 You'll just introduce yourself, Phil, and tell us how that came about, that initial book and your interest in bringing decision science into marketing. Yeah, thanks, Louise. So prior to that, I'd spent 25 years in client-side marketing. So I grew up in Unilever. I spent over 16 years there. Then I moved to Diageo and then to T-Mobile. And I thought probably quite arrogantly that I knew everything there was to know about communications, why people bought brands, how to promote a brand, etc. But I came up against a problem, which is that I had commissioned a huge piece of quantitative research when I was VP for T-Mobile in head office, Deutsche Telekom's head office in Germany. And the research was across 12 European countries and the results were just ambiguous. They just did not make sense. And I didn't know what to do. It was quite a difficult moment for me because I had invested a lot of money, company money in this and also personal equity. And someone introduced me to the guys who founded Decode Marketing, the company I now work with. And one of them is a neuroscientist and the other is a cognitive psychologist. And they completely blew my mind with their perspective on marketing and communications. And they had come at this, of course, from academia and science. And they helped me understand how and why ads were working or not. They helped me unpick some of the research and they were using frameworks and language that I had never, ever encountered before. And when I would look at them and say, how come you guys know all this stuff? They would look at me completely bemused to say, but how come you don't know this stuff? Because we're using frameworks or referencing studies that have been around for decades in some cases in the fields of decision or behavioural science. And by that, I mean a whole breadth from neuroscience through cognitive and social and evolutionary psychology. And then more recently, behavioural economics. And that was a real revelation to me. Not only did this parallel world exist that I'd never tapped into, never just been completely unaware of, but also the fact that this parallel world had been studying human behaviour for years and years and years. And in my experience, knew more about human behaviour and behaviour change than I certainly knew. And that anyone in the commercial world knew. And it was a real revelation because marketing is fundamentally about behaviour change. We want people to respond to an ad, to buy our brand, to share stuff online, to talk about our stuff, to buy more, to switch, whatever it might be. It's all about human behaviour. Long story short, I put these guys to work on the relaunch of the T-Mobile brand around Europe. And their approach had a huge commercial impact. So the first part of the relaunch was in the UK. And based on the principles that Decode introduced, we created an ad that grew sales by 49%. And to date has something like 41 million YouTube views. And then we replicated the principles across other brand touchpoints, a similar effect. We had the sales department ringing up saying, why didn't you tell us about all this activation activity? And the fact was there wasn't any activation activity. It was people responding to the advertising. And the more I worked with these guys, the more I realised this was so fundamental to marketing and I wanted to be part of it. So I quit and I joined them and I set the company up in the UK. It was founded in Germany. And then a couple of years later, I thought, this is so fundamental and it's so important. And I'm so convinced by behavioural or decision science that I want to bring it to a wider audience. So that was the genesis of the book. And I thought, if I can capture these key principles in terms that I understand, because I'm not a neuroscientist and I never will be, but if I can understand enough about the key principles, translate them into everyday examples, a language that can help other marketers and their agencies. So yeah, so that was the origin of the book. I then revised and updated it at the end of 2022 with new content. And yeah, behavioural science has become more accepted over that time, but it's certainly still not mainstream. I don't think it features necessarily in every marketer's education. It should. If I had my time again, I would probably go back to university and study psychology or something, because as I said, it's all about behaviour change. So yeah, still, it's still obviously a very active and burgeoning world in academia. There are many more courses now available for people to take but to undergraduate and masters and of course further levels to do with decision making and communication. So yeah, it's becoming more accepted, but still there's a way to go. And in the sort of in the library of what we put forward as recommended reading in behavioural science, your book for me was such an eye-opener. You go through the academic journey, the thinking fast and slow, which is hugely inaccessible. There's the childini's book, absolutely fantastic, been updated pre-Swasion. So influence now pre-Swasion. But as you were saying, and you still work very strongly in the academic field, they were very much academic psychology experiments. Whereas moving over to open your book and be actually talking about advertising, advertising, real life cases, marketing, seeing advertisements, seeing examples of such simple concepts like just moving the numbers further apart to make the price appear much wider apart. Simple things like that was really such an eye-opener for myself. I mean, since then, Jesgrim's Ripple I think is sort of along the same lines, Rory Sutherland's Alchemy. But there's still really a dearth of practical books. Ask is showing us how