This interview has been edited for clarity and length.
In Part 1 of this interview, Rory Sutherland asked about the limits of relentlessly pursuing efficiency—whether in advertising, or in other areas like transportation. A slow but comfortable train might not reduce easy metrics like journey times as much as a high-tech project like Concorde, but can have bigger impacts on how people travel and work. Does the same principle apply to other fields, like advertising? And we talked about the changes in the terms and conditions of office work. Rory continues:
Previously, you could have said, “I think I can work remotely,” and people would say, “yeah, in theory, prove it.” Well, two years of working remotely did kind of prove for quite a lot of people that it was perfectly feasible, in some cases, to be more productive, and happier and richer, working in a different way.
Now people will say, actually, I can’t really go in and ask for a pay raise necessarily. But one of the ways in which I can benefit myself is by negotiating what I call the three “free”s. There’s free time, which is leisure, there’s “free when” which is autonomy of when to work, and there’s “free where” which is where you work. None of those things were variables before the pandemic, and all of them are inevitably variables now. I don’t mind working very, very hard, but one of my things I liked during the pandemic was I could work until 6pm. Have a bit of a nap and watch TV. And then work again, between, let’s say, 10pm, and one in the morning. Which if your next working day is online, and starts at 9am, still allows you eight hours of sleep. I’m 56. I can’t work 15 hours at a stretch. But I can work 15 hours in a day if you give me a break in between.
Or you can work for an East Coast company and live on the West Coast. And you get up early, but then you have some time in the afternoon.
I have worked for a London company on the West Coast when I stay with my brother-in-law in LA. And it’s actually fantastic. You get up really early. By about lunchtime, everybody in Europe shuts up. So you can go to lunch and have the afternoon entirely to yourself. It’s kind of magical. And also, I think that you don’t have to be fantastically insightful to realize that there’s a danger in many, many forms of knowledge work of a kind of hyper competition, where you simply work hard and your reward is, yes, more money, but you become even busier. I was just off a call actually with some former London lawyers. And they said in organizations like consulting, banking or law, where you’re competing for partner, in the 1970s and 1980s, by the time you made it to partner, you were then a bit of a panjandrum, and you could kind of slack off a little bit, because your capacity to dispense wisdom made up for your need to work very, very hard. And they said, the description now of the competition to make partner in law is, it’s like a pie-eating competition where the prize is more pies.
I see the point, which is that people are starting to realize that other than the financial reward, there is no other reward mechanism which companies historically could give. Because I think people value autonomy in and of itself, I don’t think it’s just a question of how hard I work. If you’re in a creative job, you kind of know, weirdly, the conditions in which you’re most productive. Some people go to a coffee shop, some people work on a train. In the film, The Lincoln Lawyer, he always works in the back of a chauffeur-driven Lincoln, because his argument is that he’s just more productive when he’s on the move. Actually, I think most people, after a few years of work experience, get to learn that a disproportionately small number of their best ideas occur to them in the office, because the office itself has become increasingly monotonous. The open plan offices didn’t offer you the opportunity to hack your environment, according to what you’re trying to do. So you couldn’t really find solitude and you couldn’t really find sociability.
There was one study that said open plan office employees interact less than people in cubicles.
In fact, you’re more likely to email someone within line of sight in an open plan office. And increase the number of internal emails, even when the people were theoretically within shouting range. So you know, all kinds of weird stuff like that. And that’s partly because if you had an office, you had automatic privacy to hold a conversation, because you don’t know where a conversation is going to go.
Right. In this conversation, we’re talking about a variety of things.
Absolutely. I hate doing this, because I hate whinging about white collar employment, because there are many, many people who have far tougher jobs. But there is one thing about knowledge work, which offers you spectacularly little flexibility. In blue collar work, you can find work, which enables you to work around other commitments you might have in your life. If you’re a taxi driver, when you near your 60s, you don’t have a retirement day, you wind down. The idea of retiring on a single day, going from five days a week to zero, you’re making a hell of a bet that you’re going to have good health for the next 10 years, or indeed survive for the next 10 years. Whereas if you taper off, it’s a fairly good way of trading off risk between you and your employer. And of course, overwhelmingly remunerative white collar work is conducted in the world’s most expensive real estate. As a Georgist, I see the Zoom revolution is absolutely necessary. As much as happened in Silicon Valley, it’s the landowners who have ultimately made the biggest returns.
