Is AI Creative A Giant, Distracting Lie? With Alex Cooper

Everything everybody says about AI creative is alive at all of it. It's all a lie That's my that's my current position on AI ad creative I feel this funny thing Alex where like we are really aggressively building our organization around using AI tools Of course, of course, we are using AI tools to generate creative. I think of a JF growth Kind of first and foremost as a creative shop That's probably not quite right. We have a really strong media buying perspective as well And I actually think we really help people just with our media buying alone Just because I think it's better than what a lot of alternatives are I think our financial specification is solid and helpful people But I really but the place where we are continuing to try and differentiate is like we want to make ads to sell products And we want to help brands sell products And and I want us to be amazing at that. It's really hard and and I want to keep doing it and so and so my My brain Perks up. I get energized when I see all the things that everybody else sees all over the internet especially on X about You know the new creative framework that somebody made with AI that's working and it's like claymation ads and animation ads and And automated this and that and and Lostatic ad generators and and all of these things that all of these people are doing But I keep running into the same problem over and over again, which is that in my view Huge huge huge amounts of it actually don't work They act that actually like it doesn't really help that much and therefore I Have I'm coming to something like this position and I haven't articulated it exactly but right now what I'm going to is It is still the case that the message is the most important thing in in the creative that the message is what drives creative performance first and foremost Sort of separate from the offer so let's put the offer in a separate bucket because obviously that can have a really big impact too, but In the in the realm of just creative itself the message is the thing that that the drives performance the most and that therefore Many of the creative things that people do with AI are actually distractions because they're built around execution not message They're built around edit styles and like I said, you know I know you pointed out at one point on X something about I think it was you tell me if I'm misremembering this but About like the like the fake crime drama ads that were made with AI that people were doing And it was like one that floated around that was really cool and it was it was actually like I thought it was really cool So yeah, it was impressive the AI use was impressive too, by the way like the it was cool But I think you pointed out like oh actually I'm seeing these and they're not actually working necessarily I'm looking at these at a bunch of different places and etc So I'm coming to this position that That the thing that makes the AI drama ad work the thing that makes the AI claimation ad work the thing that makes your song ad work The thing that makes your static image work is the message is the message is the message is the message and therefore All of those things are really cool if they can be extensions of the message If they can make it so that you can execute a message better But if they are not it's actually a gigantic shiny objects syndrome distraction and kind of a siop that like actually Isn't the case and here's my best evidence of this. I don't know That many brands that suddenly have an Explosion of growth and profitability bit based on AI creative that was not there before I have seen some brands do this pretty well and and experience that but even in those cases It's been like a really smart person who who really focused on driving things with a great message first or something like that You know, so so maybe I'm naive maybe I'm too narrowly focused on my my ads But that is like the position I'm starting to get and you know you and I like I want you to come on every few months and and talk We're friends You have great perspective on creative. I just like talking to you all those things are all there But you are the perfect person I think to talk about this because you have so much visibility to so much that's happening And I want you to feel completely free Alex Cooper founder of Parker and ad great and all of those things right you're seeing all this great in these places I want you to feel completely free To tell me Andrew, you're just wrong. You are just on in the stone age right now and that is your problem And and so so set me straight or agree with me and and helps out the people straight. What do you think? Well, I think that if you build it through the lens that you just described It's actually the greatest opportunity not the greatest distraction like I think the reason that there is so much distraction Online in these post is that to the hardest thing in AI is to discern what is worth building what is not worth building and to do that at least in our small space You need a really unique Combination of skills of being a great marketer which is hard enough to find as it is and being AI pills There's only really someone who Understands direct response principles and is able to make great ads about AI and also knows how to build AI systems Know what is possible to build with AI that can actually look at a workflow or look at a scope or look at some of They want to build and go this is worth the time that it will take because the result is that it will generate me a static workflow That does put out winning ads or whatever it be video editing or Headlines or research or reporting whatever if you don't have someone if they're just AI peeled They're not going to understand what good ads are so they can't build the system going to make good ads if they're just great Markets is then they're not going to understand how to build it with the eye so you're looking for a really rare person that There can eight come into your company and discern like what is worth building with AI And it's just not the case when you see all these workloads on Twitter a lot of people are missing because they're really one of those two things Or neither so you just see the fact you work for you. Oh, we should implement this But without actually going to put it through that acid test of old is this actually something from the specs of like Great DR and being a pill that is someone's going to be worth a ton that takes the build so I think if you do do those things We've built workflows internally and don't get wrong I agree with you in part we have had tons of things that we've built those in hindsight not been worth it I would possibly argue that that's somewhat a right of passage Yeah, yeah, for sure building part of the process if you want to if you want to build things that are hugely you know 10x improvements, and you've got to build things that a lot of things that suck But if you do put it through that process then let you can build things you know, off static Well, I suppose have resulted in as it spent tens or hundreds of thousands of dollars Because they're built through the lens of the message been the most important thing if you send for it around that then It's not like oh the core core statics better than the human statics. There's no difference in actual output Right, it's just right the question is not It is is a core code it does it actually work is the questions? Are you a good marker because if you are right is gonna amplify you and you're just gonna make more better static So you can actually test matching quicker in through Claude code or through codex. Yeah Versus like oh does does like the set of this is the AI static's work or did the AI video as well? It's like not do they work or not because they are possible to be done It's just are you a