Rory Sutherland: On slow trains, rebel bees and the virtue of irrational ideas
Digital content team
Rory Sutherland, TED speaker and author of ‘Alchemy: The Surprising Power of Ideas That Don’t Make Sense’ talks to Marc Kahn, Investec's global head of human resources and organisational development.
Receive Focus insights straight to your inbox
LISTEN: In conversation with Rory Sutherland
Skip to sections that interest you most, or read the transcript below.
1m 47s: The power of behavioural science
Marc Kahn: I’m going to dive right in and I’m going to ask you to just talk about this behavioural science practice because the topic for today is ‘Behavioural Science in 21st century organisations and the power of that’ and it’s really kind of revolution and you founded this practice in Ogilvy, give us a sense of what you do in your unit. The kind of briefs you get in the commercial world.
Rory Sutherland: I will give you an example of one brief which transmuted into a completely different brief, it was with Kimberley Clarke and it raised the question of why nobody buys moist toilet paper. Now our theory, it’s a very interesting question, because this is actually, the fact that we’re happy to clean our bottoms with dry paper is actually a cultural peculiarity of the West, if you think about it.
In the Muslim world you will always have a hosepipe in any toilet cubicle because they assume you use water. In Japan of course you have a remote, for some reason the toilets have a remote control, I have never understood the need for that, but a little proboscis comes in and actually cleans your bottom with water at a perfect temperature and so on.
MK: I just want to say, we’re starting this perfectly, at the right angle.
RS: And so if you think about it, you wouldn’t, it is an anomaly because you wouldn’t go out into the garden to pot a plant, get some mud on your hands and go ‘oh dear, I’ve got mud on my hands, let me rub them very vigorously with some dry paper’, would you? You’d use water to clean your hands yet for some reason your bum is in the West.
MK: It is not afforded the same consideration.
RS: It isn’t, no and interestingly, our theory was by the way, quite simple which is that a large part of it is simply unconscious human decision making which is when you go to buy loo paper, there is an acre of dry paper in rolls on the shelf and then the moist paper, they are like two little bits.
It’s in blue and white packaging which makes it look medicinal and it’s on the top shelf. And so the instinctive human brain without making a conscious behaviour or conscious decision basically goes it’s probably for weirdoes or people with a medical condition. The first thing we did, which they acted on is they produced lots and lots of different scented variants so it takes up more space.
If that sounds strange, that’s more or less how for the Brits among you, how Walkers Crisps beat Golden Wonder. They noticed that more and more people were buying crisps in supermarkets rather than in corner shops, they noticed more and more people were buying multi-packs and they introduced a load of flavours not really because those flavours made sense on their own, about 80% of people it’s basically ready salted, salt and vinegar, cheese and onion and then there are a few weirdoes on the prawn cocktail side.
And then there are the niche crisps which not many people buy but the great value of the niche crisps was that it meant that Walkers by having more flavours occupied three times as much shelf space as Golden Wonder did. They won in the same way that the First World War was won, you know, so a few inches at a time.
Anyway, what was strange about this brief, and fascinating, was because halfway through working on it Kimberly Clarke suddenly came back and said, actually we’ve got a much bigger problem than the moist lavatory paper one which is factory safety. And there’s a notable problem which always happens in any, if you think about it, it’s a sawmill, they’re a paper company, there are enormous rotating saws.
What always happens is quite rightly, you give people wearing protective clothing but there’s a small downside effect to protective clothing which is that it makes you feel invincible. And so one of the things we were doing is experiments on how to improve factory safety, and I will give you two very interesting examples.
One of them which we are currently researching at the moment and seems to work is if you put the image of a skeleton or a human hand on the back on a protective glove, it reminds people there’s a hand inside and people are much more cautious. It doesn’t affect their behaviour with circular saws because those are scary anyway but with Stanley knives and box cutters, it basically makes people feel more cautious if you just have the imprint of a human hand on the glove.
MK: So can I ask you, how did you come to that finding?
RS: We looked at, there’s a very, very interesting literature on how a lot of perfect, this is ideas that don’t make sense, perfectly logical ideas often create a kind of, in psychology create a kind of counter force, so the most famous one is it looks rather as though making drivers wear seatbelts doesn’t really improve road safety all that much because people compensate by driving slightly more dangerously because the seatbelts make you feel safe.
MK: But the helmets of cyclists in London are the same.
RS: I would always recommend brain protection but it does have a counter-availing effect.
MK: Because people who wear helmets, cyclists who wear helmets in London tend to have more accidents.
RS: I can well believe that.
MK: Although if you’re going to have one, it is good to have a helmet on.
RS: There is a debate whether that’s cyclists’ behaviour that changes or motorists’ behaviour, which is that it may be, so there are interesting questions about this. And so this actually led to an extraordinary Canadian behavioural scientist suggesting that instead of making seatbelts compulsory, you should ban seatbelts and all cars should be designed so there was a massive spike in the middle of the steering wheel, pointing straight at the driver’s chest, making the point at although people would drive rather annoyingly slowly, they would be very, very safe.
7m: Analysing disfunctional behaviour
RS: The other thing we looked at which is very interesting. Korean Airlines had a massive problem in terms of its safety record, partly because Korean culture and indeed the language itself is immensely deferential.
MK: Power-distance ratio.
RS: Power-distance ratio. So what would happen is the junior pilot, they are not called captain anymore are they, it’s “pilot flying”… the reason they created “pilot flying” and “pilot not flying” is partly to get rid of this hierarchy.
