Jack (00:00)
Hey, everyone. Welcome to 4Play Fireside. Today we're talking to Maxwell Finn, the founder of Unicorn Innovations on the state of paid ads in 2025. There's a state, we're in it, but what is that state? You're probably experiencing it and Max is here to walk us through what's going on. So, Max, over to you. What's happening?
Max Finn (00:20)
A lot of stuff. There's a, it's a never a dull moment in, our industry. So we'll, we'll, what we're to talk about today is kind of where, ⁓ one, where I see things right now, and then more importantly, what you should be doing and focusing on if you're running ads, right? so we're going to, we're going to go through, have a deck right here that we'll go through, but I also want to like,
I wanna dive into a lot of stuff. So I have probably 30 tabs open and we're gonna go through a bunch of different tools that we use. Some you're probably familiar with, but some you might not be. And the ones that you may be familiar with, just more efficient and effective ways to use those tools, whether it's chat, whether it's quad, whether it's Gemini, which are the popular ones. Unfortunately, a lot of people don't use them effectively and aren't really tapping into AI. So without further ado, let's jump right in here. And by the way, Jack, I don't know how you, know,
Obviously your format on how you like to run this, ⁓ if you want to hold all questions to the end or you want to interject, I'm fine with either. I like to make it more interactive. So feel free, if you have a question or want to chat about something, dive deeper, let me know. Or if somebody in the chat, but your call totally there.
Jack (01:35)
Yeah, sounds good. Yeah, let's just keep it agile. Like, do whatever feels good.
Max Finn (01:37)
Cool.
Sounds good. So the big thing that I try to like anybody that I talk to in this space, I try to make this really, really clear to them is for a really long time. You know, I've been in this space since ⁓ early 2010 and kind of was on the early wave of Facebook and then very early on TikTok. And I've kind of seen these waves happen. And what I've seen through that time is people that are not
great marketers who don't have great products, great businesses, don't have, they're kind of missing everything, make a lot of money because they were able to find these really interesting arbitrage opportunities where they're able to find a specific hack or a tactic or a way to run something on Google or Facebook or whatever channel and make a lot of money. There's always going to be arbitrage opportunities. I'll make that very clear.
But what I do see is those opportunities getting smaller and smaller and smaller and the gains for the amount of input in terms of resource you need put in to get the gains shrinking and not really being worth it as much anymore. And so, you there were days where it's like, hey, if you duplicate this ad set 20 times and you set each one of the $10 a day budget and you run it for this long, you make money. And, you know, I still see people sharing strategies like that, but at end of the day, like, I really don't think that's what you should be spending your time on. On top of that, you have, you know,
these algorithms getting a lot more advanced. What that means is the platforms are doing more, we are doing less of the time consuming, tedious stuff, but we should be putting that time into other things that we'll get into in just a few slides of where the time should be going. What that really means is you need to be focused on how do I send the highest quality, most valuable data back to these tools because they only work with data, right? And so if you're sending crap data back,
If you're not correctly setting up both, you know, pixel and server to server and have data enrichment, all this stuff happening, you're going to get less optimal outputs. So understanding that's really important. And then in terms of just a mindset shift, right? Where should I be putting my time, my effort, my resources? That's going to require a massive reshift for a lot of people. Whether you're an agency, whether you're, you know, small business, whether you're, you know, just a media buyer running stuff, the things you spend a lot of time on.
you don't need to spend as much time on anymore. And so it's okay, if I have all this extra hours a week now, what should I be doing with those extra hours? And that's what we're going to talk a little bit about today. So just to reiterate this kind of point about, you know, the platform tricks fading and algorithm sophistication and all that stuff. I'm sure a lot of you have followed in the last few weeks, Zuck was at an event, a conference and was being interviewed. And basically in that interview, some, there's an exact quote, but essentially what it summed up to is,
We are building kind of an end to end ad solution where the ideal scenario in Zuck's mind is you come to Facebook, you tell them what your goal is, what your budget is, where you want people to go, and they do everything else. That they're going to fully run the entire ad buying process from making the images to doing the targeting, to doing the bidding, to everything in between. we all know that that's not going to happen by the end of 2026.
Right. There's so many variables there. It's just for one, like the reliability of Meta's data and their numbers, right. If those are off and everything's off, the creative is just going to be mediocre. All these things. However, it's important to recognize that that's the path they're going, right? That that's inevitably where this ends up is TikTok, Facebook, Google, Snap, Twitter, Reddit, whatever the platform is, is going to take more and more control away from you. And
give themselves more control, make you depend more on them. So again, that goes back to, well, what should I be focusing on? And begin to prepare for that today because that is the inevitable future is this kind of black box where they control everything. And so then it's like, okay, what do I do? And so what I look at right now and going into 2020 and 2025, 2026, kind of three big core success factors you should be focused on, regardless of where you fall in the spectrum, whether you're the agency owner, whether the business owner,
media buyer, strong back end, compelling offer, and then focus on creative angles. I view those three as the kind of biggest levers and pillars that anybody running any type of paid media should be focused on right now. And especially if you're an agency, media buying agency, or you're a solo media buyer, you need to start developing these skills. Right? This is where I see a big issue happening in the coming months of
a lot of people who, either agencies or just solo media buyers who have four or five, six clients, they've been able to do very well for a long time by just running ads. And as that gets more and more and more automated and business owners get more and more familiar with AI tools, you need to level up your skills and what you're offering those businesses to stay relevant and valuable. And so it means like the days of you just being like, I'm just gonna stay in Facebook Ads Manager. That's my entire scope and world is just pushing buttons.
uploading your creative and doing that. You're gonna be out of a job. If that's what you do, that's the value you bring, you will not have a job probably by the end of the year, I would say. Maybe you find some people that are really small businesses, they can pay two grand a month, whatever, but any medium-sized large business, you're not gonna be valuable enough. And so you need to develop these skills of how do I support them on offer optimization? How do I support them on landing page optimization? How do I bring new ideas to, hey,
Hey guys, we should be testing an advertorial or a pre-sell or a quiz funnel and supporting actually developing that with them collaboratively because these aren't siloed things and they never have been siloed things, but especially nowadays with all the AI advances, there is a deep need for personalization and congruency from ad to backend. And the cool thing is now with a lot of apps out there, there is the ability to kind of begin to do this at scale.