we can take these examples forward and try them for ourselves. Yeah. What I was particularly delighted about was after Decody was published and I did some work with Carlsberg, the beer company, and their insight manager in Denmark, which is where they are headquartered, loved the book, and she took some of the ideas back to her business and said, listen, we should implement these. And she was met with a lot of resistance. I said, yeah, academic stuff. It doesn't work in the real world. So bless her. She set out to test the things, concepts, and the principles herself. And she kindly shared the results with me. And some of them are published in the updated version of Decoded. So she was able to say, right, this is the theory and what Phil published. Here's the test that we did. And here are the real world results. And she found without fail that they replicated. So she's got some really robust evidence then to say, here's some really neat stuff that we should do. And often it's kind of trivial little things that you can change. If you're going to be printing, in their case, a menu anyway, then why not print two versions with a minor change between them and see what the effect is. So I think that's the real joy of it when you can take the principle and then test it and apply it yourself. Absolutely. So I'll just say welcome again to everybody who had joined us. I hope you're enjoying our conversation here. And do please feel free to put any of your questions in the chat box. And I'll give you the opportunity later on to come and put them to our speaker today yourself, to Phil. We're going to be talking in a little while about value for money, what that means, and then we're going to move on and we're going to hear some very exciting news about how AI tools are being used, I feel in particular, in their company. So we've talked very much about your sort of background for everybody who wasn't so familiar with you, Phil. Let's move on now to some work you're doing at the moment that I found really interesting when you told me about it. And it's all around the understanding, the meaning of the term value for money, which as I said to you, I could immediately excited about because I'm so often aware of subjects that aren't generally accessible and yet we all know that we walk into the supermarket, pick something up and say to ourselves, well, that's good value for money. But yet as we were chatting between ourselves, it has such wide reaching implications, both for luxury, vaingun goods as it were, and just down to a loaf of bread. So maybe you tell us a little bit more about the work you're doing and the understanding of this term value for money. Yeah, sure. Let me just share some slides. So this came up within the last 18 months because of the so-called cost of living crisis and the fact that prices were rising and people didn't have their incomes were not rising, forces behaviour change. And when you look at the main motivators or claimed motivators of purchase, whether it's during cost of living crisis or indeed a couple years prior to that during COVID time, value for money is the most often cited. I want value for money. But people do, they mention other things like product or service quality, reliability, prices mentioned, and of course that's a factor in value for money. But value for money has two components. It has money, of course, and it has this thing, this concept of value. And it doesn't always mean cheap because the same person could spend over 5,000 euros on a Louis Vuitton handbag and think that's great value for money and then they could spend one euro on a gelato and think that's good value for money as well. Or maybe that's poor value for money, depending. So it's not only about money. We need to really understand how the brain determines value and how what determines what people are willing to pay. Because ultimately, we could reduce all of our prices and sell more, but that's not what brands are about. Brands are about some sort of intangible equity and quality that makes people willing to pay for them. So if we go back to neuroscience and look at what determines a purchase decision, I think that we can learn a lot. There's a really interesting experiment which was carried out at Stanford University by Professor Brian Knutson and his colleagues. And he put people into fMRI brain scanners and he did three things. He showed them an image while they were in the scanner, an image of a product. One, Godiva chocolates was one of those products in the study. So they saw this for four seconds. Then he showed them the price for four seconds. And then he asked them to press a button to indicate whether or not they would buy the brand for that price. And all Knutson wanted to do was observe what happens in the brain at those different points in time. And what he found was very interesting because when people see a brand, their so-called reward system is activated. Now, this is part of the optofrontal cortex and the scientists knew from prior studies see very significant things. One, that the same part of the brain is activated when we see something that is pleasurable or rewarding to us. So if you show a mum a photo of her kids, her reward system is activated. Or if you show art lovers images of artworks, their reward centres are activated. So that was one thing. This is brands are inherently rewarding or pleasurable to us. The second finding from prior studies was that when this part of the brain is activated, there's a very high probability that action will then follow. Which is plausible, right? Because if we see something that we want that's rewarding and pleasurable, we want to get that thing because it provides some sort of reward. When people saw a price, something