Feb. 7, 2022

Paying Attention

This week we look at the vague way in which various groups have measured the concept of time spent or tried to parcel out the attention economy. How do you really measure attention, and even better, how do you measure the quality of attention or time spent in this attention economy? (Repeat)

Apple Podcasts podcast player badge
Spotify podcast player badge
Google Podcasts podcast player badge
Overcast podcast player badge
Castro podcast player badge
PocketCasts podcast player badge
RSS Feed podcast player badge

This week we look at the vague way in which various groups have measured the concept of time spent or tried to parcel out the attention economy. How do you really measure attention, and even better, how do you measure the quality of attention or time spent in this attention economy? (Repeat)


Will Page: Welcome to Bubble Trouble, conversations between the economist and author Will Page, that's myself, and independent analyst Richard Kramer that lay out some inconvenient truths about how financial markets really work. Today it's time to pay attention. More on that in a moment. Welcome along. Last week's trouble [inaudible 00:00:25] this week's bubble. Now it's time to pay attention. Richard, to you.

Richard Kramer: Well, Will, you've just written this book Tarzan Economics, in part about the attention economy. And one of the things I've always been really curious about is the vague way in which various groups have measured this concept of time spent or tried to parcel out the attention economy. And usually it gets neatly divided up into one of your favorite, uh, graphics, the pie chart, as if there are just no overlaps between time spent watching television, accessing apps on a smartphone, listening to the radio, or doing other tasks. So the question I always ask is how do you really measure attention, and even better, how do you measure the quality of attention or time spent in this attention economy?

Will Page: Before we get into unpacking those pertinent issues, I think it's worth us stepping back and thinking about the syntax, the language that we use when we say paying attention, and also its history. So we think about a language of paying attention as a currency to that. It's not always that way. In Spanish we land, in French we make, and in Swedish we give attention. But as English is still the biggest language in the world, the rest of the world will have to get increasingly used to the currency of paying attention. There is a monetary value to our attention. That's what the school teacher's asking kids in the classroom. That's what we're asking you to do now, is to pay attention.

I think what will help build a backdrop to this podcast is we just take a quick lesson from history. I wanna go back, Richard. I wanna go way back to the history of Mark Twain. Mark Twain's famous for many things. He once called golf a good walk spoiled, a great tribute to a Scottish word, there. But what he also did, was he invented something called a scrapbook. And I think a scrapbook is going to be relevant as we get into this topic this week.

What the scrapbook was, was a response to an oversupply of attention gathering media distractions. Because of the industrial revolution, there was an explosion in the amount of print content that was available, and Mark Twain's solution to this oversupply of content was to allow you to order it in some way. And he invented the scrapbook with tabs, with order, with the ability to lay it out in the format that was familiar just to you, to gather your attention. Now, Mark Twain's scrapbook, fine, although it's worth noting that this is not a trivial matter. He made almost as much money from the patent for the scrapbook than he made from all of his literary works. But if you look at Facebook, Richard, and you look at the newsfeed, now, let's think about that for a second. That's 1.6 billion people, all of them getting a unique source of news arranged in their own, particular way. Not one of those newsfeeds is the same.

Richard Kramer: Well, I don't think that's so different than the scrapbook in that the items you would have had stuffed in your scrapbook as a boy would have been different than your nextdoor neighbors or the, the kid around the block, or someone with a different set of interests. When we think about that shared attention, those events around which we all share the same experience are increasingly rare. Maybe they're sporting events, maybe they're political events. But many people are always seeing them refracted through a different prism, through a different lens, if you will. So I think we've always had our own, unique scrapbooks that were reflecting our own personal interests. I don't think that's so new. It's just that we've got algorithms serving up the contents of those scrapbooks for us today.

Will Page: From at theoretical perspective, where I'm coming from is how we've been here before. What Mark Twain saw was the ability to go from our one to many model of broadcasts to many relationship of a network economy, a narrow-cast where everyone's treated unique. And perhaps that's what Facebook is doing 150 years on from Mark Twain. So let's go from my fascination with the theory and the syntax of the word paying attention to your specialism of advertising. Give me the kind of groundwork. Lay out the groundwork for me.