And I think that's where we end up in the next, you know, I would say next 12 to 18 months, there'll be solutions where, whether it's people building their own stuff with replete level, bowl VO, or just somebody comes out with a software stack and somebody can join where you can do that at a very, very individual level. So the actual ad creative, the copy is speaking to a very, very specific individual. The landing page they go to is personalized for that individual, the offer positioning, even potentially the pricing can become dynamic. There's so many things that are going to be able to happen and unlock.
And so you, as that marketer, you as that media buyer, you need to know this stuff. You need to be able to become an expert at this and bring this to the table. If you bring this to the table and all the other stuff we're going to talk about the rest of presentation, you're going to be good. And I think that's the fear I see right now for a lot of people, not just in our world, but in every industry is, am I going to be replaced by AI? Right? Is my job going to be safe? And what I believe is going to happen, at least in the short and midterm,
long-term it's impossible to predict, Is the kind of bottom half, maybe bottom 60, 70 % of the talent pool is going to be at risk of getting replaced. The top percent, the top kind of core tile of talent that learns how to leverage this stuff is going to become exponentially more valuable. Because if I'm a business, I can say, instead of going to this agency and paying them, you know, 10, 15, 20K a month for all these different services,
and getting kind of mediocre talent. I'm gonna go over here and work with this rock star who can do the job of three, four or five people. It's way more talented. They're controlling all of it. So it's gonna be super congruent. Everything's gonna be aligned. And that's why think we're gonna see a big move towards.
Jack (09:43)
Yeah, I couldn't agree more. We've got a couple of questions here. ⁓ One from Jessica, which I think is quite funny. What is a strong back end?
Max Finn (09:44)
Yeah, go for it. Yeah.
That's a great question. So ⁓ basically everything that happens post first purchase, right? So the ability to kind of, if you look at the text under here, know, CLTV stands for customer lifetime value. So, and if you go more specific than that kind of CLGP, which is customer lifetime gross profit, which I think is even a better metric to look at. The reality is most people get hyper-focused on acquiring new customers.
And of course, we need new customers. You can't do anything without them. But that's not actually the most valuable purchase. The most valuable purchase is the second purchase. The statistics, I don't have this on my head, but if you just go to chat or Google and search this, you'll get the stats right away. The increase in value of that customer when they buy a second time is massive. Every metric from the likelihood they buy a third time, their lifetime value, the AOV,
of that second purchase, everything just increases and it's not linear. It becomes like an exponential. And so as a business owner, it's, you know, how do I ascend people? How do I move people from that first purchase to moving to subscription, from moving to my high ticket offer? ⁓ That's really a big, big focus because again, what that allows you to do is pay a lot more to acquire customers and be more competitive. And so
Sorry, got something I shouldn't I was like, crush your granola bar before we started talking. And I was like, that's gonna bite me in the ass granola bar, right?
And it's got locked, it got locked right in there. ⁓ So what I've noticed running a lot of media for a lot of amazing businesses over the last 12, 15 years is the businesses that are most successful, the businesses that we've seen go from seven to eight to nine figures are all willing and able to spend more to acquire a customer than all their competitors. Right? So like, for example, on it.
We've run performance media on it for over five years. We've driven 150,000 customers to them. And performance means we front the ads when they pay us a CPA. When we started running ads for them back in 2018, 2019, they were paying us $45 for a AlphaBrain customer. That's their main product, Mithatropic. By the time we, and we kind of wound up in the relationship, but...
because we moved out of performance media entirely. But by the time we were kind of at end of that relationship about a year ago, they were willing to pay us $200 to acquire that customer. So they went from $45 to over $200 for the exact same customer for the same product price didn't change, nothing changed. And that's just because they continue to refine and dial in their first time to subscription. They continue to refine their numbers or metrics. And they just want to get more more aggressive because they knew, hey, once I got a customer in here, I know what the customer's worth to me.
So if you're a supplement, if you're another no Tropic business, like that's who you're competing with. The company that's willing to pay two, three, four times more than you. Same thing with ClickFunnels, right? What Russell's willing to pay to acquire a customer is significantly higher than what they're making day one on basically all their funnels. Every funnel they have, whether it's a book funnel, whatever it is, it's all built around selling ClickFunnels on the backend without selling ClickFunnels. And so again, if you're selling competing SaaS, he's going to outspend you and be willing to pay more to acquire the customer.
So that's why it's really important that you focus on this backend, not just the front end, because that gives you a really big competitive advantage when it comes to acquiring those customers.
Jack (13:33)
We've got a cool comment from Josh here, that's because ads are an auction. I think that's a question. They're saying is that because ads ⁓ are an auction.
Max Finn (13:43)
Well, that definitely impacts your competitiveness, right? If you're in an auction environment, supply and demand. And if I'm willing to pay more than you, I win that auction. And especially if I'm willing to leverage any of the manual bidding, I'm signaling to Facebook, I'm signaling to Google, to TikTok, how confident I am in my business, that I'm willing to pay more to acquire a customer than this person. And so you also get customers put in different buckets, different tiers. I can get higher quality customers. That feeds back better data into the platform, which helps me
refine that learning, get more better customers. So it creates this really cool flywheel. yeah, that's short answer is yes, the auction environment plays a big role in why that's so.
