The Right Way to Use AI in Your Ad Creative

00:00:00] Richard: hey folks. Welcome to the eCommerce Playbook podcast. I'm your host, Richard Gaffen, director of Digital Product Strategy here at Common Thread, and I'm joined today by Mr. Joy Sharma, who heads up our global accelerator program, as well as like our seven figure brand program. And Joy joins us on occasion here to talk about specific.

[00:00:17] Issues pertaining to seven figure brands. That's primarily who he works with across that spectrum. And so we're gonna dig into some of those today. And Joy, let's you teed up for me a couple of things here before we hit record. But first off, like, I think what we wanna talk about is the mentality that the seven figure brand needs to have in today's day and age.

[00:00:36] And of course that includes things like ai, but just generally speaking, what the attitude of the seven figure brand owner, the operator, needs to have. In order to succeed in eComm here in 2026. So, joy, why don't you tell us, talk to the folks a little bit about kinda what you were just telling me around that mentality that you need as a sim figure brand owner.

[00:00:56] Joy: Yeah, it's, so we notice a trend that emerged as people started using more and more AI across their organization. And to some people it might just blow their mind. And

[00:01:05] to some people it's just like, yeah, that's supposed to happen. And I think to make it very prominently clear to people, we would need to disconnect our actions or inputs from output. To a certain degree in terms of how many hours went into doing something, because I noticed this is, this is value across the board. This is like seven figure, eight figure every figure human that you can think of. But basically people put a lot of emotional weight for some reason on something that took them a lot of time.

[00:01:34] Like this is basically the IKEA model, which is like if you make it by your hand, you will love it more, et cetera, et cetera. But it actually is, is boars for people who own businesses, which is how you win. People don't really care. It. It's like I got to where I got irrespective of how you got there.

[00:01:50] People just care about you got there. So what happens is people will go and try to achieve the same output and somebody will put in 60 hours, 60 days, and somebody will just make it in a day. And the output is the same, the result is the same. And let's say, let's talk about a landing page, like we can take any example or any work done.

[00:02:09] And this is like

[00:02:09] a general example, which is if I sit and I make a landing page and I take 60 days to make it. I will just have emotional connection to it, which let's say the landing page didn't work. I would, I would actually, this is like sunk cost fallacies, so like it didn't work and now

[00:02:23] I will spend additional money to make it work and it will still not work.

[00:02:26] And I just like lost everything. I'd lost time and

[00:02:28] money. And then there's the other people who will just make a new landing page every day. And they will win more. They will lose more, and they will win more. And overall, they'll just win so much more because in 60 days they were tested 60 different landing pages.

[00:02:41] There are people who do that also in our database. Now, do I want you to be that? No, but the point is you would actually need to just fully disconnect. Previously it was, if I have big teams, I can do this, I can't do this, yada, yada, yada. Now it doesn't matter.

[00:02:56] Now you actually have the same resources as that of a big brand, and you can choose to sit behind this idea that I'm small, so I don't need to do it. And then there are people in the same revenue size as you who are doing stuff faster with the use of ai, and they're, they're just not emotionally connected because they're focusing on volume. And I think that is the mentality shift that needs to happen for people to actually study using the thing. Because I don't think people think about this. They don't, they nobody thinks about it this way till now.

[00:03:25] Richard: Right. Yeah, no, I, I think the, in the age of ai, like the, the idea that like, you couldn't, you can get, sorry. I guess what I'm saying is like you can get great work out of, you know, sort of like five seconds of prompting an ai. So like, the actual, the excuse to not do it is sort of like falling away and there's no reason not to move fast.

[00:03:45] This is definitely something that we've like, come across with clients a lot. Is this exactly what you're saying that like your, your sense of the input. Feels like it ought to match the performance of the output, and that just kind of, never happens. So go ahead.

[00:04:00] Joy: you, you probably hear this and we hear this, and this is actually, this is the exact statement people hide behind, which is they will say, yeah, AI is good. It's not there yet. It's 90% there. The

[00:04:11] other 10% is not. And Taylor has the stake on Twitter. Anything he, he was replying this to Cody. But basically, what if instead of AI having to go and make the other 10%, you just open the brand guidelines to be 10% more open,

[00:04:26] then that means the AI is there and you can use it, and you just get outsized returns.

[00:04:30] By opening your brand guidelines 10%, your output goes a hundred times. That's a bet that in a business world, like you should take that every time and then you can. Tweak the things that work that are, that are win.

