Place of Business Logic and Irrationality in Marketing. Key ideas of Rory Sutherland

Based on the works of Rory Sutherland, Vice-Chairman & Executive Creative Director at Ogilvy Group. Ogilvy Group, headquartered in New York City, is a major player in the global advertising and marketing landscape, operating across 131 offices in 93 countries.

TL;DR

1. Preoccupation with Metrics can mislead decision-makers, as seen in the Vietnam War example where focusing on body counts led to poor strategic decisions. The digital world often values measurable data over meaningful insights, which can be problematic.

2. Human Insights and Creativity can be as potent as technological solutions. Many psychological problems could be solved without traditional media spending.

3. Small changes to the environment can significantly affect people’s behaviour, such as adding tables at airport security to reduce the inconvenience of retrieving items.

4. How a product or service is framed can alter its perceived value, such as EasyJet’s pilot announcement turning a bus ride into a positive experience.

5. Consumers’ decisions are not dictated by price optimization. People are influenced by psychological factors such as Fear of Uncertainty, Ego benefits, Fallback position, and Fear of the Negative which Brands can manage.

6. Trust and consumer loyalty are built through long-term signalling rather than short-term measurable gains.

7. Conventional logic leads to predictable outcomes; businesses should explore counterintuitive strategies to stand out.

8. Choice Architecture. How choices are presented affects decision-making, and businesses should consider multiple dimensions beyond price to influence consumer behaviour.

9. Sutherland criticizes traditional economic logic for overlooking the subjective nature of value and the nuances of human decision-making. Understanding human psychology can lead to innovative solutions to complex problems, such as antibiotic prescriptions and diversity in recruitment.

Preoccupation with Metrics stops Good Decisions

Example: Smart people make dumb decisions. Vietnam war

How many of you have been watching on BBC4 the recent ten-part American series on the Vietnam War? Absolutely fantastic. You can catch up on iPlayer. And it’s been a ten-year labour of love to produce this extraordinary documentary from the very beginnings to the end of the Vietnam conflict, which is a really, really interesting study.

Apart from anybody who is interested in just combat in general, it’s a really, really interesting study in how really, really intelligent people can end up making really dumb decisions.

One of the things that can most wrong-foot people is falling in love with a particular metric. That’s the assumption that what you can measure somehow must be important.

It is the dangerous thing, there are important things and there are things that are easy to measure, but the correlation between the two may not be all that great.

One of the problems I noticed in about episode five is all the military guys, because they were obsessed with reporting upwards, what they were doing. In some ways, I suspect if you’d left the conflict to the people on the ground, it probably could have been over within about three years. Because of the need to report information, you have to report information that can be aggregated. And in order to be aggregable, it’s got to be numerical. And a lot of incredibly important information, particularly to do with attitudes and hearts and minds, doesn’t have an SI unit or a mathematical measure attached to it.

They, particularly at the top, people like McNamara, extraordinary bright people, became obsessed with the body count. And because they were encouraged by the ratio of the body count, they missed a really, really important thing: they thought that the more bodies that you’d effectively create, the better you were doing. What they missed was a very simple fact. If you killed one VC, you had one less VC to contend with. Every time you killed an innocent civilian, however, which appeared exactly the same on the body count, you created 10 new VC.

The obsession with metrics, particularly as the higher up an organization you get, the more preoccupied you are with self-justification, the more you deal with aggregate information, and the kind of information that aggregates may not be the best information.

And it’s one of the things that concerns me about the digital world, that actually the digital world is in many ways a mathematical world. And it thrives on models which may or may not bear a very close resemblance to real human reality. Sometimes it does, in which case, great. But a lot of the time, what you can measure and what matters may be diametrically opposed.

Human Insights and Creative Solutions Solve Problems

What I’m just talking about today, this is what I’ve tried to do as far as Ogilvy change goes. I think, ad agencies, creative agencies have made a mistake because they’ve, like everybody else, become preoccupied with the technology and lost touch with humanity.

My view is an advertising agency, a creative agency, suffers from this enormous problem, which is that people only come and talk to us about problems when they have a media budget. I’d argue that for every person with a media budget, there are a hundred people with a psychological problem that a combination of human insight and creativity might solve who never come and talk to us.

We’re no longer paid on commission. We haven’t been paid on media commission for about 25 years to any significant extent. Yet the ad industry still behaves as though it is. It regards any solution which doesn’t involve a bought media component as somehow invalid. When it gets a problem, its first thought is, how can we translate this into a solution which involves giving some money to Rupert Murdoch?

My contention is, there are tons and tons of problems you can solve. The ad agency actually is a bit like a general hospital, but it has a sign out front that says cosmetic surgery. You know, we only get within our portals, five percent of the people we need to be talking to because there are hundreds of problems that can be solved with a mixture of better human insights and creative imaginative solutions.

Example: Environment impacts behaviour. How to stop people at airports trying to carry water through the security lane

Just to give one example of something we’re working on at the moment, how to stop people at airports trying to carry water through the security lane.

Everybody knows this, but the number of people who basically try and wing it with a bottle of Avian in their hand luggage hasn’t significantly decreased in eight years, despite the fact that awareness of the rule has presumably been going up and up.

We think, by the way, it’s a product of the choice architecture that once you’re in a queue and you have a clamshell piece of luggage, because over the last 10 to 15 years, as low-cost airlines have started charging for checked-in luggage, practically everybody who flies more than once every two years has acquired one of those clamshell, maximum-size, carry-on bits of luggage.

