Unlocking Creative Effectiveness with AI - Measuring Emotion Attention Recall

(00:05) So, I'll welcome you Stephanie who I've known for very long time and proudy marketer insider intelligence alumni. Ian Forrester, who Stephanie introduced me to a few months back and and got to break bread and and hear about some of the work they're doing and I just probably as we were rolling out this series and I was like I we got to find time for you to come on here and uh and they floated this idea about uh about the Super Bowl that I was really excited about and so I'll let you explain it better than I will but but just uh excited to have you

(00:43) here. welcome for and great great to see a lot of regulars back here. For others who it's your first time on one of these, it's very interactive.  Sometimes I don't have to lead any discussion whatsoever. I just get to kick back and listen quite a bit. Um, but feel free to chime in, raise hands, shout out if you need to join join the chat.

(01:14) And, and this is like just this is really my favorite way especially like within AI Marketers Guild for for just members to connect with each other and uh, and so uh, I'm just thrilled to see all you here. So uh, Stephanie Ian, take it away. Thank you so much, David. Ian, why don't you introduce yourself and then I'll give a little bit of context about about the company. I will. Hi guys. Thanks so much, David. Thanks for having us.

(01:39) So, hello, my name's Ian Forrester. I am CEO of David. I have been working on David for about six years now. And prior to that, I used to work at Unruly. So, I led the insight team there for about seven years. Um, prior to that, worked client side. So, I was at Sony Pictures for a while, Nestle, L'Oreal.

(01:59) Um, but I've always been involved in creative effectiveness and in particular the impact of emotions on creative effectiveness and essentially the work we're doing at David is taking all that work that I've done over the years to the next level. So, we're super happy to be here. Um, this work that you're about to see is literally hot off the press. Like we the team were up all night cuz we're based in the UK.

(02:19) So, we were up all night um on Sunday after the game getting everything together doing some analysis and pulling the story together to uh be able to present to you guys today. So, yeah, we're excited to to share. I'll share my screen, Stephanie, while you uh say hello. Sounds great. So, as David already alluded to, we go way back to our e-arketer days. Um I'm Stephanie Blair and I'm serving as the USMD for David.

(02:43) Um and I also am born out of the advertising world. I I started my career in print advertising, spent some time at agency within performance marketing agency here in New York, and I'm I'm here in the New York uh city area, and I'm excited to be presenting to you today. Um so, while we let um Ian, we see um a dashboard, not the There we go. Sorry.

(03:09) Yeah, this Zoom is covering the present button, but I'm on it. I'm on it. So, I know I know you're a vocal bunch, but I do think it's important just to take, you know, a handful of slides just to give you a bit of context about David and and what we're all doing here. So, obviously, we're a creative effectiveness business focused entirely on effectiveness driven by creative advertising campaigns, as Ian mentioned, born out of decades of research in this space.

(03:33) Um, but you know, let's talk about why this matters because, um, I think you know, you all in the room know this. Next slide, please. We're seeing that nearly 50% of creative performance is is impacting sales and over threearters of large large amount of brand marketers deem creative quality one of the biggest factors in success. In fact, just a few weeks ago I was at a contagious event organized by War and Ken Lions and you know creative effectiveness and its impact to drive the bottom line was one of the major themes of the of the conference. Um and

(04:05) obviously you know we can see this and we know this good creative executed well can drive profitability um to the tune of 12x. So we have power in our control but the problem is a lot of creative out there is just so so meh. So we're here to talk about the creative effectiveness um you know sort of system um and grading system that we've used today uh to help you understand who the winners are in the Super Bowl and not just uh Taylor Swift's boyfriend's team but the brands themselves. Um, in terms of what we are all about, it's also about creative testing for the real world. So,

(04:38) we've seen a big shift and I'm sure those of you who have been in the industry for a long time, you know, back in the day, it was there was a lot less content, limited number of channels, you know, there was maybe TV um and maybe one social channel and now there's a proliferation of that content um you know, more content than we can even handle. Um, and there's an always on nature of things.

(05:03) And we're seeing spend come uh, you know, sort of meet this moment as well. So, influencer advertising has doubled since 2019. It was $21 billion in 2023 alone. That's according to Statista. And obviously on social, we're seeing it, you know, continue to grow. And this has been a trend for for forever, but it's expected to double over the next four years.

(05:23) And also other fast growing channels like retail and Tik Tok, CTV. um you know this is according to e-arketer um you know we're going to see this all continue to grow and rise um so these are global figures but we know that there's a there's a shifting moment and David is here to meet that moment because gone are the days of only testing linear or hero oriented content that's important but again when we're having this mass scale of content we really need more of a closed loop always on approach to testing to help uh marketers and advertisers really understand not just what is working, but why, because that's what consumers

(05:59) demand. Um, and that's what David is here to offer. It's continuous testing at scale. And when we do this, we can really give um, you know, really sizable results to our clients, which we'll talk about in just a second. But here's the AI piece, which is interesting. So, we're continuously training David with attention, emotions, and creative attributes data.