In Korean culture you actually use different verb forms and different words to talk to people of a higher status than you than you do to people who are a lower status than you. It pervades everything, okay and so in power-distance terms, it is the most power-distant culture in the world.
The least I think is New Zealand. Now on Air New Zealand there is no power distance, the stewardess can come into the cockpit and say ’you’re going to crash the fucking plane mate’ right, that’s fine in New Zealand culture.
In Korean culture, even though you can see the plane is headed for the ground, the co-pilot or the junior person in the cockpit would say something like: "I think sir, we have found the altimeter most useful today," because that’s the most roundabout way you can actually get.
Anything more direct might be interpreted as criticism. It seems weird that we would die to prevent committing a social embarrassment. Actually social embarrassments are an extraordinarily high driver of human behaviour.
MK: But I mean we see this in organisations every day.
MK: Dysfunctional task behaviour, because no one says anything.
RS: No one says anything. What they did actually, they made the pilots speak English in the cockpit and not Korean. But one of the other things they did which we used for factory safety is we created a kind of word which was a neutral word, I think it was “pickle” in the factory and what it is, is if you see somebody else committing a safety violation, if you go up to them and say you’re committing a safety violation.
Technically in this area you should be wearing a so-and-so, it is very difficult to do that without looking like a smug, bossy git, so by having just a magic neutral word that anybody can say to anybody else, you break that social embarrassment bound of you know, the fear of kind of seeming like a git.
9m 30s: Solving problems using solutions that already exist
RS: Is anybody familiar with a thing called “Triz” which was a Soviet innovation. One of the things that “Triz” is very strong in is what we call lateral category analysis which is, one of the things is actually if you’ve got a problem, a complex problem, someone has probably already solved it, but in a different domain to yours and it has never occurred to you to transfer it over.
And so “Triz” spotted this that most problems have already been solved, but somewhere else, and people aren’t very good at making the linkage. Now I’ll give you a lovely example of this: I obsess about trains quite a lot. Now if you think about train overcrowding, right, the definition of train overcrowding is simply everybody standing up on a train.
Now (a) that’s a bit silly, because if you’re standing up for a short time, it really doesn’t matter that much, also I’d argue that if you’re resting your bum on a cushion at the end of a tube carriage, no one really minds doing that for a short time. So the definition isn’t very good to start with but there’s a more fundamental problem there in how we define train overcrowding which is, it doesn’t distinguish between a 100 people who have to stand 10% of the time and 10 people who have to stand 100% of the time.
I’m not talking about a four-hour journey; I’m talking about commuter rail. And I said actually psychologically you can solve probably 80% of the psychological irritation caused by train overcrowding by simply running two trains a day at peak time in both directions which are exclusively for season-ticket holders.
So basically if you’ve got an annual or quarterly season ticket, you can travel on these trains. If you haven’t, you can’t. The point about that is occasional travellers; 10% of the time that I travel on the Tube, I have to stand, I don’t care.
That’s just what I expect with the law of averages. If I had bought a season ticket and ended up standing every day, I’d be livid. And I suggested this. If you think about it, it’s much, much cheaper to do that than it is to solve overall untargeted overcrowding, which requires more trains, more frequent trains, faster trains.
And I pointed out this solution and someone said to me something really interesting that never occurred to me, they said ‘that’s how airlines do it’. I said ‘what? You don’t have a season ticket on an airline.’ He said no, no but if you fly to Frankfurt in economy 15/20 times in a year, okay, by about your 10th or 11th flight you’ll get put in the BA Silver level, now you’ll still be in an economy seat but your check-in experience, your luggage experience, your boarding experience and your lounge will be that of a Business Class passenger.
So basically they upgrade the experience for more frequent people so what you’re suggesting … and I went, "okay, you’re absolutely right," and what was so strange is that they’re not so different industries, are they. It’s rail travel - flying, you know, they’re both in the transportation industry but the norms of rail travel just assume the only way you can reward frequent rail use is through the price mechanism.
You give people a season ticket which gives them a discount. The idea that maybe they should have preferential treatment while they’re on the train like restaurants, you go to a restaurant every week, you’d expect a better seat, wouldn’t you, a better table and that’s what’s so fascinating about this that we’re blinded when we move from one domain to another.
MK: And it does link to what we were discussing before which is the difference between a singular experience and an aggregated scale experience and how human beings are very different to machines in that context so for example if you’re an individual as a human being and you are regularly using the train then that changes the nature of experience you expect but if you were a robot regularly using the train, it doesn’t matter about the scale.
RS: It’s exactly that, you’re absolutely right so if you think about it, economics is really bad on this because economics doesn’t understand the difference between how humans experience life over time.
RS: How economists look at the economy is in aggregate. Now it is theoretically possible if you think about it, to have a country which is getting richer over time where everybody within that country individually is getting poorer, and equally the opposite is also theoretically possible. A country could be getting poorer and poorer over time but the individuals within it were all increasing their salaries by 3% every year.
Now the weird thing actually might be that the second country might be happier than the first one, if you think about it.
MK: I mean it’s the same as the waiting room example that we discussed, at the hospital….
RS: Oh yes, the feeling of making progress.
MK: That’s right, if you keep getting put back in the same waiting room every time you see a different doctor.
