Zachary (00:00)
All right, guys, so we're live on this week's floor play fireside. Today's going to be like a pretty casual one. No, you know, super polished presentations. We have like a lot of really good minds here joining us on kind of just like a really casual fireside chat on just like AI and creative strategy and what people are doing right now to accelerate the work that they're doing in creative strategy with AI. And so one of the speeds kind of is like an open communication.
Shahbaz Khokhar (00:08)
you
Zachary (00:27)
everybody that's live in the chat, feel free to drop questions. We'll field them as we go through. There won't be like a standard sort of like Q and A section. We'll kind of just use them to prompt the conversation and kind of keep going. ⁓ And so, yeah, we'd love to just kind of like open the floor. I think everybody has maybe like that one thing that they found last week or two weeks ago that's like, they kind of keep going back to a little unlock. ⁓ I'll start with Barry. ⁓
Is there anything new or even like stuff that's like tried and true super old that you're consistently, you know, ⁓ going back to as like a tool in the tool chest for, you know, shipping more winning on.
Barry Hott (@binghott) (01:07)
Yeah, I mean, I'm always just thinking about the first second, right? Like, it's not just about that, right? The rest of the head has to be there, but I'm always just thinking about like, what's that first initial emotion feeling visual? Where is it? What does it say? How does it make you feel? That's, that's the thing. And then, you know, how can I enhance that via text overlays, um, thoughtfully, and then also like, like,
I've been also thinking about like AI hooks, not the AI hooks for the sake of like, making something that looks and feels like AI, but like, what is something I could not make physically myself or wouldn't that I could make with AI. So say that for now.
Zachary (01:50)
Like kind of the idea of just like, do I use AI to like take the craziest idea that I have and make it live or like, okay.
Barry Hott (@binghott) (01:58)
No, it's
rarely the craziest idea. It's rarely out there. It's usually like it's just something I don't want to have to work with anyone else to make. Right. Like I don't want to have to talk to another human to produce this because they're going to be like, ⁓ what if we do it this way? And they'll like try and polish it and zhizh it up fancy. And I'll just like prompt the to make the video I want. I'll either take a photo that I already have and turn into a video or I'll
Jack (02:01)
Ha ha ha.
Barry Hott (@binghott) (02:27)
get use AI to make a similar image or the image of what I want and how I want it and then use VO or whatever to animate it from there. Usually I'm making something that with that is real, like not crazy.
Zachary (02:44)
to tactically what are the tools you're using.
Barry Hott (@binghott) (02:47)
tactically what other tool am I using for AI?
Zachary (02:49)
You mentioned like Vio, but is there like a go-to or?
Barry Hott (@binghott) (02:53)
I'm going to use a
VO is my go to. VO is my go to. I use Kling a little bit. I'll play with some other tools. ⁓ But yeah, I'm mostly when I'm doing this stuff, it's mostly mostly in VO. Yeah, unless I can't get what I want. But it's kind of rare that something else would beat that. But yeah, Kling does good for non voice stuff unless they've missed something and they've added voice stuff.
Zachary (03:18)
How about you, Charlie?
Charley T | Founder & Nice Guy (03:20)
Yeah, I, for me, I think the biggest thing that I'm trying to focus on with stuff is like, I'm really trying to figure out how I can have ads that are sitting at the top of the funnel and ads that sit at the bottom of the funnel and the way that like the platform is using those ads really like, how can I make a conversation out of those things? Right. And so a lot of times what that looks like is
If I've got a video so like I think of it almost like a set of like Olympic rings like there's three rings at the top and there's two at the bottom and those are if those are three videos and then two static and that's like a fully efficient funnel that I can run up. Really what I'm trying to do is understand what are the points being made in those videos? And if I'm launching a fair amount of them and I got a good data set, I can really try to use the AI tools to understand like.
how to think about those, like somebody watched this, they went to my site, they're interested or they haven't yet been to the site, how do I get them across the finish line? Like what are the objections? What are the interest points? All of the things that I can get to have those folks see something else so that the videos are like the door knock, right? They're hanging flyers and then the saddocks are coming in as the closers. I think where I make far less
ads than I think most people do. And I actually prefer to have far less ads than most people do. ⁓ So I am trying to have way more continuity between the parts of the funnel. But for me, I think the biggest thing that I'm really trying to look at is more around the understanding of the continuity between ads that have a low CPM and a low frequency that soak up all the budget.
Shahbaz Khokhar (04:55)
you ⁓
Charley T | Founder & Nice Guy (05:12)
And then ads that have a high CPM and a high frequency that are super efficient, but can never scale on their own. And what's a way of making those ads that are hyper efficient, that sit lower in the funnel and the way the platform uses it. What's the best way of taking those ads and making them more relevant to the stuff that is first touch. And so I can take those bottom of funnel efforts, move them slightly up the funnel. So they get a slightly lower frequency, slightly lower CPM. They might not be as efficient, but
Shahbaz Khokhar (05:19)
In theory, it's usually that you're going to put a project inside the load. What happens is at the very end of the load, it's just being pulled into the market. only goes around.
Charley T | Founder & Nice Guy (05:41)
they might double or triple in spend. having that continuity of the conversation, I think is really, really strong. And that's one of the reasons also why I think I launch a fraction of the number of ads that most people do, because I'm trying to have a very congruent conversation and the few assets that I have and let them get a ton of data. I think honestly, the number one AI tool that I use is Andromeda. ⁓ To put it politely,
Shahbaz Khokhar (05:51)
Thank
Charley T | Founder & Nice Guy (06:10)
I'm not trying to hack the machine. I'm trying to give it the best chance for success by kind of staying out of the way. But I think that's where I approach things very differently than most people. I might launch 10 ads a month instead of like 100 a week. But
Zachary (06:24)
And
so, so, so you think, so you're saying that your trade off on left surface area for testing is that you're able to be much more pointed with the messaging along the entire funnel. Like it just makes it easier to be super focused on that.
Charley T | Founder & Nice Guy (06:36)
Yeah.
If yeah, like if a watch has a thousand moving parts and it breaks, it's going to take you forever to know what's even wrong or fix it or whatever. Like the more complicated something is, the harder it is to know what's wrong or what to take advantage of. The simpler something is, the easier it is to scale and the easier it is to fix. The easier it is to know what's happening. If I've got three or four ads at the top of my funnel and two or three ads at the bottom.
I have a very clear understanding of what works and what doesn't. And my entire job is just figure out what is my weakest link? How do I make it stronger? I think when you launch a hundred things, you're basically trying to grab a thousand pieces of attention for unknown number of reasons. And you might get really efficient at like that first touch, but understanding what's going to get those customers across the finish line. And more importantly for me,
what that type of attention is that's going to get somebody to buy from you more than once. Like that becomes almost impossible. If people, if people are walking into your store for a thousand reasons, getting anybody to come back is almost impossible. I'd rather focus on the most stable way of getting the highest, like I want to grow my revenue floor every month. If I can grow my revenue floor, my CPA can get worse. And I'm going to win the game on business model instead of volume of effort at the top of the funnel, which
I would rather make the business stronger than destabilize my funnel for short wins.
Zachary (08:04)
How about you, Shabazz? What is it looking like on your end?
Shahbaz Khokhar (08:08)
I mean, I have a very different take to Charlie on this one. think high volume credit production is where it's at. I mean, we've seen this with so many brands. We look up brands like all day every day and all the top brands are doing nine figures plus all the founders I've spoken to are doing nine figures plus are testing hundreds of ads every single, every single month. I can understand that like some things are situational and maybe certain types of brands. just got like a winning concept that's working and you you double down on it perhaps, but ⁓ everything that I've seen over a number of years doing this is like, yeah.