Jack (14:24)
Nice, we got another question here before we move on to marketing psychology. I just want to make sure that this one, if it relates to what we're talking about, gets answered. How do we adapt on Meta's latest algorithm update with regards to Andromeda Advantage Plus about creative volume diversity? And does that mean minimal tweaks on ads, whether on hooks, ad copies is not going to work anymore? I don't know what Andromeda Advantage Plus is, so that might be really new.
Max Finn (14:29)
Yeah, of course.
Jack (14:53)
and small tweaks aren't going to work anymore. Well, I mean...
Max Finn (15:01)
Yeah, I mean,
they're moving more towards like, you know, add add personalization. So this is just a like an advancement in machine learning kind of end of end of December with them. But it goes back to just I mean, advantage plus just the way they're moving in terms of leveraging AI.
Small incremental changes I still think are important, right? So the way I look at creative testing is twofold. One is incremental iterations to find, to either extend the longevity of a winner and or find kind of nuanced pockets of like messaging that outperform that control. And then on top of that, taking big swings. So it's for us, it's a blend of both things. It's we to find, we want to take big swings on regular basis to find that next big home run.
But I think a lot of people, what they do is they're constantly maybe taking that big swing. And when they find a winner, they just run it for as long as they can and move on to the next thing. It's like, you found a winner. There are so many ways to skin that cat. There are so many ways to turn that winner into 10, 50, 100 different winners, right? It's I can swap the hook out. I can get a totally different avatar to read the exact same script. I can do a different intro motion of something crazy happening. I can have a different text overlay.
almost an unlimited amount of tweaks you can make to that. you know, are most of those going to give big gains? Probably not because they're incremental. But if I have a winner, I don't need to even have that really big of gains on it. As long as I'm getting kind of equal performance, it just gives me more fuel to keep running stuff for longer.
Jack (16:41)
That's such
a great point. When I read that, was thinking minimal tweaks not going to work anymore. thought, well, the ad's not working in the first place. A minimal tweak is probably not going to fix it. You probably need to change things up. But a minimal tweak on a winner and iterating on that and get pouring more gas on the fire, giving yourself more swings at bat, seems like there shouldn't really be a reason why ⁓ Advantage Plus and the push towards more automation would interrupt with that. ⁓
Max Finn (17:10)
it wouldn't,
if anything, those tools are actually doing that. Or if you look at a lot of the, the kind of advantage plus the, you know, suggestions they'll make on creative, because they do a lot of different things. So I had sound to it. They'll make little tiny tweaks. A lot of those are, are relatively incremental. Like they're not doing anything when it comes to Meta's creative AI. That's like a massive swing on a piece of creative right now. It's just the tech isn't there yet. So yeah, the short answer is I still think everybody, if you have a winner, you should be
doing iterative testing on that different variations. But you should also simultaneously be taking those profits and reinvesting them into taking some big swing.
Cool. ⁓ So let's dive in here on some marketing psychology. This is the thing that like, I pretty much spend all my time for the most part ⁓ on AI and psychology. most of my day is reading or testing something in this world, right? Consumer psychology, neuroscience, neuro marketing, behavioral economics. Now the reason I focus so much on that
It's because it's evergreen. The challenge I think sometimes with being in our world, we're always testing the latest, greatest thing. it's Facebook. No, it's TikTok. No, it's this. No, it's this. And we get so focused on the medium. We get so focused on the vehicle that we forget that these are just means. Like whether it's TV, radio, print, Facebook, whatever the next big thing is, chat, CPT, search ads, whatever it might be, end of the day, they're just a means to get a message in front of a person.
Right. And so what I really care about is like, what's the message, how to create the best message possible. Because if I can do that, it doesn't matter. It doesn't matter what the next big ad platform is. I'm going to do very, very well with it. And unfortunately, a lot of people do it the reverse. They get hyper-focused on learning Facebook ad strategies and Facebook ad tactics and stuff. And they neglect the fundamentals. ⁓ And because of that, ads suck. Right. If like, if you just have generic media copy, which is just
know, product XYZ is loved by 3000 people and it has three, four bullets on features. that's like, there's no differentiator. Like nobody's going to stop the scroll and want that. Nobody's going to get excited about that. And so, you know, I spent a lot of my time on this and it's so impactful. Right. Here's just like some kind of fun, you know, stories and kind of case studies. I get really into these types of experiments, real old examples. I think it's good to illustrate that. there's a lot of things called cognitive biases.
And one of those biases is the decoy effect. And there's a famous story from The Economist. When The Economist came out with their web version of the magazine, they priced it like the left. So they came out and said, hey, you can get an economist.com subscription, 59 bucks for one year. Or you can get print and web for 125. And they assumed, hey, a lot of people are going to take print and web. They get both, a lot of value. It's great, right?
What they saw is 62 % of people took the web only and 30 % took print and web. So AOV was a lot lower than they wanted to be. What they did without changing anything else to the core product, the marketing message, anything, is they just introduced a third option. And that third option was print only. And they priced print only the exact same as print and web only. Now the reason they did this is they intentionally made this such a bad offer that now when I look at this, I'm comparing these two. I'm no longer even looking at this. Cause now I'm saying, oh, for 125 I get it.
