Let's end when welcome back to Unsensored CMO. Now, one of the primary jobs for every marketer is to capture the attention of your audience. And in today's media fragmented world, that is harder and harder than it's ever been before. So it's really important that we plan based on attention, not just based on reach. And that's why I'm delighted to be joined by Dr. Karen Nelson Fields, who is a returning guest and the world expert on measuring attention. She's also launched a new book called The Attention Economy, which has just come out. So I'm going to quiz her.
on why does attention matter? How do you measure it? And what should we be doing differently to get more share of attention? This is a great episode. You're going to love it. Here it is. Karen Nelsonfield, welcome back to the show. It's good to have you back. Good to be here. Well, you've been quite busy recently, I think is an understatement of the century. And you've just got a new book out called The Attention Economy as well, a blueprint for the category. Congratulations on that. I'm sure that's no mean feat. How long did that take to put together? I mean,
So look, it's an interesting one. So the first book was four years ago, four to five years ago, and it was kind of stepping into what's coming. Whereas this one is, okay, it's landed. What have we learned? Where is it going? So it really sets the tone for the science that sits behind it, the practices of ethics that need to occur in an attention economy.
and where kind of the product sets landing and what it will take to be an enduring category. So I'm pretty proud of it actually. So it's amazing. The thing that surprised me actually was in the beginning, the first chapter, you kind of refer to it as the boring chapter, you need to get through it. I saw that.
But then you talk about the history of attention. I had no idea that attention as a subject matter has been studied for over 100 years, going back to... Yeah, I mean, it's important to talk about that because this is not new. We're not jumping in on some sort of whim. It's something that's been important
for knowledge for a long time, right? So the boring chapter kind of defines the fact that there is already a definition in place, and we're not here to make that up. We're here to see how that then can apply in our category, which is advertising.
And just for everyone listening and watching, it's not boring otherwise. It was really... Well, we threw in some Kim Kardashian. You did, exactly. You got some googling there, there's some fascinating references. I've got some goggles and Kim Kardashian. Yeah, I know, it's very on trend as well. And presumably now, you know, when a lot of literature was written, we probably operated in a...
a world where you've got print and out of home and radio and the formats advertising were fairly basic. But as you make the point with Google, the fragmentation now is much, much greater than probably it's ever been. And that must be a big factor in why attention matters more than that. It was interesting you even raised that because my original PhD was audience measurement, right? But it was at a time when fragmentation was at its highest.
at the time. And there was a problem with measurement back then, which was, well, I know, 17 years ago. But if you fast-tracked it today, it's, I don't know, 100x in terms of what implications there are for advertisers. So, yeah, it's an interesting ecosystem.
Now, I wanted to get into some basics just to set us up before we get into some of the more complex things. But let's start with the case retention now. This might sound a bit obvious, but does attention matter? Because again, I've seen some people that go, oh, it doesn't, as long as they walk past a poster and they get the key entry points and the branding, it's all fine. But what does the science tell us about the role of attention and how, you know, what difference it makes to advertising outcomes?
It was interesting you use that example because that to me is passive attention anyway, right? So someone walking past a billboard and getting the distinctive assets in the peripheral, that's still attention. So when someone says to me, you know, is attention important, I often say, well, turn off your ads. I don't want to see you for another six months. I want you to turn off your advertising and then I want you to come back to me.
and see how your brand's going. So I think that's the case for attention. This is not new information that people need to see your ads. So I get quite frustrated with people that ask me that. And you mentioned passive and active attention. So can you measure the, well, what is the difference between those and can you measure the difference between the two? Yeah, I mean, it's interesting because when I set this business up, I mean, you know, background and academic, I always wanted there to be some sort of rigor around
whatever's next rather than you know sell some gimmick because we can and I based it on history and based it on literature which talks about active attention being someone sort of stops what they're doing pushes distraction out and literally focuses on something right to learn something whereas passive attention is unintended or you know it sort of gets you
in an unconscious, subconscious kind of way, where you don't really realise what you're looking at. So that's kind of framing and things like that. So yeah, we set it up in our own technology to be able to collect both.