Richard Kramer: Well, when you look back on the early days of the internet, it was all about monetizing eyeballs, which is a gross concept if you're not a surgeon specializing in macular degeneration or some other sort of eye surgery. The idea was that as long as you could ... And you're gonna love this word, capture someone's attention, then you could sell them something. And from this evolved a split in advertising into two broad categories. One of them was about brand advertising. And, and if you look at a traditional marketing funnel, at the top of that funnel is just generating awareness and nudging you into a consideration set of thinking that you might actually buy a product. Uh, and it's just really making you aware that this product exists.

Now, the other side of advertising which got turbocharged by the internet and the possibility of personalization was called performance-based advertising. And that was getting you to take some sort of action, whether it was clicking on something, or even actually buying something. Now, the first one was measured in a very vague concept, uh, a recall brand lift, uh, appreciate of the qualities of the brand, while the second one was supposed to measure actual sales of a product after it had been advertised. And maybe the percentage increase of sales of those products after a campaign had been run. The basic premise of all of this was that you had a finite amount of attention and advertisers were competing to capture it, to grab it, to, to take it for themselves.

Will Page: So we've got an interesting balance of how we measure attention. The first one reminds me of the term like perhaps attention lite, spelled L-I-T-E, as in Miller Lite beer. Whereas the second one's more direct and engaged in results and measurable events. There's something else when I wanna get ourselves into here, which is contestability. That is you might be competing for lite attention, I might be competing for direct attention to produce a sale. But we're all competing for your time. So how do we understand contestability?

And in the book, I, I, I lay out a framework to think about your friends and foes, your complements and your substitutes in this battle for attention. And one of the things I find fascinating there is just how some forms of attention are stackable, whereas other forms of attention are monopolizing. That is, if they win, everyone else loses. And if they win, that means everyone else has increasing less time to compete for. I just toss you out a quick example. If you think about Netflix, and let's say Narcos or Breaking Bad. If you binge watch 10 episodes of Breaking Bad, that's 10 times 48 minutes that Netflix has won that everybody else has lost. So I'm really interested-

Richard Kramer: Hm.

Will Page: ... in understanding the attention economy from a contestability framework.

Richard Kramer: Uh, I guess one of the problems with contestable markets theory and the i- that idea is that while you're sitting there on the sofa watching your 10 times 48 minutes of episodes of Breaking Bad, you could be doing other things. And one of the points about brand advertising is that it usually takes half a dozen or a dozen repeated exposures to a brand before you take any notice of it. Uh, I'll give you a good example. If, let's just say Coca-Cola launched a new pomegranate flavored Coke. The first time you heard that, you might say, "That sounds weird."

The second or third time it might start to sink in, and you might remember it. And it might only be after you've heard the phrase pomegranate Coke a dozen times before you went into the supermarket and you said, "You know what? I'll try one of those." And then you'll see if you become loyal, if you are loyal, to Coca-Cola as a brand, to the notion of soft drinks, or to the flavor of pomegranate. But that doesn't take away from the fact that you might also be loyal to other sorts of brands of drinks. And drinks are not gonna be the only category you're investing your time into making those choices around.

Will Page: So we've got attention lite, attention heavy. We've got contestability where you can have complementary forms of attention. Then you've got substitutional forms of attention, those which win, which means everyone else has to lose, which means, then again, we've all got even less time to compete for as a result of that monopolistic attention. Talk to me about recall and how we're trying to measure our attention as it currently stands.

Richard Kramer: Well, to me one of the big fallacies in measuring this whole attention economy is that we find it really hard to measure the quality of attn. And that came out in study after study showing that, hey, the average American household might be watching four, or five, or seven hours of TV every day. But that really only reflects the amount of time that the set was on in the living room, or the bedroom, or some other room in the house. How much time are people actually spending in rapt attention in front of that TV set, and do they get up all the time whenever those ads come on? So one of the challenges of the internet was through clicks or some other action, find out if you actually paid attention to that advertising message and responded to it.