basically web for free. So I've shifted where my focus is as a consumer making the decision. And you see this at, know, movie theaters used to just more, they don't do as much now, but movie theaters are a famous example with popcorn sizes. Right. The reason there's a medium sized popcorn is to push you to buy the large popcorn. Nobody, nobody, I see maybe like, I go to a lot of movies. I've seen a handful of people in my life eat a large popcorn, back for a refill. Nobody needs that giant like bucket of popcorn. They buy it because it's like,
You know, for 50 cents more, you can get a large popcorn. It's like, I gotta get it, right? You're still spending more money though. And so this is something when you think about just my pricing, my pricing page and how I'm positioning my offer, my product, understanding this is really, really powerful to increase your average order value. Or you have things like the Pratt fall effect. And so a famous example here is you look at, know, Avis rental cars, you know, back in the day when they were, you know, competing heavily with Hertz, they,
didn't really know how to compete there. They were really, really struggling. And so they hired DDB and they leaned into the fact that number two, right? Most brands would try to bury all the negative stuff. They're not going to focus on any of the bad things and the negative things. They're going to just focus on the reasons are great and why you should love us and why you should trust us. What many, many studies and experience have shown is that if you can actually kind of admit flaws, it humanizes you because everyone's flawed and everyone makes mistakes and it actually can.
increase ad performance, can increase conversion rates, increase sales overall. And that's what happened with Avis, probably the most famous example of the pratfall effect. So within a year, they actually turned their first profit in 13 years. Their market share went from 29 % to 36%. They went from losing millions a year to making millions a year. The team morale skyrocketed. And so there's so much that could be done here for your business if you understand these things, right? Decoy effect, pratfall effect.
So I recommend for anybody watching this is, Google, go to ChatGPT or Gemini or whatever you use and search for cognitive biases. Give me a list of 30 cognitive biases I should use my marketing. Just start understanding these. There's a lot more, right? There's loss aversion where people, we feel the pain of a loss more than the benefit of a gain. So typically if I'm using kind of gain framing where what you're gonna get, it needs to be at least two to two and a half times greater than what my expected loss could be from that.
So there's all these things that I need to understand as a marketer. And ⁓ if I don't, I'm going to really struggle. So again, going back to if I have more more free time as a marketer to create my ads, to create my sales pages, to create my scripts, my creative, these are things you should be spending your time into mastering and learning and working them in because you now have more free time for other stuff that you're not working on. Even just basic stuff, Like understanding how customers think. This is a pretty crazy story. So. ⁓
think it was probably the 80s or 90s, the exact timeframe, but A &W was wanting to compete with McDonald's on their quarter pounder. So this brilliant idea, said, hey, we're gonna come out with a third pounder and then we're gonna price the third pounder the exact same price as the quarter pounder. It's a great deal. It's more meat for the same price. It's great deal. It totally bombed. And they eventually brought a focus group in and the focus group asked a bunch of customers, well.
Why would you buy the McDonald's quarter pounder over the A &W third pounder? Every single person in the focus group said, well, why would I pay more for less? Or I get less meat. Because people don't do math well. So everyone looked at just the bottom number, which is the three, and four is bigger than three, so everyone just assumed McDonald's is giving me a bigger burger. And I think this is a mistake a lot of people make with their ads. They overcomplicate it. You over explain things.
You require people to put a lot of mental effort into understanding the value of your product. And the truth is that the average person does not have very good math skills. The average person does not have, you know, grad school level intelligence. And something as simple as this proves it, It's like, you know, this, this is, be pretty obvious, pretty basic math, right? We all learn fractions in, you know, fifth grade, but the average person quickly looking at an ad just does, it's, it's called heuristics, right? Mental shortcuts we do quickly. And so three is less than four.
This is less than that. I'm not going to go for it. So it flopped. So we to even just reevaluating around looking through ad copy, looking through your creative. Is this something that a fifth grader could understand? Is this something that somebody could just see it instantly get it? Or does it require them to do a lot of thinking, a lot of work? And, and we want to avoid that, right? Just one last thing here. ⁓ there's a great book I recommend everyone read called thinking fast and slow by Daniel Kahneman, Nobel prize winner, sadly passed away. ⁓ but
one of the greatest minds when it comes to this stuff and just brilliant. The book is all about how we have kind of two parts to our brain. And we have the part that just reacts quickly, makes emotional, ⁓ rash kind of decisions and the part that makes logical, more thoughtful decisions. And we want to try to get to the one that just makes instant decisions because we don't want people thinking for a long time on something. The longer they think, the more likely they're not going to buy. So, ⁓ yeah, this is really, really important stuff to understand.
Cool, let's switch gears a little bit and dive into just kind of the back half of this, which is.
psychology, understanding all that critical for being successful in 2025 and beyond. The other side of it is obviously AI tools. I think anybody that's called knows this is what we should be focused on. This is not a trend. It's not a fad. This is transforming every aspect of our day across our personal lives, professional lives. It's kind of four big buckets that we look at using AI in right now. Market audience research, creative development, ad account analysis, CRO optimization, and offer optimization.
So those are kind of the four ⁓ big buckets were in there. So we'll go through kind of each one of those. We'll talk about some examples, some tools we're using. The first one is, this is actually a great resource. If anybody's looking for something, just, I wanna try to share some free tools as well. ⁓ Prompt Cowboy, this is a great website you can go to. And basically it turns your lazy prompt, your generic prompt into a really, really good, well-structured prompt. And this is really important because what I see
The two biggest mistakes when I look at people that are using AI, especially early on and like, ⁓ it's okay. I don't get really good ad copy from it. I don't really get good ad ideas from it. I look at their chat, GBT chat log or their claw chat log and like, well, you gave it one sentence. You said, write me 10 pieces of Facebook ad copy from my business that sells collectible Batman things. It's like, okay, of course it's going to suck. There's no knowledge. There's no context.
There's no information. There's no specific instructions. You're not telling it, you know, what's, who's my customer even? How should you write the copy? You know, what should I model it off of? There's so many things you're not giving it. And so having a really strong prompt is really important. Now, if you can't write those prompts well yet and you don't understand these platforms, a tool like this is going to be really helpful. Cause I can come in here, I can just paste what I wrote and it's going to give me this prompt underneath it. That again, I can just now copy and paste this.