very different happened in the brain, the insula which is linked to, is activated when we experience pain. Both physical pain as well as emotional pain. And also when we experience disgust, the insula is activated. And again, this is plausible because it's almost like the brain is saying, I want those Godiva chocolates but it hurts because I have to give you money and giving something away is painful for the brain. And then what happened was a trade-off between the two activations. And if the reward activation was sufficient to overcome the pain activation, the person would then press the yes button, the yes I will buy, and if they didn't press no, then pain was more significant. And this is what we call, it's the neural correlates of purchase. So basically what goes on in the brain when we make a purchase decision. And because the brain does this before we actually make the decision, this is about expectation. So it's about expected reward and expected pain. So when it comes to cost of living crisis, what are some of the things that we can do? So if we look firstly at the pain side, are there ways that we can reduce the expected pain? Now we know that money is equated with pain, but it's also the behavioral cost involved in acquiring the reward. So the time that it takes, the cognitive effort that it takes, whether there is any other sort of frustration or irritation along the path to purchase, or even thereafter. So in the cost of living crisis, if I buy something and the packaging isn't suitable and I end up wasting it, then the brain learns from that experience and that will increase expected pain in the future. So it's all about not just money, but the behavioral cost involved in the purchase. Now some companies address this cost of living directly by reducing the pain. So here's an example from Nomad Foods, Bird's Eye brand. Bird's Eye pays your bills, right? It's tough right now. So we give you a chance to win money to pay your bills. You can do that directly. But there are other ways that we can work with it. Because it's expected pain, we can shape that expectation by how we visualize price information. And this is an example from Decoded, where the daily special in a restaurant was chalked up on a board and they showed the price three different ways. It's the same price in each case, 10 euros. But the menu on the top right, which just has the number 10, sold more than the others. Because the euro symbol and the euro word trigger more pain than just the number 10 on its own. The example that you alluded to, Louise, before is about price reduction and how the brain assumes that numbers that are physically separated by a greater distance have a greater difference between them. Because when we learn to count, we count in the horizontal plane. And as the numbers get bigger, the distance separating them also grows. So people rate the discount at the bottom to be greater than the one at the top, even though, of course, it's not. And how you visualize price can be even more important than the price itself. This was a study that we did where we got people to rate prices according to their presentation. And consistently, people rate the prices at the top right to be more expensive than the ones at the bottom left. Why? Because they've got codes contained in them that we learn are associated with premium. So if it's the black and white, that's quite minimalist. If it's the red on the top right, that's embossed. It's got a starburst. It's got a silver background. All codes that we learn from premium products. Whereas the bottom left is discount land. It's promotion land. It uses promotional colors. It uses very basic pricing that we've learned from discounters. Even things like how we show a price reduction. You know, the same reduction left and right, 8 euros down to 5.99, but the one on the right has significantly higher purchase intent, simply because of the way the brain processes numbers, which is based on something called size congruency, is does the size of the number, the font, the physical representation of the font, is that congruent with? Does it fit? Does it match? With the magnitude of the number. And if it does, it's processed more easily by the brain, and the fact that it's processed more easily leads to preference. And the example on the right-hand side is congruent, because 8 is bigger than 5.99. It's also written in bigger font. So this is the other example. This was from Kalsberg I mentioned before. Coming back to the menu, the Hungarian goulash, that was the example in decoded. And then the insight manager at Kalsberg said, let's test it. So they produced some menus, and on some of them they had the Danish currency symbol, Krona KR, and on others they didn't. And you can see what happened to purchase. They got a revenue increase of 11%, just by making that small change. Now a lot of people would kill for an 11% revenue increase, right? And you're not doing anything else. You're not doing anything out of the ordinary. You are going to print your menu cards anyway. So this stuff does work in practice. So that's just some examples for the cost of living from the pain side. When we look at the reward side of the equation, there's a lot more to go into. And if you read decoded, there's a whole chapter on this about goals, which is the fact that we buy brands and products because they are instrumental in helping us achieve goals. Achieving goals is rewarding for the brain, and that's how it fits back with the reward system. And goals are a mixture of functional goals. So what the category or the product helps me do, whether it's raw band speed or lean laundry or transportation or something that tastes good, but they're also more social emotional and