[00:04:42] There's a world where this used to happen. I, I forgot what the context of this was, but basically people used to test ads that were not approved. I think you should do this. It's like if you are a creative agency, if you have creative agencies and you reject their ads, you should reject them. Make a new campaign and test all those ads at $10 a day.

[00:05:01] And if they're still good ads, then you should make the idea of like, okay, I rejected that ad. But it might actually work.

[00:05:08] Do I really hate it that much? That I want to give up on the a hundred k additional revenue will bring me? And that's actually a conversation that you should be open to. That's basically the dilemma of your walking into it. Like it's 90% of the day way there. Sure. Like open the guidelines 5%, get a person to polish the other five,

[00:05:23] and you are there, you have that.

[00:05:25] But it's still better than waiting on some agency to go and do work that takes a month.

[00:05:30] Just get it done in a day

[00:05:32] Richard: totally. Particularly on, in like the, the world of like social media ads or whatever, where like. Content that's native to those platforms, like, looks kind of shitty. And that's okay. I mean, so, oh, by opening up your brand line guidelines a little bit, you just actually potentially give yourself the potential for that that ad to perform better as well, just because it seems more platform native.

[00:05:51] So, I, I, so I wanted to Go ahead.

[00:05:55] Joy: The, like, I just have so much data across these things, and that's

[00:05:59] why I, I look at them. But you should, let's say you are a business in any category, like you can choose any category and we can do this exercise live, which is like, you should take any category, let's say health, and then you should look at the biggest advertisers in, in that industry.

[00:06:12] And none of them are actually advertisers who are, who are huge brands. They're probably affiliates or stuff like that. And then I want you to go and look at their app. And their ads are actually native ads. Like this

[00:06:23] idea of native ads. I think people understand in this industry, native to a certain degree, but if you actually look at what native is like, you look at Taboola ads or rumble ads and stuff like that, you will notice that they're so hooky that, but you need to understand you are actually in a competition against those people.

[00:06:39] If you, if you go into your audience targeting in Facebook and your audience target is female above 50 in age, like, that's basically what you call the golden clickers, and

[00:06:49] that's basically what you're targeting. You are in a competition with affiliate and you should look at their ads, and people in that age region, that demographic don't care.

[00:06:58] You are branded if they actually can weigh the value proposition better than you and you should look at their ads. They're like, some of the funniest thing is like it will be a, a grocery basket with baking soda from Arm and hammer and like two more and a grocery list. And that is the best ad they have in that industry. What, what are we competing with? Like where is branding in that? And they're spending more money in a day, probably than you're spending in a month. So

[00:07:22] I think that that is actually a really good example.

[00:07:25] Richard: Okay. So you mentioned something in, in your notes kind of before we jumped on here, around the idea of like finding, and maybe you alluded to this already, but like finding the speed of your organization. So actually like taking your organization and, and push for like how, how, how far can we go in terms of speed before this breaks?

[00:07:41] So talk to me a little bit about like how that plays out or how you'd suggest people go about that.

[00:07:45] Joy: Yeah, this, this came out as a thought experiment to, if everything that I say that I just said right now is true and a business wants to achieve that, then they'll be like, okay, I can apply it to every single units of growth. We have units of growth, which is like creative landing page offer, email. I have like, these are the units of growth. And then I was thinking, okay, what is the best way to measure it and how, how you can do one thing that impacts all five of them, and then how would you measure speed should be tied to that? So I think the, the idea have is there's a learning container, which is, let's call every market moment a learning container because it actually contains the learning across all five units of growth. And then you can actually see how good your organization is based on how filled your marketing calendar is. Now, marketing calendar can be anything like if, if you're a company that sells. Anything. It doesn't matter what you sell. There is some market moment you can latch onto. Like that is actually one of the most valuable, I think, podcasts that me and Tera did, which was

[00:08:39] for seven figure business.

[00:08:40] You don't create trends, you latch onto trends and you grow your business that way. So what are the trends that are happening? And then if you can make all the units of growth to that marketing calendar, you win. And because there are trends you need to move fast, which you

[00:08:52] can move forward. So that's why I think the way you should measure how, how good your organization is. Built up culturally, if the culture is there that you can, okay, I, I, if I need this done, it will be done by the time it needs to be done. To see how good you are at that is basically how many marketing calendar moments you have. And if there's no blank space in the month, you probably won't, or you can say you are pretty good.