I can also continue to improve it. So on this right side over here, you'll see, you know, it kind of gives you some additional nudges. Hey, give me some examples, give me some ideal outcomes. And the more you give it those things, it'll keep improving and improving the prompt. And now I have a much stronger prompt to go into the, go into the tool I want to use and get much more in depth and comprehensive results.
So the first thing we really want to start with, and a lot of people skip this step when it comes to using AI, because they just jump right into it. Create me images, create me copy, is the customer research component. And so there's a lot of great tools out there for customer research. My favorite right now is definitely Gemini on the Google side of things. So if you look at like, let's go in. I actually have it running a report right now for a
an audit that I'm doing later. there's a, let's go in here actually.
So a prompt that we use for deep research for any new business we're running ads for. And basically what we want to get is just a really, really, really deep dive into who that customer is. So we essentially have, what's this up to now? 65. So there's 65 questions that we're giving the tool. And so we're saying, hey, know, context, I run this company, here's the website, this is kind of a brief bit about it.
Here's what your mission is. It's to go through, conduct comprehensive research, provide detailed answers about my target avatar so I can create the most profitable ads possible. Please answer all the following questions, specific actual insights. So it's going through everything, demographic, pain points, goals, motivations, decision-making, communication preferences, solutions, value perception, psychological triggers, emotional drivers, buying psychology. And then what we end up with here is, know, in Gemini's side, they come out with...
you know, out of 33, 31 page document that in depth answers all this stuff from, you know, gender, income level, education, location, family status. It has all kinds of, you know, it's referencing a massive amount of sources. What we'll also do is we'll run it through multiple tools. So not just Gemini, we'll also run it through, you know, Claude, can also run it through ChatGPT, because they all give kind of different types. They're all good at different things and have different trends. So they'll give different, you know, this is the exact same prompt.
But you'll see it's totally different the way it gave me back and the information it gave me back in the formatting. So we'll create this into a master doc. And now what we can do is we can take this document and we can load it into Claude or chat as a project file, which is really, really important. When you're working on, let's go over to this over here. So when you're working in a...
environment like this, you want to have, let's see, it's in this one.
So you want to have as much information here as possible. So for example, this is one account we work on. And we just loaded something today with a bunch of different research. have research in here on this is like a financial products company. So IOLs, annuities, all that stuff. So we have all kinds of e-books that have been written. We have webinar scripts. We have ⁓ deep research into the individual products. We have the pitch deck on the offer. So we're giving it all of this.
data information. So now when we go to actually create copy or come up with ad ideas, it's not just basing it off of one or two sentences, which most people do. It's basing it off of a robust prompt, and then it's basing it off of all of these documents. And then it's basing it off of this extensive research. This is actually the one we put together for this business this week, actually. This is kind of a full blown example of, you know, this has got
GPT-45, this has got Claw, this has got Gemini, it's got Grok. And again, these are all great because Grok is gonna pull things from Twitter. So you're gonna get totally different levels of data and access and information. And so now I have this this just kind of monster document here that just as a campaign brief, it has kind of high level angles that we can use. It has different creative angles that it's come up with on the image side. So voice development.
So there's just a massive amount of stuff that can be developed once you the foundation in place. And so again, think when you're thinking about using AI, it's more, I know it's tough to just like, you want to just jump in and get stuff. It's all that speed, but it's so valuable if you just take a step back, come up with a plan, map it out and actually just take your time and do it methodically. Don't just jump into the copy and create a part, do the research.
build out the knowledge base, create the right prompting structures, and then go in there. And now you can go in and say, once it's in there, now it's great. Now I can go in there and I can start kind of the next phase. And I can start doing, know, copywriting and creative development, all that stuff. Another thing you can stack onto this document, this is a tool called Appify. Great, great resource. What it basically has is tons and tons and tons of these things called Actors, which are essentially just advanced scrapers.
And so you can come in here, they have scrapers for everything, Amazon reviews, Facebook comments, Facebook reviews, anything you could think of, Apollo scrapers or CRM and B2B and stuff. What we like to use this for is ⁓ going in and like finding who are your top five, six competitors, scrape all of their Amazon reviews, scrape all their Facebook comments, and then run those through in chat or cloud or whatever tool you want to use. Have a look for
common complaints, common issues, common problems, and then identify those that you actually solve. Because what you might find is, hey, you have these big competitors, but you have a feature in your product that they're missing. And people actually really, really want that feature. So now I know that because I've gone through the reviews at scale. And now I can create ad angles around that, really calling that out. So that's one way we use it. The other way we use it is ⁓ for going through comments.
to, and it's really easy. Again, you can just paste in any Facebook ad URL right in here. You could run this directly from this interface. You just click start. It'll run it, export it into JSON or table format. Or if you're more advanced and you know how to use N8n or make or Zapier or these kind of agent platforms, you can integrate it through the API and do it automated. So we have a lot of this stuff running, a lot of scrapers running for N8n automation. I didn't want to get too advanced in this call, but. ⁓
But we basically a website can come in from a lead. We can scrape that website. We can do kind of AI deep research on it. We can build all this stuff out to do lead enrichment. Like we do some really cool stuff for lead generation, but for this purpose, you come in here, put a bunch of ads in, paste it in there, get all the comments. And then what I like to do is run those comments through a prompt to understand how my audience speaks. Because what I find is a lot of, ⁓
A lot of people that are running ads for businesses aren't the actual customer of that product, right? You might be a 25 year old guy running ads for a magnetic lash brand. So you don't speak the same way that a 22 year old girl is really into magnetic lashes speaks. The words they use, the phrases they use, the terminology, the cadence, there's all these little nuances that you're not going to pick up on. But the way that those people speak in the comments and the reviews is how they speak. So even just doing that, grabbing the comments and then saying, Hey,
look for common words, phrases, ways they speak. And so they basically create a framework and guide for me to use for writing ad copy. And now I can run future ad copy through this kind of filter to just improve it and make it more personal and sound like they wrote it, sound like it's one of their friends who wrote it. And it just dramatically increases click through rates, relevancy scores, quality scores across the board.