psychological as well. And if people are interested in learning more, please reach out to me. I want to talk to you about that in greater detail, but I don't think we've unfortunately got time to do that today. That was really fascinating, Phil, and call it what you will. Call it behavioral science, psychology of marketing, decision science. Those slides that you've shown us are concrete examples of how this really works in real life, which is, as I say, for myself sort of that groundbreaking tip-over point where we moved off from nudge, thinking fast and slow, random hundred men in university asked this question to, as you say, your very concrete examples of these practices working in real life. And as I said, since then, I've read similar stories. And particularly in Jez Groome's Ripple, he talks often about the menu example. They're such simple concepts, aren't they, Phil? Absolutely. And often it is. Rory Sutherland had this lovely expression that he's used right from his days when he was doing direct marketing before he got into this whole area. And the expression is dare to be trivial. And his example was just changing little things on a direct response coupon, direct marketing and measuring the effect. And what he found was it was often a seemingly trivial changes that actually drove a big impact. So I'm going to jump into the questions before we move on to our third subject today. There's a comment from Giles. He's not able to ask the question himself, but it's a great question. I don't know if you can actually see that there, Phil. So I'm going to read for anybody who hasn't got their chat turned on. So Giles Edwards, thanks very much for joining us today. He says, specifically if I've understood it correctly, your stance that ultimately a brand exists to reduce cognitive effort when people make decisions. There's a few iterations of this point we use regularly when presenting the case for brand or future sales as it's perhaps better articulated to bits. Can you explain a bit more about the relationship between building trust, familiarity, and reducing cognitive load as a rationale for brand building? There's a lot in there, but let's try and just tackle that quickly. Yeah, no, it's a great question. And I'm going to answer it with another slide if I can, because this slide was one that shocked me when Decode first showed me. So these are the, excuse me, fMRI scan images of the same person faced with different tasks. And what happened before this experiment was people were asked to name their favorite brands in a number of different categories and other brands that were in their consideration set as well, so their repertoire perhaps, and other brands in the category that they would never consider buying. So that's what the scientists knew. First, we got this person's stated favorite, their repertoire, and brands they wouldn't consider buying. And what they did was put people into brain scanners and they had photographs or images of the brands and they drew a couple at random and showed them to the respondent and said, look at the images and choose a brand to buy. That was it. Look at the images, choose a brand to buy. And they just wanted to see what happened in the brain. And these are the responses of the same person's brain to different choices. So when I was Decode's client, they showed me this and they said, Phil, you've worked in marketing for quite a long time. How would the brain respond to its favorite brand? And I thought, I'm trying to trick me here. And I said, it's obvious. You spend years building emotional engagement and resonance and affinity and whatever you want to call it. So if my brain saw its favorite, it's going to light up like a Christmas tree lights. And the interesting thing is it doesn't. The brain on the left is thinking. It hasn't seen its favorite brand. So it is now considering a brand to buy. Whereas the brain on the right has seen its favorite and it has decided just like that in a millisecond. And it is quite literally a no-brainer decision. So behind, you know, to back to Giles's question, brands become shortcuts to a decision. And why this is important has to do with our whole neurological evolution. In fact, our continued existence as a species because the brain on the left is burning up to 40%, 40% of the body's available energy. The brain on the right is using less than 5%. And if there's one thing, one mission the brain has, it's to keep us alive on the planet so we can pass on our DNA. That's it, basically an evolutionary terms. And so the brain has evolved the most efficient operating systems it can to conserve energy because energy is not important for buying brands, right? Energy is important for survival, which is a little bit more critical than which chocolate bar do I buy? And that's why these systems have evolved and they've become shorthanded to system one and system two and this is quite a good example of that system one on the right being automaticity, right? It's intuitive, spontaneous, automatic, geared for action whereas system two on the left is more reflective and effortful. So that I think is, you know, hopefully the answer to Giles's question. It's super important because if you can become the preferred brand you've got all this millennia of neurological evolution on your side because then it will take quite a lot to upset that and to, of course we do reconsider, of course we can interrupt and disrupt and get people to try different things and maybe with experience over time we switch preference. But if not, if the brand does a good enough job for us and continues to help us meet our goals then we'll work with the brain looking like the one on the right. Really fascinating and controlling, controlling our