good mark to understand as AI systems and if so then you can just use it to amplify What you're already doing without AI? So Patrick and I internally have had this conversation a bunch and and we go back and forth here to the exact thing You said like if I look at some of the stuff we've made that is 100% AI souped and that's right in terms of in terms of the actual visual Design or something. It's just like statics There is in some ways no obvious difference between that and what a graphic designer does, you know like it's like the design work Strictly speaking is good now Sometimes there are actually Little tells that are like that. I would say are like the m- of Visual design where it's like oh, I can see it one of them is like the fonts change slightly Sometimes let's some headlines some of that kind of stuff or like the sizing changes and we raise whatever Maybe there's maybe for good enough at your AI image and that you can like sort of get around that But so there's some little things like that that happened but basically speaking there's no question that like the designs look good, you know But then there's this but and so I'm there. I'm like I'm I'm I affirm that right like 100% I'm there But then it's still so striking to me that so many of them don't do anything and and one of the things that I believe for a long time about this is that The reason people love to play quality versus quantity right they love they love to make it so that the moment you talk about a high quantity of ads They assume what you mean is and they're also bad, but it's okay because you get so many of them That's not what I want. That's not what you want. That's not what anybody. I know wants actually everybody. I know wants good ads Okay, and they want and they want to make as many as scale But the reason that like the idea of a static image edge image generator, you know, is so compelling to me Is not necessarily that you're going to create a huge range of ads and therefore Like be bad, but it's that one way to arrive at messages that work is to just put so many ads into the ad account that eventually you'll find something It's just it's just like shots on goal basically like you actually the shot on goal can get To use and I back, you know, it's actually funny. I know I know that some of the soccer analytics people football analytics people if you're from South Africa, you know Some of those people like I think they that they That they've even talked about this idea of shots on goal or or in hockey something like this like Basically, all you should do is like stop trying to take the perfect shot all the time take as many shots as you possibly can and you'll score more by doing that and not kind of analogy I think actually That that kind of analogy I think translates here to the way that I thought about this But then what's so striking to me is that I'm not actually sure that works I'm not actually sure that like I actually think it's possible you can make a thousand ads that all don't work You know like that like it's like it's it seems crazy is if you got enough variation in there, but Yeah. It is just remarkable to me. Otherwise, like all of us should be going to make that machine right now, because it just seems like if it's purely just mass volume testing, and you can run it through my media buying setup that I believe in where like you're minimizing the spend on the bad ads and and amplifying the good ones fast. Like, yeah, I don't know. First of all, I appreciate you calling football weights, correct name. You're welcome. Happy to try to make it a welcoming space for a, I know you're, you're in America now, you've moved here. I want to make sure you feel welcome, you know, awesome. So look, what I take that is if you believe what you said about the design part to be true, and you, there is generally no or a very small difference between the design of a human graphic designer and AI. And you build it through the lens of I want to test more, more messaging, then what is the reason, what is the reason that wouldn't be more effective than have a human making a thousand ads. If you can have more, more, more shots on gold by testing more messages, and we've had a client accounts that we've run this in, and like we've had ads like static as I said that gone spent 10 or 100,000. And the thing is, like you said, like people equate a high volume to be AI slot. It's so cheap to generate statics that like you don't even have to, you don't have to run all of them. But if you ran, you know, at 10 cents a generation, which is super conservative, if you, if you ran 500 of those and only so you have 500 ads for 50 bucks and only half of them were actually usable, and you've got like high quality tasks that you want to send the account, you got 250 ads for 50 bucks. So, you know, it was like like 20 cents a generate like per ad in the account, and still way better than you'd be able to get otherwise, you know, test way more messaging quicker. If you believe those to be as are like as good as as the humans would have made the account. So this is a thing to statics. I don't, I don't understand why more brands aren't sprinting towards this static workflow. I think right now we're in a window of arbitrage where there is some brands and agencies that have got to figure out and could do mass statics very easily and I can get wins from it very soon. I think everyone's going to have this machine and it's not going to be as effective because we're all going to be testing a thousand messages instead of 10 messages. So, I don't understand why okay, so let me ask you work you have two companies that are directly related to this between Parker and Adcrate right. Yeah, Parker is the main thing you're working on, but you see that great. So, do you then create and deploy this machine for every client and Adcrate? Yeah, and it literally literally did a part a course for Parker recently that we ran. Yeah, a lot of spend behind to get people in and all we had a conversation for all of our team to see who create the best like Yaparad to promote the course and we all made a ton of ads. I'm in a ton of ads. You made a ton of ads. We had all of our engine make ads as well. We all made as we were super proud of and then we had the code code machine and the code machine was the one that. Yeah, trust you. Yeah, all of us, but it's such a simple static ad as well. Yeah, right. So like for your for your e-com clients though that you're doing that you're running this playbook for for those ones like are you seeing. We're going to explode with revenue and profit because because of this mass endless amount of static ads. No, they're not exploding, but also if we had like a thousand really good human design ads, I don't think that they would explode either. Are we getting better returns and like more percentage of winners coming from statics than we were when it was just humans making five 10 static ads a month per brand. Yes, we are. But like are we suddenly seeing brands explode? No, but I don't think even a thousand human test messages would do that. There's a reason like most of my clients are using intelligence at this point. It's because it is an incredible tool for getting more value out of every click that comes to your website very simply intelligence does that in a couple of ways. The first is with being a really robust zero testing tool that allows you to see to see exactly what every test is doing at the level of everything that actually matters in your business. That means subscription rate if you're changing your subscription offer that means discount take rate if you're changing the discount offer on a website that means the actual profit per visitor if you're changing a or shipping thresholds or discounts or offers or anything like that. You can do all those things for you at the same time it also has