All right. AI for copywriting. So I'm a huge fan of Claude when it comes to writing copy. just, in my experience has done the best job of writing high, high quality ad copy, sales page copy, VSL scripts, whatever you might use it for. Again, the key part of this is making sure you're starting with that foundation. So it knows everything about that business. But again, going in there and not just saying, write me
Facebook ad for business X about why, because you're going to get crap results, but going in and giving it an actual robust prompt. Right? So here's like a prompt we, um, know, daily wire was a client of ours. So we ran all their stuff with the razor and e-commerce brand. So going in and saying, Hey, you're, know, here's the situation, your world-class copywriter, you specialize in creating persuasive high converting direct response ad copy specifically for Metta. You draw inspiration from these legendary copywriters. You've.
you know, understand you've read and studied these books. These are all incredible books when it comes to copywriting advertising. Then I'm saying, here's your task. I want you to create this for these people. Here's the objective. Here's the knowledge, right? This is the target audience profile. Here's the brand voice and positioning. Here's some, you know, product details. Here's the brand story. Here's the website. So it's so much more information to work off of. And then in addition, you also have all of the research and all the information.
in the project file that I can pull from. And now it's going to give me, you know, six headlines, six pieces of ad copy, varying lengths. You know, this is a short form origin story hook. This is a long form emotional narrative, you know, and this stuff turns out pretty good, right? It's every morning you look in the mirror and see a man who's standing for something. You work hard, you speak the truth, you don't apologize for being masculine. So why are you handing to companies that think you're the problem, right? It understands the brand, the positioning, the tone of voice.
⁓ So yeah, a lot of variety here, a lot of good stuff. And then you create a feedback loop where you're now letting it know when you create future copy, hey, this ad copy, this ad copy worked really well. So it's able to begin to learn and evolve on that. Claude's a little different because Claude doesn't have memory across chats. So it's really just self-contained. But ChatGBT has memory. So especially if you're using chat, it's going to learn and continue to get better.
over time with everything you're doing across the board. So as long as you're sending that data back and letting them know that you create that feedback loop in each iterative version of that copy or a new form of copy will be that much better.
So let's dive into some fun stuff here, which I know everyone really cares about based on, you know, it being a four play call is the creative part of it. So this is the part that's so cool. Like I think when you combine what the four play team's doing of just having such extensive access to research and studying creative and learning from creative with the ability to now create concepts and ads really quickly for very, very low amounts of money.
It's really cool. This is the part I'm most excited about because it's just totally shrunk and democratized the ability for anybody now to produce like Hollywood level creative from their laptop for a few bucks. So these are two examples for anybody that's not seen a VO3 video. They've been going viral all over the place, but this is Google's new video generation model that can do dialogue and visuals. So here's two videos that we made and we actually run these and got good results with them.
This is like a two and a half X row ads from this video, this one we just started running. But I'll just play both real quick and we can talk about them. I should probably also say, so this is a, it's a conservative collectible business. So this is, we're running stuff for like 4th of July and we have an Independence Day product line. So we're doing like a 1776 theme for this ad of somebody, basically of a vlog style, continental army soldier fighting in the revolutionary war.
So that's one. And again, the crazy thing about this is, so you need a Google AI account, which is expensive. It's like $150 or so a month. Or you could run it through. There's tools like Leonardo AI. I think they charge about $3 or so a video if you do it through the API. So it's not cheap in the software world. But compare it like that basically cost me a few bucks, maybe 15 minutes of time, 20 minutes time going back and forth, making a few variations.
but was able to make that in like a fraction of the time it would have cost to do anything else, have an influencer go out there and shoot it, right? And then here's one other example. It's like a totally different, more modern day approach.
And that one eventually cuts into our actual birthday card we have, like a fake news birthday card. I just had the variation that didn't have it. So again, are these perfect? No. There's still, especially for me, because I'm making them, there's lots of little things that you can pick up on. But if you're just watching this, you're on a Facebook feed, scrolling through and you just see this, 90 % of people, 95 % of people are going to be like, the real video. Now, as a human, this shit terrifies me.
I'm in a weird place right now because what I'm experiencing testing all this stuff and I am an advisor and part owner in a few different companies in the space. So I see what's coming six, 12 months down the road. yeah, mean, it's scary because you can basically make a breaking news clip right now of somebody saying anything and a of people are gonna think that's real. So terrifying as a human, as a business and a marketer, ⁓ this unlocks unlimited opportunity.
Right, any idea you have for a video, you can now turn into a reality with VO3, which is just absolutely crazy. So one of the things we've been doing is using, so there's like this prompt, and I'm happy to share these docs and stuff after if anybody wants these. This prompt essentially for VO3 video generation, where it can kind of analyze the business.
And then it understands all these things, the product, the demographic, the context of where the video should be shot. And then also has kind of an output format. So you've kind of found this to be kind of a sweet spot of the format, the scene, the who, the what, and gives you a prompt. And so this makes it easy to generate and quick to generate good prompts. Because again, with VO3, especially because it's expensive, you don't want to waste prompts because it can burn up money pretty quickly. So you want to the effort in to come up with good prompts right out of the gate.
And then on top of that, using a tool ⁓ like Cloud or like Chat to just come up with really, really good ideas. So let's see. This was a.
I'll come in here and basically say, know, you're a world-class comedy script writers, written some of the hilarious, most popular comedy TV shows and scripts of all time using VO3 to do this. This is the video I'm making. Give me 10 ideas. And so I'll come up with the kind of 10 concepts for videos, know, kind of Foundry Fathers, Vegas Weekend, time traveling, know, gift shop pitch, know, colonial pawn stars, different things. And then from there I can say, okay, you know, turn that script, turn that idea into a script, detailed prompts, character description.