And as it turns out, some of those clicks may have been accidental. Some of those clicks may have been generated by deceptive means. It doesn't really speak to the quality of attention because a click is a poor substitute for real engagement with a brand, or a topic of interest, or an advertisement. So it's why it's so interesting to see the skip ad button on YouTube ads for example. Because if you actually take the time to skip the ad, it may tell the advertiser you're not interested in the product. It may tell Google or YouTube that you just find the ads intrusive. But it doesn't necessarily speak to the quality of your attention, how quickly you might press skip ad, or whether you'll let the ad play because you were distracted by someone else in the room at that time.

Will Page: We've come a long way from worrying about whether the TV viewers putting the kettle on during the commercial break, is that fair?

Richard Kramer: And equally, all of the ads that come on during the, the breaks in sporting events are usually when people will get up, and they'll go to the fridge for a beer. They'll go to the toilet. They'll do something else in those 15 minutes in between the two halves of a football match.

Will Page: How does that work in a Netflix era where a lot of our real deep attention to television content or content that happens to be on that screen? It could be TV, it could be Google Chromecast. Different means to the same end, but if it's Netflix that we're watching, we're no longer used to TV commercial breaks. So A, I spend less time watching TV content with commercial breaks. And B, when I do, I'll probably be more reluctant to even consume them in terms of what's this irritable interruption doing to my TV viewing experience?

Richard Kramer: It's a great comment by Scott Galloway saying that, "Advertising is really a tax on the poor." In the sense that those-

Will Page: [laughs].

Richard Kramer: ... who can't afford to strip the ads out of their content consumption are subjected to them.

Will Page: You-

Richard Kramer: And one of the real question marks around ad free offerings of content is whether the value that those content owners get from direct subscription payments are gonna, in the long run, be greater than the value that they would get by selling advertising space around them.

Will Page: This why they say soccer will never take off in America because you have 45 minutes of uninterrupted sport where you can't insert commercial breaks.

Richard Kramer: Well, and you have, again, you see those little banner ads that flow across the bottom of the screen while you're watching the match that are branded content from an advertiser. In the content of American sports, you see so many of these excessive replays from multiple camera angles of individual plays. And unfortunately, football doesn't lend itself to having the extra time in between plays. The break between face-offs, or down in football, or in between innings in baseball where you would replay the same scene again and again, and introduce commercial breaks.

Will Page: On my end, what I see as the most fascinating trend in attention is to learn from gaming, especially gaming during lockdown. Richard, do you know the game Animal Crossing?

Richard Kramer: I ha-

Will Page: So when that launched in April of last year, just as the global economy went into lockdown, the average duration, the average game time spent playing Animal Crossing was a whopping nine hours. It's insane. Music, it's about two hours a day tops. Separate between your commute to work, your commute home, and whatever breaks you have for audio listening time in between. But for gaming we were seeing people play that game for nine uninterrupted hours. The were forgetting to eat. So I, after the break, I wanna come back on gaming, explore what we can learn about attention there.

Richard Kramer: Well, Will, having read through Tarzan economics and thought about the attention economy, I guess my biggest question mark would be around ah- how does one find value both as a consumer and as an advertiser in those long duration sessions? If you can get someone to watch a four hour art house film, they really have to be paying attention to it to the exclusion of all those advertising messages. And within Animal Crossing in a nine hour duration of gameplay, how does an advertiser get their message in? Or will the game itself, in paying for it, be enough to say, "I want to have a space that's free from advertising for a period of time."

Will Page: Coming up in a moment, we'll look at the importance of understanding the quality of attention that's being bought and sold. For Richard Kramer, I'm Will Page, and you're with Bubble Trouble. Back soon. Back again with part two of Bubble Trouble. This week, we're asking you to pay attention, note the currency reference in that language, there. Paying attention. I'm seeing a lot of trends in attention, particularly in gaming, where I'm noticing that the average time spent playing a game like Animal Crossing was nine uninterrupted hours. People were forgetting to eat. But that's the new world. That's, that's, b- attention that you can measure very accurately, because there's an engaged person participating in its time. I, I wanna turn to the old world, which dominates so much of the ad dollars today, the linear world, Richard. So much of that linear world of broadcast, of radio, the way that they're measuring attention, it doesn't feel very modern, doesn't feel very tech savvy. Would that be fair?