a robust upsell cross sell tool built in so the level of the product page in the cart at the checkout even post purchase you can be adding upsells into the flow of your traffic and therefore generating more value per click and of course you can see the output of all of that testing as well at the level of profit per visitor. Intelligence is just awesome it's one of those things where if you're trying to get more value out of each click that comes to your website if you're trying to do something like what we call CRO you should be using it it is awesome it is easy to use you don't need to develop or go to intelligence dot IO and use the code Ferris 20 F a RIS 20 and that will get you 20% off for first three months or go to intelligence IO slash audit a UDIT audit and see if play with the tool free tool that will allow you that will have intelligence AI tool tell you what test you should be running right now on your website really cool really fun to put with good intelligence get it. Start making more money with the traffic comes to your site. Yeah and that's so interesting why why not if you have endless limitless creative on static images like. See I think that's a fascinating thing to me shouldn't you be able to is it just that there is some capped upside on static ads like you just have to win with video is that like. What why is that if you have endless and you have you have Alex Cooper's brain and operations workflow on your team like you you know like I just. What is it landing pages offer why is it that that doesn't work better if you have that kind of firepower um well I think it like this this so many factors that go into it I I think that the the variation will be a lot greater for video like if you managed to crack if you managed to truly crack. Unlimited video in the same way that you can crack unlimited static I think now would have a much more significant impact on on business growth overall. And okay so the principle so the principle let's just be clear I just want to get to the get to the principle I'm sorry to interrupt you the principle there is that video has more creative variation possibility and more potential impact so so part of the bet that you're saying there the principle that you're saying there is video just has more impact than still. Uh yes I would agree with that okay yeah okay I think I probably do too yeah but yeah depending on the category but yeah yeah um. And then you also take into account like how much how much more of an improvement is this overall for our creative the mirror we do it we're in some like pretty. Some preview accounts that have multiple agencies in there as like they're making some really good at so is us going from 10 to 1000 ads even really improving the account like at large by that much. I mean it's from testing measures perspective but is it going to be from a number of when he adds. I mean it's it's again it's an improvement on what we were doing previously which was using human graphic design as to make it. And in that case it's worth the investment that you know we spend on the credits on the tokens inside of cord. Sure but like I'm I promise in that people who build this are going to go and something be able to generate unlimited money and. And explode their out accounts no but I don't think that should be the goal like the goal should be to improve on what you currently have and if that is the case then what would be the reason not to do it. Yeah but it's it's still so remarkable to me that we're saying there's that much of an increase in static out of fire power and it doesn't really have that big of an impact. I'm not I don't want to put words in your mouth like why not why can you like is it because yeah just an ad only has so much potential power or. It just should be that if we can all 10X our static ad jet output and it only improves things like like you know incrementally and I or you know marginally let's say. At both of those words have multiple meetings but I think you know what I mean only somewhat let's say it is still striking to me like that says something about. Like it says something to me about where an operator and even an agency ought to put their time and attention you know like where if that's true I mean what we're saying. Another way of framing this right is that there is a time and money investment some of the money is easy to see and in quantifiable in terms of like tokens or whatever but. But some of it is just like the time to improve the process is there is there is something here where like we just all should stop putting our time and effort into that. Because because it just doesn't it just doesn't actually like like all of your efforts should go towards something that's going to move the needle more you know and this is actually the big thing for me in this whole conversation is like. AI is is so promising in so many areas and further even when AI is not very helpful when it doesn't when it when it sometimes when it actively misleads you or when it or when it looks like it's going to be really useful but actually doesn't have that much impact it's incredibly convincing right so like it's like the it's the greatest liar of all time like you just use and I look I mean as much as anybody but the thing I'm watching so closely right now is like. What is actually creating these unlocks for these brands and is that is it just about like product category 10. and unit economics? Is that really the only thing that matters? Or is it just about making video? Or is it just about off? Like what actually does it? And therefore, what ought we all to do? Who, particularly people like us, or agency people, and tool people who are trying to make a creative and help brands grow, if what you're saying is true, then it's really cool sounding that we can go make all this stuff. But maybe just like, I don't know, spend as little time on that as possible. Or even like, honestly, part of me thinks like, like cut that part of our service, let somebody else go make the static ads for really cheap and move on. You know, like, and I'm not saying I'm not, I'm not saying, sometimes the people listen, I'm not talking about firing anybody today. Like, I'm just trying to work back the first principles, you know? Yeah, so, where I'd push back there is I don't believe getting a static workflow to put out good statics as scale for your brand takes as long as vast majority of people think. I think the time investment that we took to build what we've got, and I didn't spend maybe a few hours building the workflow, and I just build feedback mechanisms for the teams to go in and prove it. And they've been the one training it over the last few months. I don't think it was like a not good use of my time at all to build that. I think everyone watching this. I don't think it's not worth three, four hours of your time to go and put together a swipe bar of templates, get it to some of the build system to connect to customer reviews, add comments, all different voice of customer sources that you've got, and get to start putting out static ads. Like that for me, even if the overall effects on the ad account, which can be, can be small incremental, also can be huge, like again, I'm at the market where it's actually our top forming ad. I think that's a very worthwhile use of time, or just get someone on the team who's AI building a good market to go and build something like that. But I agree with you, it definitely sounds a lot sexier and sounds like it's gonna have a bigger overall impact on the business than it actually does, but I think so is a lot of AI. Yeah, yeah, I mean, 'cause what it points me to is like, what actually does then have a really big impact on the business, you know? And how should our service be designed to support that? You know, like, it's funny. There was a conversation yesterday on X about