RS: There’s a huge difference in happiness if you go to A&E and you see the triage nurse and the triage nurse says okay, you have to wait a couple of hours for the consultant, but he’ll be in in a couple of hours. If she then shows you into a new waiting room, you’re totally happy with the experience.
If you’re sent back to the original waiting room, you’re livid.
MK: A massive difference in the experience.
RS: Now, the objective experience is no different. The duration of your wait is absolutely not different at all.
14m 27s: Variance reduction and understanding human experiences in the commercial world
RS: If you look at the world and understand that the individual life is
lived is path-dependant and non-ergodic, variance reduction is rational so economics tends to view utility as an additive function. You gain a bit, you lose a bit and the order in which you gain or lose doesn’t really matter, whether you have three losses in a row or three gains in a row, it doesn’t really matter.
All it is is that gains should exceed losses; it’s a very naïve view. Now if you think about that, really, good fortune is multiplicative because as you said, if you’ve already won more in a casino you can gamble more ambitiously if you want to. People play more ambitiously with the house money.
Life in most ways is highly path-dependant and the difference between what’s good in average and what’s good over time, is not very visible to human beings so I would argue that financial crashes are… well Ole Peters of London Mathematical Laboratories is the guy you need to speak to here, he published a paper with Murray Gilbon the physicist saying there’s a fundamental error in economics which is it that it assumes that ensemble probability is the same as
If you think about it, in any multiplicative thing, and I argue this is really interesting self-serving ad-man’s argument but this is what I talked to Ole [Peters] about. Economists don’t really understand why people pay a premium for brands, right, my argument is what people are doing when they pay a premium for a brand is that they are paying for variance reduction but we never, when we talk ….
MK: Just explain variance reduction…
RS: All of you in the past year have been to McDonald’s, is that fair? Now, you can all afford to go somewhere better, I’m not disputing for a second that there are better places to eat than McDonald’s, the reason McDonald’s is the most popular restaurant in the world is because it’s got exceptional low variance.
You know exactly what you’re going to get. You’re not going to get ripped off. The loos are going to be clean, your kids will eat the food. You’ll find the food mildly enjoyable at worst and you won’t get ill. You can’t say that about a Michelin-star restaurant. McDonald’s isn’t going to give you the best meal you’ve ever had in your life.
You don’t take someone on a date to McDonald’s - that’s what Nando’s is for, obviously. But at the same time you want something where the risk of downside is very low,.
MK: It’s the predictability.
RS: I had a friend who worked for Schroeder’s years ago and everybody there was saying: “surely we want to get our fund into the top four performing funds for this year”. Schroeder’s had a completely opposite view, which was variance reduction.
They said no, what we call a good fund is a fund that avoids being in the bottom half for 10 years in a row and actually once you have compounding effect, not being terrible for a long period is much more important than being brilliant once and yet, if you think about how the financial industry is bonused and how it’s measured and rewarded, you think about it, it tends to encourage over-optimisation.
MK: Well, can we pause on this point. I’ll tell you why because it really brings into view the most critical lens that you bring to the commercial world and I will even have a crack at summarising what I think that lens is and you can comment on it. It’s to say that we have been using the wrong paradigm to understand human experience in the commercial world because we are treating human beings as if they do not have psychological life.
The rules are different when you understand that the people consuming the service, producing the service, making the product, innovating the product are living, feeling people as opposed to machines and when you look at the rules from that point of view, things seem irrational hence your book.
RS: Absolutely right. An economist’s definition of rationality is almost certainly not only wrong psychologically it’s even mathematically wrong, just to be clear about that. The whole of business particularly when it became turbocharged by IT, has spent much of the last 30 or 40 years trying to pretend that it’s a science.
Now that’s fine by the way. I I’m entirely in favour of things becoming more scientific. I’m not opposed to the scientific method, unfortunately the science they’ve all seized on, which what they think science looks like, is effectively Newtonian physics, you know, it’s schoolboy science where there was always only one right answer and the opposite of a right answer was wrong.
And one of the things I have to argue about by the way in psychology is the opposite of a good idea can be another good idea, so the reason people love this sort of mathematical modelling, they love AI they’re desperate for this stuff, is when you provide someone with an ambiguous recommendation, they can follow that recommendation and they won‘t get into any trouble.
19m 34s: Permission to fail: why defensive decision making should be avoided
RS: This is one of the things we really have to watch in business decision-making because one of the most common behaviours in business is effectively no one ever got fired for buying IBM, it’s called defensive decision making. You make the decision not which is likely to be best, you make the decision that is easiest to defend, that won’t get you into trouble.
And this is a very interesting thing we discovered. Working with British Airways, we couldn’t understand why there are only two flights, they are all Business Class and they only seat 48 people from London City to New York. Now we thought logically since so much of the traffic from London to New York is financial, we thought the flights to be over-booked every day.
Weirdly they are now down to about one flight a day and they are wondering about cancelling that. That’s very interesting, because you’re much closer to the city and you’re at Heathrow, it’s a really wonderful service what might be going on and this is what we surmised is this: If you’re a PA or you’re a corporate travel agent, you always put your boss on Heathrow/JFK, you don’t try Newark, you don’t try London City, you don’t try Gatwick.
If you put your boss on a flight from Heathrow to JFK and back, if anything goes wrong, he blames British Airways. If you do anything original or imaginative, he might blame you. You can’t ring your PA and say “what the hell do you think you are doing putting me on a flight from the world’s second busiest international airport”, but you can ring your PA from London City and go “if you hadn’t booked me from this sodding toytown airport, I’d be in New York by now.