[00:09:16] And I think that should be the measurement that everyone should have. And then a lot of people realize they have 20 days, they're empty, and then

[00:09:23] migration would be Y or 20 days empty.

[00:09:26] And that I think is the best thing.

[00:09:28] Richard: Well this is interesting 'cause it suggests this idea that like there's no more, evergreen periods or something like that. Obviously you'll have evergreen creative that's running all the time, but there's no sort of like moment when there's nothing special on, right? Like, so if you have a week, let's say you have, I don't know, a product launch one week and then the next week's blank, and then you have, I don't know, whatever, it's April Fool's, April Fool's Week or something, the next week, whatever that blank week needs to be filled up with something, or three days or two days or whatever the case is.

[00:09:55] Like you just come up with something else.

[00:09:57] Joy: it indicates two things, right? Like it indicates how fast can your organization move, because to latch

[00:10:02] on to a trend, you need to have that ability. So it measures that. It also takes all the shots available on the chess board, which is like, I can only do five things. I can do landing pages, offers, creative emails, and I'm doing all of them. So your execution is actually multiplied by that degree. So it can't be, oh, I'm really good at offers. I'm not good at landing pages. You're actually good. You need to be good at all of them. And then you get the highest chance of success, which is the most beautiful thing, which is if I were to individually if you think about this, if you were to go and test five emails, right, and you were to test it in Evergreen, it would just not make sense in terms of return on invested capital because there is a cost to everything. If you were to go and do the same thing combined as a marketing moment, because your chances to win are athletes three times better. then it becomes like, if, if I have the best chance of success, I can go and execute every single unit of growth. That means it gives my organization the highest chance of success, and then I do it in a way that I can move by the, by the second, like every second.

[00:10:59] That makes the most versatile organization and we are trying to define the goal center of culture of an organization that makes it set up to win. And the difference in a person, like a difference in an organization that has 10 black days and an organization that has zero black days. Is is white. Like you can't even imagine how it is.

[00:11:17] Like, I think probably I'll run this through Claude and make a graphic for this and it might, it might show you how big the difference is. It is worlds apart

[00:11:25] and this idea that no evergreen moment exists is actually wrong because if you think about it, or it might be execution, I don't know if people know how to execute this well, and I think that might be different.

[00:11:33] My, my reality might be different from yours,

[00:11:36] but if, if say there's a marketing moment, I'm talking to baseball moms. I'm not talking to just baseball moms.

[00:11:42] I'm still talking to my core demographic and I'm trying to unlock something new and it might be unlisted landing pages. Like if you think about creative list, think about the entire customer journey.

[00:11:51] I will launch ads. Facebook will not serve that ads to people who are not relevant to that. That just goes fundamentally against the auction mechanism, so that will not happen. Meaning the people who are not that demographic will never see that. And also you need to think about this way. They never knew it did something so according to them, you are still ever know you're doing nothing.

[00:12:10] And then that goes to unlisted landing page and the email is sent specific to it and people only open it when it's there. It's interesting to them. So the, so to you as a brand, you are doing something every day to the end customer demographic. You are probably doing one thing a month,

[00:12:25] which is pretty good.

[00:12:27] Richard: Yeah, yeah. That makes sense. And so, and if you can sort of think of like the diversity of the different sort of. Elements of your customer pro or file, you're going to, there's, there's people that you can hit in such a way that they experience you once a month. So it's like, okay, we're, we're doing a launch once a month.

[00:12:43] It's like, okay, from your experience, you're doing a launch once a month, but there's a possibility that every single day could be somebody else's once a month when they're your customer. Right? So

[00:12:52] Joy: It happens with sales all the time. Like

[00:12:54] if you have a winning offer, you should test it all the time. And a lot of time brands are like, oh, but we

[00:12:58] did the same thing on Prime Day and we did the same thing on Mother's Day and yada yada, yada. And it's like, okay. But how many new customers are there?

[00:13:07] 80% of your revenues, new customers.

[00:13:09] That means they just haven't never experienced your brand. So it

[00:13:12] doesn't really matter what you're selling

[00:13:14] at that time period.

[00:13:15] Richard: totally. One of the hardest things about marketing is the fact that like what you're doing will be new to somebody almost all the time. And actually exhausting things kind of takes 'em much, much longer than you actually think it's going to. Let's talk a little bit about well, okay, so, well there's two things.

[00:13:28] So one, you mentioned the, the difference in culture between the brand that has no white space on the calendar and the blank and the the brand that has like, I dunno, two weeks or whatever. So what are like the major culture differences between those two organizations?