Yes, the whole Silicon Valley economy only makes sense if the landlords and the limited partners in the venture funds are the same people.
Very, very interesting indeed. So that way, they make money either way.
Exactly. You remember the famous job ad, which was something like “Come join our fast-moving team to build the technology that enables creative people to collaborate from anywhere in the world. Must be willing to relocate to San Francisco.”
I would argue very simply that actually technology only means something when you can see its effects from the air. You know, all the really meaningful technologies had effects around where people worked and lived, or how work was organized. And actually, Ogilvy must have made literally tens of millions from laptop ads showing people looking at a spreadsheet next to a lake. That was always the promise. In Microserfs, Douglas Coupland says the ultimate premise of what you might call the nerd economy was to make location irrelevant to the performance of any task. And therefore, it’s completely countercultural, in a sense, for this to require you to be in San Francisco to actually perform the very work that is supposed to be enabling this.
One of the things that drives me nuts is fatuous comparisons between video conferencing and real world meetings, because they don’t take account of the opportunity cost. If you and I had wanted to have this conversation in the physical world, it would have cost $4,000, or it would have had to wait nine months for one of us to cross the Atlantic. There’s a huge opportunity cost to requiring colocation, and I’d also argue there are certain functions where I think people prefer online meetings. And I think for your relationship with say, a financial advisor, Zoom is perfect. I don’t want him to come to my house so I’m going to have to buy biscuits, or put on a tie. I don’t want all that, but at the same time, I don’t want an email relationship. There are quite a lot of business relations which are best conducted this way. Most of the time. I’m not saying I won’t buy the guy a pint every now and then. But nonetheless, most of my business with my financial advisor, I’d like to conduct over a video link.
There are certain employers that say, our people need to be co-located to do these, whatever amazing things together. And one of the employers that’s most famous for this is Apple, but their latest round of laptops was probably their best yet. They came out with a MacBook Pro that has all the ports put back on. The product manager at Apple who was in charge of taking ports off of the laptops must have left their dongles and adapters at the office in March 2020.
There’s a difference between someone moving 50 miles out of London and someone moving to Austin. Someone who’s 50 miles out of London can still travel to London for the day, perfectly feasibly. Someone who lives 100 miles from London can do it every couple of weeks. So there’s undoubtedly a need for colocation, but not all the time. And whatever the right ratio is, for most knowledge work, if it’s completely measurable, what the people are doing, why should you care where they are? And in things where they aren’t measurable, you should probably allow them some discretion to work as they see fit.
There were a few outlier companies that were extremely remote or extremely decentralized before the pandemic. Have you seen those companies doing particularly well with the current environment, or did the knowledge spread across organizations faster?
I was weirdly a pioneer of this because back in 2018-2019, I actually encouraged my team to work from home on Fridays. People will probably work better if you give them some variety of setting. Some variety of their general habitat, if you like. And so luckily, all my team were fairly well geared up to know how to do it. Because it’s worth noting, I’ve made a few radio programs. I was perfectly familiar with the idea of just speaking to a microphone, but it takes a few months to get used to that. My hunch is that most of the required skills are acquired fairly quickly. Not least, because it was no longer forgivable to be crap at it. There was that slight mode among senior people where being rubbish at video calling was almost a status marker. I don’t have to concern myself with this stuff. And that was killed pretty rapidly. There’s a great paper, actually, by Noah Smith, the economist, which talks about distributed service sector productivity, and argues that you can achieve now, through Zoom, the same kind of productivity gains that were obtained through electrification of manufacturing, when it was realized that not everything had to be centered around the steam engine. In other words, when you replace big steam engines with big electric motors, you don’t see much of a productivity gain, what gives you the productivity gain is lots of small motors in different places, you can turn on and off.