sort of like brand consideration versus, you know, the DR Demons. And one of the things that came up in that conversation was somebody saying like, I know all these kids under 25 who are just making all this money selling their DR Demons and they're making all this money selling DR supplements. That's where all the money's going. And my response to that is like, it's like, okay, so is the answer just supplements? So I'll come back to supplements. (laughs) It's the answer that we just, because the economics are so good and all that that like, in that case is what's driving this forward not even the DR Demon part of it, but actually just like selling the right category. And there's obviously some truth to that because having like category channel fit or category business model fit really matters, you know? But yeah, it just makes me think about what how should we be designing our service to actually help people the most possible? And what should we be telling people? And how should we be prepping people for what we're going to do? Like, you know, I don't think I overpromise to our clients and say, hey, we're going to bring in this creative process that's going to triple your business or something. I resist to those kinds of claims on sales calls at IGF growth. But it's almost like we need to more fully reinforce the idea that the way you really win. And I do think this is true. And this is like another thing I'm circling on a lot right now is that you is by just mattering a lot to your customer. And that happens at every level of the way your business works and great brands do that with great creative for sure. But that's only one part of the process, you know? And it's about a lot of other things that you do as part of that. And creative execution then becomes not the magic way your business triples or something, but actually becomes one part of a larger team effort where you have to have it and you have to do a great job of it and hiring a grad agency to do a good job and push it forward has real wins and has a real impact on the business. Just like product development does, just like affording your supply chain does, just like everything else does. And the expectation should be aligned to that possibility. And in the midst of that nobody should expect AI to come in and become rocket boosters on it in like a crazy revenue profit way, right? Like so it's still as important, it's operationally significant and I just want to be really clear. We are talking about thinking about investing in AI generated and powered workflows every day all the time, right? So I'm not saying like we're going to be the anti-AI agency far from it, we're the opposite. Like we are very AI forward, I think, but yeah, you know, just it's just interesting how much all of this stuff happens. And yet then maybe there's some principles in place about what makes a brand great, what makes a brand grow that actually are completely the same as it was 12 years ago when I got into e-commerce and nobody had ever heard about AI. - Yeah, I think that's right. I mean, I mean, like I'm not gonna dispute that. Like, you know, wrong there is like that, that's why it's just gonna be an amplifier of people who are great marketers. And also it's funny on the conversation about the stuff that I suppose is someone who launched around recently and struggled with it. And then he just texted me, he was like, yeah, like if you have to win this game, you just have to start a software and that's what I might just do. So maybe, maybe that's just the move for all of us. - We all have to just go start some of the brands and that's pretty much exactly. - Exactly, exactly, but going back to the conversation. - Okay, so what like, what is a use case for AI that like can move the needle? I mean, I would argue the stats, you guys can move the, or not to the extent of some of the other things we spoke about, but like, what's something that I've found as an unlock and like has been huge for us? Well, I think in the research part, and I think this is, it goes back to the conversation about like, what is it to be a good marketer and come on as an agency and like actually help brand care about what matters? Find like doing things like crawling through all your different voice of customer data sources, looking through the customer reviews, looking through the ad comments, looking through the Reddit forums, for emotionally loaded phrases, for edge cases of people using the products in a different way than you thought. We had a client not so long ago who was sort of nose patches and all of their ads were around nose patches and we found in ad comments someone mentioned how they actually prefer to use them on their chin. And you're going to do some deeper research on Reddit. We have looked through customer reviews using AI 'cause they have synthesized, you know, thousands, tens of thousands of reviews at a time. We found that this is not an uncommon use cases actually our second or third most common use case behind like the nose patches. And we started making ads around that and that's completely new angle for us to hit and the ads went on like and go on to do really well. That is something that like now they spend a lot of money on ads around chin patches. Like that is something that you can do. The, it's like, it's using AI but it's using AI for the lens of like, this is just, we would do this anyway. We would do this if we didn't have AI and this is just what be a good mock drive for how the client see what with. Yeah, I think, I mean that really does come back to the message being the most important thing 'cause what you're saying now is you're framing the content, you're framing the use of the product in this really different way and that actually comes down to words. And I would actually say one of the reasons, so that's a really good distinction because what you're saying is you're using AI to generate the research. And by the way, by the way, I should say in all this, when I think about what Parker is and does for people, which you should clarify at some point in this conversation in case anybody wants to dig into it. But when I think about that, I actually see you building more towards the kind of thing that we're talking about here. I see you building towards like this ultimate tool for supporting a more robust approach to ads as opposed to, you know, just endless generator or whatever. You know, like, hey, this is going to be easy. It may be the way to think about it and not trying to, whatever, is compare the way you talk about Parker to the way icon launched where it was like get rid of everything, we're just going to make you all the money. Basically, like, I see you building tools towards it. So I think we're on the same page here. But anyway, to use the example you just gave, AI as research assistant and angle generator, message generator becomes a starting place. But then somewhere after that, you have to go and actually generate the ads that make that claim. And then that is how you push really far forward, right? Is like the ads actually do that. And now you can use the tools as a way to like the actual creative tools to then service the message, right? To support and reinforce the message about, in this case, the thing on the chin or whatever, right? As opposed to, as opposed to just saying, like, it's going to make all the ads and we just win, you know? There's still some element of like, you have to get in there and think about what that process is and AI can be this incredible support tool for surfacing that kind of message. One of the interesting things in all of this is that as many people have pointed out, it's very, very hard to get AI to write extremely well, in a way that doesn't sound like AI. I think, yeah, tell me if you think that's wrong. But like, I still think a lot of AI writing is quite bad. And