You've got to be testing a lot of ads and a lot of diverse ads as well to get different audiences. I think so for me, like really it's about high volume credit production and production is expensive. Like there's a hidden cost to production that people don't really talk about, you know, in getting content creators, but not just that. It's like the faff of like hiring somebody to go and write the brief to then send it to the content credit to get the content back, to then get it through revisions, to go through that process, deal with the payments, deal with disputes, blah, blah, blah. So there's like the raw cost of the brief, but there's also the management time involved in doing this.
And usually when you're doing more content, it becomes less efficient. If you're putting out one brief, it's relatively efficient. If you're a single founder, putting out a brief, if you've got a team and you're putting out hundreds of briefs, your efficiency is coming down and down, no matter how hard you try and beat that, the efficiency is going to reduce. for me being able to produce and I'd say produce an experiment using AI is definitely enabling us to produce more volume of creative. I think like big thing for me really is it's just about getting the mindset of the team.
correct to actually experiment with tools and giving them the freedom to do so. Because like, I don't know about you guys, but every time you log into LinkedIn or X or whatever, right, there's a new fucking tool and like, you're just like, you get this overwhelming anxiety of like, oh shit, I've missed out on another one, right? So you've got to unpick which ones are like really new tools and new models or which ones are just like a workflow that's been strung together by somebody. like, everybody's always making these posts like, hey, comment this and you're going to make a million dollars by tomorrow. I'm like, mate, just fuck off. Like there's load of nonsense. Like it's just.
Charley T | Founder & Nice Guy (09:59)
that.
Jack (10:00)
you
Shahbaz Khokhar (10:00)
It's
it's it's a load of crap. So what we normally do is we sort of how to get people in the team to actually pay attention to this. Honestly, it's a bit of like gamification. So we have this app in Slack called matter matter app. If you guys know I'm not affiliated, just check it out. It's really cool. You can give each other tokens. So it basically means you can incentivize people to do certain things. And each week we run a challenge, an AI challenge with a new tool or a new process within an existing tool. And we say to the team, Hey, produce whatever you want. And the winning one at the end of the week.
And it doesn't have to be the winning ad, but just the winning. Somebody judges it, gets a bunch of tokens. And those tokens can be used at Starbucks, Amazon, whatever. They can be used in lot of different places. Occasionally, I'll get passionate about some new tool, and I'll put a bounty there. So I'll put out a big prize. When I say big prize, I'm talking maybe the biggest one I've done is $100 or $200. And I'll just put it out there and say, hey, somebody make me an NA 10 workflow that does this. First person to do it gets the coins.
And there's always a lot of chat and there's always a lot of experimentation happening around that. And I think that's how you create an ecosystem where people experiment. Also, I single-handedly approve all the softwares in the company. So anytime somebody requests a new software, I approve it straight away. And I see that all the different tools that being, I see the rate at which we're experimenting with new tools. And if I don't see new requests coming through, like at least a few times a month, I'm like, we're not experimenting fast enough, but we also have to cancel things. So we kind of rotate and test a few different things. And then just touching on like what we're using at the moment.
I'd say like, I mean, the staples is mid-journey for images for sure. Flux is great. We access Flux via file.ai. So if you guys know file.ai, that's like probably the best site where you can experiment with lots of different models in one place. The UI is kind of fixed, but it allows you to access like almost every model. And you basically rent their GPU space. So definitely check those guys out. They just got their Series C, which they raised like 125 million in, which is, I find, insane because it's essentially a front-end wrapper for a bunch of different models.
So we experimented in FAL, we used Mid Journey a lot, we've been playing around with Omni Reference and things like that. And then, yeah, VO3, course, where videos just completely change the game. And then Runway References we started to use recently, so we're kind playing with that. Kling, again, as Barry mentioned, we've played on with that. So we're continuously experimenting with different things. Drumina we played with, Archads as well for UGC, which people have mixed feelings on. But yeah, we've got a whole stack of things that we use, but we're continuously testing every single week.
Charley T | Founder & Nice Guy (12:24)
I got a shout out for flux too. think it's great with the rep. use replicate forget the tokens for flux or whatever. And it is a super phenomenal. I'm a big fan of that. think it's so clever. You just I think training it is the hardest part. I wish I could pay somebody to just train it for me.
Shahbaz Khokhar (12:34)
Yeah.
Yeah,
there's a new thing that came out a few weeks ago called Flux Context. I don't know if you played around with it, but basically you can train a model. It's the fastest way to train things. You get a bunch of images of a product or a person or an environment. You put it into Flux Context, and then once the context is built, you can actually use that to then generate things in Flux Realism. So I find that's working pretty good.
Charley T | Founder & Nice Guy (13:02)
Nice, maybe I'll use it for my YouTube. Just throwing all my thumbnails and never have to work at that stuff ever again. That'd be nice. ⁓ No, I think that's all. Can I ask you a question real quick? ⁓ Yeah.
Shahbaz Khokhar (13:07)
Yeah.
Barry Hott (@binghott) (13:10)
can I give you an idea real quick? Wait, can I give you an idea real quick for that, Charlie?
Something I did for a YouTube thumbnail to get like Mr. Beast-ish ⁓ facial expressions. Just take a picture of yourself, a frame of yourself like this, right, in the setting of the video, and then just put it into view and show a bunch of reactions. And literally, it'll take your face and react however you want it to. I haven't... Yeah, just take a frame.
Shahbaz Khokhar (13:16)
you
Charley T | Founder & Nice Guy (13:35)
And then you just click, and then you just snapshots.
All right, I'm going to put that in my notes here. Hold on a second. ⁓
Zachary (13:40)
That's pretty cool.
Barry Hott (@binghott) (13:40)
See, yeah. I told, yeah, I had Dara,
I had told Dara that trick. I think she used it too, but I'm not sure. It's fun. At the very least, it's fun to see your face. It's also horrifying, by the way, to put a frame of yourself and then make, ⁓ really? Yeah.
Shahbaz Khokhar (13:47)
That's interesting. I'm going to play with that. Well, one thing I'll say as well is.
I can imagine.
Charley T | Founder & Nice Guy (13:54)
Yeah, it never gets my hair right. That's my only problem. I
Shahbaz Khokhar (13:57)
Yeah.
Charley T | Founder & Nice Guy (13:58)
am more the arc ads is that UGC tool, I think we've all used them. And again, yeah, mixed mixed feelings. when you nail it, it's great. ⁓ I have a question. I mean, I have a couple points, but like, something you were saying, Shabazz.
I totally have worked with nine figure brands. I've scaled brands to that high and we've had tons of ads in there. I totally get on board with that. Captions AI is dope by the way. I have a question though. Why are we copying nine figure brands for strategy? The idea is if I wanna start a sneaker company, all I gotta do is pay Michael Jordan and then I'm gonna be rich. Sure, here's my point.
Shahbaz Khokhar (14:45)
Yeah.
Charley T | Founder & Nice Guy (14:47)
don't copy a nine figure brand to become them. That's not how they got there. so like, like, so let's extrapolate this out a little bit more. Like the payment structure for a creative production, is that a better investment than having more singularity in the customer journey? Because the first sale is the least profitable. Do I want to be good at performance marketing to get my CPA low?
Shahbaz Khokhar (14:53)
Mmm.
Charley T | Founder & Nice Guy (15:12)
Or do I want to have a really steady customer flow of people that are likely to pay me again and again and again? And when we launch new ads, inherently any new creative, organic or paid launches at essentially the bottom of the delivery funnel. Like the machine shows the ads to the warm audience to see if they like it. And then based on that reception, it decides the frequency and the CPM and how it's delivered. So every time you launch new ads, you're forcing fatigue on all of the old ones.