And now it'll give me kind of the scene by scene broken up eight seconds because that's the max time frame that I can use for each one of these scenes and now I can create a cohesive video in in vio3 I have to make the clips individually still because the feature to extend them isn't available yet for the vio3 model But I can just put them in CapCut and easily just clip them all together pretty easily pro tip by the way if You are making multiple clips with the same people in those clips
make this and include this in each prompt. So basically a character description. So you notice the kid in the army video I showed you, like again, he's a little different than everyone, but he, for the most part, very, very similar if you're just watching it quickly. And that's because I included a character description in every video and it was the same character description. So the outfit, the fact that he was missing teeth, all those little things were the same. So you definitely want to make sure you include that in your prompting.
And then the same thing with, you want to, cause with VO3 you can have products in the media, but then there's tools out there like, like Create-ify that we, we use. actually a full disclosure and advisor and a ⁓ minority owner of, ⁓ where you can actually generate video clips of avatars with the product. And you can do, you can use easy there in their suite of tools to narrate and basically put a dialogue behind it, or you can do voiceovers. There's a of things you can do with it, but just, you know, from their tool.
It can turn any, it can generate the image and then it can turn that image into a video. So I can now have different clips of a product being, you know, picked up unbox. I can have a doctor kind of holding the product and reacting to the product. And again, this is a hundred percent AI generated. There's no real actors in here. So this was just uploaded the product. And again, is it perfect? No, not yet. Still little things. You saw the, the text kind of disappear a little bit when he tilted it. ⁓ But you can obviously trim those things. And, and yeah, so if.
If I'm in, know, Createify, I can basically create a new product shot, product avatar. So whatever avatar I want to test for my product, I can create them here. I can create my own avatar. can create a, you know, if I have a fitness offer for grandparents, I can have a 65 year old jacked, you know, fitness dude who I can make and I can have him holding whatever I want to hold and create these different video clips of that, that I can then use for.
future video production. So there's just, there's so many incredible resources out there for developing video creative. Again, either as entire videos, clips, or these additional kind of just product avatar clips that I can then narrate or put dialogue in. And then the same thing with AI static, right? You know, with, with chat four, and four, one, it does an amazing job of image generation. Again, if you give it the right prompting,
and the right source material. But both of these, again, were created in probably a minute of total time to get to that point and are totally AI generated. There's no human intervention here. So even this, like let's go over here, you don't even need that sophisticated of prompting to get started here. So what you can do is feed it a bunch of templates. There are so many resources out there with just lots and lots and lots of
ad examples. I don't know why it's not popping up here, but this is basically just like graphic of like a bunch of it's like a board of a bunch of ads. So I literally just paste that in here and said, Hey, this screenshot is a ton of winning static meta ads. I want you to templates going forward. Can you identify them and then use them as a templates for different business? So then
Jack (49:07)
I love what you're
doing here with using a board as the input into chat because you need that research, right? Garbage in, garbage out. You need to give chat something that it can use ⁓ that's actually gonna be adaptive, right? You're gonna get an output that's worthwhile, right?
Max Finn (49:15)
Exactly.
Yeah,
that's where like four, what you guys have built is so great because I can build that board and then I can export either those individually or just I can even literally take a, like this is a full screen screenshot and it's good enough that I was actually able to identify every single ad in here just from this kind of mass screenshot. So again, definitely whoever's watching this, like keep adding more stuff to your boards, keep building that out because it's so valuable for creating ad creative at scale.
but it can come in here and identify all the ads, like really specifically.
It then creates kind of templates from those different before and afters, UGC testimonials, identifies all those. And then it kind of says, hey, add adaptation to different businesses, give me a niche or business and I'll write out three to five fully adapted ad templates. So then I basically gave it this, said, hey, okay, conduct extensive research on this business, new update loss. And this is a rushed one. Again, if I was doing this for, what we've been doing essentially for our offer is for any leads coming in, we're like quickly making.
statics and videos for them to just show them how quickly we make stuff because we just we've really Templatized it so we make us a view of three videos some chat GPT images So we obviously don't put as much effort into each one of those if this was a client We would do the deep research. We'd have a much more extensive prompt here But the cool thing is like you can still get great outputs even with relatively little input So hey conduct the research understand the business extract the content customer reviews product descriptions all that
then use the research to determine the best templates to use and give me four high converting metastatic ads for the business. And I gave it the logo and I gave it the product image. You want to give it that so if it needs to use a logo, it's not going to pull something random. And again, it's not going to be perfect each time, but it does a pretty good job in a lot of these of mashing the logo pretty well. And so went through, did the research, key takeaways, pricing, benefits, core offering, reviews, ⁓ best ad templates to use based on that before and after. So again,
Good, it's an Ozempic type product. It recognized that. Number one recommendation before and after transformation. Then it gives me the different ads that I can approve or change if I want. If I like what I see, I can say create the ads. It'll go through. It'll fully build them out. And then it'll create them in 1080 by 1080. So then it made the transformation ad, a this ad.
This one, not that great. It kept messing up the list there. So this was kind of a wash. ⁓ But then it made, you know, ⁓ one for men. I could just go and say, create, you know, create a male version of this. So it does a really good job of this. And if I ever want to change anything, another kind of pro tip here, sometimes it'd be like, ⁓ the ad, looks perfect, but it misspelled this one word. And it's like, crap, I can't use it. All you have to do is just copy it.
Go over to Canva.
could drop the image into Canva.
Come here to edit, go to the magic tools, go to grab text.
select the text that's misspelled. So let's just say misspelled transformation. So I'm just going to click that, click grab. It's now going to make that text edible.
So now I can just fix that word there and export it and it's good to go. So you can easily fix any of little things there.
Jack (53:06)
Nice.