You can organize the machines by whatever is most productive, not by having to run a belt up to a big shaft.
Now, interestingly, one of my arguments in defense of Zoom, is that there are people you only want to employ some of the time. I said to a client of mine, I’m the classic example of someone who would be quite generally, without blowing my trumpet, if you have an interesting problem, I could be pretty valuable to you for an hour, and I could be valuable to you for a day. What you wouldn’t want to do is to take me on full time. I’d start setting things on fire or creating chaos. So that ability to inject eccentric expertise into organizations on demand, rather than having to make the decision, we either have no access to this, or we put it on the payroll. If you think about how filmmaking works, it effectively works through assembling teams on the fly for the duration of the film, you use the people and even the equipment.
And I’ve been generally proselytizing this at Ogilvy. Maybe this brief requires half an hour of a Nobel Prize winner. There’s probably someone somewhere on the planet who could transform our understanding of this problem in a 20 minute conversation. And most people on the planet, I think, would give you an hour of their time for 1000 bucks or less on Zoom. A hell of a lot of people on the planet would not give you an hour of their time, if it required their physical presence. Because if I commit to an hour in Amsterdam, that’s actually a day and a half out of my life. And actually, it’s a week out of my life, because I now can’t work on a week-long project because of a trip to Amsterdam right in the middle of it. The opportunity cost of a two hour meeting in Amsterdam is actually three days. Where’s the opportunity cost of an hour Zoom? I could be on holiday that week, come in from the beach, do the hour, go back on the beach. It prevents me from doing nothing before or after that particular hour.
And you can bring in the organizational tools to best manage that specific person’s time. Is that what the Ogilvy behavioral science group does, when some company has a problem?
I always offer what we might call surgery. Before you commit to anything serious, let’s have a conversation for a few hours on where we might be able to lend useful insight into creative solutions to your problem. Now, you can do that physically, if those people are based in London. If people are based in Palo Alto, and I go and say to my boss, look, there’s probably a 10% chance of half a million dollars worth of business here. Will they say, “That’s great, four of you get on a plane, and stay in Palo Alto for two nights?” I don’t think they will, because the odds just don’t stack. Whereas if it’s an hour on Zoom, you don’t even require permission.
What we’ve got to remember is that we tend to feel stupid bigging up technologies that are 20 years old. So saying, “I think video conferencing is really important” makes you look a bit like a Johnny-Come-Lately, because it’s been kicking around in the shape of Skype and goodness knows what else for quite a long time now, certainly since the turn of the century.
There was WebEx in the 90s. There’s definitely been flaky video conferencing since the 90s. And at some point, it got to be stable enough that the kind of senior people you’re talking about can no longer justify not being able to do Zoom, because now it’s just a button in your calendar.
I think Zoom was a perfect example of that, which is that by giving the meeting a URL, they effectively demanded of someone a behavior which was completely instinctive, and essentially system one. This is the location of the meeting. But the other mistake of Skype was to use the model of the phone call, not the model of the meeting room, because in a B2B context, it doesn’t really work. If I were dealing with the chief executive of Rolls Royce Aero Engines, I’d no more phone him proactively over video, than dress up in a clown costume and jump into his office shouting “Surprise!” It’s just not an appropriate way to actually convene people. And so by making the space kind of neutral, and a space that people join, that was actually really quite an important psychological leap forward in terms of the comprehensibility of the technology, I think.
It coincided with cultural changes around business use of the telephone. Now you don’t just pick up the phone and make a phone on someone else’s desk ring.
I have a couple of friends who still do that. Look, it’s three o’clock on a Thursday afternoon, there has to be a 50% chance I’m on a video call. There’s a 10% chance I’m on a video call speaking to 75 people. You really can’t do this anymore. And if we get back to the advertising question: how we pay for content needs to change, and the whole advertising model needs to change to acknowledge the value that’s created. That may be immeasurable, but it’s still very real, in terms of the convening power of a great media brand.