noticeably AI, and therefore, it creates this challenge of like, it won't take the step for you to find the angle and then immediately write the perfect script to do that. Like, you actually have to still get in there and think about, okay, how are we going to talk in a way that makes sense for our brand and all that stuff? Yeah, again, we use a lot of AI assistants in our script writing, so it's not to say that it's not there. But yeah, it feels like what you're saying is that same thing. That like, now there's a way to build the tools around that message first process and you go to first principles there. I have one more thing to say about that, but let me pause and let you respond to that. And somewhere that's helpful to Parker does. Yeah, I mean, on the thread of. like execution versus research. This is kind of why we built Parker 18 months ago. We saw that a lot of the tools in the space were going after execution and going after like the unlimited as generator. And I just don't think that is the best use case for AI. The best use case for EIs is a research assistant and no one was really talking about that. So we built this thing that allows you to connect to all the different data sources, allow you to connect to add comments, and I just connect to Royate to customer reviews, poke purchase surveys. And basically trying to create the data and see what are you missing? Like what's in your customer reviews? We haven't tested what are personas that have come up on on Reddit forums, but like haven't really got that much spread in the ad account yet. So that you can actually as a mark to go, oh, okay, I want to make some answers to this and we're not claiming to be like the 1000 as generator. Because I think if you're optimizing for the best outputs, I would rather have someone who is doing like intentional volume versus someone who's going for pure volume and just trying to like crack that I first forget it will one find the waiting message. So I agree with you. It's always going to be or not always going to be like a research assistant. And this is also just knowing when to use AI and when not to like you just and it's briefly like we we internally add create are the same. Actually a lot of our strats and a growing testament I've kind of come across as I speak to a lot of people who I think are very good copy right in. It's just they're like not using AI to write scripts. And I think again, people see like my constant or see people in the industry and go, oh, like he's trying to say one person created seeing you could click a button, you get a thousand ads. Yeah, like no, I strongly strongly oppose that. Yeah. It's just knowing when to when to use it when not to. I just I don't think the AI copyright and has got to the point where it can be reliable. I think for static headlines, if you generate enough, it will give you some good ones. But script writing is just so so difficult. It is. Like just to understand like for even a human to understand the flow of like, oh, why do you use a like this? Why do you make it flow from this line to this line? It's just so difficult to explain that process to AI. And I don't think it's going to get better anytime soon because I think that's your regress since this point last year. So yeah, it's just knowing when to use a when not to not to sprint towards execution or or copyright in and use it as a as a research partner or anything else. Part of the reason I think some AI script writing and I'll say I'll call it AI assisted script writing works actually is precisely because of what you just said about human script writing, which is that it's really hard. So a lot of it is really, really bad. And therefore like you can actually probably get some bump just by using some AI tools in the process somewhere to help you with some parts of it, having some structural ideas and some of that stuff just because the baseline the floor is so bad. But to get great, I don't I agree with you. Like you have to they have to be more human to it. It's interesting because we've literally I speak about a lot the idea of like we're moving towards the world where the the creative team is just a never to be shrinking or it was so much more output with AI and the person in the best context is going to win. So we talk about a lot internally like can we just get everyone to record their work as they're doing their work and narrate what they're doing as they're doing it. So we have our editors as they are like making cuts or choosing piece of b-roll or like working out what the pacing is going to add that they're talking about that process as they go through it. We have our creator sources saying like oh why this creator, why does creator not in that career. Just basically for context, we're going to feed into into agents. We literally spent the like a we did a six month project where we had strategies film themselves while they were writing scripts and cooperating to try and solve this problem and we had tens of hours of transcripts that we fed into our internal model and that basically concluded that I mean the the uptick in script writing was marginal and it was not to the level of like a human copyright. So even with tens of hours of description on like what good copyright is or what we believe to be good copyright is still didn't get the point where we're consistently happy and our our strats don't use that tool to to to write their copy which is a which is saying something. If you love this conversation you want great creative coming to your ad account with people who are thinking these ways about their approach to creative you don't need a full agency you don't need someone to take your media spend you just want a really good creative service especially as an extension of your team for the kind of business that has good ideas for your messages good ideas for where you're going direction but you want great execution of the actual creative concepts with a thoughtful framework about how to turn those ideas into a wide range of creative diverse creative assets for your ad account and you want all of that at a reasonable price you should look to my friends at behind the scenes studio BTS studio.co behind the scenes studio is a Philippines based creative agency that is really in a lot of ways an extension of what I do at IGF growth each and every day because they're working very closely with me and my team but they have a separate side business where they staff not only my team but they're staffing other businesses as well specifically in that world of creative if you want stills and videos there's messages driven that's again focused on creating a wide range of creative diversity BTS is a great fit for you they're working with me every day all the time but getting ideas from me at the same time they're providing those to other clients so BTS is just a great option for you again they're best positioned for brands who are trying to extend their own creative resources and want designers editors and people who are who are able to take your best ideas and turn them into a wide range of concepts so if that's the place you are at really recommended seven figure brands who are trying to go their creative not ready to pass the whole thing off to a full agency or even eight figure plus brands who just want again extension of their teams resources at a reasonable price go check it out BTS studio.co they do a great job I'm working with them all the time you should do behind the scenes studio at BTS studio.co. Yeah I think it's I think it's very I think it's very fascinating that in all of this uh one of the challenges is that if we have like a rule in a gift growth which is it's a soft rule but I try to reinforce it anytime anytime we're doing ad inspiration of any kind we ask the question why did this ad work whether it's a client's account or just somehow that