One attribute of those pieces of luggage is it’s impossible to open them in public without spilling your underwear all over the floor. One of our simplest solutions to this does not involve getting any money from Rupert Murdoch. It says, put some tables in the run-up to the security lane so that people who know they’ve got a bottle of water in it can actually retrieve it without hideous embarrassment. You know, an awful lot of behaviour is driven by the fact that it’s impossible to open them in public.

Not by information, not by persuasion. It’s just driven by the environment in which people find themselves. That’s one of them. There are about five other solutions we’ve looked at. But what’s patently the case is this isn’t a case of awareness. This is a case of something about the environment, something about the choice architecture is driving people to behave in the wrong way.

Create Value in the Head of the Customer

What we also do is we look at extraordinary things about what we call innovation. Now, innovation is spelled I-N-N-E-R-vation. It’s a close relative of innovation, but it works slightly differently. Innovation is where you innovate not by doing something new in the physical world, but by doing something differently within the brain. You can create something within the brain without necessarily creating it in reality.

I’ll give you an example of that, by the way. That yellow is totally an optical illusion. I assume this is an RGB pixelated screen. There aren’t any yellow indicators. They’re all red, green, and blue. LCD screens, RGB screens are species specific. They’re designed to work with humans. If you’re a bird looking at that, you’d think it was a shit representation of reality. Humans just have red, green, and blue cones that detect colour. And the screen, unlike birds which have ultraviolet detection, but they don’t detect red.

This is a hack. There’s no yellow. All it does is mix red and green in a mixture which your brain associates with the color yellow. And the color yellow is not created there remotely on the screen. It’s created entirely in your head.

Our contention is you can create value, you can create great experiences in the head without doing much to reality.

When I go to digital conferences there’s always someone there with a sophisticated technological solution. There’s always a guy who has this really clever beacon solution which detects when you have a past customer walking within 100 yards of the shop. I’m intrigued by this. Contextual communication is useful. And they always say: “You can serve them a 20% off offer.”.

Look, I don’t need a PhD to help me sell things by dropping the price. If you want to find someone who can sell things by dropping the price, just go to a car boot sale. Don’t go to Harvard, right?

The art of marketing is surely to charge more for the same thing by making it seem better. Just as you can synthesize yellow in the human head, you can synthesize value in the human brain.

Example: Synthesize value. EasyJet pilot announcement

The best example of this was an announcement by an EasyJet pilot. You know that feeling where your plane draws up about a mile from the terminal and you hear the engines wind down and everybody on the plane thinks to themselves silently: “oh shit, it’s going to be a bus”.

You know that feeling, right? You kind of paid for the flight and you’re just expecting to taxi to the bloody terminal and then the plane stops a mile away and everybody goes: “oh shit, it’s a bus”. I was experiencing exactly that same emotion.

The pilot was a genius. He gave an example of a piece of fantastic innovation. He said: “I’ve got some bad news and some good news. The bad news is that we haven’t been able to get an air bridge because there’s a plane blocking the gate. But the good news is the bus will take you right next to passport control so you won’t have far to walk to collect your bags.”

That’s always true, isn’t it? When you have a bus, it takes you right next to passport control. You don’t have to walk 400 yards before you can get to passport control. The bus takes you right next to it. Actually, I’ve got quite a heavy bag so I’m glad there’s a bus.

Just by changing what people focus on you can make people you can make something that people thought was bad seem good.

Shakespeare said the same thing. “There’s no such thing as good or bad but only thinking makes it so”.

“Yes, but” trick

If you look at a lot of great advertising end lines: “Good things come to those who wait”, “You either love it or you hate it”, “Reassuringly expensive”, “Naughty but nice”, “We’re number two so we try harder”.

A lot of great advertising end lines have that “yes, but” in them.  It’s a very clever trick because what the human does when you’re presented with bad news and good news is, to minimize regret, you disproportionately focus on the good news to cheer yourself up.

I’ve actually gone to the government with this and said look, part of the problem of a lot of things government does, tax for example, is there’s no upside. If you created a tiny upside to tax, like you got a badge. My own suggestion that higher rate taxpayers should be allowed to drive in the bus lanes was rejected but there we go.

If there were a tiny upside, we could make ourselves much happier. Things like speeding tickets and tax make us disproportionately angry. There are people who’ve gone to jail for five years who can tell a positive story about it: “It was a formative experience; I met a lot of interesting people I never would have met”. There are a lot of appalling experiences where you can tell the upside story but with something like tax or nobody provides the upside story.

Example: People pay for a downside. Ferrari factory tour

You can make people pay for things which are a downside. If you buy a Ferrari, they’ll deliver it for free to your local Ferrari dealership. Or you can pay 500 pounds, have a tour of the factory in Maranello, to which you travel to at your own expense, and then drive the car home. By framing it as “a tour of the factory” they’ve got you to pay to collect your own car.

Example: Frame of Reference. Framing. Nescafe

You can make things cheaper by changing perception. I love my espresso machine like a child. It is, in fact, insanely expensive. If you had to buy an espresso coffee in a jar like Nescafe for an equivalent dosage of caffeine a jar of of Nespresso would cost about 70 quid. You’d look at it next to the Nescafe and think: “this is completely bonkers”.

