The Exchanges

Q Do you think we're looking for efficiency in the wrong place?

A We usually do, uh, in the sense that when you pursue efficiency, there are quite a few problems, but when you pursue efficiency, generally you start looking at numerical or mechanical factors. And of course, in the process, you disregard psychological factors where the greater gains may be found, and so you focus too heavily on cost reduction and too little on value creation. I mean, one of the greatest forms of efficiency, by the way, is employing a human being who's really, really nice. Now, this is complete anathema to people in tech who love to define business processes so as to make them susceptible to automation. You know, person X does this. We will take that function, we will replace it with algorithm Y. And it's a very beguiling message, because it usually comes with cost savings attached. You might have heard of my thing, the doorman fallacy, did you?

AI assessment note: “We usually do, uh, in the sense that when you pursue efficiency”

Q Is that an example of where the costs are really visible, but the benefits are? Absolutely correct.

A Yeah, so you, it's very, very easy. Management consulting firms, uh, if you, if you're in a business and some management consultants come in, go to the management and say, are they on a gainshare agreement? A game share agreement is a management consulting scam where you claim a certain percentage of the cost savings you identify in year one. Now, as Roger L. Martin, your fellow Canadian and my own personal Spengali says, any idiot can cut costs, ok? What takes real skill is cutting costs in a way that doesn't destroy value. One of the things that I don't think is understood by tech nerds, and if I'm being really cruel, you know, males in general, in many cases, is the extent to which in evaluating any business or experience, the human component of it, the face-to-face component, Does really, really heavy lifting. I've got a lovely story to illustrate this, which I think is fantastic. It's the absolute perfect example of misalignment of optimization through quantification bias. So wonderful man, Alex Batchelor, used to be the marketing director of Royal Mail, but, you know, similar to what you have with Canada Post, right? USPS. Yeah, sorry, not US, yes, just USPS. And they couldn't make any sense of the fact that the brand perception of Royal Mail bore no relation to service levels. So there would be districts and areas where, you know, every single first-class letter arrived …

AI assessment note: “any idiot can cut costs, ok? What takes real skill is cutting costs”

Q There's so much opportunity here, right? Oh, if you think of you competing in a world where we're moving to AI-driven call centers, uh, if I took the exact opposite approach, which is, by the way, you should do both.

A A really, really good service organization allows for very streamlined, efficient service for people who know exactly what they want, and extremely empathetic service for people who are undecided, uncertain, or find themselves in an unusual situation. And so you need to do both, and that's, that's where I think the tech bros have got it all wrong, because they see the opportunity of tech as being a one-way street towards ever greater efficiency, streamlining, and let's face it, tech bros are not neurotypical in terms of what they want from things. When you allow tech bros too much power over decision-making, along with their running dog lackeys in kind of management consultancy, You're optimizing for something which may be very, very distant from what your real-world customers really care about.

AI assessment note: “And so you need to do both, and that's, that's where I think”

Q So what would be an example, I guess, in your mind of how that will play out with self-driving cars?

A Well, it's an interesting question, of course, um, One of them, which would be that if you're in a very large number of parts of the world, you employ a driver or a taxi driver, partly for security, not, not only for actually maneuvering. I mean, okay, it tends to coincide with countries where you have low labor costs, but most people in say, you know, large parts of, you know, let's say Pakistan, you wouldn't hire a car and drive it yourself. And part of that is just navigating local customs, it's local knowledge. Um, now, to an extent, some of that can be replaced with technology. But the fact that actually, if you have a local driver, no one's going to dick with him. Whereas if you've just got two American tourists in the back of the thing, um, uh, you know, they're quite vulnerable. A second one, by the way, with human driving is that Google's mostly tested this thing in California. Two things about Californians. There are very few pedestrians, and I would argue that, that Californians have a very low level of mischief. Okay.

AI assessment note: “you employ a driver or a taxi driver, partly for security”

Q What is it you're representing when you say tech pro, though?