we see on the internet or whatever if we have good reason to think it works we try to ask the question why did this work and my what I do in that process is I put my thumb on the scale of that question and say I am coming in with a predetermined answer and the answer is the message okay that's what I say and that everything else works in service to the message um and I and I I am aware that that's actually not totally always correct that like sometimes the execution or the editing or the there's cleverness and the engagement or the hook whatever it's all so good that the message um actually like uh it's a decent enough message but it's not remarkable and therefore it but it's just delivered so well that like it gets there actually that's really hard to do but uh but maybe there's maybe there's something something to that or or or or maybe another way of saying it is that sometimes the message is a part of it but the execution is so good on the message that like if the execution was worse that same message would not have worked I fully get that that's all actually true in the nuance but what I always want to say is in the analysis the message is the most important thing because then I want to always try to analyze I just want to sort of force us to try to analyze ads through the lens of this is a communication act at core and therefore we have to communicate clearly and get something get a point across and change somebody's beliefs um or feelings about an ad uh about a product or whatever in order to or entertain them or whatever we have to we have to change somebody's beliefs or feelings in some way to get them to make them make them to want to buy the product and that's about that's about message that's about uh you know what you say and and how you communicate things okay so that's but that's what I always I tell people internally thing is if you don't have some very strong opinion like that one of the challenges in this conversation I see right now is that very often brands or operators or whatever we'll see some some ad that gets a bunch of of like hype on the internet or something like that and they'll go uh they'll go oh look at that song ad like the rise one that went around you know the rise one yeah yeah yeah the one about how if you don't take rise your husband will cheat on you and leave you a lot of single um just just just like terrible honestly anyway but um yes you can sell a lot of products by appealing to the worst fears in people so congratulations anyway the point is what everybody focused on when they looked when they analyze that ad was the song and they focused on the animation uh you know it's like a six-minute thing right but and of course there's something right about that the animation is striking because the the most noticeable thing about the ad when you first look at it is that it is animated differently than almost the ads that you see so it's an animation and secondly it's a song and that's not most of what ads have been for the last 10 years or whatever and so so they really stand out none of us have had the ability to hire an animator to go animate our ads and whatever and so now you suddenly have this power to create these animations and nobody none of us have the ability to go hire a band to go record or a music artist to go record a six-minute song about our product you know but so they really stand out but what if that what if I am right and actually the reason that the ads sell sells products presumably I think it sells products I don't know I don't have access to that account what if the reason that it sells products is actually because uh is actually because the message, and that's it. But those things as an execution of the message are actually like 20% of the equation and that you could have gotten that same message across a really different way and sold a lot of products a really different way. In that case, everybody's attention goes towards the AI side of things, but actually it should, the lesson you should take from that is something really, really different as the core message and AI just becomes an execution of that one thing and that's it. And of the reason I think people get distracted is because of the way all of those other features, the song and the animation stand out so much in a way that's so different and so it really does stand out. And at the same time, I'm not sure that's actually what's driving the performance of that ad because if you go make a song and an animation that looks just as good, your ad is very likely not to work as well unless you're willing to prey on people's worst fears. You know what? I don't know. So yeah, I think, or am I undercounting the execution there? Am I underweighting it? I think what you're saying is actually proven by the fact that, I mean, I'm sure you guys have added something similar. We've had internally, we've built our animation song ads and we've taken that exact script and had like a real human creator like tell that story, I mean, not to the same extent of like the, the Rosilia, Raya style ads, but similar. What I took away from those sorts of ads is like trigger event based stories, they resonate with people. Now, I personally don't have a tolerance to do something like that that's just made up. I think that if you find them in customer voice, then like those can be great stories. If you go into customer reviews, if you go into your ad comments, if you go into Reddit and find like real stories of real trigger events that people had that drove them to go and buy your products, I think it's totally fair game rather than not like as strong as those are some ones that you see online to go and create ads out of those. So we've done some of those and had those air animation ads recreated by real creators and they've just performed just as well or even better in cases. We've also had them put users native ads like the primary copy of native ads. So you've taken the same message, you've taken the same concepts and you apply it to different formats. There's less sexy looking than an air animation song ad, but they do, I mean, it's that's what's driving it to convert and not I guess it's easy because it's visual and it's very sexy to say that online, but yeah, no, you're right. It's the concept and message that's there's driving that and it's driven by the fact that like you can go repurpose that to loads of different formats and it will work just as well. Yeah. The reason I'm so focused on this right now is that I find it very hard to not be distracted by the shininess of the tools personally. I see that stuff online all the time and I'm like, oh, we got to be able to create that. We're going to get behind or if we can't do that for our clients or whatever. But I also find this thing happening, and by the way, by the way, I know the dynamic we should definitely call out speaking of that dynamic specifically is that I know for a fact, for a fact, for a fact that there are ads that are going up online and tools that people are talking about and look at this. We use this tools and this to generate this for our brand and the brand is not actually selling anything. I've seen it multiple times when I've gotten looks and inside out accounts that are going public recognition. Alex, I assume you have seen that same thing at play many times. So just be careful. Just be careful as long as I'm saying. Yes. Yeah. Sorry. Go ahead. Go ahead. Go ahead. Go ahead. Go ahead. To play devil's advocate, like the other side of that would be I do think there is a level of like arbitrage when a former is new to get on it before the rest of the advertisers get on it. Like, for example, when the AI animations first came around, it just goes back to first principles. All we're trying to do it adds is recreate the experience of the organic feed or get AI animation videos start going off the