Richard Kramer: I think nowhere do you see this battle between reach and conversion between brand and performance ads play out more than in traditional linear TV versus online video. Traditional TV seems to capture attention mostly around big tent pole events, notably sports events, which is why there are such huge sums put on the table for sport rights. It's one of the very few places national advertisers, in any country, can go to reach a mass audience reliably every week.

Online video is far more fragmented, where advertising could be far more personalized even if the ability of the advertisers to surgically select highly interested audiences have been consistently overstated. Now, sure, if you go onto a part of YouTube which specializes in fly fishing, you'll be able to find highly targeted ads for rods and reels. But generally as you can see by the number of wildly inappropriate ads which are run as pre-rolls ahead of content on any number of the online video services, at best you can think that advertising might be selected on the basis of geotargeting, where you live, or some other vague characteristic. Because generally that targeting has been proven to be a myth. And sure, Will, when you're up at your parents' house in Scotland, you might see a different set of ads than you do down in London. But the idea of it, they could identify you as a Hearts fan based on your online activity and figure out what the propensity of a Hearts fan is to buy a certain product. So let's just say that's been really overstated over the years.

Will Page: I wanna pull out some smoke signals from this linear advertising world just now. I mean, firstly, before we actually start smoking those signals, let's just look at how radio and television is measured in 2021. Is my understanding correct that we're actually measuring recall diaries?

Richard Kramer: Look, there's a whole range of attribution and measurement techniques for every form of media, from digital fingerprinting and counting clicks from your web activity, to building up third party cookies in your browser, some of which maybe, no longer be passed on under new privacy policies from various big tech players, through extremely vague panel studies, which find a sample of 1,500 to 2,000 people in a given country, measure their behavior on the basis that they are a representative sample of the demographics of the country, and then extrapolate that up to the entire population of the country.

Will Page: I've never understood that idea that 2,000 people's a great way to understand the motivations and preferences of 333 million Americans. But if somehow some statistician justifies it.

Richard Kramer: But, Will, talk to me about how you feel about these diaries, about listening diaries for radio, or, or watching diaries for TV.

Will Page: Sure, I mean, and radio I think it's the most embarrassing. If you think, here we're doing a podcast, and what's been developed in podcast analytics is really impressive. Time of day, city level, base level, targeting activity as well. Yet radio as it currently stands is still measured on a recall based diary, manual, filled out digitally, but still a manual process which asks you what did you listen to last Tuesday? Well, six days ago. How am I supposed to remember? I'll just tick the box for what I usually listen to on a Tuesday and assume that was correct.

And that's literally how the British radio industry is measured by [inaudible 00:18:29] and a very similar story applies to [inaudible 00:18:31] new service samples are not that great. Tiny fractions of the population, a drop in the ocean. Not just that, but they're measuring based on a containable data set whereas what you've seen in the s- explosion, this proliferation of media content, it's very hard to contain. I mean, if I don't ask the question did you spend one hour watching TikTok last night, then you're not gonna give me the answer that you spent one hour watching TikTok last night. So there's a lot of boxes not being added to those manual diary questionnaires. It's a worry for me that we have such diversification in what's drawing our attention these days, but the measurement tools are still defined so narrowly.

Richard Kramer: Well, I'll tell you, there's another dimension to that, which is that if we fast forward in the future and assume that all content will be digital, and all of our consumption will be on a limited number of highly identifiable devices that can be associated with our own personal consumption, then that gets you to the starting point of saying, "We can tell all t media that you might have been engaged with at any given time." However, the notion that you can constantly monitor the consumption of 330 million people and slice and dice that, much less process all that data simply, and then map that onto an equally diverse or frighteningly larger set of potential products that could be advertised to those 330 million people, that's a big ask. And it becomes a question of how much computational power, and algorithms, and coefficients you want to find among those groups of potential customers to pick those many, many, many needles out of the giant haystack?