Because I was fascinated by in the Gilchrist work, isn't it, the kind of left and right hemisphen? It's almost like the right hemispheres are aware of what's around it. But the left hemispheres engaged in actually focusing on something. And they're two different modes of attention, aren't they? It's absolutely the same. So that's why we collect what we collect. And it's interesting because both have different value from an outcome perspective. So there's three states, and one is active, one is passive, and one is non-attention.
which makes up 100% of your viewing experience across an ad, for example, content or whatever. So once you understand those three states, it's really interesting to see what kind of outcomes occur as a result of that different state.
Obviously, depending on which state you're in determines the outcome. In terms of active attention. Well, it's nuanced because it can be related to if it's a new news or not new news, big brands, small brands, things like that.
But yeah, it's important to capture both. Yeah. This is a question I want to ask you, actually, because instinctively I'd imagine if there's a brand that I'm really familiar with, it's got very strong distinctive assets. Maybe it's KitKat, for example, it's red and got the white logo and got the bars. I probably don't need as much reminding to be aware of it compared to a brand I've never seen before. Does that play out on the data? It does, except...
your children haven't been aware of KitKat, right? So what I often say to brands is that is the truth, that a highly distinctive brand you can pick up.
as you're walking past or as you're not looking directly, but you have to make sure that you have this flywheel because the next generation category buyers don't even know who you are. So it's kind of easy to say you don't need as much, but you can't rest on that current group of buyers forever.
And I suppose if we take the classic Ehrenberg bass principle of light buyers are where the growth comes from, that probably becomes even more true. Well, and if you take their philosophy around the distribution of buying, people are constantly moving through, right? So light buyers are constantly coming through and then become heavy and then maybe get out of the category and there's this kind of
literal flywheel. And that's how we look at the attentive states that, you know, sure certain campaigns don't need as much because it is expensive to buy active attention. But at the same time, you have to be careful not to rely on passive only, which is
unfortunately what a lot of people do by sort of pushing their money into some of the long tail of digital. Yeah, yeah. I'm sure I saw Peter actually presenting at Think Box some attention adjusted kind of cost per millions and it completely changed how you look at the
the world. I mean, I think TV ended up becoming like the second cheapest media on that basis. It's interesting you say that years ago, one of the first pieces of work I did when I left the university was around the value or the cost or the CPM. Because to be honest, I, and it's no surprise in the book, I call it cost per meaningless thousand, right? We'll talk about that in a minute. But the point being that it's interesting because expensive
CPMs have pushed their CPMs down to match the low CPMs that are coming out in the digital space. And so when you actually do the analysis and work out, not just how much attention, but the outcomes, the CPMs are probably too low relative to their impact and the CPMs that are already cheaper, probably not cheap enough.
So I interestingly enough I got picked up by Wall Street Journal quite a few years back now and I think I talk about that in the first book. So Pete the whole concept of CPM is so flawed because it used to be that was
the unit pricing. If you pay more, you get more. But these days, you don't know because not a reach is equal. So your perceived value of what you're paying for isn't there. But on top of that, we live in a bidable system. So you're paying what you think you have to bid for for something you don't know what you're getting. So it's like this double whammy.
So CBM I think is one of the hardest hitting metrics given what we deal with at the moment in the measurement ecosystem around attention as well as obviously viewability and things like that.
Now, can you also link attention to memory as well? Because that's the other thing, as marketers, we're trying to create memories and associations with our brand. So what's the evidence of the role of attention has in terms of what we remember? There's so much evidence in our own, but not just ours. There's this actually effect called switch cost effect, which basically means that the more switching you do, the less likely you are to retain information.
And we see that in our own data because we can track attentive states. So if you're switching between passive and active and non and active and passive. And so we can tell. So the more switching you do, the less likely you are to remember. And then there's also an impact on longer plays. So time is important. So the more time you spend processing something, the longer
time you will remember it afterwards. So we've seen that across time as well. So plenty of evidence, plenty of evidence. And I think I've seen a pretty incredible chart from you which talks about, is it the memory threshold of around two and a half seconds? Yeah. That something like 80% of ads never gets that point. 85.