And because you have more inconsistency in the funnel that's making growth more expensive. So for me, it's like the more ads you launch, trying to copy how like, I mean, I did this stuff for, you know, Fabletics and those guys. Yeah, we were testing a ton, right? When I got three, 10 nutrition to a hundred million, like we were launching a ton of ads. Absolutely. But I'd rather have an ad like an under outfit that gets to two million.
rather than launch a thousand ads that lasts for four days and constantly cycle up paying creative teams instead of letting the machines AI tell me these are the best people. let me give it, let me give a small example of this. And I feel like I've stepped over three questions. And so I apologize. I'm gonna try to wrap it up in a simple thing. If I have a steakhouse, I could try to put everything on the menu to get you to buy from me.
And let's say I get you a wedge salad, it pays me, it cost me like 40 cents and I sell for $40. Great profit, right? It's lettuce with cheese, done, right? ⁓ Shout out to Chelsea living room and their wedge salad. Not a fan, but great markup. ⁓ Fan of the business, not the meal, but whatever. They're good stuff. ⁓ Or now you might enjoy that, but odds are I paid 40 bucks
Jack (16:41)
Hahaha
Charley T | Founder & Nice Guy (17:02)
I'm never going back again. But I could have served you like the best filet mignon you've ever enjoyed. I might only make $20 on that, but you might come back 10 times and bring all of your friends. The point is the first transaction is the least profitable. The stickiest action in any customer journey with the business is what makes somebody want to come back again. And if I can elevate the leverage, if I can elevate leveraging the business model,
Isn't that far more important than paying a bunch of people to help me get more individuals who are more and more unhappy with my business, which raises my CPMs, creates ad fatigue and makes acquiring new customers more and more difficult. I get that you need to do that when you're spending 150, 200, $300 million. But most people, if you're not spending over six figures a day,
The more ads you launch, the harder you make it for you to be able to reach new people, and you're creating ad fatigue as like a standard practice.
That that's a very long winded, poorly asked question. I apologize for that.
Jack (18:06)
about the
Shahbaz Khokhar (18:09)
Yeah, is that a question or a statement? I don't even know.
Jack (18:10)
I might be missing the point. I could go.
Charley T | Founder & Nice Guy (18:12)
Okay. So the question
is, why are we copying nine figure brands?
Jack (18:16)
Right, I might be missing the point a little bit here, but I think that...
Shahbaz Khokhar (18:17)
because they make nine figures.
Charley T | Founder & Nice Guy (18:18)
Yeah, but
the creative isn't the reason they make nine figures.
Shahbaz Khokhar (18:22)
Yeah.
I mean, there's a product in the offer before the creative, but the creative is definitely driving the product and the offer to the market, right?
Charley T | Founder & Nice Guy (18:29)
I would disagree
with that. I don't think it's the product or the offer. It's the business model. There are a ton of athleisure brands. Fabletics and textile is massive because the business model is better.
Jack (18:33)
Maybe it's for the same reason you get this.
Shahbaz Khokhar (18:36)
Hmm.
What do you mean business model though? What does that even mean? That's just an abstract term.
Charley T | Founder & Nice Guy (18:47)
The ability to get a customer
to pay you repeatedly.
Shahbaz Khokhar (18:50)
That's my scene.
Charley T | Founder & Nice Guy (18:52)
No, no, that's not marketing. That's mean, it's marketing. But most of it is like you don't pay. I don't care how good your marketing is. If you spend money and you feel ripped off or the product doesn't meet your needs or you're buying it because of price or some trick like the ability to get 100 different people to all want to buy from you again and again is is is a lot more around how your business operates than like the quality of the ad. But like
Shahbaz Khokhar (18:53)
That's not business model.
Charley T | Founder & Nice Guy (19:21)
I don't want to derail us because I know this is about like, I guess my point is I find that a lot of people are using the AI and using a lot of these tools to massively ramp up production value. But I think that one of the greatest values of AI, especially with Andromeda is that you have less and less of a need to do that. And you can take far more educated shots than throwing way more spaghetti at the wall. And that's to be fair, one of the things
that I was just going through the lens stuff. That's one of the reasons I'm a big fan of it is because I can see what's working, what's not, and what's sticky and what's fading. And if I just make more stuff like the stuff that's sticky, I don't have to make nearly as many things, and I can spend the rest of my time getting more people to buy a second time rather than trying to trick more people into buying once.
Jack (20:10)
I can see the cogs turning, the shabazz right now. ⁓ But I think one of the reasons why people might be copying the nine figure, eight figure brands is because they think they've got something figured out. And maybe it's for the exact same reason that Alex has shared this and this screenshot of Dr. Squatch's ads. Is there something to learn from this? And when you were talking about focusing on the message and when Barry was talking about overlaying text and coming up with ideas for overlay text, I thought,
Charley T | Founder & Nice Guy (20:10)
Fair enough.
Shahbaz Khokhar (20:13)
I'm just, yeah.
Charley T | Founder & Nice Guy (20:21)
Mmm.
Jack (20:39)
This reminds me of what Alex was saying. Maybe there is something to learn from the brands that have spent so much money to pick on you out.
Barry Hott (@binghott) (20:44)
I wonder where Alex got that
from. I wonder, know, Alex and I, for those, for those who don't know Alex and I are business partners. And so we talk about this stuff. I own a piece of ad create. Um, but like, yeah, like, first of all, I want to address that. Like shout out Alex. That was a great post. One thing we, him and I talk about all the time is like creative waste. Um, and like not like if
Jack (20:48)
Yeah, where'd you get that from?
Charley T | Founder & Nice Guy (20:48)
It does look a little familiar. I'm going to go ahead and say that it
Barry Hott (@binghott) (21:11)
A lot of times we'll like put our energy and effort into making something and then not and then it'll fail, which is fine. But then we don't reuse it and retry it. So it's about like, no, how do we like go back at it? Take that thing that like we had a good hypothesis about and just put that life back into it with a little change. It's it's really not that crazy, especially now that you can use a eye to analyze it, you know, and also add to it is definitely something like it's kind of silly to not.
do a better job reusing it. Yeah.
Shahbaz Khokhar (21:43)
Yeah,
Zachary (21:44)
It reminds me
Shahbaz Khokhar (21:44)
I agree.
Zachary (21:45)
of like kind of like the idea of like finding there's this concept of finding sawdust in a business like let's just and it's like pretty much you know, Ikea is built on this idea is like Ikea just buys a bunch of sawdust to make bullshit furniture. ⁓ But you can implement that idea and they have like a bunch of margin on it. The idea is like at any point in your business, everything you were doing has sawdust. And so like what can you if you are building a beautiful table out of hardwood, you generate all this sawdust like what can you do with the sawdust that just exists. And I think
That's like one of the most powerful things with AI is like it connects production to strategy much closer. like what Shabazz was saying was like, yeah, we, if you go like the traditional route, we have all these people in the way it takes a long time and your efficiency goes to zero and there's like zero use, there's zero use in that, in that sawdust. But if you can have someone in the shop sweeping it up and just trying to figure out what to do with it. ⁓ and you, and they're the strategic thinker.
your actual return on that is so much higher than what it could be because your hard sunk cost and the materials is zero. And so I think that's what people should be thinking about a lot is even just that concept of where is there sawdust and how can we use AI to transform this into something that's valuable ⁓ or an iterative engine that's just consistently going ⁓ because the hard work's already been done.
