is a cost in effort and time that's attendant on that. And what we're doing is we're making the fantastic mistake which you can exactly make if you're one of those bees and it's very, very common to make if you're looking at an accountability, an unaccountability machine.

If you're trying to dodge blame, you pretend that efficiency is effectiveness because efficiency has easy maths and effectiveness has complicated. That's basically what's going on here. Now, very finally, rather than trying to change the world, it's a lot cheaper to change how people see the world, because when people see the world differently, they behave differently, and that then changes the world.

I'll end on this, actually, before I think I've run out of time. I always show people this for two reasons. It's really, really interesting, and it might save somebody's life.

Okay? The thing on the outside you're familiar with, that's called a speedometer. It's for the young people who don't bloody drive anymore.

more pathetic. Okay. Okay. But the older people among you will recognize the thing around the outside as a speedometer. We all know that.

10 miles an hour, 20 miles an hour, 30 miles an hour. We're very familiar with that way of presenting speed. The thing around the inside is technically called a paceometer.

And that shows the number of minutes it will take you at that velocity to go 10 miles. It's exactly the same information, right? 10 minutes per 10 miles is 60 miles an hour.

They're exactly the same. That's why you can see that next to 60, there's a 10. I got that right. Because at 60 miles an hour, it takes you 10 minutes to go 10 miles.

But you notice, they're not remotely evenly distributed, these things, around the desk. And And the reason I show you this is physically those speeds are exactly the same. When I present road speed to you in the form of a paceometer, you will very quickly realize something that you've never noticed at all yourself when driving. Maybe you have if you've got a GPS or sat nav.

You've noticed the fact that if you're driving at 70 and you think, ooh, I'm going to be three minutes late for that meeting, and you accelerate to 85, it actually makes fuck all difference to your arrival time. You've noticed that, right? This is really important because going at 30 miles an hour rather than 20 miles an hour over a 10 mile distance, that'll save you 10 minutes, right?

Okay, going at 80, 90 miles an hour rather than 70 miles an hour will save you two minutes. It's not worth it, okay? Now, do you see what I mean? So once you've seen that and you've processed it, and assuming you're going on a five-hour trip, finite journey, okay? Basically, it's a really good idea to go quite quickly, but going really fast is a total idiot's game, okay?

If you're going 90 on the motorway, if you're going 85 on the motorway, okay, basically, you're imposing a huge risk on yourself, a huge risk on other people, you're completely disrupting the flow of traffic, and the net time saving you'll make is so small as to be completely irrelevant. What I'm saying is that the way you present information to people affects how they behave. And it can be the same objective information.

This is mathematically 60 miles an hour, 10 minutes for 10 miles. They're exactly the same speed. A physicist can't tell the difference.

Psychologically, if you present the information in two different ways, it has completely different effects. Now, I realize I've run out now, but this is the point I'm going to make. We are making rationality, in other words, the construction of a rationality. rational optimization argument a requirement for any decision making In other words, it's a prerequisite unless you can build a completely robust business case built on pre-existing data We can't spend any money on this thing.

Okay, it's happening everywhere literally everywhere The problem you have is I think and this is why you need to experiment why you need the rogue Scout bees is Is there are more good ideas we can post-rationalize, then our good idea is we can pre-rationalize. Most of the really interesting business ideas, most of the really interesting marketing ideas of the last 40 years did not make sense in advance. Try making the business case for Red Bull or, if you were Dyson, 750-pound vacuum cleaners. No one would take you seriously at all.

Ultimately, we have to allow, alongside rationality, we have to allow imagination, speculation, instinct, common sense. We have to allow testing the opposite of the thing that we think makes sense. Sometimes speed is a... good. Sometimes speed is stupid because the time spent is exactly what makes it valuable.

Now, here's the interesting question I'll end with. Jeremy Bulmore always said that even if you have no media budget and you're not planning to actually... run any ads.