But it's digital so look that that's an interesting slide because we see that across every single collection I've never published on it But we see that across every single collection I've ever seen where there's a point at which Under that this doesn't mean it doesn't
that under that isn't valuable, but from a memory perspective, particularly related to mental availability, that's kind of the line in the sand where people go, oh, now I get it. And I mean, that makes common sense. It does, yeah. But every regression, every study, we sort of see roughly around that point. It does vary on the size of brand, but largely speaking, that's threshold.
So you can link attention to memory. Can you link attention to kind of advertising outcomes as well in terms of results? Oh, absolutely. So apart from we in our own technology lead people to a virtual store straight after the viewing experience, we
through our ad tech product, which is optimization and verification, we link it directly to a customer's outcomes. So, you know, it's funny because I often talk about outcomes being too directional. One is called meaty proof and one I call marketing proof. So marketing proof
is reasonably spurious because, you know, recall, brand uplift, you know, all those fluffy things that, if your baseline's not strong, then you've, it's spurious. But then the meaty proof is what you want, which is profit and cost of goods sold and, you know, acquisition kind of. So the things that CEOs want to see or CFOs want to see versus the things that marketers want to see, and we've done both.
So we've done, you know, modeled work around mental availability and short-term advertising strength with baselines, but we've also then connected our attack to, as I said, increase in conversion of applications and, you know, more web traffic and reduce cost of goods sold and improved profitability and reduced resources and you name it.
So both brand and business effects can both will be lifted by a greater attention. I often say the reality is attention and outcomes are not perfectly related and the reason why that is is because if the creative is poor, then that attention will be wasted. So if you put a very poor ad, not even from an emotional perspective, but from a poor branding perspective,
it can easily be misattributed to your larger competitor. And that's why anyone that sells you the dream that it will always lead to a sale is driving you into this short term rubbish.
No, I don't mean, you kind of believe the data and then you're just spending a lot of money on, you know, yeah. Basically what you're doing is, but also you're bringing people that were already going to buy your brand forward, which is not good for profitability anyway, not growing your brand. Yeah. And that's probably why the brand element and the business element needs to go together, doesn't it? Because you're also trying to build buyers of tomorrow. I mean, back to our friends, Ehrenberg Bass in terms of 95.5.
I talk about that all the time. It is really important to think about the buyers of today and the buyers of tomorrow. You mentioned the creative there. That's quite interesting. You introduced this idea when we last spoke about attention elasticity that actually varies, doesn't it quite a lot on different platforms in terms of the variability, I suppose, that the creative makes.
So just to step back, it's probably also a controversial, particularly creative versus media, right? So the way that the user experience works is that it trains you how to use Facebook, it trains you how to use YouTube, it trains you how to use TikTok. So the nature of that user experience defines how much attention you will pay. It gives you this boundary, this ceiling and floor.
And what, if you think about an average which sits right in the middle, good creative from an emotional unexpectedness relevance play will sit slightly above that average, whereas those that are less appealing will sit slightly below that average. So that's kind of how creative and media sit into the attention equation.
I hadn't thought about the concept that you used to say of that the platform's almost training your attention. Is it that way around then? The platform's almost creating behaviours that drive your attention. There's lots of books on it. There's lots of books on it for fear of our teenagers. So it works the same way for ads. So most marketers think if we build it, they will
they will watch it. It just doesn't work that way because of the nature of the experience that's built by the engineers that keeps you scrolling. And is that because they've got the data on me? So they're basically going, right, we know what's capturing its attention. Therefore, we're going to serve up this kind of thing. Is it that way around? Is it kind of playing with the data or? Well, look, at the end, I am not in San Francisco to understand the nuance, but there's a science behind
keeping you watching the content or scrolling through to see ads, for example, more ads, but less time. But they control our user experience within an inch of our lives. For example, you all think you're in control of your own attention, but the way you use Facebook is the same as where I use Facebook was the same as where they use Facebook. They use Facebook within a very small margin of error. And that's not, it's largely because it's designed that way.
So if you're seeing small margin of error within platform, you're seeing differences between different platforms and how they're doing it. So to your point about this elasticity, so the capacity of attention is much greater on this platform and much smaller on this platform. And that's why not a reach is equal. Because you could put a 30 second add on here.
and half of it will be seen. You can put a 30 second end on here and it won't. No matter how good the creative is, even if it's a can-line award winner. So most people don't understand that that is your roadmap to successful advertising or not. Yes. Regardless of how beautiful and creative the ad is.