So the decision taken is not necessarily the best one for the passenger, it’s the best one for the intermediary.
MK: That is spot on. Let me give you a little bit on an insight Rory, and acomment on the application of this inside our organisation, I will maybe give a little bit of a secret away here. Okay, here in the UK we have this compliance training requirement based on the regulator and everyone in this room is constantly getting these emails to say you’ve got to do your compliance Training and you log on and you do your thing, right.
Currently we hit a 98 or 99% success rate in terms of people completing it by the deadline. Prior to a particular intervention of behavioural science, we were at 67%, and I will give you the intervention that we did and its little secret and we were struggling and our compliance, people came to myself and the team and said look, we don’t know how to get people to do this.
They are not doing this. We send them threatening emails and so forth and we said the following, this goes to the little bit of understanding psychology. We said okay, what you do is you do the following: if someone isn’t doing their training don’t just send them an email, send their manager an email copying them.
It says the following, it says: dear Mr/Ms Manager, your employee Simon, and Simon is copied, is now over deadline. If this is not remedied by this time, both of you will be, yes you’ve seen this, those that haven’t done it, both of you will be reported to the Remco and this may affect both your variable reward.
It went from 67% to 98/99% in a single month and why was that the case? It wasn’t because of the money. It was because people are far more avoidant of upsetting their manager than they are of being threatened. The idea that your manager is now on the hook for something that you did and the speed at which the manager forwards the email to you and says ‘look, won’t you deal with this’ mobilises people to behave.
RS: Absolutely true.
MK: And it’s based on simple insight. The most scary thing you can have in the working world is your manager cross with you because you did something that affected him or her.
RS: And actually similarly the most valuable thing a manager can often do is provide permission and air cover for people to, as it were, for people to fly from London City. Do you see what I mean? Because the second the guy says to his PA actually to be honest I prefer City and Gatwick, she’s totally fine to book from Gatwick and City and this always happens with me.
I get actually can we have some flights from the City and Gatwick please, loads of Gatwick and City flights come back and I go on one of those. The second I have instigated it, there is no fear anymore.
And very interestingly by the way, do you know when you take penalty kicks to decide world cup soccer games, the person taking the penalty is more likely to score if he actually kicks the ball straight down the middle than if he tries for either side, statistically. And the reason this happens, the reason more people don’t, is because if you kick the ball down the middle you’re more likely to score but you look much stupider if you fail.
And the reason that works in the first place is because the goalkeeper is disproportionately disposed to dive left or right because if he just stands still and the ball goes either side of him, he looks kind of rubbish. If you dive left and the ball goes right, you’re unlucky, and likewise vice versa.
But if you stand in the middle and the ball goes either side or equally you are the person taking the kick and you kick it straight down the middle and the guy just picks it up, you look like a total dork, so the avoidance of social embarrassment actually reduces people’s propensity.
Now what you have to do if you want to overcome that is very simple, now obviously you don’t want everybody to kick it down the middle or they get wise to that, I think after the third kick, so what the manager can do is simply go to the second and the fifth person to take a penalty kick and go ‘I’m telling you to kick it straight down the middle’.
Now at that point if that person fails, he has plausible deniability because he can say ‘manager told us 2 and 5 have to kick it straight down the middle’.
MK: Permission to fail, I mean this is Elon Musk’s you know in the States, where they say "failure is not an option", very American idea, "failure is not an option". At Tesla their primary directive is "failure is an option".
MK: And that’s Elon Musk because of this "permission to fail" allows for the possibility of very different application of ones talents and outcomes and the degree to which there is permission to fail is proportional to the amount of innovation in a company.
25m 50s: Investing in the unknown: a better way to judge success
RS: You have to define areas where, I think everybody should be judged on more than one matrix which is, this is basically where you’re supposed to do your job.
Your job here is to do a low variance thing, then you equally have some space where failure is permitted because that space is a where…. A very interesting thing that bees do, I think I probably told you about this before, bees have this wonderful waggle dance system where they go back and inform the other bees where there is nectar and pollen, in which direction and how far away it is, amongst the bafflement of people studying this, about, it varies, but about 20% of bees or bee journeys ignore the waggle dance, they just go off at random.
At first this drove bee scientists absolutely insane because they said that you’d think of over 20 million years of bee evolution we would have had bee compliance officers who would have enforced higher levels of compliance with the waggle dance to bring on the nectar collection target in line for the forecast for Quarter 3, right, this kind of thing and then they modelled it as a complex system down to your point, if you look at it as a complex system over time in an unstable environment, if every bee obeys the waggle dance, you get trapped in the local maximum and the hive starves to death, because you get over-dependant on what you already know because you haven’t traded off by investing in what you don’t yet know.
The point of the random bee journey is that they need a very different ROI, an obedient bee journey must be energy recovered must exceed energy expended in recovery. It’s a fairly straightforward sort of what you might call an appraisal system for those obedient bees but then the other bees are kind of engaging in a RND activity where most of their journeys are a total waste of time.
But on the other hand if they discover a completely new pollen source, you know a field that’s just come into flower, that information is actually worth thousands of times that of an individual obedient bee journey.
MK: And we don’t do this in organisations.
RS: We don’t do this in organisations; we have a bonus which covers your whole performance. We should have these bonuses - here’s your obedient bee bonus and here’s your rogue bee bonus.