Shahbaz Khokhar (23:03)
Yeah, I totally agree. I think to that point, kind of agreeing with both of you really there. I think one thing I've been doing more recently is getting existing ads which didn't work and just trying to improve them. You know when you get a gut feeling where you're like, you know what, this should have worked and I don't understand why it didn't work. Just putting it back in Premiere Pro, sitting with the editor and just like hashing it out again, doing another version. And often what we'll do now is like, you know the typical thing of launching multiple different variations in an ad set and everyone has their own different testing methodology, I'm sure. we still do.
testing in that way with auction selection. ⁓ But I find that once you've got a version that we're happy with is it's compliant and the spelling is correct and everything, we'll just accept that as one of the final versions. And if we try and do revisions, we won't then go and only launch the revision or launch both versions. So we go through this iterative process of improving something and then just launching all of them at the same time. Because you never really know which one is going to work. And I think just to conclude on the point around nine figure brands.
I totally agree. You've got to look at where you are in terms of your own brand. If you're at seven figures, eight figures, whatever. But ultimately, think as your revenue goes up, you need to be producing more creative to appeal to a wider audience of people. So naturally, you have to actually produce more creative across a more diverse range. So I think it's not about copying a nine-figure brand. It's more about aspiring to what they've done to get to that point. And all of them have got a diverse range of creative, all of them, every single one of them. If you go into all of their ad libraries, they're all running 1,000 ads.
or at least in the hundreds of ads. And it's generally quite diverse. I think like, yeah, for me, I think it's like just looking at what resource you've got at the moment and getting the maximum yield out of that resource before then leveling up to the next step. But the nine figure brand is just like, it's just like something to aim towards as opposed to, okay, we should do exactly what they're doing.
Barry Hott (@binghott) (24:48)
Yeah, I just I mean, look, I agree with both of you. Like, ⁓ I'm kind of more with Charlie. There's so much that large nine figure brands are doing that no one listening to this, maybe one person listening to this should replicate or be paying attention. But like, it's like from again, from a business perspective. Yeah, like or product perspective. Maybe I go. I'm not I hate when people just copy other people from that. I think it's lazy. But
Makes sense, right? From a business perspective, but from an ads perspective, a lot of the stuff that like a larger brand is doing, A, is they're at a different stage in their life cycles of business and what customers know and understand about their brand, about their product, etc. But more so like just from there are some bigger businesses that are absolutely still have the the the motor in place. But most, especially the ones that have gotten acquired, it's fucking gone. The people
The motivation, the drive, the hunger, all of it. The machine is dead. And we don't know that as outsiders, but I've been through enough of these and worked with enough of these to know. I don't. I know a lot of brands that are big that I wouldn't. I would look at what they do and actively not do that. So there are some that I would look towards that I think still understand what they understand what needs to be done from a modern performance mindset. But way too many mega brands.
that have scaled and gotten to those points still have the drive and motivation and understanding of their modern customer and the content that they consume that it's worth it's worth replicating. It's not worth paying attention to them for almost anyone who's a six figure brand or a seven figure brand.
Shahbaz Khokhar (26:36)
Yeah, totally agree with that.
Charley T | Founder & Nice Guy (26:36)
I want to
jump on this too, Barry, real quick with a quick example of that. I made a lot of mistakes once and wasted $17 million of Activision's money and Call of Duty was still the number one game.
Barry Hott (@binghott) (26:49)
That's a great story that most people would never share, Like most people would, or actually most people who'd be in that situation would have actually said, no, I'm the hero of that story.
Charley T | Founder & Nice Guy (27:01)
no, no. I
I staked my flag in the ground for canvas ads had developers. I was four teams at the time. Hyphen. ⁓ I think they're dead now. Who knows? ⁓ Hyphen was the senior ad ops. We made so many mistakes that still worked. Why? Because the product was amazing. The positioning, the brand, the everything around it like it existing worked. But the thing is,
there's 17 layers or whatever of like infrastructure of people that are just trying to keep a paycheck, including me at the agency of like, okay, great. Well, they're willing to spend whatever we want. What's the way to raise the billable as high as possible? Well, they said yes to 13. They rejected 22. We got 17 and everything was an abject failure. And two weeks later, they deprecated the canvas ads anyway. So
There are just a little bit of context to that stuff. ⁓ There are wins and losses. Now being said, Dr. Squatch brought up earlier a great example of a brand knowing their customer, finding a spot in the market and going all in on something that worked like the Sweeney stuff. That's brilliant. That's a huge win.
Barry Hott (@binghott) (28:24)
Interesting, interesting. Let's actually talk about Dr. Squatch for a second. What just happened to Dr. Squatch? Does anyone know the news about Dr. Squatch that happened in the last, what, month or two?
Charley T | Founder & Nice Guy (28:32)
I mean, they made a lot of money. Or they got a lot of money. I don't know if they made a lot of
Barry Hott (@binghott) (28:34)
Didn't they just get acquired? Yeah, they got a lot of money. They
Jack (28:36)
yeah, they got a quag,
Barry Hott (@binghott) (28:37)
just got
Jack (28:37)
yeah, Unilever bought them. Yeah.
Barry Hott (@binghott) (28:37)
a lot of they just got acquired. So yeah, a great brand to have studied up until pretty much tomorrow. me like. No, maybe look at what they did three months ago, but three years ago wouldn't even be relevant to now. But like what they did three months ago was probably really hard work because they're still hustling getting it. I mean, I'm sure they'll still put out some great stuff for a while. ⁓ But often after these acquisitions.
Charley T | Founder & Nice Guy (28:39)
Yeah.
Yeah, look at what they did three years ago.
Barry Hott (@binghott) (29:04)
The acquiring company is like, we'll take it from here. Squirt. We'll take it from here. Sport. We got it. You guys great job pat on the back. We we're the big business people now. We got it from here and then they'll. I forget who, but then maybe that won't happen and maybe that won't happen and maybe these giant mega companies have learned from from these mistakes in the past, but I've seen it up close and personal both through acquisitions. I've been through and other clients I've worked through worked on through acquisitions.
Charley T | Founder & Nice Guy (29:16)
Is it Unilever?
Jack (29:18)
Yeah, that's a lot of it,
Charley T | Founder & Nice Guy (29:19)
It
is? Yeah. Okay.
Barry Hott (@binghott) (29:34)
Like post acquisition a year later, I will be very and someone mark my words and come throw this in my face. I dare someone at Dr. Squatch team to hear this and come at me. I bet you they'll be very making very limp boring ads a year from now, but tons of people will still be trying to replicate.
Shahbaz Khokhar (29:38)
you
Charley T | Founder & Nice Guy (29:42)
Time stamp.
Let's hope they do better than Jaguar.
Shahbaz Khokhar (29:54)
Yeah.
Zachary (29:55)
It really just depends on the corporate strategy of the acquisition in most cases, right? That's the real truth here is that performance got them there, but the corporate strategy of Unilever potentially isn't to 10x the D to C, right? And I think that's just the truth of what happens here is the mission ends up being completely different and you can end up reading the wrong lines. ⁓
Jack (29:57)
Jackboy.
Barry Hott (@binghott) (30:23)
We
should do another separate thing about talk about that. And I think that's not what we're here to talk about today, but it's really an important topic.