And this feels like the Holy Grail bit, doesn't it? This feels like the bit that everyone should be talking about. This is literally the Holy Grail piece. Because once you understand that, you can do your job. Yeah. Because you can't change it, right? Everyone thinks, well, I'll just build this beautiful and motive piece of creative and it will change the way that you watch. It doesn't work that way. That makes sense. So, and we know that because of the technology that we have, we can control for creative over the years. It's been nearly eight years. So we can intercept
Facebook with an ad that we want you to see without you knowing in your own personal feed by the nature of our app, right? We've tested on this forever. So it is literally the Holy Grail. Now, you gave me a demo of this earlier, right? So I'm really excited for you to explain this because you showed me on your phone how this all works, but explain how you measure how you measure attention. So it's quite sophisticated and it's amazing the data that you can get from your app.
It was designed originally as a research tech, which essentially, if we want to go and collect a thousand views, for example, in Finland, our tech is connected to a panel exchange and programmatic panel exchange, we can tap on the shoulder of 5,000
finish people and a thousand will say yep next time I'm on Facebook I get five bucks or whatever it is and what happens is you download our app and then the next time you're on Facebook or TikTok or YouTube or depending on the setup it allows us you give us its GDPR compliant in that it gives its consent based but what it does is it turns on the camera
which she knows is normal for us now, and it films you, but then it also grabs the back end of the view, right? So this is really critically important. So it's not just collecting facial footage, which I can then turn into gaze points, but it also collects reference points behind the view. So I can tell how fast I've scrolled, it can tell how fast I've scrolled, it can tell if I've orientated my phone, it can tell if I've got ear pods in and how
hyper-volumears or not. And it does that at a sub-second level, so 0.2 of a second. So what that does is it gives us a reference point to how to anchor the gaze. So whether it's active or passive or non-attention is really accurate because we have the coordinates at the back of the device. The other beautiful thing about that is that's the information that allows us to build ad tech products because ad tech world lives.
against device data and viewing data. So it's been serendipitous in that we built it that way because now we can move into which we have into product which allows us to predict attention across the programmatic ecosystem.
And presumably you take all that data and you can train the AI, then can you to learn from that data as well? We've done that for years, yeah. Start to then predict how people will behave. I mean, we were built as a proper ML business, so we built the stack first and product second. Yeah. That's a great way to do it again.
One data point I found fascinating in the book as well, and maybe explain it, is the difference between what people think they're buying in terms of you think you're buying lots of impressions versus what you're actually buying as an advertising. There's a big difference between those two things, isn't it? Yeah, I mean, it's a versus scene, right? So look, that's been a big part of my category growth agenda for a long time. And, you know, as someone who
is from the measurement side. I struggle when an impression is supposed to mean an impression when it doesn't. So basically the reality is you can buy a thousand impressions here and a thousand impressions here and the level of scene versus served is so significantly vast.
but you still pay the same a million dollars per thousand, but do you know what I mean? And that is literally the core problem of our entire ecosystem. So when measurement is not equal, what that has an effect on things like viewability, because viewability is based on this equation that time and view is equal or impressions are equal. It impacts reach-based planning.
You know, you talked about Pete, it impacts his pet peeve, which is Esoph. So, you know, my whole agenda for the last sort of eight years has been if we can build signal-based data.
that makes reach equal again, then my job is done. And that's where it's going. Yeah, you're quite right. I mean, Peter's just saying, like, here, save you a minute, we've all come to know and love 10 years ago, doesn't work if you can't make an assumption that will reach is equal. He has literally slides that I've seen and used that describe the point in time when everything changed. And it's really, really striking that it's at the same time that the digital spend
increases. Again, not to suggest the digital is bad, but you're buying impressions, but not getting the full 100% of what you're buying. Yeah. And that variability depends across all different formats. And I think from the data in your book, you think you're buying around 78% impressions and it gets down to about 15% when you look at viewing more than two seconds.