(06:18) Then we can show David ads and explain why that advertising will work or not. So again, in a world of a lot of what content and metrics, we're really getting to the heart of the matter by using an emotional lens um to understand that why. And this is not just for video. It's really for any media.

(06:38) So of course, in today's conversation, we're talking about Super Bowl video ads, but this could be applied to, you know, copy, uh imagery, um really any anything that people can react to. Um and so we're enabling advertisers to understand why the creative worked or not so that you can set the strategy for the future. And we bring the Y as you can see here and allow people to test uh affordably at scale. So what does this mean? You know, we can see here some of the clients that we're working with.

(07:01) So whether it's about um making creative effective for a certain platform like what we're helping Snap understand um helping to add a layer of excellence to modeling for a certain agency such as Essence Mediacom or to really look at the influencer space and improve briefings or effectiveness of that of that strategy.

(07:22) Uh David's here to help all through our scaled and afford affordable testing. So we're excited to bring this to you firsthand because what we've done is we've run a number of the Super Bowl uh ads through through our platform um and to help you understand, you know, what the trends were and who won. So I'll pass it over to Ian to give you a little bit more uh you know the insights driven approach of why it works and then we're going to take you through the rest of the time in a more com um you know uh conversation around the the winners and

(07:47) what we all think. Thanks, Stephanie. Yeah, so I mean, just to add to that a little bit on this slide, um we're not saying that it's no longer important to test hero content. Um and in fact, we've tested a lot of hero content with this Super Bowl study, so you're going to see some analysis on hero content.

(08:06) Um which is is totally fair enough. A lot of money is still spent on that content, of course. But what we're seeing is also in addition to that hero content there is this huge body of work which is being put out by brands across social the various platforms uh influencer campaigns which isn't seeing the light today in terms of that insight and understanding process it's just getting pushed out the metrics which are being collected on it are social metrics things like engagement rate like shares comments uh maybe reach and impressions and brands aren't really

(08:37) understanding if that creator is doing anything for them and and and if they should continue doing it. And this has really come to fruition in the last I would say 9 months where we've been banging this drum for a while but Branson asked to really come to us and say this is a major problem but I was in a call with Unilver last week um one of their homeare guys and he said to me Ian we've got a disgusting amount of content and I've got no idea if it's working or not. We just keep on pushing it out across all the social channels and I've got no idea if it's good or if it's not. if it's damaging the brand, if it's

(09:08) helping. And this is where our major clients, the ones that are working really well with and who are using data to our absolute best, these are the clients which are really leaning into this and testing. Yes, they're testing their hero content, but they're also testing these large bodies of work which they're producing or perhaps which other people are producing as well to inform their strategy. And that's where we we really come into our own.

(09:36) And so for this study, which you guys are going to see, there's a bunch of hero content which we tested in there, but also a bunch of Tik Tok content because prior to the big game, there was tons and tons of activity on Tik Tok. So, we've kind of merged those two things together with the presentation we're going to show you guys today.

(09:55) How does it work? Let me walk you through at a high level what we're doing at David, the data which we're collecting and why uh and how we're bringing all this together because that brings some some context for you guys and enables you to see what we're doing, why so um okay, everything that we do at David is based on this effectiveness process. This is the foundation for all that we're doing.

(10:17) And the cool thing about David was when we started the business, we had literally a blank piece of paper. We didn't have any legacy data which we had to shoehorn into the business or use in some way. We literally had a blank piece of paper and we said to ourselves, what data should we collect from the ground up which is going to best help us answer the question for our clients, why is content effective or not? And to arrive at this effectiveness process, we did a meta analysis of creative effectiveness studies from both academia and industry.

(10:48) So we've looked at the work of the IPA and the IAB and all the work we did at Unruly over the years and Erinberg Bass and Wharton and so on. And each individual study that we looked at was tending to look at a small part of this process, but we've taken this helicopter view to see how they all fit together in the round.

(11:06) And so um it's quite interesting. I'll walk you guys through it. To be effective in the wild, a piece of content first of all needs to capture attention. And of course, attention is a major buzz word in the industry right now. A lot of people are focused on attention. And absolutely, you're rightly said because without attention, you're dead in the water.

(11:23) So attention is critical to creative effectiveness and it's a big part of what we're doing at David, but it's not the be all and end all because once you've captured attention, you've got to do something with it. And that something is evoking an intense positive emotion because it's the emotion which creates the memory and the memory which drives the action.

(11:45) A lot of content these days which brands are putting out emotionally is just a bit average. So you watch an ad or a post you're like yeah that was okay. You feel a positive emotion but at like a four or five out of 10 in terms of intensity. The issue with that is that 2 seconds later you've forgotten that ad. It's just wallpaper. This is another average thing that you've been exposed to that day.

(12:04) So the memory is not created. The action is therefore not driven. However, when a brand can elevate that emotion to an eight or a 9 or a 10 out of 10, the viewer remembers the way the ad made them feel, they attach the feeling to the brand and it's that feeling which causes that person to do something either immediately or at a later date.