Charley T | Founder & Nice Guy (30:30)
I do want to say something about AI and the iterations going back to something where I really agreed on what Shabazz and Barry were talking about, because because I do think while we have differences on some approaches, there's also a lot of singularity and what we're talking about. And I really want to call something out and get to a little bit behind it. The iterative process and the replaying of winners and Shabazz, you were talking about going into Premiere and just cutting stuff up. There's an element to the AI and again, I'm going to come maybe more from the machine learning version of AI than the tools to make more stuff
because I'm more of a fan of the, I'm more of a fan of data science than I'm a left brain, right brain guy. I'm on the other side of the fence of where you guys are at, which is great. One of the reasons the ads really work, especially on third, fourth, fifth iteration is because you're giving the machine more and more data around who likes this stuff. So for instance, the under outfit ad I called out earlier that spent $2 million. That ad was, we were spending 2000 a day before that ad launched.
And we went from 50K a month to a million a week in like 18 months. The reason that worked wasn't because that was the ad that was great. We did a photo shoot. We launched a bunch of image ads. We found the ads that worked. We then found a bunch of headlines that worked. We then took the headlines and threw them in the images that worked. And the machine is getting better and better because we didn't launch a thousand ads. The data set that the machine has on the ads that we're deploying and who likes the brand.
is getting more more focused. It's getting smarter and smarter doing this really valuable thing well. We then found out another headline that worked and then we took the best images and we put it into like a slideshow. The video of that slideshow built on the image testing of inside the product and outside product shots, people in it, all sorts of stuff. And then all the text overlays and everything. The fourth iteration of that process.
sat on top of the mountain top of all of the data, all of the spend that had come before it and all of the modeling behavior that the AI had done around who likes this messaging. And because we really drove deeper and deeper and deeper with that singular sort of approach, which was show the construction and talk about like, know, the positioning of the business over and over again, by the time that video hit, yes, that video was a good ad, but it's
a slideshow of six images with a text overlay that stayed a persistent text overlay in the upper left-hand corner. I did that with the slideshow function on the Facebook ad tool. Like there's no text.
Shahbaz Khokhar (33:05)
Yeah, but just
a question. Is that the AI learning or is it you as a creative strategist as a human learning what's working and therefore producing the next round?
Charley T | Founder & Nice Guy (33:14)
It's both. It's both. It's us. So it's me as a data scientist understanding what people respond to well and what they don't. And it's also the AI getting more more data around these are the people that respond well to showing the construction of the bra rather than a woman in it. These are the people responding to this type of vernacular, this type of language of, know, of an us versus them, you know, kind of thing of like, you know, not to get too much in business, but know, skims is for one kind of woman, we're for everybody else. Like,
those types of things. And as the machine gets more and more data around this customer persona, the more hits you take at that customer persona, your ability to strike gold gets a lot better because you're digging deeper and deeper, but also the machine is getting wider and wider at what you're teaching. And so it's, I think it's honestly, it's both. and, ⁓ I don't think you could get there without, without leveraging both sides.
Shahbaz Khokhar (34:09)
Hmm. Yeah.
Charley T | Founder & Nice Guy (34:11)
And I will say that
Jack (34:11)
Your approach.
Charley T | Founder & Nice Guy (34:12)
almost every time I've gotten an ad that spent seven figures, it looked a lot like ads. was built on the learnings of other ads that dropped off at four, at five, and at six. And it was that mountaintop they were building, not only through getting better at the ad, but also understanding the reason that ad performs is because the machine is getting better at knowing who to serve it to.
Shahbaz Khokhar (34:32)
Yeah, I think we can reach a limit on that point where like you don't want to produce when I say produce a lot of ads, you don't want to randomly produce stuff, but that's just a bad idea. You do want to do sequential credit testing, right? So you want to like figure out what worked and go to the next step, right? And it should always be based on a hypothesis of what previously worked. I will kind of caveat that and say there is space for like human randomness and weirdness. And you need to sometimes just like go all in and lean into that and just like ignore everything else in the ad account and just like zone in, go into a quiet space. I have a tent in my office. I literally sit inside.
and just like, it's, have like, so I have these like things on the side of my desk, like panels, and have this like big sheet that goes over the top of it in the back. This is like a four, I'll send you a picture afterwards. Oh yeah.
Zachary (35:03)
Camping tent?
Charley T | Founder & Nice Guy (35:11)
It's like a fort!
Jack (35:13)
Hahaha
Barry Hott (@binghott) (35:14)
Sign me up.
Charley T | Founder & Nice Guy (35:15)
amazing because i was thinking like you
know when they're selling a house like this is the kids room and it's like this little teepee thing but now you've got like a fort like i'm so jealous i gotta build a fort
Shahbaz Khokhar (35:21)
Yeah. I'll send you a picture afterwards.
It's pretty good. But for me, that's my like creative space. If I like switch off all that, there's no light basically. It's just all dark and I have this like LED light. And if I feel like being creative, I just kind of lock myself in there. And I feel like you need that human weirdness and creativeness. And that's like where you end up not copying what everybody else is doing, where you just come up with random weird things that work. But I feel like that only works when you really understand your consumer psyche and when you spend enough time.
reading the comments, reading the reddits, reading all the things that your customer's saying, speaking to them, going on video calls with them. ⁓ I think then watching the ads.
Barry Hott (@binghott) (35:56)
Watching the ads, like literally just watching
the ads that work and don't work, spending that time instead of just naming them and studying, ⁓ did the ones with, the ones that were ⁓ UGC work better than the ones that weren't labeled UGC? Fuck that, go watch your ads.
Shahbaz Khokhar (36:02)
Exactly.
Yeah, I totally agree with that so much. Like we spent a long time trying to figure out naming convention and getting it perfect. And like now AI is coming in, it's tagging everything anyway. So like, what's the point? And even the AI tools that do tag everything, you kind of consider like, what is the usefulness of that? Because like, okay, well, what are you going to tell me that this is one consistent element, but it doesn't watch the ad for you, right? I think one thing that is more useful, think maybe Barry, you alluded to this earlier about like improving ads. One thing we've been playing around with is just chucking an ad into Gemini and say, Hey, can you just improve this for me? Like, where would you improve it?
And also raw footage, asking it to, we send it like a long raw footage that's uncut. And we're like, okay, find me the hooks. And actually I'd say like 50 % of the time it's finding some pretty decent hooks. know, bear in mind it's looking at the visual and the copy as well. It's not just the transcript, it's the visual as well when you do it in Gemini. I found that's pretty useful.
Charley T | Founder & Nice Guy (37:01)
Can I ask what's the actual operational process of that? Like, are you just opening up Gemini, uploading a video file, and then saying, find the hooks? it, just for people watching this, like, what is the executional process to leverage that? Because it feels simple, which is great, but if people don't execute it right, they could maybe miss it. I'm just wondering, if you walk through that real quick.
Jack (37:02)
Ken things?
Shahbaz Khokhar (37:20)
Yeah.
Yeah.
Yeah, I don't want to misquote, but we've got a Gemini plan for the company. And around about half the users in the company have access to it. And I think they get like a certain token usage, which like we've never had any additional billings on. But there's definitely a lot of hard costs involved in like running the different models and the different tools. And it does get expensive, but then you consider the cost of making real content. And like, you're like, it's just still a fraction of the cost. So like,
It's just a non-argument basically. like, well, this cost a few cents to make or analyze or whatever. I think it's just, it's worth spending every single time.
Charley T | Founder & Nice Guy (38:01)
Gotcha.
Shahbaz Khokhar (38:02)
Where
things can get out of hand is when you've got a workflow, like an NA 10 workflow that just starts to loop on itself and you can end up spending a lot of money by accident. So those are the kinds of situations you want to avoid. We've had that happen a few times. But generally I'd say the Gen AI stuff is like relatively not cost prohibitive.
Charley T | Founder & Nice Guy (38:21)
Yeah, I'm production. Yeah.