If you looked at a jar of coffee costing 70 quid you couldn’t buy it. But it doesn’t come in a jar. It comes in a pod which gives you an individual dose of espresso coffee. Your frame of reference isn’t Nescafe, it’s Starbucks. And you put the 29p pod into your espresso machine and you think, well, it’s 29p but that would cost me £2.20 at Starbucks.

Example: Frame of Reference. Framing. Luxury cars

The Rolls-Royce and Maserati made their cars cheaper through framing. They stopped exhibiting them at car shows where they look expensive and they started exhibiting them at yacht and plane shows. If you’ve been looking at Lear Jets, Gulfstreams and Sunseekers all afternoon, a €250,000 car is an impulse buy.

By the way, if you allow people to pay for something contactlessly it feels 15% cheaper.

Psychological factors of Customer Experience and Consumer Decision-Making

When you look at queuing there are about nine psychological factors that affect people’s experience of a queue which are not the duration they have to wait. Worst thing you can do is have a queue alongside your queue which is moving faster than yours is. Anybody stuck in traffic knows the same phenomenon.

Example: Fear of Uncertainty in Consumer Decision-Making. Uber

I would argue that Uber is very largely an innovation. Regardless of what people talk about it as disruption, economic model, blah blah blah. The reason it disrupted was not because it got you a car quicker but because of a very simple thing. Just as with the human brain you don’t produce yellow, but you can generate yellow. You can reduce wait time not by reducing it but by reducing uncertainty of it. The thing that was appalling about minicabs apart from the fact that they always lied about how long they were going to take.

Minicabs were a frustrating experience: “Look, where is the minicab? Where the hell is this guy in the rain? I’ll look around, maybe, he’s already arrived, maybe, he’s parked around the corner. I’ll go out in the rain looking like an idiot and see if I can find our minicab. Oh god, he’s not here. Maybe, he’s already left”.

With Uber you can look at the screen and go: “Look, he’s stuck at those traffic lights. I’ll have another pint”.

Example: Fear of Uncertainty in Consumer Decision-Making. London Transport

The single thing that London Transport did that most improved passenger satisfaction per pound spent wasn’t faster, more frequent trains it was dot matrix displays on the platform. We’re happier waiting nine minutes for a train knowing that it’s going to take nine minutes than waiting five minutes for a train in a state of uncertainty.

Example: Ego Benefit in Consumer Decision-Making. Uber

There were other emotional benefits to Uber. I think there’s a small ego benefit. Does anybody else do this where you time your arrival onto the pavement to coincide with the arrival of the car? At the end of the Usual Suspects, you come out of the police station and a car draws up. You feel like Louis XIV.

I also think the fact that you didn’t have to pay and fart about with a wallet. The fact that you just got out of the car and said thanks and that the receipt was emailed to you made it feel like a service and not like a transaction.

Example: Fallback Position in Consumer Decision-Making. Uber

London Transport is very silly being nasty to Uber. Uber is actually incredibly valuable to people even when they don’t use it. I’m much more likely to take a train now because Uber exists. Because I know there’s a plan B, a fallback position.

Previously, if you got stuck in the wrong part of London at one o’clock on Saturday morning, you know, you’re basically completely stuffed.

A lot of the time I used to drive into London, not to take my car in as a form of transport, but as a kind of fallback position which is “I need to have a car there just in case”. Last year I made 40 fewer journeys into London than I did the year before. Largely because of the existence of Uber and yet, instead of those journeys, I only used Uber to get home once. It was knowing it was there that made the difference.

So, generally in human behaviour there’s much more subtlety going on than economics and standard mathematical models really understand. The trick for innovation is as I said: look for things which are objectively similar but subjectively different.

Example: Mental trick. Chilli pepper

The right mixture of red and green is the same as yellow as far as the brain is concerned. Similarly, nature has been doing this for a few million years. The chilli by the way isn’t actually hot. There’s nothing hot about a chilli. It doesn’t burn your mouth or other parts of your digestive tract. It actually hacks perception. Capsaicin makes your senses in your mouth hypersensitive to heat so anything that comes into contact with them feels really hot. It’s changing the way we perceive something rather than changing the reality.

Example: Power of Positioning. Ed Sheeran Peep show.

I’d just like to give you an example of this. Depending on how something is presented you can make something brilliant seem absolutely terrible and you can make something pretty bad, as Ferrari did, seem brilliant. That’s why marketing isn’t an optional extra. It isn’t added value. It’s an intrinsic part of the value of anything unless you’re a Marxist and you believe there’s some weird objective labour-based theory of value.

If you accept that value is subjective there are two things which determine value effectively. How it’s perceived is a factor of not only what it is, but how it’s presented. I always wanted an example of something absolutely brilliant which was terribly presented.

I used to use a theoretical example which was a restaurant serving Michelin-starred food but which smelt of sewage. If you assume that we’ve got to improve objective reality you decide that we must make the food even better. No, get rid of the smell.

I managed to find an even better example of this from two Australian comedians who took the hottest property in the entertainment world at the time and presented it in a way that did not inspire trust or belief or conviction.

Peep shows have a pretty bad name normally associated with lewd content, but by definition they don’t have to be. So, in an attempt to change that they took one of the world’s biggest performing artists, kept all his clothes on and set up an “Ed Sheeran peep show” that was written outside of very dodgy looking venue.