(12:27) So another important thing to say about this is that attention and emotions are the drivers of output. Either from a a brand building point of view or from a direct response point of view. The only difference with direct response is that memory doesn't really come into play. You get get capture attention, you evoke an emotion and that person takes the action immediately.

(12:45) They click through, they they search, they they go through and buy whatever it is you're you're trying to ask them to do. So this stuff is equally important for brand building and for lower funnel campaigns. So with that in mind, we're training our AI using attention and emotions data because those they are the drivers of effectiveness. Now another thing about David is that we're bringing lots of different methodologies together.

(13:10) So I've learned over the years that if you do one thing and not another, you get part of the picture but not the full picture. So the more techniques you can bring together, the more complete a picture you get. So this is what we're doing at David in terms of attention to create the attention data which we're using to train the AI.

(13:32) We're showing creative to people and we're asking those people to turn on the webcams as they're interacting with the creative because that allows us to gather facial coding and our tracking data. So we're understanding are people looking at the screen or not? When when are they tending to look away and also what are they looking at on the screen? So what is capturing their attention? So that's a dual attention metric which in itself is is pretty unique in the industry. In terms of emotions, again, we've got a couple of methodologies here.

(13:55) Facial coding, where we're filming people's faces, allows us to see where are people smiling or frowning in the different moments of the video. And that's what facial coding is really good at, pinpointing those moments which are working well or not so well. What facial coding is not very good at is telling us which emotions are being evoked because we're only we only output six expressions universal expressions and there are loads more than six emotions.

(14:21) So that's why we've developed the data 39 which is our own emotional categorization which I'm going to come on to explain in more detail in a second. Final part of the puzzle are created attributes. So at scale we're sending content to various computer vision APIs which are telling us what is happening within that content both in terms of what you can see and in terms of what you can hear and then we mix all this data up to produce our outputs which I'm going to explain to you guys in a bit more detail in a second but first of all let's double click on the David 39 and um because this is this is really

(14:54) fundamental to what we're doing now this categorization has been taken from academia There's a huge body of work that exists within academia around what's driving uh what constitutes an emotion that lots of debate between academics and lots of gray area nuance.

(15:12) And essentially we've taken this big body of work and we've condensed it right down to this list of 39 emotions which is just really useful for practitioners. And we've done that because so much of that why why is an ad working or not comes down to the emotions that it's evoking because we're all humans. we're al responding to everything that's put in front of us in an emotional way.

(15:31) That's just that's how humans deal with the world. Um so the why is really really driven once you capture attention once that hurdle has been overcome then the emotions that you're evoking come into play in a big big way as to whether you're going to drive the result which you are which you had intended to drive.

(15:53) So if an ad isn't working it might be because it's evoking negative emotions. So in the bottom right here, boredom, confusion, those two are super super common. We see those come through all the time. Or more extreme negative emotions might be being evoked as well. Things like awkwardness or embarrassment or shame or anxiety or contempt, distrust.

(16:12) We're testing in creative around the world every single day. And every day these emotions are coming through. Um on the flip side, when ads are working, they're working because they're evoking intense positive emotions. So that intensity is super important.

(16:31) You've got to be going for an eight or a nine or a 10 out of 10 because that's what raises that ad up above all the other stuff we're being bombarded with on a daily basis is what makes it memorable. So intensity super important. Also, the kind of emotion you're evoking can be really powerful as well because if you can evoke an unusual emotion versus your competitors just by being different, you're going to stand out.

(16:50) You add intensity to that difference. That's when you get those those fireworks. And that's particularly put in today's to today's conversation actually because as you'll see and I guess you guys got the sense anyway like lots of the creative at Super Bowl this year went for a a single kind of trope like they really kind of clustered around amusement as as the the lead emotion here.

(17:14) um which is fine if you do it well and there's certain ways which which which allow you to do it well and I'll come on to explain those but what I would say is the first thing before you even get into it like and these great if you can do it well brilliant you you've got to be going for a nine or a 10 out of 10 is that's what makes it not just mildly funny but then forgettable but actually very funny and then and and then memorable but look at all these other emotions that brands can play with like look look at all these wonderful

(17:39) emotions here or excitement surprise pride, gratitude, hope, belief, knowledge, nostalgia, a romance, all these other things which can be evoked and you know a large mass of the brands at the Super Bowl went for amusement which basically made it very hard to stand out but I'm kind of spoiling the presentation so I'll move on.

(18:02) Um but yeah so what I've just described in terms of the AI and the training data set is one set one side of the coin in terms of our training data which is the how are people responding to stuff part the other side of the coin is what is happening within creative and to gather that data at scale we're sending assets to various computer vision and computer listening APIs which essentially tell us what is happening in that content both in terms of what you can see and in terms of what you can here on a frame by frame basis and essentially that becomes

(18:31) a very deep understanding of what is happening within content and our system then understands these two data sets and understands the connections between them and that's really where the magic of what we're doing lies. So I'll give you guys an example. If you just imagine that's a single frame in a video just from that one frame we're picking up loads of information.