organic feed caught up a year ago. Advertises pick it up earlier this year. And if you made ads at the start, you could make anything like as an animation people and that's true to that work for you, you could make anything as an animation. It worked. We had, I mean, I say anything Lucy, but like, you know, they're like the higher I'm a tomorrow. And like the like all that stuff, like they work really well. I don't seem as many of them working anymore because then advertise that sketch on. And then people go start to go, oh, actually, these animations that like oftentimes, these are ads. And I think the same advertisers are already ruined that's for about exactly exactly. I do think to some extent that is true for the song ads. I think now, like everyone's doing them, but I why completely agree with you and that the message is a much bigger determinant in why those are resilient rise quality style ads worked. I do think there's also times we said for if you do catch something before the rest of Twitter or before the rest of, you know, the DCS starts making it. There is an arbitrage because the the person on the other side of it does not equate that to being an ad versus now everyone knows that yeah, animations can often turn into ads. Yeah, yeah. Yeah. I think that's fair. I think there are moments like this, part of my own sort of inclination is to not pursue arbitrage moments like that very much because I find that it's like a often a large amount of effort for a very short term impact when I want to build more first principles things all the time. That might be a mistake. Maybe like we should have a unit all the time. It's just like scrolling and finding stuff and just like injecting for our clients like that, you know, that might be a good way to handle that problem is of operationalize the like arbitrage and you know, you like Alex Cooper. I like Alex Cooper. That's why I'm having him back on every couple months to talk about creative and AI and all these kinds of things. That's why you should subscribe right now where we're watching listening so you don't miss future episodes with him and other repeat guests like Nate's Abogal and Taylor Holiday and other people like that subscribe one of the things I was going to say before was that in the midst of all those things, unless you are really disciplined about doing this, you're just going to be distracted by all of these things that everybody's telling you. This is the next thing that works. Why don't we do this for our brand when instead what you should be doing is generating much, much more focus on these sort of more core actual needle mover type things that start with words, that start with ideas, that start with your customers, you'll be getting on the phone with them and the and the brass tax on that is that therefore when your ad account is struggling, you know where to look first, which is like, oh, probably unless my media buying is a mess, that happens for a lot of people, unless that happens, like, you know, you actually have to turn around and say and say like, okay, what am I not saying or communicating or even making at the level of product? What is not resonating for people in the way that it should because that's what I actually want to do if I'm going to actually sort of move units and find a way to sort of like push all of the other things away from your consciousness a little bit because otherwise it becomes, it becomes a way, I guess, you know, there's this back, there's this love hate relationship with the phrase, you know, you're always one out of one out of way. You know, you know, people talk about that, like you're one out of way from sort of like a success. In one sense, that's true. And in another sense, it's like the, it's like the biggest problem to believe, you know, because it's like, it's like, if you do that, then you're just going to sort of spin your wheels so much. And a lot of times you're actually not really and you're really much farther, you're really much more towards thinking about your core promises and your brand and that will show up in one add at some point, but the work is not just going to make another ad, the work is to go think about how your brand is resonating with customers and those sorts of things. So I don't know if I'm being clear enough, but yeah. Yeah, I mean, this is definitely a struggle that I've gone through this year. Like we were, despite how a position now, like we kind of resisted, I get on to a gentle work called code encode X for the first few months, but I was kind of in this state where I do weekly training for the team. And I'm like, well, I want to train them up on on being a gentick and like working, like as an AI for an agency, but the same time, I also just want to train them on like research and position. Like, and I found myself doing a lot of training on a gentick work and some of the copyright in our, you know, our ads started to slip and like, should we even be worried about this stuff if we're not like a plus copyright as a plus researchers and actually understand like the first principles. So it's very tough because you want, you want to change the shiny thing and you know, it's like works on you and get more agenda and you do need your team to be upskilled into it. But the same time, like the most important thing is to be a good mock, that's understand research, position and copyright and all these things. So it's a tough balance. And I feel that too at the level operationally where I really, I really wrestle with this notion of like, where should I put my, my own time? I'm not naturally super AI forward. Like, I've gotten better, but I'm not great with it in the way that maybe you are spling. And, and, and anybody I think about where I want my team's investment time to go, you know, there's this part of me that says like, like, you know, put them into more stuff to become more agentic. And other times it's like, actually, we still just need to sell products first and foremost. And agentic things, like we got to get better there first. That's the next place to go. And until we're better there, like it's the only reason to use agents or anything AI is to support that goal. And that's it. And, but again, I was just like massive credit to Patrick here, who's basically built all of our tools for that exact purpose. Like he is just, he's pretty laser focused on that. Just like, how do we actually support the, the real work of that humans are doing? You know, we're still in place. We haven't fired anybody from because of, you know, AI, you know, taking their work or whatever. So anyway, there's, you know, so anyway, the point being like, there is a real trade off here. Because again, it looks like you see see people safe. things like, well, if your people aren't AI forward, that's like hiring somebody doesn't know how to use a computer at this point. Don't try to train them to teach them. You know, people will say things like that sweeping, you know. And I'm kind of like, yeah, but I'm not sure that's what our job is. You know, like, yeah, is the job even not much different than it was three years ago or four years ago or whatever? Like, I just I'm not sure it is. I'm not, I haven't been proven to me yet. Maybe it is. I keep going back to the the concept in my hair when I'm thinking about hiring of I would much rather. If you just like, fuck out, if you have two people, one person was a a nine out of 10, 10 out of 10 marketer and a six out of 10 AI peeled person and the other person was the vice versa. They were a nine out of 10, 10 authentic work on a six out of 10 marketer. I would hire the first person every day of the week. What if that person's a three out of 10 agent, a person, I would, yeah, that is a little bit more difficult. I still think that it's a lot more difficult, a lot more difficult to get someone from a a six out of 