By the way, this is pretty much how it feels if you’re a financial services marketer. Because it doesn’t matter how objectively good your product is if you don’t have basic trust, if you can’t manage to create that idea which is “this is probably going be okay”. And I would argue the whole sector regardless of merit or desert really suffers from a huge shortage of trust. That’s what ultimately happens now.

Notice here that the first thing they think when they start failing. What’s their default? They start blaming too high price.

One of the things that worry me most about algorithms is they’ll automatically start testing the price of things rather than the other psychological factors which may cause people to buy.

Interestingly when he finally persuades that couple to go in, he uses two or three really well-known tested techniques of human behavioural science: he says “it’ll probably get pretty busy later on”, “he’s only there till midday”, “you both can come in”, etc. Those things basically work.

I’ve been studying behavioural science for 15 years. When I’m on an airline website and it says “only four seats left at this price” I know exactly what they’re doing. They’re trying to exploit my scarcity bias but I still buy the bloody tickets.

The same can be used to solve important problems. This isn’t just your typical brand managers problem. The over prescription of antibiotics has been solved through behavioural science. by a post-dated prescription. Essentially, you give people a bit of paper which is a prescription for antibiotics. They leave the GP completely happy because they got a bit of paper. I think 80% of the people given a post-dated prescription don’t actually cash it in for the antibiotics.

Otherwise, if you give people a prescription and say “don’t take this until Friday”. What they do is: they pass the chemist on the way home, they get the antibiotics anyway, because they’re planning to go home and wrap up in a duvet. They spend 8 quid on their prescription charge and sunk cost bias means they might as well start taking the pills.

I’ve got an even more controversial theory. With the NHS you could reduce unnecessary visits to GPs by about 50 or 60% by just having an answer phone message which lists the 3 most common ailments that people are suffering from at that time. Psychologically why do you go to the doctor now? The logical answer is “we go to the doctor to get well”. 90% of the reason we go to the doctor is to be reassured. When you have an illness, what’s the best thing the doctor can ever say to you in terms of reassurance? There’s a lot of it about. You immediately feel fine.

It’s rather like when there’s a power cut. When there’s a power cut you go out on the street and when you see all the other lights are off, you go: “thanks for that, it’s not just my problem, it’s a shared problem! I can relax.”

What’s the worst thing a doctor can say to you? “This is an extraordinary condition. I’ve never seen anything like it before.” So, if you had an answer phone message or an on-wait message an on-hold message at the surgery which just said: “these are the common viral conditions that are going around at the moment”. My hunch is that 2/3 of the people wouldn’t bother going in at all. They’d just go: “oh I’ve got that.”

I think also choice architecture, understanding human psychology can solve problems like the question of diversity and recruitment. An awful lot of the problem with diversity probably isn’t due to conventional prejudice. It’s due to a thing called status quo bias which is: if you’re making one decision you go very close to the expected boring norm. There’s a solution to that: if you hire people in groups without any bidding whatsoever, people will automatically make diverse hiring decisions. Because now they’re looking for complementary skills. They do this instinctively.

When everybody had one car per household, everybody had a saloon car because it was like right close to the kind of middle default boring unusual in no dimensions thing. Once people bought 2 cars per household, they had 2 completely different cars, and probably neither of them was a saloon car.

Just by changing the way that choices are made you can completely transform the decisions that people make.

Calculations Destroy Nuance

One of my problems with maths is that when you multiply 2 things together or you add 2 things together you lose information. What does Amazon Prime do? It encourages you to buy lots and lots of things from Amazon. You make one payment and then all your subsequent delivery charges from Amazon Prime are free. That’s designed to encourage you to make lots one-person commitments.

Why does Amazon Prime have to exist? The answer is because 1 times 7 isn’t the same as 7 times 1. 7 people buying one thing a month don’t care about a delivery charge. One person buying 7 things a month is not going to buy the 5th, 6th and 7th thing from Amazon because he’s already spent 12 quid on bloody delivery and that’s the limit.

If you wanted to make the London congestion charge more psychologically effective, you’d make everybody’s first 3 journeys into London free, then the next 5 journeys are 10 quid, and after that it starts going up. You do the opposite of Amazon Prime, but they haven’t because the congestion charge treats 7 people going in once the same as one person going in 7 times. It’s not optimized to change behaviour.

I would argue there’s a fundamental problem that Amazon has which it doesn’t know how to solve. E-commerce is a very good way to sell one thing to 7 people or 15 people. It’s not a great way to sell 15 things to one person. That’s kind of Walmart. And the two are not the same.

If you start accepting that 1 times 7 is not the same as 7 times, 1 plus 7 isn’t necessarily the same as 7 plus 1, you can suddenly start solving problems creatively. What the multiplication is doing in your model, what the addition is doing in your model is actually stripping out information. It’s stripping out human nuance.

I was talking to transport ministry the other day. I said: “you’re talking about overcrowding in trains. And the idea is anybody who has to stand on a train is bad. I agree it’s kind of bad. But we’re never going to produce a train network where no one has to stand. Let’s look at it differently. If 10 people have to stand 10% of the time, it’s kind of overcrowding, but meh. If we had to stand on a train every 10th journey: “yeah, shit happens, whatever”. If you bought a season ticket and you had to stand every day you’d be livid. So one person who has to stand 100% of the time is not the same psychologically. It’s the same in the model as 10 people who have to stand one journey in 10. Once you split it up like that, once you accept that commuting is not commutative, you can solve the problem in an interesting way.