(18:54) So, we're picking up the fact that there's a crowd there and there's traffic noise from the road and there's a girl between the age of 25 and 34 and so on. All of these things become data points which we can correlate with outcomes. So maybe people are paying attention to the girl because of what she's wearing or maybe people are looking at the crowd and they're feeling a bit anxious or perhaps a bit excited and so on.

(19:19) So our system is continuously learning what it is about content which is resulting in a certain outcome. We're continuously training our algorithm with more and more live human data where we're assuring content to people to gather that data to then train the algorithm. So it's continuously learning, continuously updating and that then forms the basis of everything that we do.

(19:37) So in summary, we're using AI to handle this complexity. We've got the various data sets which we're pulling into David. We've got facial coding, eye tracking, David 39 listening, computer vision, and then we mix all these things up to produce our outputs. Um, okay.

(19:58) So, what does this mean for our Super Bowl analysis? By the way, guys, if you got any questions at any point, please do dive in. I can't actually see you guys at the moment. Um, so if you want to interrupt me midflight, all good. Just please dive in. Um, okay. So, Super Bowl, Super Bowl 58. How did this shape up? Well, as I said on the previous slide, a a large large number of ads were going for amusement.

(20:24) Like they had amusement as their lead emotion. In fact, twothirds of the brands which launch Super Bowl campaign had amusement as their lead emotion, which is totally cool if you do it well. If you evoke intense emotion and really really with amusement I'm when I say intense I'm talking about a nine or a ten out of ten because you really have to move the needle to with particularly when there's so much funny content being produced you really have to move that needle to um create a funny ad and some some brands that did it really well. So Popeye's and Paramount Plus did it well. I'm going to

(20:54) I'm going to go through the content guys and play it. So, for those of you who didn't um see all the ads, I know there's like a mass there's just so much content released on the day, but it's worthwhile kind of just having a look at what's working and what's not. So, Popeye's to start with this tested very very well.

(21:11) This piece of content is actually second in our overall list of of hero content. Let me just play this for you guys. Yes, there's a better wing pot lip. What else is cheese? Here we go. You're crazy. Two dogs in one. Hooray. I finally had wings. Just checking you guys can hear the audio. Yes. Of the video. Yeah. Cool. Awesome. So, um, yeah, Popeyes did really well.

(22:01) That amusement score is very decent. 29.9% of the audience found that intensely funny. Um, so nice big success for Popeyes. It's also brilliantly branded. Um, very strong brand recall. So that's what was driving the the performance of that particular crate. It also opens in an interesting way. So it it opens this kind of lab scene.

(22:25) You're not entirely sure what's going on. There's a bit of intrigue there. So in terms of attention in the first three seconds, it's it's doing a really good job. So in terms of our creative effectiveness score and how we rank this content, it's attention. So potentially people paying attention after the first three seconds and it's potentially people feeling intense positive emotions and potentially people remembering the brand which the combination of those three things which we really focus on based on our creative effects process. Paramount again this

(22:52) was um a very successful ad um featuring lots of different characters and u and celebrities as well. So um Chris Perkins wanted to know more about the brand assessment context. Does it consider any relational dimensions to either the category or current consumer impressions of the brand? So um when we test creative particularly when we are um building an algorithm we test that creative in isolation because we need to understand the impact of the creative itself on people's responses.

(23:30) And so you imagine we can't possibly control for um all the different environments that it might be shown in or what it's shown next to, what's what's seen immediately before it, what's seen immediately afterward. Um people's impression of the brand as well. So when we're testing creative, we need to make sure that a sufficiently robust number of people watch that creative.

(23:58) So that sample is representative of um of the the global population within within a country. So our standard testing is 350 nationally representative people for each asset that goes into an algorithm build because then that that irons out a lot of the um different preferences around brand. kind of like some people love a brand, some people don't.

(24:22) So you need a a fairly large sample size to kind of negate that. And the creative is tested outside of context because we want to understand the impact of a creative itself. Of course in the real world that can be either heightened or dampened down by the um by the way it's distributed but our baseline is the creative itself because we need to understand how that compares to other creatives and then and then we can make recommendations as to how the creative is working or not which then the media part of the campaign either dampens down or or amplifies. So thank you for that. And then there's

(24:56) a couple of other questions, but I'm going to cherrypick them because some we're going to get to. Um, but just very quickly on this slide, what does a 29.9% amusement score indicate? Kate Crook would like Cook would like to know. Is it a good score? Is it a slice of the pie? Emot emotional attribution? And then we'll get into how that compares to some of the others like State Farm and uh Dunkings, etc. Yeah. Yeah. So, that means that 29.

(25:21) 9% of viewers found that ad intensely funny. So they were they they felt amusement at a nine or a 10 out of 10 when watching Paws and 29% when watching Paramount. And yes, that is an excellent score versus the US norm. So the US norm for amusement is about 9%. Um so that's like a a large overindex like a 300 index overindex. So it's a decent score.