10 marketer to a 10 out of 10 marketer is to get them from a three out of 10, agentic walk up to a 10 out of 10 agentic walker. So I guess I just want to know, I might, I'll tell you, I might take a one out of 10, agentic worker, if they're a nine out of 10 marketer or a 10 out of 10 marketer. Like I, that's, that's kind of, and like that might be, I'm not sure I think that, but I, I think I might, I think I might think that that's the case, that the person could be so good at the core things of copywriting and all that, that I actually don't really care how much they say, as long as they can do the things you need to do now, the game may change such a way that like, you literally can't accomplish what you need to without that. But like, it's, I think, I think that little thought experiment you just gave is, is it really helpful with thinking about where I'm at? And if, for me, that is about what I believe the core problem to solve is for the clients. That's the reason why I think that is like, is like, the, is that, what is the core challenge that we are facing? And how does, and what is the way to solve that? Like, I don't know. Yeah. Yeah. And maybe another thought experiment is if you think of people in your team, because I, I was thinking about this recently, and like, we have a couple of straps on the team, like, super, super strong, really good marketers. And I would say they're maybe six to seven out of 10, agentically. And I thought to myself, how much less valuable would they be to be a team if they were one out of 10 agentic Walker? I don't think, I mean, I don't think it would be significant. I wouldn't look at them as, like, wow, they're contributing a lot less. I got to get rid of them. Yeah. Right. Yeah. Because you would never, I would never even dream of it. If they were 10 out of 10 walks of, there's no way, even for one out of 10 agentically, that I'm getting rid of them. Yeah. You know, and part of that is about the marketplace. Like, this is, because like, even if that person can serve us less clients because they are not as agentic or something, let's say, let's say, they just, they just can output less total stuff. I can guarantee you, you could charge more with clients they do serve to make up for the, for the, like, gap, like that person has so much earning potential because that is the actual course skill. Like, my clients would pay, actually, there's a way I can quantify this. I can tell you, I, I can, I can guarantee you that my clients would pay way more money than they pay us right now. If we could definitely sell them more products and grow their business and they'd pay way more than, than what we're currently charging, if we could do that really consistently. And here's how I know this is a fact because tribe exists, especially if you're willing to price the way tribe does, right? As soon as you invert the economics so that, like, so that the payment is only related to actual success, people will pay so much money for the success. They're so happy to do it. They don't care how agentic a tribe creator is. They do not give a crap how agentic it is. All they want is ads that sell products. And by the way, I definitely thought about our pricing model relative to the, well, relative to the tribe phenomenon, you know, because if a tribe creator is making 50 grand a month, 100 grand a month because they're 10% of their ad spend because their ads are spending a million dollars a month, my goodness, that, that brand will turn around and be like, absolutely, how do I write you a bigger check next time? Just please keep making the ads, you know, because it's such a hard problem to solve. And yeah, and I think they don't care what AI they use or not, as long as they're within legal grounds, you know, I think you made a good point also on like having someone like a Patrick or we have alien our team who is a good luck to an AI pill to kind of shield the rest of the team from having to go deep and get super agentic and build like a custom app, I'm pretty sure you guys have done. Yeah, yeah, we do. Do it a custom app that like is really friends. Yes. So they're kind of like taking the, taking all the bullets and and shielding everyone and presenting it in a way that they can get more done, but in a way they don't have to spend a ton of time getting really agentic. Yes. Yeah. All right. Tell people again what Parker is Alex. Parker is a tool that can help you make better decisions as an advertiser. It connects to all your different resources. It connects to customer views, ad comments, Reddit, we only told it does that consistently to post purchase surveys. And just basically allows you to make better decisions and understand what your customers are saying so that you can make better ads for them. It's hayparko.ai and we don't have a big launch, probably have done that by the time this comes out where we're launching a big marketing brain. So if you want to go check that out, go to hayparko.ai follow me on Twitter. If you won't done a few ad creative as well, adcrate.com. And yeah, I just, uh, my, my job at moment just chief tweeter. Sorry. Come on, come on. Uh, links for all those are in the show notes. As you heard Alex say right there, he guarantees that with Parker, you'll just have an endless ad generator that makes you all the money in your sleep. Press one button. It does it all for you. The one man creative team. Is that, did I hear you that right? Yeah. It's been a tough road to get there, but we've finally got that. Sorry. And if I don't, then you get your money back. Amazing. You heard here first. Um, okay. Uh, all the links for that on the show notes. Alex, uh, come back soon. Thanks for coming. Thanks for coming. Exand dealing with my dealing with a little bit of ranting from me today. I've next time next time you come with something that's like you're just on one about, I don't know. Does that Africans get on one about stuff like that? No, we don't. We don't. We don't. I might get on one about the pod race, uh, and losing to the group. You see that? You see, you see, you see come off the roof? I did. I did see that. I was going to run about that. Okay. Good. Suddenly, the guy goes to one pod race game and suddenly he's a big pod race fan. Oh, do you? That's awesome. That's awesome. That's great. I'm glad. I'm glad. Like I said, you're, you're becoming a true American. Oh, yeah. You're on this call from Iowa. You've gone into a pod race game and bought a baseball shirt. You're talking about baseball playoffs. Like you're, you're in. You're good. Heck yeah. All right. See you. Alex Cooper is great. I love him. He's awesome. And he's going to come back soon. Hopefully and keep coming back. So like I said, subscribe wherever you're watching and listening. Don't forget to do that. Um, big thanks to him. Go check out Parker. Go check out I'd great. Uh, if those are the tools for you, especially Parker, really cool tool. You should subscribe to this show. Like I said, wherever you're watching and listening, if you want to work with AJF growth is we try to build an incredible agency for all the things that we talked about in this episode. Go to AJF growth.com. Tell me a little about your business on the intake form there. And we'll go to conversation started or just email me podcast at AJF growth. Uh, AJF growth.com. It's an easy way to get a hold of me and tell me a little bit. Like I said, about your business, uh, a sense of what's going on in it. And then I will get back to you as soon as I possible again and see if there's a fit there. Thanks to IntelliGems. Thanks to behind scenes studio, both great companies that I work with a bunch. Thanks to them for sponsoring this episode. And I will, uh, I'll see you next time. Don't miss any episodes. Subscribe. I said that already. I said that like three times. You get the idea. See you soon.