excitement is, you know, that's the major, that's the human task that overtake, we overtake to stop ourselves being impulsive. I did a bit of research into dopamine. I have Parkinson's coincidentally and when you start taking medication for Parkinson's disease which is a form of dopamine, there's racks of warnings about getting people to watch your behaviour because as your body's becoming adjusted to the extra dopamine there's very, very sad cases of people who've been involved in, you know, impulsive shopping, impulsive gambling, actually losing their homes from impulsive gambling. So it's a really, really fascinating subject and as you say, the core to all this argument all just comes down to our human behaviour. Yeah, absolutely. So let's move on to our second subject for the conversation today. Giles saying there that's the slide I used. So there we are. I'm stuck to Giles using the appropriate material. Let's talk a little bit about something you were telling me about, AI tools that have been built. You are using the behavioural science principle for creative optimisation in advertising. It's, you know, it's the glittery toy at the moment. Everyone wants to talk about AI. We all want to quickly upskill learn how to use AI. So let's hear how AI is being used in your own particular company. Yeah. Yeah. And I think I was looking back over my career at memorable moments and the whole revelation from behavioural science is certainly one of them. This is another. And the reason is as follows. One of the founders of our company, he is probably one of the most intelligent people I've ever met. He annoyingly did a double PhD, not content with one, at Caltech as well, California Institute of Technology, like, you know, the foremost academic institution worldwide. He did a double PhD, one PhD in neuroscience and the other in artificial intelligence. He co-wrote a book, Understanding Intelligence, which is still used in academia. So this guy has been waiting patiently for the computing power to enable his vision. And his vision is to have AI tools that are trained with human data and built around behavioural science principles. And now, thanks to, you know, the NVIDIA chips and the sort of leaps and bounds that have come on in machine learning, we're able to do that. So I'd just like to show people some of this because I think it's really, really exciting. It starts with the fact that, and this is from Nielsen, but if you talk to Cantar or econometric modelling companies, they all put the figure around the same contribution, which is that the effectiveness of creative assets is the single biggest driver of sales and profit. So you can get your, you reach right, you can get your targeting right, but if you get your creative wrong, then you're leaving a lot of money on the table. So this Nielsen put it at 47%, Cantar say it's 50, data to decisions, a big econometric company put it around 50 as well. So it doesn't matter. It's the single biggest driver. So what we thought was how could we use AI and behavioral science to optimise that 50% to get it as good as it could be prior to going to market? Well, firstly, the question was, what's the barrier to doing that? And the barrier is often companies don't do it because it's hugely complex, particularly given the velocity of digital. Now, you know, how many different creative treatments do we have? How many versions do we have for targeting? So we haven't got time to do it. We don't have the budget to do it. Often we don't have the knowledge to do it. And you also find pockets of excellence in a business, particularly international businesses, but no real consistent way of doing this. So there's a lot of barriers that are very, very real. So we went back to first principles and said, what do we know from behavioral science are the drivers of creative effectiveness? And they are effective. They are, in essence, these six things. So firstly, you've got to get into the brain. That kind of goes without saying, but often say attention doesn't get the attention it deserves. People jump to things like emotions and forget the fact you've actually got to get into the brain to start with. Secondly, you've got to make sure that your creative is correctly attributed to your brand, otherwise you waste your money. If it's not linked to your brand, if it doesn't get encoded to memory, then game over. Thirdly, you want your creative to be processed as fluently, as easily as possible by the brain. So coming back to the two brains there and burning energy, we don't want the brain to burn energy. We want this to be processed as easily as possible. Fourthly, you will probably have some sort of message, an intended message. It's written in the brief that you want people to take out. You also want the message and the execution to be on brand and to support your brand values. Fifthly, we do want emotional response because we know from science that ads that evoke an emotional response get processed more deeply and have better memory encoding and recall and a higher chance of being shared and going viral. And then lastly, you will probably have some sort of call to action, something where you want to motivate and persuade. So we took those six drivers of creative effectiveness and what we've done is build a digital consumer brain which mirrors all of them. And we call this Brainsuite. It's an online platform. Let me just give you an example for one of the models within Brainsuite. So we have over 100 different models that sit behind it dealing with each of these component parts. So when we look at attention, for example, what we want to replicate is what the scientists call ground truth. And ground truth is where human beings look actually and in this case, measured