(25:48) I I didn't want to put lots and lots of data on these charts because I wanted to keep it quite conversational. But yeah, the reason I'm showing this both of these um examples here is that both of them did very well, but these these were some of the best in terms of amusement. As you'll see as we go through the deck, there's some less good examples as well.

(26:08) Okay. And then lastly, um there was a quick question 30 seconds or less. Are we using natural language processing to look at copy? I think we've covered that a little. We are. So um so the copy on on the creative attribute side the copy goes through um Amazon recognition which includes natural langu l language processing.

(26:35) So uh understanding the the copy so the words being used by the actors we're transcribing those we're transcribing anything that's said in terms of subtitles or any text which occurs on the screen. Um and then we're also adding not just NLP um but also an analysis of the tone of voice which words are said in.

(27:00) So we're using different API to add the tone of voice analysis and that adds a lot of color to like a simple text read. Um, when you add the tone to the NLP, then you get a much deeper understanding of not just what's being said, but also the context and the tone and also the back and forth between uh protagonists as well. Okay. And there's a lot of other good questions. So, let's just take one or two more here. We'll share a few more examples, go deeper.

(27:24) Um, we'll talk later about uh ads that broke a lot of rules this year versus last. We'll talk about that in a moment. Um there are a couple questions around how do we align a particular set of emotions back to what's right for each brand and I think this speaks to the way in which we collaborate with brands and also um you know industry by industry there are different standards or emotional whites space that we can help brands figure out.

(27:49) Um so for example you might not care you might care a lot about performance for BMW but not so much about funny um so help us answer that question that's a lot of people in the room are are curious about that. Yeah. Yeah. Yeah, great question. So, um I'm often asked like what's the best emotion for uh for sales or for uh for recall or search or whatever it might be.

(28:16) Um and the reality is there is no one best emotion for each brand outcome. Um uh a more interesting way to think about it is what which emotions are right for my brand? Um, and we often do these kind of category studies for brands. We did one for Chaseedo, for example, looking at US cosmetics on Tik Tok. Chaseedo had a problem.

(28:40) They had lots of new direct to consumer competitors who had launched on Tik Tok. And these guys were really eating Chaseo's lunch. Like Chase were nowhere on the platform. Their engagement rates were well, well lower than their competitors.

(28:57) And they said to us, why is this the case? So we analyzed a load of their creative and their competitor's creative and created this emotional positioning map which showed which emotions were being evoked by which brands in the space. And what that allowed Chaseo to do is identify the emotional white space. So these are underused emotions in that particular category. And as I said before, if a brand can move into that emotional white space, if it can invoke an unusual emotion versus its competitors, by doing so, they're going to stand out. And then you add intensity to that and that's when you get that those great results.

(29:27) So there's no one right positive. All all the positive emotions can be used effectively by a brand. But I would encourage brands to understand their competitive landscape. What are others doing within that landscape and where's the emotional white space that they can move into and own? Um and it's not just about picking any emotion which people aren't using.

(29:50) It's also about understanding which emotion is right for the brand. We help brands go through this process and this is really the crux of this matter as well in terms of the Super Bowl with everyone being with everyone focused on humor is really hard to gain cut through because if everyone's basically doing the same stuff I these two did well but so many ads like only seven of the ads the the big game ads but which went for amusement only seven of them evoked intense amusement among more than a quarter of the audience. So loads were were just not hitting the mark. And when you consider how much

(30:26) money is being spent on these celebrities as well, it's um it's a shame because brands can there are so many other avenues you can go down to gain differentiation as we show we saw on the open slide. So let's go through some of our other findings because I have the benefit of knowing what is in store here and there's a lot of juiciness.

(30:45) Um somebody did ask about our model and does it you know how do we break out by demographic and are there relevant differences? So, where possible, if you have any color uh to add on that relative to the other examples, please add them. But I did uh type in an answer. Yeah, but we want to see the results and Exactly. Let's go. All right. All right. All right. Cool.

(31:02) So, uh let's have a look at the next one. So, if you're going to use humor, cool. But remember that humor is polarizing, right? So, as with any emotion, you've got to be getting to to a 9 or a 10 out of 10 to be memorable. To get to that place with humor, it's often necessary to go to go to a place where a brand isn't comfortable.

(31:31) Because while you're making some people really, really laugh, you're really turning other people off because humor is very polarizing. And that was the case with um Caravee um which I'll play a little bit of with Michael Sarah talking about the Save brand. I'm Michael Sarah and I'm pleased to announce that this is my cream. Oh, you didn't know. The truth has been idated.

(32:04) So, it's slightly creepy. Um, he's talking about he basically is making out that he produced the brand Saravee. And this was supported by a lot of content on Tik Tok as well, which I want to talk about in that section. Um, and it was polarized. Some Some people find this creepy, they feel awkward when they're seeing it.

(32:28) It's or some people are just confused like, what the hell's going on? Some people understand that it's a spoof and it's ridiculous and they find it hilarious. And that's great. that that is this is what you want to be doing with humor. Like own it. Don't don't be afraid that you're going to turn some people off. As long as your target audience are finding what you're producing intensely funny, then all good. Um because the the opposite of that can be just a bit average.