MK: Yes, I mean Ricardo Semler, you know his stuff, recommended this sort of like a period of work in which you do nothing. Your job description is okay, XYZ, ABC today but a good percentage of what you do must not be defined and you should not be at work and why that is, is because if there’s a, this is the disobedient bee scenario.
RS: Well if you think about it, if all the bees were disobedient also, this is the really important part which is a large part of the future is unknown and I argue this being a Vice-Chairman is a bit weird as a job title because you can define it as much as you like, and one of the ways I’d describe my job title is I have to spend a disproportionate amount of time doing the kind of random things that no one else would do.
So occasionally I would go to a conference and my PA would say why are you going to that? It’s that actuary conference; you know what are you going to that one for? The thing is if I go to a marketing conference, they are full of advertising people so if a prospective client is there I probably won’t even get to talk to them, but if I go to an actuaries conference, I am the only advertising person they have even heard of.
Now this is actually a really interesting point, which is I think we become so instrumentalist in the way we justify business activity that we’re preventing ourselves doing things which are obviously work but for undefinable reasons.
So if you’ve got any teenage sons or teenage daughters for that matter, they always go out on a Saturday night don’t they? If you asked them why they basically say, I might get lucky. That might mean sexually, that might mean having a great time, they don’t really know in advance. What they do know that if you stay home, you won’t get lucky.
And I said there’s a perfectly good justification for advertising which is just the more famous you are, the more likely you are to have lucky things happen to you. Lucky things that result from fame are more likely to be positive than they are negative unless you’re a crooked company or get involved in a scandal.
So don’t advertise if your Chief Executive is planning to have an affair with a goat, okay, it’s better to be not famous under those circumstances. But if you’re planning to be a kind of honest, decent, you know fair dealing company; make yourself famous because indefinable lucky things will happen to you more often.
Now a lot of this will never be measurable, right? If you’re a famous company, I guess that your Chief Executive could ring all but 30 people on the planet of importance and they will return his call, now if no one had heard of you the number of people would be, I don’t know, 10% of that.
People come to you with good ideas because they’ve heard of you. People work for you for a bit less, because you’re quite famous and it’s better for their CV to say they’ve worked for someone that’s a bit famous. Right, okay it’s much easier to recruit really good people. When I launched my book, a most extraordinary thing happened.
I never would have said look, it’s a book about behavioural science and it’s partly about marketing, partly about business. You know it’s not very specifically a business book, but it’s not a mass consumable necessarily and I would never have said, the first thing I’m going to do is get on the Chris Evans Breakfast Show because that would have been ludicrous, right?
It so happened that Penguin’s publicity bumped into someone who was a friend of Chris Evans and I got on the Breakfast Show and for five days I was outselling “The Hungry Little Caterpillar”. I was at number seven on Amazon after I was on that show. Now that’s never something that you can plan for specifically, it’s something you can only plan for generally which is if you’re a bit famous and make a lot of noise, more people hear of you and some asymmetric upside will happen to you through luck.
Now if you talk to really successful businesses, I was talking to the guy who founded Box.com and he was basically being despised by his mother-in-law while running a business selling lavatory paper out of a garage, when he happened to get on a TV programme about his business which he was running out of his garage selling loo paper online and weirdly the TV programme then got picked up by someone else on national TV that was syndicated and he went from effectively as a result of that totally happy accident, he went from working out of a garage maybe planning to get a second garage to a company worth a billion dollars.
If you talk to most billion dollar companies, a freak thing happens, timing for example for all those things. Also maybe being famous you’re actually given the benefit of the doubt more, that is also impossible to quantify.
32m 25s Measuring the unmeasurable: The value of doing things that are impossible to quantify
If we’re only allowed to do the things which we can predict specifically and quantify exactly, one of the weird things about it is it prevents us from doing things which are just fucking obvious, so I remember. I don’t know if you’ve used London Bridge Station or Victoria, they’ve got those nice quite wooden Scandinavian slatted seats they put in, well that’s a really good idea, proper seats at the station so if you’ve got to wait for your train and you can actually sit down.
They’ve got little tables so you can use a laptop or you can put your coffee down. What a brilliant, obvious thing to do because part of what will happen now if a train is delayed, is that instead of getting out of the way and sitting down and doing something, people stand like cattle in front of the departure board and going ‘ahhh’ and they get in the way of all the people trying to catch a train.
I thought how the hell do they justify that? How can they justify spending money on seats? But they had some sort of seat budget. At the bottom I saw they had put in “we found that the addition of tables increased the sale of takeaway food by 17%”. Now, we created a world where unless you can come up with a statistic like that you can’t do something that’s blindingly obvious like provide people at stations with a place to sit.
MK: Now this is what you know, drives me a little insane sometimes. We get these very hard-core operation minded lenses that say if it can’t be measured, you mustn’t do it and it’s a very popular business school thing, people out of MIT coming to talk to us and say you have to measure it otherwise it’s not gonna happen and you know, I always sit there and I think, you know it’s good to measure stuff, it’s very useful to measure it
RS: Your teenage son would, if you had a teenage son, [and asked him]: "Why are you going out on a Saturday night?"
"Oh I might get lucky."
"No, sorry I need a key metric. what’s the ROI? How much are you planning to spend on drink? What’s the value of a long-term life partner? You haven’t even done the maths have you? What’s the cost-benefit analysis of meeting a girl?"
Teenagers don’t need to do that because they go “if I go out tonight, I might get lucky and if I stay home I won’t”.