Jack (38:22)
I like your approach,
of using AI to drive quality and not quantity and using that internal competition between you as a creative strategist, understanding the data to understand what you need to iterate on, what's working. But I do agree with you, Shivas, that quantity is also important. You need swings at bat to hit a home run. ⁓ But I think at the core of this is that prioritizing one can't sacrifice the other.
Can't prioritize quality to the extent where you're not doing anything new. And you can't prioritize quality to the extent, or I should say quantity to the extent where you have just a bunch of garbage ads.
Shahbaz Khokhar (38:58)
Yeah. Yeah,
I totally agree. I think the underlying principle is what Barry said. It's like, watch your ads. Like, why aren't you watching ads after they've been produced? The whole team should be watching them and trying to improve them. And that's the key thing.
Barry Hott (@binghott) (39:09)
And I don't think I like someone, someone called me on thank you for saying it, but like also someone called me out for being a dumb dumb. Like I really think we have some time before AI can actually beat human brains in studying and winning in terms of video assets and storytelling. And I'd love to be proven wrong. ⁓ but I that's, I'm just not seeing it yet.
Shahbaz Khokhar (39:33)
I mean, I think we've had ads winning, which have been completely produced with AI. I'd say we've had more images working than videos, but there have definitely been some videos. But I'd say, usually, it's not the entire video. We have had entire videos made, and they've gone to spend $100k plus. But I think it's mainly production of hooks, B-roll.
Barry Hott (@binghott) (39:46)
Right. Same, same, same, same.
But there's still a human prompting
that, there's still humans, you know, involved, no? Wait. Wait.
Charley T | Founder & Nice Guy (39:56)
Can I make an analogy to this? it's one thing to
Shahbaz Khokhar (39:58)
Yeah, I agree.
Charley T | Founder & Nice Guy (39:59)
talk about this is all, sorry, Barry.
Barry Hott (@binghott) (40:01)
I
want to hear what Subhash was saying. Hold on.
Charley T | Founder & Nice Guy (40:04)
Yeah,
Shahbaz Khokhar (40:04)
I've got Barry,
Charley T | Founder & Nice Guy (40:04)
sorry. Sorry.
Shahbaz Khokhar (40:05)
sorry. Barry, you've finished. What was the question?
Barry Hott (@binghott) (40:08)
No, I was just saying, surely there was a human involvement, is what I was suggesting, right? That you still had to prompt for that AI stuff that you've had made, right?
Shahbaz Khokhar (40:18)
Yeah.
Yeah, I think we end up in a situation where there's like, we're all using reinforcement learning models, which basically, if I was to kind of explain that in simple terms, like imagine you have a Tinder for making ads where like, can swipe left if you don't like the ad and you can tell it why you don't like the ad and you swipe right if you like the ad and you tell it why you don't like the ad. And I think that's where we're all ending up within 12 months, 24 months. If somebody doesn't build it, I'll build it. But if somebody is building it, me.
Barry Hott (@binghott) (40:29)
Nice.
But that's biased.
But that's biased right by our human beliefs.
Charley T | Founder & Nice Guy (40:48)
You could also
Shahbaz Khokhar (40:49)
Yeah.
Charley T | Founder & Nice Guy (40:49)
feed the data though and see performance metrics as being the bias of it. I, I, I want to kind of make a bit of an analogy here for it. Cause we're talking pretty technical stuff, but like the way I look at it is like, you know, I'm a musician. I've been in bands, a bunch of stuff. And it's like the AI led stuff all feels like boy bands. Right. It's the next replaceable hit song from no name pop group. And what's that? Nothing.
Shahbaz Khokhar (40:55)
Yeah.
What's wrong with boy bands?
Charley T | Founder & Nice Guy (41:18)
I love it. Don't get me wrong. I've got like a party playlist and I will dance and sing along to the in sync stuff. There's nothing wrong with it, but there's a place for it that also creates opportunity for other folks that are, ⁓ that are manufacturing their own, their own stuff. think the more that folks are leaning in to the AI, the more that creates the opportunity for person. And I think ultimately there's this life cycle of
Some brands are built in a way where their end goal, they have one product and it's the business and they're doing everything possible to exit that business and they're going the boy brand route from day one. There are other brands that are more founder led that might go in the very human side of things from day one. And I think that both playbooks can work. I think you can even do a little bit of both, but I do think that
The more you lean into the manufacturing of just what works and just having the reinforcement of just everything being very much the same, the more there's an opportunity to come in with a different idea and introduce humanity to certain things in a way that might ultimately get lost. And to your point, Shima, think taking both shots really works because you can have people on your team that are just doing the, like the person doing the fun thing.
Shahbaz Khokhar (42:36)
Hmm, I'll it. Yeah
Zachary (42:37)
I
think.
Charley T | Founder & Nice Guy (42:44)
And then other people doing that reinforcing nature and every human wind might end up feeding that nature. so the ground is being like a broken.
Shahbaz Khokhar (42:44)
Hmm.
No.
Zachary (42:52)
I think like a
I think a really, really good way to like frame this and like think about it is if you looked at the billable invoices of like traditional agencies over like the last 50 years and you compared the production budgets to the strategy budgets would feel silly. But the truth there is that like strategy was always like incredibly important. And like people paid a lot for strategy.
The number of strategies you could execute on was limited by budget because there was a certain production budget. What's true now is that production doesn't have as many constraints on the strategies that you can execute. And so what does that now mean to be true? It means that more strategies are now, they now have the option to be executed on. And so like, I think it's just like, if we look at the reconfiguring of
Shahbaz Khokhar (43:42)
Yeah. ⁓
Zachary (43:50)
where money needs to go to make something happen. The opportunity now is like you can now execute more strategies potentially. And so that's probably what you should do because like that's where I think now that the arbitrage is because there's been a reshuffling of where money needs to go. It means it unlocked the other thing, which is like, we don't need to choose one. ⁓
Shahbaz Khokhar (44:12)
Yeah,
I think that leads to a natural flattening of job roles as well. So at the moment, if you think about the process of making ads, right, you've got a media buyer, a creative strategist, a video editor and a content creator, right? Usually you have like three, four people involved in, let's say making an average ad. I'm excluding like graphic design stuff on whatever, right? Just let's just say a standard video ad. I think what will end up happening is you'll just have a media buyer or I don't even know what their job title is going to be going forward, growth strategist, and they'll just be prompting and making and launching themselves.
So think basically three or four job roles get flattened into one, at least for a certain type of ad production. I think that gets interesting because all of a sudden, people who weren't necessarily, didn't have that creative outlet or didn't have that ability to produce, now have the ability to produce and everybody in your company becomes a producer. I think that's the biggest shift, right? In my opinion.
Barry Hott (@binghott) (45:00)
But those
people might not have taste or understanding.
Charley T | Founder & Nice Guy (45:02)
Yeah, you don't want me
doing it. I'll be, I'll be flat out. I'll be the best person looking at the data. But you asked me to do the prompting. mean, Barry has challenged me repeatedly to make more ads and I do appreciate that. But like, they're our kid. Any kid with a Snapchat account is gonna beat me on making video nine times out of 10. 99 out of 100. Like, so I agree with your approach. I have one question.
Shahbaz Khokhar (45:11)
Yeah.
Charley T | Founder & Nice Guy (45:29)
with what you just said, Shabazz, and what Zach, you just said, because I agree that production is now not the constraint. And I agree that there's opportunity on arbitrage there, but there is now a new constraint.
which is inventory. You could launch a million ads and they all get 10 impressions. The stats sig, the confidence in the data set that you have, there is a direct and inverse relationship. The more shots you take, the less confident you are. Basically the more ads you have, the dumber every one of them is. And so,
Jack (45:47)
to not just a simple.