(32:50) Um often what we find is when brands try and be funny, they're not comfortable going to a place where it's going to really make people laugh. And so it just gets the ideas get watered down to the point where they're just a bit meh. And that's what happened with progressive insurance here. Um again, I I'll just show play a little bit of it.

(33:08) You know, it's just a bit it's just a bit average unfortunately. And um there was a lot of creative in the Super Bowl this year which followed a similar kind of thing like um featured a celebrity tried to be funny and just just was I mean not not terrible but just was okay.

(33:28) And and amidst all this noise, and I guarantee you this will be forgotten almost immediately. And don't think that celebrities are going to get you out of jail free. So this obviously ridiculously expensive ad Beyonce, the Verizon one. Um I mean again it wasn't horrendous, but it was it was just okay. We had a fairly average this thing is covering yeah 12.

(33:53) 8% 8% um amusement score is this Beyonce again. I'll play a little bit of it. Um you'll see what I mean. Broke the internet again. Did you post this? Oh, well, not on purpose. Well, it's coming in hot. It's rise 5G. The network is crazy powerful. I bet you can't break that. I bet I can't. Get it.

(34:18) But can't she break it? Broken? So, Beyonce tries to break the internet in various ways. um you know really really expensive and and again the back backdrop of other brands doing very similar stuff albe it not with Beyonce but you know essentially the same formula it just didn't gain cut through um one thing you can do with humor is mix it with other emotions so that's what this blueberat ad did brilliantly when you do that the ad can work in different ways for different people. So, you're not relying on it being funny. You're evoking other emotions and that broadens its appeal. At the end of the day, we

(34:56) want people to feel intense positive emotions. That's what's going to drive that memorability. It doesn't really matter which emotion it is. Um, and so if you can combine amusement with different emotions, that's what's going to be very effective.

(35:11) And this one, this particular ad with Jennifer Aniston, David Sha, and so on did that. Thank you. I didn't know you could get all the stuff on. I remember that. Well, you know what they say. In order to remember something, you got to forget something else. Make a little And that's how I remember Ubis has coffee by forgetting something else. Have a seat. Yeah.

(35:34) What? Remember when you used to be a pepper lady, wasn't it the cinnamon sisters? Basil paprika girl. No, that's absurd. That's posh as well. But so this is working in lots of different ways. funny, but it also is generating in interest in the product itself and excitement in the product itself. It's got nostalgia coming through friends and partial and all these all these kind of um people that we've known for years.

(35:58) So, this is a really good way to incorporate humor. If you have to go for humor, cool, but mix it with different emotions in that is exactly what one of the greatest Super Bowl ads of all time did. Uh Volkswagen's the force. Um, I won't play this cuz I guess we've all seen it like a million times.

(36:18) But the the brilliant thing about The Force was that it combined lots of intense emotions. Yes, it was funny, but it was also heartwarming and adoration in there, inspiration, nostalgia, pride. Um, and it worked in in a lot of ways. So the such that nearly 80% of viewers felt at least one intense positive emotion watching that ad which was what makes it so iconic and so brilliant because it was working for a large proportion of the audience.

(36:45) Um, now the winner of the Super Bowl with regards to the data which we collected was this NFL ad which didn't try and be funny. Um, and yeah, featured celebrities, but they they weren't kind of the the main event in this particular piece of content. Um, let me play a little bit of this to just remind you guys what it was all about. in Las Vegas. Quincy, it's 2:00 a.m. You have school tomorrow.

(37:15) Get some rest. See you tomorrow, Sequan. You got it, Quy. So, it's a lovely ad. Um, we have inspiration in there. It's about this this kid who's growing up in in Ghana. He's dreaming of being a a American football player. And it's really heartwarming. It's inspiring. There's admiration coming in there.

(37:52) There's hope and it's just succeeding on lots of different levels. And it's not trying to be funny. And yeah, there are little humorous bits in it, but that's not the main thrust of this ad, which made it stand out from the crowd. When everyone else is going for amusement, going for different emotions can be very, very effective. And that's exactly what this ad did.

(38:10) And remember that attention plays a big role. So this Google ad was interesting because it did brilliantly at evoking positive emotions. Like 66.2% is a very good score for intense positive emotions. So a large proportion of the audience felt intense positive emotions from watching the ad.

(38:29) But what let it down was this attention score because it opens. It's about um someone whose vision is um is impaired and it opens from his point of view. So if you guys remember, for many people with blindness or low vision, there hasn't always been an easy way to capture daily life. One face dropped. Move your phone down.

(38:54) Which is getting across the point of the guy's vision visual impairment. But as a viewer who doesn't know this creative and doesn't know where it's going, if you see that, you're like, "That's blurry. That's weird. I'm going to look away." And a lot of people did.