I can't remember the numbers exactly, but to your point, it's like real-class. The difference is massive, isn't it? If you treat two seconds as what's required to actually... Well, this is true. So less than around about 15 to 20% max, get more than this threshold, which is what you speak about.
But there's tools now. There are products, including your own, that use human data to essentially normalize this error. So we are definitely, and I call it Fab 4 in the book, so there are products for every stage of the ecosystem. So whether it's planning and strategy or buying or
you know, whether it's optimization, there is creative testing, there's a stage where people can jump in from an attention perspective. So, yeah, it's becoming quite a category. Yeah, it is, isn't it? And much needed by the sounds of it. And Amara is saying, so each different platform will have different definitions and impressions as well, as we were saying before. So, impressions on one platform might be different to the definition of impressions on another.
It's not as much about definition of impressions anymore because the MRC have kind of come in and the regulations around what that means. The Achilles heel of our industry is this little thing called time in view and that's what they rely on so but the relationship between time in view and actual viewing is what's
unstable. So it's not that impressions are different because that would be MRC against. It's more that the relationship between
the ad being on screen. And so the time and view attention relationship varies. Got you. Okay. I've got you. So you've been able to marry up what people actually respond to versus what the beach platform are saying is time in view. And those two things differ in terms of relationships. Yeah. So that's why not all reaches equal. Yeah. So if the error that sit underneath
the delivery was equal. So let's just for argument's sake, say, served and not seen or seen but not served was 75% across all impressions and you could live with that because it's consistent. But the issue is that that number changes drastically across, depending on the device and the platform and the format and the ad length. And that's the problem.
I really like, by the way, your definition of served and seen. That's really, for me as an, you know, I'm not an expert like you on this kind of thing. I get that, right? The difference between the two.
One is a much more human approach. Served is an ad-tech approach and its binary. Did it happen? Did it not? Human approach is, did it get served and how much time do we spend with it? That's the reality. And I think, and will I know, that's where the future is going in terms of human-led synthetic data.
You've spent a long time thinking about this research here, you've built a business around it, you've written three books on it as I think. Are you seeing the industry move in the direction it should be? Absolutely. Give us a sense of the progress. I'm pretty proud. The MRC have stepped in this year and I'm pretty proud of the IB, it's done a great job, but I refer amazing and everyone in the MRC now
kind of going, okay, we need to build some standards around this. What I love about that is its validation for its value. So I can see it moving. And I think they're talking roughly, let's just call it middle next year. I think that will change the industry again, because a lot of people, and that's what happened with with your ability. It took a lot of years for it to be standardized, and then it became standardized, and then it took off.
I just see this as the next iteration of that. Now, viewability measures will always need to be there because you need served and seen. So yeah, I'm pretty proud of the industry, to be honest. I can see media planners getting pretty excited by applying this tool to their thinking and planning. What about the platforms themselves? Because presumably there are winners and losers in the mix. And it's hard because that is the truth.
So if you use a metric like attention or value or if you use a metric like human involvement or whatever it is you want to, when it's not binary, you are always, there's always going to be a winner and a loser. The trick is that
when you set up a four, that, because all platforms have its value, but what this does is it normalises the CPM again. If you get an outcome, but you pay relative for that, then it works. So it's less about more attention matters. It's more about
how much do I pay for the outcome relative to the attention I'm buying? So that's where it's going. Which raises a really good question, because ever since I started my career 30 years ago, whenever it was now, CPM's been in my lingo, right? When you talk about CPM, it's one measure that allows you then to compare lots of different media and how much you're spending and it's kind of efficiency measure, I suppose. What could we call, what do we name the new, what's the new thing going to be now?
Where does it go next? Is there a metric we can then create? No, it comes back. Well, look, CPM is going to come back, right? Because at the end of the day, there is a value equation. John White, who is amazing in our industry, talks about this all the time. So the value equation for an advertiser is, from a publisher perspective, how much reach can I get?
you know, what does it cost and what is the quality and or attention, right? So those three things are important. And at the end of the day, you're always going to need the CPM will come back to be whole again once the reach component is fixed. So one thing that has been interesting with attention is, you know, we're all learning, right? And so I think we're ready for
metrics that are simple for advertisers to understand, to kind of get their head around what is it that they're buying, and that's kind of next phase. Yeah, because that feels to me, because I think most marketers out there will go, ah, yeah, I get there's a single attention. I get seen and served as a different thing. I want to move to them. It almost feels like we need a language, a bit of help with the language, don't need to go, ah.