Maybe, what we need to do is to have two trains a day in each direction which are exclusively for quarterly and annual season ticket holders. We’ll make sure that if you’re on that train you get a seat. If you want to travel on that train and you haven’t got a season ticket, maybe, you’ve got to pay a supplement.

Restaurants would do this. If you’re a really regular customer of a restaurant they give you a better seat. Why don’t railways think like restaurants? The answer is because they’re looking at information in the aggregate. They’re not looking at it at the personal disaggregated level.

Models look clever, anything that uses maths makes you look clever. It demolishes the argument because the guy with the spreadsheet in the room always wins the argument. But the very process of adding things up and multiplying them together may be actually limiting your creative ability. Aggregate information looks clever but can make you act differently and dumb.

I genuinely think that marketing needs to be the science of knowing what conventional logic is wrong about. The problem about conventional logic in a marketing setting is: if you’re conventionally logical, you’ll never get fired. Any decision which is made based on the pretence that people are logical and rational will never get you into trouble. The only problem is: one, you might be wrong, two, you’ll often use logic and end up in exactly the same place as all your competitors. You’ll end up in the worst part of the market which is a kind of bloodbath. All because logic tends to take you to the same place as other people’s logic.

For example, you want to compete with Coca-Cola. For the last 80 years, it’s been the most popular cold non-alcoholic drink in the world apart from water. And you’re Mr. Logic. What do we need to do to compete with Coke? You say: we need to produce a drink that tastes nicer than Coke, costs less than Coke and comes in a really big can so people get great value for money.

No one will kick you out of the room if the drink you subsequently launch fails. You won’t get fired for doing that. In logical terms “costs less, tastes nicer, comes in a big container”, you’ve totally slam dunked it. The only problem is the most successful attempt to compete with Coke is Red Bull. It costs a fortune and it tastes disgusting.

There probably is by the way a deep psychological reason for this. If you want to believe that a drink has medicinal or psychotropic powers it can’t taste conventionally nice. With a lot of medicines, they add a shitty tasting thing to it to maximize the placebo effect. If Nurofen tasted like black currants, it probably wouldn’t work very well. We need it.

That’s why people think wheatgrass is good for you. Because you’re not drinking the stuff for pleasure. You can just go and lick the underside of your lawn mower and create the same effect.

Marketing logic is different from conventional logic

In a business there’s kind of finance department logic there’s operations logic, there’s what you might call air traffic control logic. You want the people in operations to maximize efficiency. All those things which make a business efficient and good. But there’re certain spheres of business which shouldn’t operate on the same logical precepts. HR is one of them. R&D would be another one. Anything to do with experimentation and R&D. HR because the skill of HR is to find people that your competitors aren’t finding to work for you. And marketing is a third one.

The other one here is military logic has to be different from conventional logic. If you’re conventionally logical in the military your enemy knows what you’re going to do. It kills you because they set a trap for you. They know what you’re going to do. You become predictable. If you become predictable, you become dead.

okay this if you want to over intellectualize a joke in the film Airplane I think it’s called Kramer if I’m right and he’s in charge of the control tower he’s a Korean war veteran he’s in charge of the control tower for those of you who don’t know Airplane there’s essentially a stricken airliner with a very inexperienced ex-military pilot on board who has to land sort of 200 civilians after dark because the pilot and the co-pilot have been taken ill. The quote is:

Air Controller Macias: Captain, maybe we ought to turn on the searchlights now.

Rex Kramer: No… that’s just what they’ll be expecting us to do.

The point is you don’t want military logic in air traffic control. You don’t want the people who tighten the wheel nuts on your plane when you next fly to be wildly creative people going “screw it, let’s try anti-clockwise”.

There are spheres in business, and military, and governmental, and political decision-making where you’ve got to abandon conventional logic. It’s okay to have procurement people buying military hardware. What you don’t want is procurement people deciding military strategy. If you’d done that, they would have insisted that the D-Day landings took place between Dover and Calais to minimise fuel costs. The whole point of the Normandy landings was to create the expectation that that’s where the landing would take place while actually planning to land somewhere much more expensive and much more difficult. All military strategy is actually the cunning deployment of wasted resources to wrong foot an enemy. If you ever do what’s most efficient, you’ll come up against the most resistance.

I would argue the same applies in many cases to marketing. If you try and make marketing merely efficient, you’ll end up predictable and in the same place as everybody else. The problem that we face is that finance dominated decision making. Business is overwhelmingly dominated by procurement logic which is cost-cutting, maximise efficiency. The default mode of any business is to essentially reduce the price of things. No one who goes into any business meeting, who says “I’ve found a way of doing X cheaper”, no one will have to face any hard questioning.

In some cases, the problem arises that actually marketers suffer a kind of Stockholm syndrome where they pretend, they’re like procurement. They shouldn’t be doing that. Because marketing isn’t that kind of thing. It doesn’t use that kind of logic.

Example: Discounts Reduce Demand. London theatre chain

Someone I knew studying for a qualification in behavioural science made their money doing email marketing for a large London theatre chain. They discovered very quickly that advertising a discount on theatre tickets reduced demand. I can understand that the fact that you’re discounting the tickets suggests the play isn’t all that great.