with eye tracking. So measuring real behavior, right? You've got an image on the left. Then you've got the eye tracking data so you can see where people look when they look at that image. So that's what we're trying to replicate. So up until, well, it still exists today, but there have been some systems around. In fact, we used to use this ourselves until we had our own from 3M, the visual attention software, which was based on a set of rules. You know, our people present is a contrast, our edges present, et cetera, et cetera. So it gave you an approximation, but you can do better than approximating it these days. So this is our visual attention model. And if you compare it to the ground truth, you can see how accurate it is. And the reason why it's accurate is very simple. It's machine learning trained with human data. So we've got something like 15,000 hours of human eye tracking video that helps us predict the attention to video. We've got eye tracking on something like four and a half million static images as well, which enables you to do this sort of thing. And you can do this in a minute. It's just ridiculous. You know, you don't need a sample of humans if you've got something like this. But especially if you're iterating creative or if you've got, I've got six different layouts. Where are people going to look? You just chuck it in the machine and it'll give you the results in a couple of minutes. So Brain Suite, as I said, covers the six effectiveness drivers. It works across any asset, whether it's social media video or a bit of pointer sale material, email marketing, anything you want to throw at it. It produces ready-made reports with metrics for each of the six effectiveness drivers. This is a real report for Budweiser. We can help as well by overlaying our expert understanding and telling you why you're getting the metrics you're getting and what to do about it if they're not good enough. And this is a real example for a PepsiCo launch in France of an energy drink called Rockstar. Let me just show you the results. So they gave us this out-of-home ad. They also used it for point-of-sale. We ran it through Brain Suite. It got an overall score of 68. You can see at the bottom we can do cut-through in the environment as well. And then we made some recommendations. And based on those, they changed it. And this is the second image. So the score went up to a much more impactful score. But when you look at the detail, there's some really interesting stuff. So not only did cut-through in the environment increase, but this is a new product launch, right? So attention on the product and attention on the brand name, which are critical, of course, they rose dramatically. And then in terms of the response, because this is trained, it's all trained with human data. So sentiment, which is one of the components of emotion, trained with human emotional response data, that went up by 20%. And we've also trained it with human responses to images and text so we can measure the associations that something conveys. So this was an energy drink. That's pretty important to convey. So energy associations went up by a third as well. Now, PepsiCo said, well, this is great. It sounds very interesting, but it's AI, right? We want to test this with real humans. So they then put both of these visuals through their own shopper research panels. And they found that the improved version had a purchase intent increase of 5 percentage points, not 5%, 5 percentage points. And that was enough for them to say, yeah, this is good enough for us to use. And they're now using it globally to test all of their POS materials. Yeah, they're spending hundreds of millions of dollars a year on POS and they'd never really tested any. So this enables them now to do this very, very high response, sorry, low response time. I mean, like they can go online and get the results, get a full report in about three minutes. And the cost, because you can buy this on an ad hoc basis or a licensed basis, a licensed basis enables all you can eat. So the cost just tumbles. And obviously the more you put through the machine, the cheaper it becomes. So this is this is really impactful. We've had lots of clients obviously wanting to validate. And why wouldn't you? It's new. It's a bit scary. But we're always very open with how it works. You know, we tell people like the predicting visual attention example, how it works. And these are some of the results that clients have shared with us in terms of validating market impact. So 83% correlation with sales uplift, increase in video view-throughs and stronger branding and social media video. Those are all, you know, they're good metrics. But it doesn't stop there because the other benefit, real benefit of AI as well is what it can do inside a company. Coming back to the reasons why we don't test all that creative because it's complex, because it's time-consuming, because we don't have the capability and because we don't have the budget. We've had some amazing feedback from clients saying, this is just enabled us. You know, we're making decisions 10 times faster. We've saved 95% on our testing budget. It's 80% less effort. I mean, just, you know, astonishing capability improvements and also sort of throughput and efficiency improvements. So I think that's why you can see, you can tell from the way I'm talking about it. I'm hugely excited by this because I just think it's, you know, this is an element of AI. People get wrapped up with generative AI and chat GPT. And this isn't an area of AI that may not be hitting the headlines for the same extent, but the machine learning side. And when you, you know, you can see what you can do with it. It's having tremendous impact. So it's so