34m 30s Using relevant metrics: what gets mis-measured gets mis-managed
MK: In one of your first Ted Talks you give this wonderful story about the Eurostar. I’m not going to paraphrase you, I’d like you to just give it and then I want us to talk about the Gatwick example.
We met a couple of years ago when we discussed that Gatwick brief you had to compete with Heathrow on the security, but if you can give the Eurostar one and then we can apply our minds to Gatwick as an example and then we can open.
RS: It’s very important because what gets measured gets managed, is the famous phrase. But there’s a counterpoint to that because what gets mismeasured gets mismanaged and if you’re measuring and obsessing on the same matrix, and if the only matrix you’re allowed to use are what you may call scientific and objective and not psychological, I think you can gets things badly wrong.
So I had, before I realised they were going to build High Speed 1, I had some issues with some issues with spending £6bn to £8bn reducing the journey time on the Eurostar from London to Paris from I think 3 hours 20 to 2 hours 45 or something like that, even though it was significantly slower than flying end to end, everyone was travelling by train anyway and the reason why everyone was taking the Eurostar rather than flying wasn’t about duration, it was about the quality of the time.
Because on a train you plonk yourself on a big comfy seat and you can walk about, go to the loo and have a meal, and you stay there for about 3 hours 20 minutes until you arrive in central Paris and you get off. So you have 3 hours which are either productive or enjoyable, depending on what you want them to be.
On a plane the journey may only take two hours 20 but basically the problem with plane travel is every 20 minutes you have to do something else. You know you turn up with your luggage, you check your luggage in, you go to the security line and then you go to the thing, then you go to the shop, then you go to the gate and then from the gate, you board the plane and then the plane goes for a drive around Heathrow for about 25 minutes, and then you take off fly for about 35 minutes and then the whole pain in the arse stuff starts again at the other end.
So it’s not surprising that the people would much prefer the time and I said the whole question of reducing journey time is not always the deciding metric. Now bear in mind, I will be clear here, this is particularly important that in your complex systems world of both psychology and complexity, lots of things aren’t linear.
I’ll make a big caveat here. Before the French introduced the TGV it was four hours from Paris to Lyon, Lyon being France’s second or third-biggest city, the TGV reduced that to two. That is a really, really big deal, because now you can go to Lyon for the day from Paris or vice versa. You don’t need to stay in a hotel and there’s a big difference between getting up at 7 o’clock in the morning, arriving in Paris at 9, doing a day’s work and getting home at 7 o’clock versus the four-hour alternative.
In the case of High Speed 2 with Manchester, I’d argue the difference between 2 hours and 1 hour 10, particularly when you factor in the overall end to end journey time, isn’t a game-changer in the same way? Because a trip to Manchester before High Speed 2 will involve, just as it does now, a unit of time which we businessmen technically call ”a day out of the office”, and that’s not going to change, whether it’s an hour 10 minutes to Manchester or whether it is 2 hours 10 as it is on Virgin Trains?
Maybe all that will happen is maybe they will start the meeting a bit earlier, because there’s a faster train so as humans we will benefit not at all. Also I don’t know if anybody else travels by train a bit but the journey by train is actually the most productive part of your whole week, it’s the meeting that you’re going to that’s a waste of time generally. But my argument was very simply: look, if the purpose of the train is to be better than the plane, trying to compete on speed which is the one area where the plane will always win, is to some extent a mugs game.
You’ve already lost that battle. If someone is in a desperate hurry, they are already going to fly. So what you have to do is arguably, you’ve got a budget of £6bn and you think the best use of £6bn is genuinely simply to reduce the journey time by about 17 to 18%? Then I said, I could take 1% of that budget, £6bn, what’s that? £60m, why don’t you just put Wi-Fi on the trains? I’m not making this up.
That is a far bigger comparative advantage over flying, the ability to actually work with Wi-Fi for three hours uninterrupted is a far better way of competing with British Airways than trying to make it less slower than what it was. And it’s 1% of the cost. After they spend £6bn building the new railway it was another 12 years before they actually introduced Wi-Fi on the new trains.
They still haven’t got it on the old ones. I said if you really want to spend a billion pounds, I said basically hire all of the world’s top male and female supermodels, get them to walk up and down the train handing out free glasses of Chateau Petrus to all of the passengers.
You would have saved £5bn and people will ask for the trains to be slowed down.
39m 44s Social pressures: insights underpinning behavioural science
RS: There’s a really interesting thing about queuing, which is fascinating. One of the, sort of four or five insights underpinning behavioural science, one of which is conventionally economic logic is not a predictive guide to human behaviour in the real world, because humans are more complex than that.
They are not a kind of Newtonian thing. The second one is that we don’t genuinely know why we feel the way we do about things and the explanations we give tend to be post-rationalisation, not accurate descriptions, so one of the strangest things about queuing which we only discovered laterally and I partly discovered it because I had an accidental experience and then I Googled it, nobody will ever tell you why don’t you like queuing, we always talk about the wait. It wastes my time.
I really, really hate just standing around like an idiot and a bit of that is true, what nobody will tell you is that a large part of the stress of being in a queue is the people behind you. Would you believe it? Because when you go through airport security your lizard brain is conscious of the fact that there are between 15 and 20 people looking at you and you’re convinced that if you make the slightest mistake going through that x-ray machine there are 20 people saying ”dickhead, he’s forgotten to take off his belt. What a loser.