If you think about it, going to a not very good play still costs you 140 quid after the discount by the time you’ve paid for babysitting, a meal out, car parking, and a taxi. So, she used to tell people this. They’d come and say “I want you to advertise the matinee on Friday for such and such a performance and I want you to tell people it’s 25% off” and she said “I’ll happily advertise the available tickets but I won’t say 25% off because it’ll reduce demand and reduce the price, so you’ll sell fewer tickets and at a lower price”. In every case the people came back and said “I want you to do it anyway”. Why? You’ll sell fewer tickets. If I pretend that economics is true, I’ll never get into any trouble.

Let’s say, I didn’t sell all the tickets and I didn’t put the discount on. He might say “but if you put the 25% discount on economic theory tells us we would have sold more tickets, therefore we would have done a better job”.

Effectively economics is the new IBM, no one ever gets fired for using it. This is a fundamental problem with the way algorithms are designed, with the way business decisions are made which is that basically pretending that economics is true never gets you into any trouble.

Imagine if you wanted to go and say “I want to give our staff a pay rise because I find that generally jollying them up and making them happy increases the sales of our product”. Let’s say that was a call center. You could make that case. In fact, in many many cases it would be absolutely true. One of our clients had a very good call center and I asked him “how is your call center so good? Why are all the staff so brilliant?”. He said “”to be honest, we probably overpay them”. They stayed for years and they got really good at their job.

To justify that it would take you a year of hard argument and you’d still probably lose. If you went into the same meeting and said “I’m going to put the staff on zero hours contracts and outsource them to somewhere else”. That would get through on the nod.

There’s a fundamental bias in business which is effectively towards making everything 20% cheaper than consumers want it to be and 40% shittier because those business decisions are just the easiest ones to make.

That’s why there are no curtains on the Eurostar. Because of some scrote in procurement. You pay 280 quid. There are hooks for the curtains. You can’t see your laptop in the sun unless you pull the blind down which is a bit passive aggressive because you share the blind with someone else. Someone pointed out that we save 200,000 a year in dry cleaning costs by not having any curtains. The corresponding argument which is “curtains will make it a lot nicer and more people will come back” is probably true. But it’s much harder to prove and it’s much harder to argue.

These are the assumptions of economic logic that human behaviour is objective. It isn’t we construct ideas of value and price in our heads from various cues. It’s not objective. It’s not perfectly trusting. It’s not proportionate. It’s not ergodic. It’s not status free, context free, individualistic. It’s not maximising. It’s not path independent.

Example: Companies Manipulate Customer Decision-Making. Restaurants selling wine

Look at how a restaurant gets people to buy wine. Wine is very profitable. Why is wine very profitable? The reason is that, basically, it’s kind of bullshit. How many people would really rather drink cocktails, or gin and tonic, or whiskey, or beer for that matter? I think quite a lot. But the great thing about wine is you can markup wine to a huge extent because it’s bullshit. You can’t charge 50 quid for a glass of johnny walker because people know what a bottle costs in the shops. But you can buy a case of Chateau d’obscure 2008 for 6 euros a bottle charge 50 euros a bottle for a never-ending crap on about the scent of black currants and all that sort of bollocks. That’s what you’re expected to do so.

Restaurants really want you to order wine so they manipulate you every single time. First of all, when you come into the restaurant there are already wine glasses on the table which says this is the kind of establishment which expects you to drink wine. If, indeed, you say we’re not drinking wine they take them away with a bit of a huff.

They give you a drinks list which isn’t called a drinks list. It’s called the wine list. Add the choice architecture of the wine list is: you have four pages of a totally insane and ridiculous range of different kinds of red, white and rose wine and then this little back page for the perverts and deviants who’d actually prefer to drink something produced by a culture that’s mastered brewing or distillation rather than just trampling on grapes and leaving them to go off alright. You feel you have to.

But there’s an even cleverer piece of manipulation than this. They only hand out one wine list, if you think about it, there’s only one drink that you can kind of share. So, the guy with the wine list turns to the table and he says “red or white”. At which point it is game over for the gin drinkers, the beer drinkers, they’re forced to go along with everybody else. You think you’ve chosen to drink red wine but the whole choice architecture has been fundamentally geared up to flog you wine. That’s the product of years of genius in how to run a profitable restaurant.

Example: Customer Decision-Making with just the price. Airplane tickets

Let’s look at how we choose flights. Not hotels. It’s fair to say nobody when they book a hotel goes online and goes “I want to stay in Barcelona for three nights. What’s the cheapest hotel? You’d end up in some rat-infested place”. Hotel choosing is a bit nuanced.

What they do in the airline is they make price the only salient thing by which you can choose. If you’re an airline that said “I’d like to be 10% better, have a bit more leg room and we’ll charge everybody 10% more” you’d probably disappear off page 1 or 2 of the search rankings and you’d go bankrupt. Even though lots of people would probably prefer that trade off.

There’s a very important thing. Daniel Kahneman, who’s the first psychologist to win the Nobel Prize for economics, made the point that his most important discovery is “nothing is as important as you think it is”. While you’re thinking about it, what you’re looking at, and what is salient becomes important because you’re paying attention to it.