And today we are in the process of changing how web advertising works. The web browser is becoming the marketplace for delivering the ad to the user, and then sending information on the delivery of that ad back to the advertiser. And a core principle of the World Wide Web Consortium is that the browser works for the user. So what is the value that advertising provides to that user? And how can the browser help them get more of it? What is it about an ad medium that makes the user willing to participate in advertising?
That’s an interesting question. Because if you ask consumers, they all say, I don’t want any remarketing at all. I don’t know what I think about that, because consumers would say it’s absolutely of no value to them, because they notice it when it’s not relevant. So we’ve got to be slightly careful here, because consumers will disproportionately notice remarketing, or, for example, recommendations when they’re spectacularly irrelevant. But personalized advertising when it’s relevant, they just think of it as a service. So we’ve got to be a bit careful about asymmetry of perception. I’m not sure that consumers are capable of saying what is most valuable to them from an advertisement. A signal of value is actually of value to a consumer, but I’m not sure many consumers will say I like advertising, because it’s an honest signal of value.
You can’t go just by what people say about advertising, you have to go by how much value does it provide to that user in the form of ad supported content or providing useful information about some market.
And actually, of course, if the browser actually came with a pay for content ad free function that would undoubtedly meet the definition of the consortium, wouldn’t it? If I would rather pay for this content in its purest form.
The problem is that if some people pay for ad-free content, then the ads only go to the people with no money, and now the advertiser won’t pay.
So in many cases, the package deal might actually create more overall utility than the ability for people to opt in or out. Hence what you might call the positive sum game that used to exist and got eroded.
So from behavioral economics first principles, how do we reinvent the positive-sum advertising business?
We didn’t invent it. Ironically, we didn’t understand it to begin with. It just evolved, I think.
Of course, of course. This is absolutely fascinating. So does the consortium think it is close to reaching a consensus on how to do this?
Well, one problem is who’s in the room. And who we’ve got showing up to W3C is a lot of those links in the chain in the middle, but not so much the actual content creator, the actual consumer, or the actual advertiser.
Do you have a vision for how you can actually ensure that essentially, content creators are adequately rewarded, whether by payment or by advertising or both, for producing content? Broadly speaking, actually, what you might call the convening of audiences around relevant content did quite a good service for advertisers, didn’t it? It didn’t allow for remarketing, it didn’t allow for cookies. It didn’t allow for that kind of thing. But in terms of producing fairly relevant, believable, convincing, advertising to accompany the content, it didn’t do a bad job.
Any ideas? What is it about an advertising medium that can make it more constructive?
A better medium would definitely have a better balance of what you might call the probabilistic and the reductionist. Undoubtedly, looking at people’s past behavior is a sensible thing to do in determining what they’re going to do in the future, particularly for niche products. And there is a danger that we create an advertising mechanism, which is totally ill suited to special interest products, unless those people will read special interest magazines. So there is that risk, you know, the fact that in The Daily Telegraph, you can advertise fishing equipment to fishing enthusiasts, but to replace it with a malt whiskey ad for whiskey enthusiasts is of value. But at the same time, the net effect may be that the value of serendipity, of discovering customers you never knew you wanted, gets destroyed if people take it too far. And it’s difficult to know how to kind of calibrate that, I think.
And as a user, in the new world of web advertising, you’re going to have ultimate control over that because the browser works for you. So what choices would make the user better off?
Or could you make the whole thing completely transparent? What would people choose, if they were given a kind of feedback mechanism? What would that tell us? Would there be a mechanism where actually there was a trade off between how interesting the ad was and how much they had to pay to serve it?
And then the whole other question is attribution: which information about what ads did I see is a user willing to share? In the old school customer survey days, the brand could say, we see you just bought a washing machine, where did you see our ad? And then they check a little box and say, I saw it in Good Housekeeping or whatever. And today, there’s a lot of discussion. Do we want the browser to report that kind of thing automatically? And when is it in the user’s interest to report that they saw an ad?