And as I said, fear of social embarrassment is massive. One of my most important findings is that man will never order a cocktail in a bar unless the menu has a picture of the glass in which the drink is served, because if the male brain thinks there’s a 1% chance it will turn up in a pineapple or with a sparkler and an umbrella, he’ll order a beer instead.
And if you think about it there are some really weird psychological things, the most ridiculous drink in the world when you think about it is the Martini, because in objective terms it’s a seriously hardcore drink. It’s basically neat spirits and they put it in a glass where, unless you’re auditioning for James Bond, no guy will feel comfortable in a bar holding a Martini glass.
So that’s why there is really clever innovations in hipsterisms which is put it in a Mason jar. Guys will drink anything in a Mason jar. So that’s a really clever idea, the Mason Jar cocktail idea, but it takes away that huge male anxiety of oh please god, don’t let it be a pineapple. But with Gatwick the fascinating thing is that when you split people, you know that business now when you tend to go through airport security and there’s six of you in parallel with at most one person standing behind you, the stress is totally diminished. It’s a really, really weird thing.
The psychology of road rage is quite fascinating because if you think about it, the guy in the car who’s annoying you mostly has his back to you which unconsciously we perceive as a kind of insult.
RS: It’s kind of aggressive and so a lot of these things are really fascinating because we basically know how we’re feeling but the reason for the feeling we can’t attribute, and so that Gatwick experience was fascinating because one of the first things you’ve got to do is when psychology is involved, ask a really dumb questions, because nobody has asked them before.
So one of the questions I asked rail companies was why don’t people like standing up in trains? That sounds weird, but actually is it that they feel ripped off because they paid for a seat? They don’t know by the way, you can only experiment. Is it because part of the reason you’ve got to stand on a train, you’ve got to hold on to something to stop yourself falling over. So now you’ve lost the use of one arm, you can‘t read a book, you can’t look at your phone, you’re basically staring into space.
You’re worried about your bag which is on the floor. Now, I just said a very simple thing. If you redesign trains so that there’s a trade-off between sitting and standing, so when you stand you get a bum rest, you get a table, 2 USB chargers, a view out of the window, a cup holder and hook to put your bag. Put the seats of the train in the middle of the train so you don’t get a view, all you get is a cup holder and a place to put your bag, you don’t even get a table.
Standing actually becomes a choice not a compromise. And you can actually make people happy, and this is a really important thing: when I say the opposite of a good idea can be another good idea, that’s because the way humans perceive things depends on context and not on objectivity. For years in China Town there was a restaurant called Wong Key which was famous for being the rudest restaurant in London.
Now normally being rude is a bad thing in a restaurant unless you’re famous for it in which case guys on Friday night go “hey let’s go to Wong Key, they were really rude last time“. And so I always call this the bus to the airport thing. Tell a story about something and you can turn a weakness into a strength. This goes back to Chaucer who, Chaucer invented the phrase in English to make a virtue of necessity - and humans like to do this - if there is something that is a requirement we like if we can tell a positive story about it.
A fantastic case - you land at an airport and you’re expecting an air bridge and the plane stops and the engines wind down and you’re still a mile from the terminal and everybody on the plane goes “oh shit, it’s going to be a bus” and “oh my god, I’ve been cheated, I’ve been short changed”.
And then there’s this pilot who said something ingenious. He said “I’ve got some bad news and some good news. I won’t be able to get you an air bridge because there’s a plane blocking the gate but the good news is the bus will drive you all the way to passport control so you won’t have far to walk with your carry-on bags,” and we all looked at each other and said: “Hold on, that’s always true isn’t it?”
MK: Yes, mostly the process is faster.
RS: But no we weren’t aware, we just framed it as the poor man’s air bridge and we framed it as the inferior version of the air bridge so it must be pretty bad and suddenly this guy had told a story which made me see it as a convenience, as a conveyance and not as an inconvenience.
MK: It’s why people stand up the moment the seatbelt light goes off, they stand all on top of each other when in fact it’s totally unnecessary.
45m 50s Reinventing the wheel: balancing short-term and long-term decision making
How should we balance short and long-term decision making?
RS: A great book by Robert Cialdini, he makes the point that the human brain disproportionally pays attention to what’s new. The way the human brain works is we disproportionately think important that which we pay attention to, and so when new technology comes out, when you think about what happened with the Kindle a bit, which is an interesting technology because it’s kind of reversing in its popularity, it hit a kind of asymptote of adoption and people were actually buying more books for that while again.
And part of the reason is when something is new, we notice the things that is better about that thing than the old thing more than we notice what might be worse. So we go “wow, that’s amazing because it allows me to do all these new things, therefore those new things become very important”, and we go “wow, this new thing is amazing”. In the process we forget what was perhaps better about the old thing.
Now that obviously over time what’s new changes and I always call this again. In cheddar cheese, for a long time all cheese had rind on it and someone in the 1970s, maybe the late Sixties, chopped off the rind and sold it as a premium as rindless cheddar, and about 20 years later all people copied this and so all cheddar had no rind on it and so someone left the rind on and called it authentic farmhouse cheddar with added rind, and charged a premium. And so in a weird kind of way one of the reasons fashion is so strange is because of this weird human distortion of noticing and I think we do an awful lot of that, we notice what is better about the new thing.