(39:12) So, where this fell down was like in the real world, it you'd be losing people after the first 3 seconds. Um, but also the brand recall was was low as well because it's a lot of the content is just showing the phone in this kind of blurred vision. Again, it's it's getting a point AC across the point, but not that many people were attributing this particular ad to Google, which is a another problem, of course, when you're spending so much money as this is an example of so um Squarespace big game spot directed by Martin Scorsesei. So, you can imagine how much this cost. Um and it's beautiful, you

(39:49) know, it looks it looks lovely. Again, positive positive emotions were pretty decent. attention is good, but that brand recall is just horrendous. Like it, this was part of a larger campaign. They put out a lot of content on Tik Tok where there's a guy who's designing his website um designing website for aliens and the aliens actually arrive and they use the website which he's designed.

(40:13) But if you've not seen the buildup content, then this is just a bit what well, you don't get it. And then the branding right at the end, I can show you it goes through. It's like a minute and a half long. I told you Broadway this always happens. It's super quick and then it's gone.

(40:37) And you've not seen any of the buildup content where they're creating the site and they're showing the aliens and like and so it just didn't really make sense. And that brand recall score 52%. You know, when you spend God knows how much, let's call it seven, $10 million on this spot when half the people watching that spot don't know it's a Squarespace ad. It's not a great result.

(40:56) Um, so Ian, we've got minutes left, so let's um let's not use any more video because there's also some good debate and discussion going on in the in the chat, our findings. All right, cool. So, one good way to make to really hammer home your brand and your brand and to drive that really great brand recall is to make your brand integral to the ad. And that's what Pringles did.

(41:21) Um, they had Chris Pratt who um has got a Pringles uh mustache and it basically was became the brand. And so all of the the ad is very much centered on Pringles. And if you remove Pringles from the ad, the ad doesn't work anymore. And that's a best way honestly to draw a brand recall. Absolutely. That brand recall score you can see from Pringles is superb there.

(41:40) Okay, let's talk a little bit about Tik Tok and what brands were doing on that platform. Paramount um they had a lot of really nice polished content to support the main ad. So it wasn't the main ad, they didn't repost the main ad. They had other ads supporting it which performed very very well.

(42:04) They they added color and amusement to the main spot in a in a really effective way. Um VW also did this brilliantly. They had a an unusual piece of content as well, very much focused around nostalgia and um the soul that the that Volkswagen brings to to its cars and the the Tik Tok content supported that brilliantly. So it's really powerful. 15-second ads all very strong.

(42:26) Um one thing that Tik Tok can be used very very well for is to tease trend and extend. So prior to the ad, tease it. And this is what Dunkin Donuts did with a lot of Ben Affleck content ahead of the of Game Day. Lots of teasing, building on last year's campaign as well, like what are they going to do? How's it how's he going to what's going to be the story line this time? And then they trended.

(42:49) So they put actually the main ad on the on Tik Tok, but they did it in a format which was made for Tik Tok. It wasn't a horizontal video, vertical video, you know, mastered and and and edited for the platform specifically. And then they extended. So after the game they showed um they showed content of what happened after um he goes in and tries to sing from pop star and he's kind of like shamed.

(43:16) So this is a really nice example of tease trend extend and also Sarah V te teased trended and extended really well but what they did in addition to Dunkin Donuts was they did so with unique content. So they built on the main spot. I didn't take like stuff from before it. It wasn't it was shot in very different ways. Lots of different kinds of content.

(43:42) So, influencer content um spoof content from the the brand like it leaked internal memos that that Michael Sarah has been trying to claim that he was he uh invented the brand and they shouldn't be he shouldn't be allowed in the building and these influencers being sent weird boxes by Michael Sah. Very very good. um all joining up together to really support the main message of the brand.

(44:04) Um so I won't read through these. Basically, it's just what we've what we've discussed in terms of our our key insights. Um hope that's helpful and we've got seven minutes and there's like 40 messages in the chat. So um does anyone want to come off mute and ask a live question? If not, I'll pull out a couple of the remaining questions from the chat.

(44:25) Thanks for being such a great audience. Yeah, thank you. This is great. I wanted to jump in on one of the rule breakers uh that I put in the chat about Bud Light with the genie uh because they ended with their wish being to go to the Super Bowl and then exactly cut right at their back at the game and the genie has brought them uh in the seats and and how that scored in terms of at least it was back to back and right there and it was a real good rule breaker.

(44:49) Yeah, interesting. Um I have to remind myself how Bud Light did it and certainly not not stand out. I think they were mid-table from memory. Um then they did that extension bit that um you know where they just flashed to them very quickly in the audience. So yeah but again into real time the focus was amusement right.

(45:13) Mhm. Like it. So, yeah, it was a slightly different creative device, but it was still playing in that similar kind of space to the other bands. And that's that's really the major takeaway for me. Like, fine, do funny if you want to, but you've got to really turn up like you really turn up the noise if if you're going to cut through.

(45:32) And actually, there are so many other emotions that you can go for which are which are going to get you stand out, make you distinctive. just to poke further and please other people jump in, but the the the fourth wall breaking is really what I'm getting at here. The ad was fine. I don't I agree it was middle of the road.