Charley T | Founder & Nice Guy (46:12)
I'm wondering how you feel about that. Like you could launch a hundred ads, a thousand ads, a million. But unless you have the budget to back up getting an understanding of how the machine is using that ad and how people are responding to it and what that ad does to your bottom line. I think. You it is very easy to essentially over leverage that arbitrage position to the place where you are just.
hope you're just pulling lottery tickets hoping to get lucky.
Zachary (46:44)
So is your thesis that there's potential winners that will never see the light of day?
Charley T | Founder & Nice Guy (46:49)
No, my thesis is that you're looking for winners when it's actually not about a winning ad. It's about having a well-constructed funnel that is stable and consistent. And let put it to you this way. If that ad gets spent, what happens to your bottom line? And if you can't answer that question, your system is too complicated.
And so ultimately there's only so many eyeballs. There's only so many places to show an ad and that inventory is getting more expensive. I'd argue the more ads you launch, the more expensive that inventory gets faster and faster. So it's not free, right? Like there's not this infinite world where you can just get data.
Shahbaz Khokhar (47:33)
I think you have to look at what is a total addressable market. You're not just comparing against your own ads. It's the whole market that you're going after. Right. So if there's a market, I don't know, sleep supplements, you're not just competing against your own sleep supplement ads, you're competing against everybody's. So ⁓
Charley T | Founder & Nice Guy (47:39)
Sure.
Yeah, ads and organic
and just people walking around in the world. Absolutely.
Shahbaz Khokhar (47:50)
Yeah. So you can't say with, nobody can say with certainty because we don't understand how their algorithm truly works. People at Meta can't even tell you whether or not you launch another ad is going to reduce the frequency on one of your existing ads or somebody else's ads that sells sleep supplements. We have no idea.
Charley T | Founder & Nice Guy (48:07)
But you can launch an ad and see what happens.
Jack (48:07)
Well, the inventory isn't purchased by individual.
Shahbaz Khokhar (48:10)
Sorry, Jackie said.
Jack (48:11)
I was just saying the inventory isn't purchased by individual ads, it's purchased by dollars. The dollars is what's going to determine the cost of the inventory.
Charley T | Founder & Nice Guy (48:21)
Yeah, but and also the user experience, right? Like if people have an inconsistent experience with your brand on their website, or if they don't like the product, like an easy way of reducing your CPMs and making customer acquisition cheaper is by having a more consistent customer journey where more people are happy with your business. Like one of the easiest ways to reduce your CPMs is just have more people buy a second time.
Zachary (48:44)
Yeah, but you have to assume that like in this, in this like simulation that we're running in our heads, we have to assume that we figured out the business. And then now our objective is how do we scale the business with advertising? Cause that's our main goal.
Charley T | Founder & Nice Guy (48:57)
Sure.
Yeah, and I'll say this, okay, if we take that, I think we then have the, you both sides of the coin of when you iterate, you can probably do better at what you're good at, which is optimization. And when you take the shots, like Shabazz was saying, and I'm a big fan of that, then you're gonna get to the spot where that's bit of a risk, right? Like it might work, it might not, but I think you have to take those risks if you wanna get.
bigger instead of just getting more efficient with the money you spend.
Shahbaz Khokhar (49:30)
I
think this comes down to testing strategy as well. And Barry, I'd like to get your opinion on this as well. I know that some people go down the route of just have one ASE and chuck everything in there and just see what happens. From my point of view, that doesn't give a fair test to the new ads that coming in, because existing ones have got traction. So we still go down the route of launching with cost gaps in new testing campaign with different ad sets for every single test that we're running. And we migrate things across to ASE. So that's our pretty typical setup.
Jack (49:52)
and I used the...
Shahbaz Khokhar (49:58)
I guess my question is, we do testing with cost gaps, but when we have confidence in a creative, so let's say we're like, we want to push this one a bit more because we just got a gut feeling, a sense, a marketing sense that this is going to resonate better with the audience. We tend to bump up our cost gaps or we switch to highest volume. I don't know what your thought is on this because I think this impacts into like what we're saying here around like volume of creative, because if you're doing cost gaps, you're going to have to produce more to get something that works naturally. If you're doing highest volume.
It's high risk because you're putting a lot more money behind an ad, but if you have high confidence because you've got previous traction and you know what kind of ads work, maybe there is a space for that. yeah, I've seen all your posts online. I'd like to just get your perspective on this one.
Barry Hott (@binghott) (50:37)
Sure, I love that question. It's my favorite topic perhaps. ⁓ Aside from making ugly ads. ⁓ You know what you said makes sense. I like the sound of that. There's just the, you know, I really try not to go against what the system wants to do, assuming that I have the system set up in a way like that the data ⁓ that I trust that it's optimizing on my behalf, which it's not always doing. But to your point, like if you're saying like,
Hey, this ad that we made, it's not getting a lot of spend, but we think it's really good for if that's sometimes bad. But if it's for good reason. That's sometimes good, like, so if you just like it because you guys like it, like you're fucked, right? Like, please don't write. Just, you know, try not to. I'm not saying you write. I mean, everyone listening, right? Like trying to have an ego about it, trying to try to be aware of your biases and try and break those down. But if you're like watching this ad, you're like, I don't like.
This should be scalable or like it seems like it has other positive downstream effects. Then yeah, crank it. And usually I like to think of that as like.
The the the I think of kind of like a marathon like the motivation to get you to complete a marathon right is very different than the motivation that it gets you to like sign up for a marathon. But the ad that gets you to complete the marathon is the one that gets the credit for the conversion right gets credit for you. So like is the last cup of water someone drinks during a marathon. Is that what we fully attribute the completion of the marathon to right.
Or was there a whole slew of things that had to happen before they got there? And now that depends on the product that depends on the thing. There's entire ads that can get someone from zero to buy right now for certain things, right? ⁓ But other you have to acknowledge like some ads and some products that they just require more education, more time. And the less the more it gets, the more your head gets colder people into market.
Shahbaz Khokhar (52:17)
Yeah.
you
Barry Hott (@binghott) (52:42)
or colder people to become aware. Yeah, the higher it's going to cost. And you either should be doing that via understanding your ad creative and then increasing your cost caps to allow for that or, you know, intentionally like excluding people so that that ad is only being shown to those colder to colder users. And then also building your tolerances. It's not just about cost cap. It's like your tolerances because right. You're saying like highest you can go to highest volume to
It's about having a tolerance for ads that you're you're OK with being less apparently efficient. And I I think that nuance is really, really, really important.
Shahbaz Khokhar (53:23)
Yeah, I totally agree. You get brands where they run a VSL, it's like 30 minutes, 40 minutes on the landing page. You get them from zero to a hundred and they convert, right? that's a completely different, it does happen. does happen. It's a completely different strategy. Well, I mean, one thing that we do, don't know what your take is on this, it'd be interesting to hear, is that when we know that a particular style of ad has got a higher new visitor rate, so, you know, we check GA and see like, is that ad on average got like a higher new visitor rate? Sometimes what we'll do is we'll bump up the cost gaps. We'll allow it to spend more just because we know it's bringing in.
Barry Hott (@binghott) (53:31)
It's possible, yeah.
Shahbaz Khokhar (53:53)
new traffic to the website, whereas some of your offer statics, your product focus things, which are more bottom of funnel, getting a higher return of this to rate, will bring it down. And what we do is we work out the blended average, and we try and predict, this is where the cap should be roughly for this type of ad and this kind of ad. Of course, what's happening is they scale differently, so you're never going to get it perfect. So we're tracking this in real time. So I don't know what your thought is on using that process, or if you have a different way of doing it.