Will Page: Let's come to what I think our listeners find is the most valuable part of our podcast, which is our smoke signals. And they've been paying attention now for 20 odd minutes. Let's give them a couple of smoke signals they can look at for to work out where the bubble troubles are in this world of paying attention. Whereas Herbert Simon once said in 1961, "A wealth of information leads to a poverty of attention."

Richard Kramer: Hm.

Will Page: Where are those smoke signals?

Richard Kramer: Well, I'm gonna ask you the same question since you've written a book on this. But for me one of the big smoke signals to look at when I'm looking at companies that are listed on the stock market, is instead of their talking about time spent or monthly active users, which are very nebulous concepts, and they're often defined in a very wooly way by the companies themselves. So if a company sends you a push notification once a month that gets you to open up their app, which you then realize is garbage and you quickly close, they will certainly include you in the population of monthly active users. Instead, I think the, the question that needs to be asked is what sort of quality of attention is being bought and sold by the advertisers themselves or being obtained by the companies?

Will Page: I'm feeling that quality's important. For me I use a different term. I think about the connections. So you can get time spent, that could win your heart. Doesn't necessarily win your mind. It's not necessarily the case that the more time is the better. You know, you go back to the history of the browser, the original Google versus Yahoo wars, Yahoo viewed time spent using their browser as a positive, more time the better. Then they would try and flock you lots of other forms of distraction while you were there. Whereas Google said, "Time spent is a negative. Less time spent the better it is for the browser. Let's get you onto Google, then get you where you want to go." And we all know who won those browser wars because Google took a more productive, efficient use of [inaudible 00:22:01]. So I agree. I mean, I think time spent in, you know, we have to get beyond that. But I think it's connections that we need to grow towards as well. How do you get the-

Richard Kramer: Well-

Will Page: ... hearts and the-

Richard Kramer: ... well-

Will Page: ... minds?

Richard Kramer: Will, talk to me about the area you know best, which is music. So music i- to me, strikes me as something where you capture a lot of time spent, but you capture a very shallow pool of attention because you can obviously have a lot of kinda music playing in the background, you can really listen hard to that music if you're an aficionado, but a lot of us frankly put that music on in the background and we don't give it our full attention. So how did you, as chief economist at Spotify think about that attention economy with respect to the music that you were trying to get people to pay for?

Will Page: I think about it exact same way as the late, great David Bowie, who left this planet five and a bit years ago and said famously around the time of the millennium, he said, "The way that music is going it's gonna be like running water coming out of a tap. Always there, running in the background." And what I've always said about music is it's achieved two things that nobody thought it should, could, or would. A, it's become more valuable. And B, it's become less relevant. You find me one person living in London today who can tell me who's number one on the singles charts. Nobody knows and nobody cares.

Now, it's so many playlists there that are there to fill your time, but not there to grab your attention. If you get on the Peaceful Piano playlist as a jazz pianist, great. You've got to three million new fans. None of them want to know your name, 'cause that's not the purpose of the playlist. It's ... Back to David Bowie, we have to understand that music is capturing a lot of time, but it's not capturing a lot of attention.

Richard Kramer: But that speaks to the fragmentation that we were talking about earlier, and the fact that we're losing touch as a society with those big, shared events. The number one song, much like the number one television program that was watched by the majority of the population. And there are fewer and fewer of those events or pieces of content that get widely shared across the generations, across demographic groups, and so forth. It's that attention is increasingly parceled out into ever smaller bits.

Will Page: Plus one to that. And I have always described it using the, the visualization or imagine that long tail chart and you understand that the tails got longer, it's got thinner, there's more and more competing distractions on the market right now, which means you need more relative investment to stand above that increasingly crowded space. It's very interesting to think about the dynamics of where some choice is better than none, but more choice is not better than some. That's a classic paradox of choice that we're dealing with today, and how do you compete for attention in a market like that? I, It's, it's one that we've gotta watch playing out.