I pointed this out to British Airways. Look, the way you get people onto the 12 o’clock flight rather than the 10 o’clock flight is exclusively by dropping the price of the less popular flight. I pointed out that the 12 o’clock flight has got a brand new 787 Dreamliner on it but I can only discover that by clicking several times. That plane’s cost them 150 million pounds. But in my actual choosing it’s not even visible. The fact whether there’s Wi-Fi or not isn’t visible.

You can use loads and loads of things to get people to change their behaviour as well as the price mechanism. When you make price the dominant thing people automatically will choose on price.

I think the airline industry is killing itself through choice architecture. I don’t know how many other people have done this. I’ve done it myself. So, you’ve got a family of four, you’re leaving Gatwick to go somewhere on EasyJet in like five weeks’ time. There are various flights and you see that one of the flights is 10 quid cheaper than the three other flights going to the same destination that day. There are four of you so you multiply it by four. Saving 40 quid, I’ve got to do that, fantastic! You book it and then two days before you travel you have a look at what you’ve booked again. There you realise it’s the bloody 6am flight. A day before you go “there’s no way I’m going to get the family out of bed at 3.30 in the morning”. So you end up booking a hotel the night before and having a meal in Pizza Express which costs you about 200 quid which completely eradicates the 40 quid you saved by choosing the 6am flight. But at the time you chose that was the salient factor.

If airlines can’t find other things to advertise, you’ve got to make choice architecture more multi-dimensional. Maybe the mid-day flight has two kids for the price of one. Maybe the one o’clock flight has “free Wi-Fi on this flight”. Maybe one flight says “take this flight and get a free upgrade to premium economy on the way back”. But you’ve got to mess with the choice to not make it all about money. Otherwise, the entire world’s airline industry is just involved in a massive race to the bottom simply because of the way they present choice to people who are choosing a flight.

If you want to buy art, don’t buy architecture. The way we choose a house is “where is it? How much does it cost? How many bedrooms has it got? Has it got a garden? Has it got a greenhouse?”. Then we narrow it down and we look at five houses and we choose the one we like the look of most.

I did it backwards. I said “let’s look for a fantastic house architectural first and then worry about the other stuff”. I ended up in a flat in a Robert Adam grade one listed. There are only 2000 grade one listed houses in Britain. I asked a friend economist of mine: “how much extra do we pay for the fact this is a Robert Adam house?” He said: “somewhere between nought and two percent”. You don’t pay two percent extra for a Picasso versus some crappy thing done by a guy on the Bayswater Road. If you bought art the way we bought architecture that’s what would happen. If art websites were like property websites, you went “I want a painting that’s about five by three. I want it to be painted in the south of France. I want it to feature two cows and maybe a tree. I’d like it to be mostly green but with a bit of blue” and then the bloke said “here are five of them” and you said “I’ll have that one”. In those conditions Picasso’s would be really cheap.

The order of elimination of attribute has a huge effect on the actual decisions people make. The airline industry is effectively trying to commit suicide because of choice architecture.

Fear of the Negative in Human Decision-Making

Every model of human decision making assumes that people are trying to maximize. It assumes they’re trying to optimize something. Economists call the thing they’re trying to optimize utility. I would argue this makes no sense given that our brain is the product of evolution and given that we’ve evolved to live in a world of uncertain information.

The reason by the way that finance tends to hate marketing is because finance is very heavily influenced by economics. Economics assumes perfect trust and perfect information in every decision. In the mind of someone in finance, or in the mind of an economist, in their perfect world model advertising shouldn’t exist at all because consumers would already know exactly what they wanted and how much they’re prepared to pay for it. Finance tends to hate marketing because it sees it as a cost to be minimized not as a source of value creation. They don’t really accept subjective value; they don’t really understand it.

There are decisions like archery where it’s all about maximization. You aim for the bullseye. It doesn’t matter if you can’t see, if you’re drunk, if it’s windy, you go for the bullseye. Why do you go for the bullseye? Because there’s no better strategy in archery. If you miss the 10, you get a 9, if you miss the 9, you get an 8. That’s a maximization problem. I would argue that’s why archery has never made it big as a televised sport. There are no trade-offs. If you’re an archery commentator it must go something like this: “Barry, what do you think he’s going to do next?”, “Well, John, I think he’s going to aim for the 10 like he’s done for the previous 500 bloody goes”. There’s nothing else you can do. Archery is not an interesting sport.

Darts by contrast is actually quite interesting. If you’re not very good at darts, aim for the southwestern quadrant of the board. You won’t get a triple 20 that way, but you won’t get a 1 or a 5 where there’s a messier outcome to a decision. When there’s a degree of uncertainty to a decision you have to consider two things: you have to consider the average, but you have to consider the variation as well.

The human brain has evolved to make decisions that are much more like darts than they are like archery. Do I climb an extra 10 feet up that tree to get the extra cherries even though there’s a small risk I could die by falling? Those are the kind of messy uncertain decisions that we’ve evolved to make. We have to consider both how good is it on average because if it’s really good on average but one time in 10 it’s fatal then we don’t want to do that. We’ll take a trade off. We’ll have something that’s less good on average but has lower variance. It’s never going to be terrible.

That’s why, I would argue, McDonald’s is the most successful restaurant in the world. It’s not because it’s really good, it’s because it’s really good at not being awful.

An awful lot of human behaviour is driven by this. It’s not only asking the question “how good is this thing optimally”, it’s also asking “what’s the worst that can happen” and if there’s a very low probability of something bad happening but that bad thing is very bad, we won’t take the bite.