Now I know what I'm buying. But what is next is metrics around the volume of attention relative to the reach components. So big on our agenda. And also how do you do it? So if you describe what you do for media planners, how do you go around the planning process? What does that look like?
We have optimizers and we have SaaS platforms where you can kind of jump in, upload your media plan and then optimize to different objectives or optimize for more or less attention using CPMs and it's recommended, it's got recommendations engine or that sort of thing from a planning perspective. But a lot of our customers in the planning area will use either an API of our data or, and they can do their own normalizing. So a lot of people
like to do, some people like Richard Kirk, he's quite famous for the stuff that he's written around, equalizing reach. So he's used our data before where it kind of adds as a signal for their planning systems. But for generalized brands, a lot of people will just use the platform and kind of upload media plans and go from there. That's what I was trying to get to really was like, if I've got a brand objective,
I know when my crate is good or not, you know, I've done that. I know how the last campaign is. So presumably this can help me optimize for my reach objective. There's two ways to do it. So there's the planning in, which is pre flight, but there's also in flight, right? So you can also add a tag to the back of your creative and then it will
return how much attention you're earning and then you can make changes mid-flight to the myriad of impressions that you're buying to basically cut out the long tail of rubbish and improve your objectives. Because often these days we have multiple bits of creative and multiple different platforms running at different times, different promotions, different cycles.
Taking all that into consideration, I could optimize all of that based on a kind of... Yeah, I mean, planning in basically does one part of it and kind of gives the buyer a roadmap to use, but the optimization live
kind of then tidies it up even further. So it goes down to the URL level, which is super fun to watch, because we do it literally instantaneously. So it's super fun. That's quite fun. The graphic must be cool. It is very fun. That must be really fun. It's very fun. Now, talking about implementation as well, not only have you written a book, you've also launched a course, as if you weren't busy, right? She says, oh yeah, there's that one as well. What's the course all about? Well, the course is kind of an extension of the book really. I mean,
You know, the book is words, but this is interaction, tasks, exam, one-on-one with us, like we do a team thing at the end where we all talk about the industry and, you know, what's happening, the last one was about JAG.
but also they must see. Look, it's kind of, I guess, for people that want to have life in the delivery. It's me lecturing, really, but then having interaction with there's some exercises to do and some fun bits. So yeah, so that's been kind of fun. The cohort was sold out last time and the next one is, I think, in a few weeks time. Fantastic. Yeah, it's kind of fun. And people want to do it where they have to go.
And you can find it on the work website under lines learning. I thought I'd finish up as well by asking how amplified intelligence is going, because it's expanding rapidly. You're growing all the time. Your roles changing. You launch your books and courses. How's that journey been for you? It's been amazing. I mean, to be able to put in a new CEO has been super valuable for us. You know, Annalise has a long history with extreme growth.
in technology, so she was the right person to take the business to the next level. It gives me the opportunity to double down on what we're building from metrics perspective and making sure that our product is absolutely spot on in terms of
It works. I mean, you know, when you're a measurement business, it's supposed to work. So that's the role change. But yeah, we're across three markets. So we're in the UK, the US and Australia. It's going really well. Brilliant. Oh, Karen, thank you so much. It's always exciting to catch up with you. I love what you're doing. Thank you so much. And it's so much value. Yeah, I should say this, actually, for anyone listening, we now have an exclusive Tony's chocolate only branded chocolate bar. Indeed. Well, there you go. In case of emergency open,
There we go. Karen, it's brilliant to see you again. Thank you so much. Thank you very much for listening or watching Uncensored CMO. I hope you enjoyed that. If you did, please do hit the subscribe button wherever you get your podcasts. If you're watching, hit subscribe there as well. I'd also love to get a review. Reviews make a big difference on other people discovering the show. So please do leave a review wherever you get your podcasts. If you want to contact me, you can do. I'm over on X at Uncensored CMO or on LinkedIn where I'm under my own name, John Evans.
Thanks for watching and I'll see you next time.