And then of course, there are ethical questions. Obviously, for things which might be compulsions, like gambling, the ability to turn off advertising for certain categories might be very, very important. Because that’s another thing that genuinely concerns me, which is that it’s fiendishly difficult to remove yourself from exposure. I don’t know what recovering alcoholics do. But are they served a huge amount of alcohol advertising? That’s slightly problematic, really, isn’t it?
Yes. The online casinos are very good at picking up on people who recently quit gambling. So that’s a nasty problem. And sometimes people’s employers have an interest in their privacy as well. If your employer is making drones for the Department of Defense, your employer may say, we don’t want to have your interest in infrared cameras shared with who knows what country.
Really interesting challenge, because you’ve got three, hopefully not necessarily divergent, interests to serve. But creating the mechanism that serves all three is not easy to do.
The print advertising industry did it when they were literally breathing the fumes from molten lead all day. How hard can it be?
Fair point. They made it work. So what the first question I suppose to ask is, what aspects of the old traditional advertising system actually worked better than we realized? In other words, what we tend to do when something new comes along, we see where the new thing is better than the old thing, because that’s typically very salient. But we often lose benefits of the old thing without noticing it. It’s rather like the Kindle question, which is, people like me made the crashing mistake of assuming that 90% of books will be ebooks in 10 years time. And in fact, it hit a sort of ceiling, a surprisingly low ceiling. I’m not even sure that more paper books are produced and sold than ever before, even after the advent of the Kindle.
Personal computers vastly increased paper consumption at first, but now a lot of people do home office without a printer.
The famous Jevons paradox is that when steam engines became more efficient, coal consumption actually went up. Because there were more applications for steam engines.
If a light bulb becomes more efficient, and you spend fewer dollars per month keeping the light on, then you’re going to put your light bulbs all over the place.
So one of the slight curses of LED lighting is probably that we will see quite extensive use of pretty wasteful outdoor lighting, for example. The strange thing was that content brands, the general advertising media, threw away a lot of the value they added. They were so desperate to make money in any shape or form online, that they were happy effectively to sell space to the lowest bidder in the process, effectively failing to command any premium for both the audience they had convened, and what you might call the halo effect of the surrounding brand material.
As a publisher that’s participating in online ads, you’re not just enabling cheap, low budget advertisers to get onto your pages. But you’re also enabling the advertisers who would have paid more to pay less.
So you’ve created a race to the bottom. At the extreme level, if you tried to place an ad for incontinence pants in the first 50 pages of Vogue, they’d basically refuse. I did want to ask a media buyer just as a joke to try that out and see if they’d eventually ring up S.I. Newhouse or something, if you offer them half a million. So in other words, the advertising space ceased to be differentiated, I suppose.
And don’t forget that long chain of intermediaries. So every intermediary has an incentive to get a little more efficiency. So that means if you can get the same audience or credibly claim to get the same audience at a slightly lower ad rate, then that’s a win for that link in the chain.
Of course, even though it’s not in the interest of any of the three parties who the ecosystem intends to serve. Ultimately a quantification bias problem, I suppose, isn’t it?
It’s got a bunch of behavioral economics issues in it.
Of course, it’s worth noting that print publications, kind of partitioned their advertising inventory, didn’t they, in that you had big pages towards the front of the publication, classifieds and small ads towards the back? And then there’s the final question, your idea of buying Twitter and running one ad a day, which unfortunately, Elon has kind of preempted you in that?
There may be an acceptable way to actually impose that partitioning, because there’s also the problem that there isn’t a digital equivalent of what you might call the first double spread in Vogue or the back page of a newspaper. The internet doesn’t yet allow for what you might call the big display play.
It’s not obvious to the reader if this is a big ad buy or a narrowly targeted one.
So if you don’t have that mechanism of costly signaling made visible and unconsciously visible, you’re losing out. I don’t know whether you’re spending a million pounds messaging this to 20 million people, or whether you’re spending 57 cents just picking me. If you don’t have audience size as part of the heuristic, then the mechanism doesn’t fully work, or spend size. But as you quite rightly said, people who were high on lead fumes managed to solve the problem. So is it impossible for us?