(45:49) It's that the genie took them to the game and they were at the game and if that registered I'll have to have a look. Honestly, I'm not sure. I like the the campaign itself did not it certainly wasn't one of the top 10. Um, so, uh, it makes me think a little bit of the Spongebob Squarepants tie-in that was all over the place. That was never really I don't know.

(46:15) Did they have an did Nickelodeon have an ad that tied into all the Yeah. during channel sim simal cast was Oh, okay. It was absolutely much better at some points than the actual uh broadcast itself. Wow. Okay. I missed that. It was the first time ever Nickelodeon also I mean they've done it before but they had a a box with commentators live.

(46:38) Um so oh see I just saw the ad part and that was cool because I didn't know what was going on but I'm like wow they paid a lot to be right there so it was actually a another show simal casting. I had to fight my seven-year-old not to watch that version and watch the real version so I could see the commercials because it was kind of like the puppy bowl you know has its own vibe. Okay. Thank you.

(46:56) No problem. Um, question. Oh, go ahead, please. Question about kind of the background. Um, this is fantastic. Stephanie, Ian, this is so fun that you worked your ass off for this for us. And David, thank you for for having this meeting. It's great. A AMG um jam. So uh the question I have is the hard question which is about you're talking about outcomes and do you have any insights and how you take this magic that you're providing to engage with that huge leap and jump across the company between what the marketing people are saying and what the sales people are actually checking or

(47:35) SKUs that are moving and what what actually as you say outcomes deliverables um the the question really is about you know the sales loop back. Do you have any experience or insights? Cuz I I bet that is the steepest hill that you're climbing. Yeah, 100%. So, um, when we work with clients, uh, on these strategic projects, they're often bringing their own data to the party and often that's sales data.

(48:07) So, for for instance, DirecTV, um, we are analyzing all of Direct TV's social content. Um and one of the outcomes that they care about that their seuite cares about is net sentiment score. So that that's one of the metrics that we they really care about. But also they have just brought to like we did a project on net sentiment score to to kick off and now they've just surfaced a bunch of website traffic and signups data.

(48:32) And so we're we're then correlating the social performance with traffic driven to the website and then people who actually convert. That's a super rich data set. Like clearly we're not normally privy to that data, but if the brand brings that into the process, we can absolutely correlate like our data with their data and give them insights not just around how to provoke great emotions, but how to then result in how how can we drive the results that are you are you talking to them about literally Direct TV new subscribers? Yes. Exactly. Exactly right. Yeah. Because they've got that data. Yeah. Yeah. So, that's a really powerful way

(49:09) to use like when the brand brings a lot of their own data and we mix it in with ours. Then you you get the why. Why did this content drive the outcome that you're looking for? In that case, it's new subscribers. Yeah. Yeah. Let me I I have a question to uh I can squeeze it in. Uh Ian, Stephanie, great presentation.

(49:30) I super super valuable information. Uh I have a question around uh David itself. Are you planning to make this available as a product as a subscription service or is that an internal tool that you're using uh for your services only? I could take this and a few people have commented above and so I added some um you know pricing for managed service. So we have two ways to buy managed service or self-s serve.

(49:59) on the self-s served side, it's giving you some directional emotions, but the managed service is really these deeper dives that we've been looking at um you know, analyzing all elements. And so for a single asset test, anywhere from 5 to 7K to really understand why the why behind it um 20 asset tests in the 18 to 20k range um to really get figure out that content at scale and that that deep dive whether it's your own things, a competitive landscape um and then we can do bespoke projects beyond that.

(50:28) So, uh, drop us a line. Let us let us connect on LinkedIn or I'll I'll put my email. We also have a giveaway. Um, anybody who wants to look at, uh, you know, a fuller report of this, you can download that. I'll I'll drop the link in now as well. So, David, you should introduce them to Donovan because they're pretending to build this.

(50:47) Yeah, for sure. I mean, literally, they're pretending to build this and why should they bother? They should talk to these guys. I mean, like, call now if you, you know, or or I'll email Donovan if you don't. So, uh, let you do it cuz this they're literally as I in tongue and cheek pretending to build.

(51:05) This is what's happening, right? But we have a built. So, and we're talking about Oglev is what I'm talking about. Yes. From last week. Yeah. We would love to be connected. Please. Um, let's continue. So, um, on that on that point, Rory Rory Sutherland is uh has just become an adviser of the of the business.

(51:24) Actually, we was on a call with him like an hour ago. Um so yeah he could help um move that along with as well. Well well thank you so much Ian for presenting all this and and Stephanie all your insights and and great to connect with you both uh appreciate everyone coming today. Yeah stay tuned for the updated uh info. Um anything else anyone needs from this then uh just let me know.

(51:54) Uh happy to connect you with everyone and uh um me and Stephanie hope to see you again soon. Thank you all. It's great to be appreciate it. Cheers. Happy Valentine's. Bye-bye. Cheers. Happy Valentine's. Bye for now. All those emotions. I know. Perfect day. Very emotional you got. You got an 89.5 by the way. Just just so you know my emotion. I just tell you that. Thanks David. See you soon.