interesting, Phil. I can see their people are putting in the chat. They're finding everything that you're sharing with us. Very interesting, particularly interesting for me. I worked with 3M Vaz practically every day in market research. We were giving feedback on new product design on efficacy of websites. So to see it honed in there, the difference from the 3M Vaz, which I always thought was amazing as it stood. But seeing the specification that come out about new model and also I suppose thinking about, we don't all have the budgets to sign up for these tools. But thinking back to your earlier conversation that there are very easily cost effective ways. You know, we've gone to the extreme. Some ways like this is the AI solution. It can be free to you using the tools. There are amazingly built tools, but also back at the groundwork stage. Sometimes, as you say, just doing an A-B test. It's hugely, hugely effective and results driven, as in the examples of two different adverts or three different styles of menu tested at different times, different days of the week, things like that. And that's one of the real benefits we've found clients have said, often the agency will come back with a few different layouts or, you know, it's up to you. You pick one. And what happens? Debate. Right? I like this. Well, I like that. Show the boss. What does the boss say? Well, I like this. And it's just time-consuming and it takes a lot of energy. And in the end, it's very subjective. So who's right? Who's wrong? And why? So this just helps make it objective. It takes the time and the effort out of the discussion. And it gives you a valid answer as well. That's the really important bit. So, you know, if you throw six different layouts at it, you can very quickly get one dashboard back saying, you know, it's this one or whatever. Or you might want to then iterate. So, okay, if we tweak this or tweak that like PepsiCo did with Rockstar, what's the impact of that? And because it's so fast, you can turn stuff around. You know, I mean, that report I showed you, less than five minutes. Right? You can do a visual salient thing in 90 seconds. It's just ridiculous. And in terms of cost, I mean, if people want to reach out and contact me, we can go through it in more details. I don't want to share cost publicly like this. But suffice it to say, we've got clients who are using it via a license. All you can eat, who are testing visuals for less than five euros. Fantastic. Fantastic. Fantastic. We've had such a wealth of information from you here today. We're coming very close to our time to wrap this up. Thank you, Giles, for your super question and all the other chat that's been going into it. If you've got really driving questions that you want to put to Phil, please send them via an email to me. You have my details and I'll pass them on to Phil. Yeah. Can I just answer the question? I love you too. I didn't have time. So please go ahead. Yeah. So Tal's question is, can you talk about Framie in terms of eye tracking testing as a creative will be next to other elements and various environments? Yeah. Great question. Yeah. The fact is you will never be able to replicate every permutation of reality, right? You don't know what your competitors are doing. You don't know what the lighting's like, what you're up against, et cetera. So what can you do? You can certainly test the image in isolation to see when people look at the image. How will that get processed? The other thing you can do is you can give the machine different contexts. So you could say, yeah, let's put it on a poster in a street or let's put it on a digital site just outside a store. Let's put it online, you know, on the left or the right, whatever it might be. So you can provide the environment and the machine will give you an idea of cut through within that environment. So it's a guide. It's to help you. But as I said, it will never replicate all the possibilities. Thank you so much. That's fantastic. And as you said, Bill, attention is number one. I kind of laughed when I saw you bring that up because coincidentally I was watching Zoe and Gary Glenn Roth again last night. Eric Baldwin, are you paying attention? So absolutely. When all said and done, everything that's been said today, attention is key. Yes, absolutely. You've got to get into the brain. It's, you know, without that, nothing else gets processed. Absolutely. Harry asked a question. How come in the brain scans, we see more efforts with boredom or boring brands? So the reason why there was more effort was because the brand that had been shown to the respondent was not their favorite. And there are a pair of brands and neither of them was the favorite. So they were then having to think which of the brands will I buy. And it's just that reflection, that extra thinking effort when the favorite was not present. Thanks. Thanks for drawing attention to that, Mary. And will the slides be available? Will you be able to share that deck with me? I can share a selection of them for sure. Super. That's super. Thank you so much, everybody, for joining us today. I hope you've enjoyed this whistle-stop tour about the science behind price with Phil. Thank you very much for joining us today, Phil. Always a joy to speak to you to hear all about the work that you're doing. Thanks, everybody. And do join us again. We usually host these events once a month. Watch out towards the end of the month who we'll be speaking with in March. Thank you so much, Phil. Thanks, everybody. Sorry, Louise, just very quickly. You can share my contact details with people. So if they want to follow up, feel free. Thank you. Thanks so much, Phil. Thanks, everybody. And have a good rest of the day. Yep. Thank you. Bye. Bye.