Max Finn (53:12)
All right, think we got one more slide. We're running out time. And then obviously, questions. I know we went a little long. And kind of the last piece of this really is the analysis part of it, right? Using these tools to analyze lots and lots of data. And again, for this tool, for this purpose, Claude, we use, because again, Claude is just really, really good at ⁓ coding tasks. It's very, very good at ⁓ data analysis.
There are some AI tools out there that specifically just do data analysis and some are pretty good, but I'm not like, you can kind of get in this endless loop of just new tool, new tool, new tool. So what I find to myself is I really get hyper-focused on just the big ones, because they're just moving so much faster. And then if there's something really, really cool out there, like, okay, I'll explore that, where it doesn't mean very, very unique, that's not done well by the other tools. ⁓ But with Claude,
you can come in and give it a, there's two ways to do this. there's something called actually MCP servers, which you can use now to directly integrate and connect Meta, Google, TikTok with Claude, with these tools. It requires a little more know how to do that. So the example I'm sharing right now is something anybody can do. The other option is to just bulk export CSVs. So go into your Facebook ads manager and export your different reports, your gender breakdown.
your age breakdowns, your just overall performance. And then running a prompt like this, where you're saying, hey, you're a PhD data scientist. You specialize in analyzing large volumes of advertising data. Here's the CSCs I've given you. I want you to go through them and examine performance metrics, identify obvious, not obvious correlations. I want you to give me a recommendation plan, all this stuff. It'll go through it, analyze all the data. It'll come up with different findings, age performance gaps, device opportunities.
different video completion correlations, ⁓ hidden patterns you might not notice, some immediate recommendations, reallocate budget, optimize videos for this length, increase Android device targeting. ⁓ So you can get some really great insights in this stuff. And you can obviously, you would want to have a conversation and go back and forth and say, hey, we'll dive deeper into this and look more for this and stuff. But you get this really nice kind of initial report on just breakdowns and recommendations and kind of.
sweet spot for video completion rates. Even little things like this, these are things that you just as a human wouldn't be able to really get to is that fact that hey, videos that achieve 70 % completion by 15 seconds show 30 % higher purchase rates. Morning placements show 18 % higher ROAS for 45 plus demographic. And now some of this stuff, it's like, well, maybe there's nothing I can do about it. It's like, it's interesting to know, but I can't do anything. You might find one or two things where it's like, oh crap, I didn't even think about that. And now I can refine my messaging, refine my creative.
refine my bidding, my budgeting, my pacing to just get better performance here.
⁓ Yeah, this will again, it'll break it down into kind of short-term strategies, medium-term strategies, long-term optimization strategies, giving implementation timeline, give you kind of success metrics, risk mitigation strategies, secondary KPIs to look at. So really good stuff in here for anybody looking at kind of analyzing the ad performance data.
Questions?
Jack (56:43)
We've got a less topic-based question here. Will there be a recording shared later? And will you share the presentation document?
Max Finn (56:53)
I imagine there's a recording, you guys are recording it and I'll definitely share ⁓ the deck with you guys so I can get the link over right after this.
Jack (56:55)
Yeah.
Justin saying, can we get those docs, the docs they shared earlier?
Max Finn (57:04)
Yeah, I'll include the ⁓ deep research prompt and I'll include the VO3 prompt as well. So I'll put a doc together for everyone.
Jack (57:16)
Jacob's interested. He said, do you work with Jeremy's razors?
Max Finn (57:21)
We did. They brought everything in-house and basically paused their ad spend at the beginning of the year. Unfortunately, with Trump winning, it actually had a really negative impact on the business because their whole positioning as a brand has been the woke. That messaging works really well when you're not in power. But when you're in power, all of a sudden, it's like that messaging just totally died.
They brought everything in house, of reinvent the brand. But they're amazing that the team over there, like such a cool company to work with. They do a lot of amazing stuff. ⁓ But yeah, was, that was a fun one. Cause I had, there was like a week period where I had gotten access to Ben to Shapiro's page, like Walsh's page, all their pages I could post and run stuff. And then we also did an audit for PBD, so, but that David. And so I like was in Facebook and I was like, I have access to PBD's page and Ben Shapiro's page. And it was.
It was a pretty ⁓ crazy, crazy day. Yeah, that was fun.
Jack (58:21)
It's fun.
Awesome. Do we have any more questions, guys? I've added a link for Createify if anyone wants to check that out. That's in the chat. Chloe says, nice. Thank you, Po. Yep. Short and sweet.
Max Finn (58:36)
Sure and sweet.
Jack (58:40)
We've got one more question.
Max Finn (58:41)
Cool, I'm
sure there's people that watch the replay recording. Feel free to pass those questions along to me.
Jack (58:46)
All right, all right, and where can people reach you if they want to ask you a question after this?
Max Finn (58:51)
⁓ So our websites, unicorninnovations.com is the website. then I'm on all major, if you just Google Maxwell Finn, it's any of the social channels that are there, ⁓ facebook.com slash Maxwell Finn, instagram.com slash Max Finn. any social channel is good. I post like every day I post kind of deep dive marketing case studies and experiments and lots of stuff.
Jack (58:53)
the best channel.
Max Finn (59:19)
If you're interested in that stuff, give me a follow.
Jack (59:22)
Nice, Awesome. Well, guys, ⁓ we are one minute over time. If you have another question, feel free to drop it in the chat. ⁓ But then we're going to close things out. Max, while we wait for that last question to come through, Mary says, thank you. Yeah, I wanted to say thank you for coming. This was really interesting. I learned a lot.
Max Finn (59:43)
No, of course it was a blast. I had fun, I'll do it again.
Jack (59:47)
Awesome. All right. Thanks so much, everyone. Talk soon. Bye.
Max Finn (59:49)
Cool. Thanks guys. Bye.