You should still produce an advertising campaign for your own company. True of very many B2B companies. They're never going to run an ad campaign. They should still produce an ad campaign. And Jeremy Bulmore's argument was the questions you ask, the process you go through, the journey of arriving at that final endpoint is what delivers 80% of the value.

Even if you never run the ads, the process of producing an ad for a company forces you to ask questions like, what are we for? Who do we serve? All those things which... And it forces the company to answer questions, open-ended questions, which it will otherwise duck out of. Because they're open-ended.

You heard what I said earlier. Okay? Now, what I think we'll do with AI is potentially catastrophic.

Many of you, when you were students, if you were students, wrote student essays. Right? I'm fairly confident that none of you, very few of you, have kept them and you certainly haven't reread them.

In fact, I'm confident in saying that nearly all of them have. are completely shit and of no value whatsoever to anybody, even if they survive in your garage anywhere. But that's because the value of the essay wasn't in the essay.

It was in the process you had to go through to write it. It was the effort you had to make, the questions you had to ask, the books you had to read. The effort you had to go through to get to the end point was more valuable than the end point.

It's a bit like a cruise ship. Nobody boasts about how fast a cruise ship is because the whole point is the time on the ship is supposed to be part of the holiday, right? What we will do, I think, with AI is we will leap straight to finished ads.

We'll congratulate ourselves on the incredible efficiencies and the acceleration. But rather like AI-written student essays, AI-written student essays may be better and certainly faster than human-written student essays. But in overall terms, they're completely worthless because you've actually bypassed the painful process, the time-consuming process, which is what the whole thing is about in the first place. Okay?

So that's me, okay? I only have one final suggestion. This comes from Herodotus, and he describes Persian decision-making.

I made this example that creative people always have to present their work to rational people for approval. It never happens the other way around. Okay, you never get accountants going and saying it seems like it's 7.3 But I'll just share it with some really wacky people to see if they've got alternative ideas So there's this complete Ian McGilchrist left hemisphere one-way street where everything right brain is ratified by the left brain But the left brain is allowed to run completely riots without actually being forced to actually frame questions in different ways There is a possible solution Herodotus records that the Persians whenever they had to make an important decision they debated it twice among themselves, once sober and once drunk. And only if they agreed in both states, if it appealed to sober self and pissed self, and if pissed self didn't manage to come up with a better idea than sober self, then they'd enact it.

I think we could borrow from that. So my final question is, you know, we're assuming, we always optimize for reduced time, despite the fact that... a lot of human activities, never a good idea to make sex, by the way, a high-speed activity, you know, oh, 3.2 minutes, that's my personal best, okay?

Right, there are lots of things which, of necessity, take time, where the time is part of the value of the whole process. I don't see anybody who's looking for a single rational optimisation, the definition of a problem, I don't see anybody out there effectively going, but what if slower's better? It's never being asked and I think that's a question we all need to ask.

So thank you very much indeed. Especially thanks to the people in the front. And there is a fantastic course called Mad Masters which I produced in conjunction with the people who run MadFest.

With both live sessions, pre-recorded sessions, a whole gamut of content which I would really, really recommend you actually sign up to. For the very simple reason that if you don't like it, don't do it. That's fine, okay? I don't mind. If someone comes to you with an idea, okay, it's absolutely fine to go, that's interesting, but it doesn't work because of this.

But if you're rejecting the idea, not because it's not a good idea, but because it messes up your decision-making model, that's when you have a problem. And I think that latter thing happens everywhere, all the time. Here are some lovely publications from the Ogilvy Behavioural Science Practice. We do these in our spare time as a pure labor of love, and that's been me. Thank you very much indeed.

Thank you. That wasn't too bad on timing, was it? Thank you, Ulrich. Thank you, Talia. Wonderful.

Okay, that's a wrap, folks. Thank you so much for coming to MadFest. I hope you had a well of a time. There's a party going on out there.

There's loads of really fantastic soul music being played, which is my flavour. But Madmaster students, you're in for a big treat. You're going to have lunch with Rory.

Have a really great time. And hopefully see you all again next year. Thank you very much.