Richard Kramer: The Paradox of Choice was a terrific book by, I think it was Barry Schwartz and for me, my classic example of the paradox of choice is going to a U.S. supermarket and standing there, trying to choose a mustard. And you have literally hundreds of variants of mustard to pick from. You have honey mustard. You have horseradish mustard. You have German mustard. You have every manner of mustard you could imagine. When it's ultimately paralyzing. You just wanna be able to reach onto the shelf and pick one thing without spending all the mental energy. So I think that quality of attention focused on each of those individual decisions can end up being quite a burden for people.

Will Page: It's very true. And y- the flip side to that is our famous curry house in Edinburgh called Khushis, which offered only three dishes. And the start of their menu said, "You know about Indian restaurant where they offer you 30, 60, 100 dishes? We put all our energy into making these three the best dishes you could find in the city." So again, the scope of limitations of choice.

Richard Kramer: So my other big smoke signal to bear in mind, and I think we're all incredibly guilty of this in the world of digital distraction, is the way in which our attention can be divided by the many internet services we've come to m- rely on in our lives. And a lot of those services are simply distraction engines waving, "Look at me. Look at me. Look at me constantly." And they can actually have a lot of negative connotations for a brand or a services. And it's incredibly difficult to model the human psychology, the customer journey, whatever you wanna call it that says, "This is the right time to actually insert a clever marketing message." A lot of it is done by brute force, a lot of it is done very crudely, which is why we find so many of those ads so intrusive. But the notion that we're all paying rapt attention to whatever is in front of us all the time is just a fantasy.

Will Page: So with our smoke signals in place, here, I think what we're able to do for our listeners is just give those calling cards about understanding claims of our attention. Find there's a trade off between the quality and the quantity. The fact that time spent has its limitations in gauging how much valuable time was spent, and there's shortcomings in how we measure the big attention merchants of TV and radio now in 2021, when there's so many advancements happening elsewhere. I guess one of the big undercurrents that we have to deal with in avoiding bubble trouble is the fact that there is only 24 hours in a day. And i- if you stacked up all the attention claims of these merchants out there, be it selling advertising, or putting out IPOs, or trying to build us, uh, an investor story is there is only 24 hours a day and something's gotta give. What's the one way our listeners can start to understand what gives? How can we, we help them put that constraint on there? There is only 24 hours in a day, and only so much can be shared around.

Richard Kramer: Well, again, I think when I go back to the two smoke signals that I've described, the first is to take a look at the metrics that any company is providing about itself, for example, monthly or daily active users. And say, so ask the basic question, "What is daily active user? Does it mean that they've interacted with your app for more than one millisecond during the day? Does it mean that they opened the app, or the app was opened for them in the background?" Because a lot of apps are obviously open on the background of your device, and if you forget to close them, are you a user from day-to-day?

And the second point we, with the second smoke signal I try to lay out is that our attention can be divided in many ways. And how many of those services are really capturing the full extent of user attention if they're even relying on advertising services at all? Some services like Netflix, they're trying to capture your attention, but not for means of advertising. They're trying to capture your attention for means of ensuring that you stay a monthly subscriber. And by the way, they do have an advertising business, it's all about product placement in Netflix shows.

Will Page: So we bring this week's episode of Bubble Trouble to a close. I am reminded of one of the rules of how you measure radio in the UK, which, you know, you were talking about time spent and how you quantify that time spent, how you gather some sort of qualitative evidence of that time spent, and on the radio measurement in the UK, there's this rule which is if you listen to an accumulated 15 minutes of radio a week, you are therefore classified as an avid radio listener. And they extrapolate your listening habits out based on that heavy listening power of the imaginary heavy radio listener. And I'm always reminded that for n- Gen Z and millennials today, it probably takes 15 minutes to switch the damn thing off.

So you have there a good example of time spent on one side, but a false positive from another. So that wraps up this week's edition of Bubble Trouble. We've been learning about paying attention, the currency of paying attention, the contestability, the quality and quality trade offs. I want to thank my colleague Richard Kramer. And remember, this week's bubble is next week's trouble. I'm Will Page. If you're new to Bubble Trouble, we encourage you to follow the podcast wherever you listen. Bubble Trouble is produced by Eric Nuzum and Jesse Baker at Magnificent Noise. And you can learn more at bubbletroublepodcast.com. See you next time.