You can see that in people’s behaviour on things like eBay, it only takes one really dodgy thing like bad grammar in the listing, two recent complaints from people. If we were a logical species on eBay: if someone had a 95% approval rating, they should be able to sell their products for 5% less than someone with a 100% approval rating because they’re 5% unreliable. Actually, what happens is if someone’s got a 95% or 90% approval rating they sell the bloody products for about a third of the price or they can’t sell them at all. There’s that uncertainty thing, there’s too much variance in there.

That I would argue is why people pay a premium for brands: not because they think brands are better, but because they think they’re less likely to be shite.

Cats are showing exactly the same thing. It’s a cucumber, put it outside your cat’s line of sight, wait, and there you go. There’s not much chance it’s a snake but why take the risk. It’s green, it’s got a pointy nose, it’s low to the ground. There probably were cats early in evolutionary history who were calibrated the other way around they went to check that cute little cucumber and those cats tended to exit the gene pool. So, most cats are descended from the slightly paranoid cat.

This affects human behaviour. When you have penalty kicks to decide football competitions even like the World Cup. People would have scored more; they’d get a higher average if they kick the ball straight down the middle. The reason they don’t is although you’re more likely to score, you look much stupider if you fail. If you kick the ball left and the guy saves it, kick the ball right and the guy saves it, you look unlucky. Kick the ball down the middle and the guy saves it, you’re more likely to score, he’s less likely to save it because his propensity is to dive, but if he does save it standing in the middle you didn’t even try.

That drives nearly all business decision-making. Fear of the negative is a much more powerful force than desire for the positive and it can lead to business decision-making which is dangerous.

We also discovered in drinks there is no way you will ever sell a man a cocktail unless the menu contains an illustration or a picture of the glass in which it’s to be served. Because if the male brain thinks there’s a 0.05% chance it arrives looking like that he’ll order a beer instead.

Long Game in Marketing: Meaning Over Metrics

There’s what you can measure. In digital advertising, what’s easiest to measure? It’s a purchase or transaction or a click from a message that is highly, that is very close to the point of purchase. I would like to argue that a significant part of marketing must rely on doing the opposite of this.

You go to a car dealership and the bloke buys you a cup of tea. You don’t think “what an incredibly generous bloke, he’s just made  me a cup of tea. This is fantastic. What amazing customer service.” Because he’s  trying to sell you a car. It doesn’t instil trust because the cars cost 20 grand and the tea bag cost 2p.

If, on the other hand, you buy a car from the guy and a week later he rings you up and says “you forgot the log book, I’ll drop it off at your house on my way home this evening”. You think that’s really cool.

Because the first is self-interested and nakedly short-termist. The second gesture is patently long-termist because he isn’t going to be selling you another car for three years. The meaning and what’s easy to measure and what has meaning to consumers may be very different.

A bit of game theory. Generally, people are honest, decent and nice and this applies to relationships with brands just as relationships with other people. To the extent they expect future interactions. A lot of altruism is basically long-term selfishness. It’s worth me being nice to this guy because I’m playing the long game. And people who are playing the long game who are looking to actually engage in a series of repeated non-zero-sum transactions over time tend to be decent and trustworthy. People who just want to flog you something and then disappear, we don’t trust those people.

I would argue that an extensive part of marketing is doing things that only pay off in the long term. Now the way we’d signal trustworthiness to another human whether is by doing things that involve effort now or cost now which will only pay off over a series of repeated interactions because that shows we’re playing the long cooperative game.  Not the one-off cheaty game.

The problem is those things by definition are the hardest thing to measure and they’re the things that every algorithm will miss because the payoff may come three years hence.

Example: Building Long Term Relations. Rope Handle Bag in cosmetics shops

An example of a generosity post purchase would be when you spend 200 pounds on cosmetics or women’s fashion, they have to give you a rope handle bag. It’s not better than the polythene bag it’s just more expensive. It says “we’re not stingy here; we’re giving you a proper bag”. Measuring the efficacy of the rope handle bag would be really difficult because the only time it will show up might be two years or three years hence or six months hence.

The effectiveness of the cup of tea is really easy to measure but the meaning is inversely proportionate to the ease of measurement.

Engagement ring is really costly long-term signaling. Upfront expense as proof of long-term intention. You wouldn’t buy an engagement ring if you’re only planning a one-night stand. There’re cheaper alternatives. If you notice an engagement ring also has to cost an amount that is actually painful to the man buying it. Doesn’t matter how rich you are, it has to involve actual pain as a sincere gesture that I am entering into this arrangement on the clear and proven expectation that it has a long-term future not a short-term expedient.

Similarly, Five Guys giving you extra fries. After you’ve bought the fries, they give you an extra scoop which is above and beyond the scoop that fills the cup. Again, that’s long term signaling.

The fact that Selfridges sends you your stuff from Selfridges.com in much much better packaging than you have any reason to expect. That’s long-term signaling.

My hunch is a very large part of marketing and particularly the establishment of of trust involves a behavior which, by definition, is very difficult to measure.

By focusing more and more marketing budgets not on where they’re effective, but where you can prove they’re effective, there’s the danger that we’re misdirecting investment dangerously.

Avoid conventional logic, because your competitors will all use that. Test counterintuitive things, because your competitors won’t.

References

1. Conference speech “My Advertising Is So Efficient It No Longer Works“