A Um, anybody in a tech industry, just as anybody in advertising, myself included, is Overweighted towards a belief in the ability of advertising to solve all problems, and I will freely admit that I am guilty as charged. Um, however, and people don't get that much opportunity to convince other people of the rightness of their argument. What has happened is that in marketing, for example, the tech companies, the consulting firms that sell stacks, and I owe a lot of this to a guy called Adel Borky, very interesting, self-taught Libyan marketing writer, former boxer, Anil Borky, and I coined the phrase technoplasmosis by analogy with toxoplasmosis, which is tech people and consultants have taken over the finance department so that the marketing metrics the finance department has faith in and demands from marketing are not those metrics which are most conducive to building brand and customer value over time. They're the metrics which are most conducive to selling tech solutions to the companies.

AI assessment note: “anybody in a tech industry, just as anybody in advertising”

Q What's different about today? Why is everybody on such a hair trigger? Why do we identify so strongly with extremes?

A I'm gonna say, actually, that everybody always blames social media I would argue that the mainstream media, it is always in your interest, and this is a problem in journalism, it's always in your interest to provoke a fight, because then you have something to cover. Nobody's interested in reading about peace and harmony, they're interested in reading about discord and argument and dispute. By the way, that's an evolutionary tendency. If we're sitting here, and we hear, we heard a fight break out across the street, we'd be there with our noses pressed up against the glass. If, if we had, you know, people amicably discussing the weather, we wouldn't pay it the slightest attention, you know, where, where we've fundamentally evolved to pay attention to, uh, conflict.

AI assessment note: “it's always in your interest to provoke a fight, because then you have something”

Q I see what you're saying. I know we're coming up to time here. I want to get to a couple of questions. What are the rules of good copy? If I asked you to teach me how to write good copy... How would you do that?

A I think a large part of it comes down to, you And his idea of a tacit skill, which is we know more than we can tell. There are some generally good rules, which is write conversationally, much more conversationally than people think they should write, because everybody thinks they have to write a, and actually, I'll give you two examples of this. David Ogilvie, all his books are incredibly readable, because I think he was a copywriter first and an author second. His prose style is very good, and he also adopts a clever trick Which I've stolen, and which, uh, a few other people, is that he writes extremely plainly for the most part, but will throw in the odd sesquipedilian long word, just to remind the reader that, you know, you're not an idiot. You know, it's almost there to flatter the reader as much as it is to flatter the writer. Conan Doyle, I was talking to Rick Rubin about this, he and I both grew up on those Sherlock Holmes short stories, which are not only models of thinking and deduction and, uh, Uh, you know, and a fantastic lesson for, uh, mental gymnastics, I think. They're also, um, Kingsley Amis believe this, that he's one of the greatest prose writers in the English language, because, bear in mind, a lot of that stuff's written in, you read a lot of stuff that's written in, you go, fuck, does that, fucking, what's that mean? Yeah. Oh, God, hold on, I'll have to go…

AI assessment note: “There are some generally good rules, which is write conversationally”

Q Are you defending Bud Light? I want to hear this.

A Yeah, yeah, no, no, no, no. They could not have necessarily anticipated that because it was confected outrage. Look, I'm politically on the right, ok? I get just as angry from confected outrage on the, on the right, which is, oh, they've gone woke, da-da-da-da-da-da, when, you know, you're merely showing You know, a mix of ethnic groups here advertising or something. This is bull, right? And the right do it, and the left do it. Uh, there's confected outrage on the left, I would argue, about, um, American Eagle, and it's confected outrage on the right. It's done for signaling purposes by a very narrow group of people. Do those cases actually damage the business? I don't know the figures. Gillette did. There was an extremely, Now, this wasn't confected outrage, because Gillette ran an advertisement. They realized they had to move on, perhaps, from the best a man can get, despite the fact, obviously, that their customer base is overwhelmingly male for fairly obvious reasons. And they produced an ad which I would argue, and even my wife would argue, was needlessly, um, insulting towards men, in that it conflated the Me Too movement with barbecuing.