Barry Hott (@binghott) (54:16)
It's noble, it's honorable. I like the sound of it. I just really want to be doing. I don't want to be doing that much to get that much manual granular control. I'm not saying I'm right, but like that's just generally a little too much involvement for me. I'd rather the system do some more of it. But if you have human context that the system cannot have, right?
Shahbaz Khokhar (54:27)
Hmm.
Yeah.
Barry Hott (@binghott) (54:41)
then yeah, I would apply it. Just know that your context or your bias might be wrong, but. In this case, I don't know. I don't think it's bad. It's definitely not bad.
Shahbaz Khokhar (54:50)
Yeah,
yeah, I think where we find, yeah.
Jack (54:53)
What are your guys' thoughts on
the abundance of new tools that are out there? Are you testing all of them? I know I heard from you, Shabazz, that you're checking out quite a few of them, you're encouraging your team to check them out. But what are your guys' thoughts and feelings about some of these new tools that are coming out and making lots of noise?
Shahbaz Khokhar (55:15)
icon. He's blocked me on all social media because I had to call him out for making a product that doesn't work.
Barry Hott (@binghott) (55:22)
Yikes.
Charley T | Founder & Nice Guy (55:23)
I'll say this.
The tools are only as good as the application and the value for the business is only as good as how much it amplifies the way that you acquire future cash flow. And what really matters by that is like,
It's only as good as how you can use it. I think whether it's icon or arc ads, whether using VO or making images in chat, what matters most is the ability to do the job in a way that you are attracting and earning the attention of the folks who are the highest quality customers for you. And going back to something Jonas said here,
Are you guys using AI to create customer profiles? I absolutely do. I'll look at all the people that are my second, third, fourth time buyers and I'll try to understand what products are buying and I'll talk to them. And if I can find ads, if I can build creative, that is more likely to get me somebody who's going to buy two times or three times. That insight is exponentially more important than an ad that reduces my CPA on the platform by 5%.
⁓ And so I am absolutely looking at who are my ideal customers, who are the two or three ideal customers that I have? If you get two or three customer profiles of the folks that will come back and give you money more than once, like just shoot for second purchase rate. Those customer profiles is the easiest way to get to $50,000 a day if you're stuck at like 500. Chasing more than that is doing a lot of expensive work poorly.
And I think having that end goal in mind is really important. So Jonas, to your point here, what I'm doing there to take an action item is I'll talk to the customers. I'll have like a zoom call or something. I'll take the transcriptions of those conversations. And I'll also take all the customer journeys of people who bought more than once that like my number one goal is let me sell the one thing in the business that has the highest propensity to somebody to come back. Right? Like I'm not running 10 different offers. I'm not launching a thousand different ads. I'm trying to do one thing exceptionally well.
instead of a lot of things like mediocre, because that one thing is what's going to get you to 50 or 100. And so I'm taking that feedback from all of those people, phone calls, Zooms, customer journey stuff. And I'm saying, what is the singularity? Like, what is the what is the way that if I were only going two or three different directions, how do I get the most of that person? Because those are the people that are actually going to grow.
your business. There's a very different thing from being good at some ads are great at being a salesperson. Other are great at acquiring cash flow. And I think I'd much rather try to grow the business than make a sale if that makes sense. So I'm using AI to understand who those best customers are and then using that.
Shahbaz Khokhar (58:28)
Wait, what's the difference
between growing a business and making a sale? Explain.
Charley T | Founder & Nice Guy (58:32)
Okay.
Great question. Selling any product at a profit, selling anything you can for losing caps and thousands of ads to get the lowest CPA is inversely related to having the most amount of people coming back and buying from you again. There's a big difference between selling a product and buying customer journeys.
Jack (58:34)
Yeah, I think if any...
Yeah, I've had a similar experience to...
Shahbaz Khokhar (58:52)
Jack, sorry, you were saying something.
Jack (58:52)
Yeah, I've
had a similar experience to what Charlie is describing. I used to work with a cheese brand. They were lovely. They were featured on her favorite things. Cheese. Yeah. And ⁓ their best selling cheese on top of funnel was this Kinta cheese, but it had terrible reviews. didn't get returned customer purchases. It would turn up and people would see the ad again and then comment on the ad. Mine turned up like a brick of mayonnaise and that would get a lot of, you know, likes and comments and people agreeing. They get bad reviews online because of this best selling cheese.
Charley T | Founder & Nice Guy (58:58)
Cheese!
Jack (59:22)
Right? ⁓ But their blue cheese was, had a higher CAC, right? It was more expensive at acquiring new customers. But these people who bought the blue cheese would come back sometimes for years. Their lifetime value to the business was so much more. We lost Charlie. ⁓ But yeah, so I've had a similar experience to Charlie. That wasn't, I guess, a move for efficiency, right? Switching over to focusing more on the original blue cheese. It was a move for the growth of the business.
Right? Not just the reputation of the business, not just lifetime value.
Zachary (59:53)
Well, I think cheese
is a good spot to wrap this up. Also guys, well I appreciate you all taking the time, people on the call, chatting about this and then everybody that tuned in live. For everyone that's asking if it's gonna be, the recording will be shared, it will be, it will be on our YouTube channel probably in the next few hours, not tomorrow. But yeah, definitely go follow everybody on this call. All of their socials are in the description, everything like that.
Barry Hott (@binghott) (59:59)
could talk about cheese for a while.
Jack (59:59)
you
Zachary (1:00:22)
Does anybody else have anything? Lastly, want to talk about AI or anything like that or wrap it up?
Shahbaz Khokhar (1:00:29)
Talk to Grok, it's insane.
Barry Hott (@binghott) (1:00:32)
Was it talks a grok? Really? No, no, no, I'll do that.
Shahbaz Khokhar (1:00:33)
Yeah, have you tried the unhinged
Zachary (1:00:35)
yeah, I love it. It's
unhinged.
Shahbaz Khokhar (1:00:36)
mode on Grok is absolutely fucking crazy.
Jack (1:00:39)
That's incredible,
Barry Hott (@binghott) (1:00:39)
shit
Jack (1:00:39)
it writes all our newsletters. It's amazing.
Zachary (1:00:42)
Yeah.
Barry Hott (@binghott) (1:00:43)
I can't
tell how serious you're being.
Shahbaz Khokhar (1:00:45)
You need to try it Barry, it's worth the $20, $30 just to test it out for a month.
Barry Hott (@binghott) (1:00:48)
Oh, I definitely pay for it. I pay for it for
sure. I Elon Musk to have all my money. But I'm just kidding. But yeah, I'll play with it. What should I do? should I? What's what's up? Wait, is it? What's so crazy about it?
Shahbaz Khokhar (1:01:01)
Go into voice mode, go to unhinged and just start talking to it. You're going to be entertained for many hours.
Barry Hott (@binghott) (1:01:03)
Sure.
Alright, is there anything productive I can get from it?
Jack (1:01:09)
Ha
Zachary (1:01:10)
No, absolutely not.
Shahbaz Khokhar (1:01:10)
Probably not. Probably
not.
Barry Hott (@binghott) (1:01:14)
Good. Well, that's why I just want to go play. I don't know. I don't know anything else. Thank you so much,
Zachary (1:01:17)
Alright guys.
Jack (1:01:19)
Thanks much everyone. Talk soon guys.
Zachary (1:01:20)
Peace out, thanks
a lot.
Shahbaz Khokhar (1:01:22)
Take care, see you. Bye bye.