Online shopping - actually I think Amazon is hugely overvalued. The natural assumption of people who sort of look at the growth of Amazon is that online shopping will be 60% of all shopping. I don’t think that’s true. I think it will actually hit a sort of threshold or a ceiling sooner than Amazon investors say.
MK: Well we have seen that with books and we have seen it with cinema because I remember consulting to a kind of Barnes and Noble type shop that was terrified ….
RS: The fax machine is really interesting because it’s the one thing that did really die.
MK: That died.
RS: Except in the NHS where apparently it still has use but …
MK: Then you still have the Ronier, the Gestetner machine, but as I say with books, the end of books was the digital book that came out, because now we see more books being sold, physical paper books being sold than ever before because people want a book.
Cinema: video was going to kill cinema but it hasn’t, you’ve just got a very particular type of cinema.
RS: The novelty thing is also true with airports, when I was a kid the rich kids at my school had been to Schiphol, then it was Changi, then it was Dubai Airport. They would come like, it’s amazing, like there’s shops and restaurants and everything, and walkman is absolutely incredible and apparently blah, blah, blah and I think okay, and then suddenly all airports became a shopping centre didn’t they?
And suddenly people started going London City is brilliant, there’s hardly any shops and you get through the whole thing in about 10 minutes and in 1970 London City would have been a crap airport, because there weren’t many shops and that’s also interesting by the way, the whole airport design question is quite interesting because the retail environment of an airport doesn’t distinguish between a 100 people who fly once a year and 10 people who fly 10 times a year.
And actually that’s a really interesting question because the people who fly 20 times a year do not want to buy luxury goods. It’s not a special occasion and so what’s really interesting there is aggregate information often sort of provides nobody with what they want, because if you look at an airport, it is actually two very different kinds of customers
MK: It’s all things to different people.
RS: If anybody has got a BA Gold Card they have bypassed that whole, you’ve got your own security language that takes you straight into the lounge because they basically realised that people who fly that frequently don’t want to go shopping. And so I think that’s the kind of interesting thing that was developing.
50m Calibrating risk: the avoidance of downside variation
Why do people prefer variance reduction?
RS: It is partly the avoidance of downside risks, because if you think about it, if you take a risk with a one in 100 chance of dying and you take that risk a hundred times, you’re quite likely to be dead, so it’s partly the word calibrated for taking a risk a series of times, not once. The other thing is very simply what Ole Peters would say is a very simple thing really, this is the most childish maths in the world okay.
There’s no difference between 3 + 1 + 3 + 1 + 3 + 1 and 2 + 2 +2 + 2 + 2 + 2, right, there’s no difference in fact between 4 + 0 + 4 + 0 + 4 + 0 and 2 + 2 + 2 …. When you multiply 2 x 2 x 2 x 2 x 2 x 2 is a much bigger number (64) than 1 x 3 x 1 x 3 x 1 x 3 x 1 x 3 which is 27 and so Ole argues and we really need to know this because apparently it’s really important in finance.
I think it’s interesting in marketing because it explains why people pay a premium for brands. They are not paying for the fact that the branded, the Samsung TV is better, they are paying for the fact that when you buy a Samsung TV, you pay 100 dollars more, but it’s less likely to be terrible.
You know it’s somewhere between good and really good, whereas there may be a cheaper television which is somewhere between brilliant and goddamn disastrous. If you think about it, who shops on eBay? Interestingly, if you were totally logical, if someone had a 96% approval rating on eBay and was selling something for 20% less than somebody with a 100% approval rating, in a completely rational world, would you go well, there’s a 4% chance he doesn’t sell me my goods, but I’m getting them at 80% of the price, therefore I’m up on the deal.
For whatever reason nobody does that, if you’ve got a 96% approval rating on eBay you can’t sell anything to anybody, even at half the price. You have to go lower than that. So we have painfully calibrated our approach to that, essentially, if you want to maximise the time average growth rate 2 x 2 x 2 x 2 x 2, a series of distant long gains is better than a high variance thing around the same means. It is difference between an arithmetic mean and a geometric mean.
So the fact that humans essentially like McDonald’s, you know, basically I mean there isn’t normally a McDonald’s at the airport, now that sounds strange because I can claim for my food at the airport, why don’t I want to go to Jamie’s Plain Food. My reason is very simple, I’m going to be on a plane for nine hours in a pressurised metal tube, the most important thing is I don’t want to get food poisoning. Okay, I don’t want an Adam Bolton incident, I don’t know if anyone saw that on Twitter.
MK: We’re back to where we started.
RS: Yes, exactly, full circle.
MK: That’s right. Did he have the wet wipes though?
RS: No, I think Beth Rigby had to supply them,. But anyway the point is that there are many conditions where actually what you want is low variance, constant growth rather than what appears to be a better than average return with a high level of variance.
And so to some extent you know, he would argue that what behavioural scientists have called ”loss aversion” and have labelled as a ”bias” isn’t a bias at all, it is actually really, really intelligent and so that has huge implications because apparently the distinction between time averaging and ensemble averaging is becomes really, really acute in the financial system where there is a lot of credit available with very low interest, essentially when there is just loads of money switching around because the difference between the two become more acute under certain circumstances.
MK: So Rory thank you very much. I think we will have to bring it to a close now. Thank you everybody for attending and Rory especially to you.
RS: It’s a pleasure. Thank you.
Focus and its related content is for informational purposes only. The opinions featured on the site are not to be considered as the opinions of Investec and do not constitute financial or other advice. The information presented is subject to completion, revision, verification and amendment.