AI assessment note: “They could not have necessarily anticipated that because it was confected outrage.”

Q into some ethical stuff about, um, behavioral economics and even advertising and creating intangible value, but, um, First, let's talk about reading a little bit. I know you're a big reader. Uh, what's your process for selecting what you read? I mean, we're all struggling these days with filtering a massive amount of information, and, and you, out of all people, I, I know we've had conversations on this before.

A I, I'd like, I, I'd like to take credit for this. I have to admit that in the last two or three years, I've probably fallen behind. Um, I don't know why that is. Partly, um, I, I partly blame email. I had a friend who was a barrister who, as part of his job, would effectively have to read hundreds of pages of sort of court papers and transcripts, and the biggest resentment he had about this was he said that when you have to read for work, it slightly kills reading for pleasure, because by the time you get home at the end of the day, your, your mode is to zonk out in front of the television, uh, like a zombie, um, Because simply because you've, you've read enough that day, and you've been paid to do it. I'm, I'm less good than I was. I think I was very, very lucky in who I chose. Uh, the early inferences, Robert Frank, Bob Cialdini, um, uh, Nudge was obviously a very important book. Um, Some slightly unusual things. Quite a lot of evolutionary biology and evolutionary psychology. Geoffrey Miller's The Mating Mind was one of those game-changing books. Uh, then he wrote Spent. We, we invited him actually to, uh, give a talk at the, the Institute of Practitioners and Advertising on the links between evolutionary psychology, sexual selection and signaling, and consumerism. Um, that sort of stuff, uh, uh, has fascinated me. Some of the happiness literature, uh, Um, but no, I probably…

AI assessment note: “I think I was very, very lucky in who I chose.”

Q So how do you, how do you take people one step away from that tribal loyalty and more towards a middle?

A Well, there's an interesting one for a start, which is that if you take the protests, there's a total asymmetry to the business of protest, isn't there? Which is that basically right-wing people don't do the demo thing, do we? I mean, the only reason a right-wing person would stand on the street holding a placard would be to advertise a golf sale, right? We're not going to, you know, we're never going to write weird messages like stop the Tory cuts or whatever it might be, and just march around. I don't know why conservatives don't do that. Height may say it's deference to authority. Okay, we've chosen the damn government. We've had the decision. That's it. It's all systems go. It's why, in some ways, conservatives Right wing political parties, certainly the British Conservative Party, one of the magic ingredients it has is that it falls in line. So there are a lot of people who have a huge disagreement with the leader of the party, but there's a conservative kind of deference to, uh, Arbitrary. I mean, you're Canadian, ok? You've got this, both of us have this totally arbitrary head of state, right? I mean, it's absurd. I mean, no one would have designed it that way. Actually having an arbitrary head of state is actually really good, because it means that if the election doesn't go your way, it doesn't force you to reassess your loyalty to the country. The fact that, I mean, t…

AI assessment note: “having an arbitrary head of state is actually really good, because it means that if”

Q So what makes Dyson so good at marketing?

A Well, the story I heard from someone who worked at Dyson for 12 years is that he was sitting in a presentation, James Dyson presiding, and they were presenting the call center sort of efficiency statistics. By the way, I'm absolutely fanatical about call centers because I'm terrified that in an AI age, people will try and effectively automate them. Uh, I think what you should do instead is make them slightly smaller, but a lot better. Mm-hmm. Because I don't think you can substitute for the human in cases of unusual problems, special circumstances, empathy, generally. Don't get me wrong, I think you can get AI empathy. I'm not suggesting that AI is going to be completely unempathic, but at some point it's rather like the equivalent of I want to speak to the manager. There will be situations where you want to speak to a person who can understand your specific situation and has the power to intervene to override the normal Rules and regulations.

AI assessment note: “By the way, I'm absolutely fanatical about call centers because I'm terrified”