[00:13:42] Joy: I think it, it's a culture of the organization or the heartbeat of the organization. And this is like where like you should have performance coaches, like people used to have performance coaches to see how well, like how good performance of every single high value individual is. But I, I would say that was needed because we were working under the premise that a players probably produced twice the output of a B player.

[00:14:03] And now you can probably produce a hundred times more output. And the only way you can fathom that reality is when you can actually think about, okay, I can just disconnect the input from the output. And for the longest time, people actually thi thought about this issue. Like, if I put a hundred hours in, I will build something better than if I put 10 hours in. And that is, that is how we grew up. That's like a very important part. And I think disconnecting that is how you would win. And. Disconnecting that as a business owner is just hard. Like it is. It goes against everything we stood for and we stand for for a long time. It might not be applicable in every scenario is the other part is like this is why it would make your brain hurt,

[00:14:42] and that's why people go to like performance ities and stuff like that.

[00:14:44] This is like more of a mindset of how you would win if you were in this industry. But yeah, it is hard because human brains by default, like to have binaries and the binary of do more input. You get more output.

[00:14:57] What's the best thing for our brain? And now you need to actually think about every single individual instance and be like, doing more work.

[00:15:04] Will it actually help

[00:15:06] and are my hours actually correlate to success? Or is there something I can just produce a hundred times more output with one hour of input? And that actually becomes the,

[00:15:14] Richard: Yeah, no, it's, it's difficult to, to create a culture where people are tied to results rather than tied to effort. And I think one way that you get more results in such a way that you can actually begin to see the effects and tie yourself to those results is by getting more reps in and doing more things.

[00:15:31] And I think that's kind of fundamentally what we're saying. If you wanna see, if you wanna be more output focused and input focused, focus on getting as much output as possible rather than foc as much input as possible. And I think that's kind of what we're talking

[00:15:41] Joy: But management for the, for the longest time is actually opposite, right? Like management is, oh, you can't control the output. So you, you should force people to put more input. And now I'm actually going and saying like, that might not be the best way for every single thing. And that's what, it's going to make it hard where people,

[00:15:57] I also think it affects this ideology of it. I, I hear this, which is like, oh, this company that is our competitor. Has 50 people working in the department for the same thing that I am just here with my two best friends. But what you're forgetting is that it doesn't actually matter and they're probably thinking I'm firing half of the people they have on their 50 people team anyways,

[00:16:20] so they're redundant to them and you are small and you can't compete in a business like TE said this really good statement when I was small and we were talking to a business, but it was basically. What growing is, is basically you as a brand going into environments you have never entered, competing with businesses and or people who have better sources, better supply chain and better competitive advantage than you. And you go there and you win and you can't win under the same pretext of they have something that's better than me. Because I know they have something better than you. That's why they're there. And if you need to win with them, you can't win with the same exact A to Z copying. You need to have something different. And that different comes from can I as three people beat them with 25 p?

[00:17:06] Richard: Yeah. And, and like we were kind of saying before, like, now is the easiest time in some ways to do that. Like obviously like a larger brand is, is probably technically more resourced to have more workers producing more with AI or whatever. But I think like particularly with ai, it's how you use it as much as how much you use it.

[00:17:24] And with three people, if you're being smart with your like, implementation of these tools, like your advantage could be closing that gap would potentially be enormous.

[00:17:32] Joy: I, we, we were on this internal CD call yesterday where we were talking about token usage across the organization, and I think that's something you can track with, like how many tokens you're using. But when we were talking about our ops and GNA and like all the costs, we were saying, okay, now if our token usage becomes a certain number as a percent of revenue, it needs to reflect somewhere.

[00:17:52] And the, the flip in opex to token usage is actually going to be interesting, which like.

[00:17:57] Now can I replace how many people for how many tokens is actually the point? Like just go and spend on tokens. Like token is just like, is just actual like energy. Like that's what a token is, right? Like it's, it is literal energy and you can just transfer over energy if it's actually used in the best way.

[00:18:13] Richard: Yeah. Yeah. Yeah. No, no. It's, it's a, it's a hell of a time to be alive for a seven figure brand. I feel like things are hard. Things are harder, things are easier. There's all kinds of stuff going on. One. One thing I wanted to talk about is like, this was in your, in your notes as well, which is that the most cost effective way to test these things is the learnings container.

[00:18:29] So talk to us a little bit about the learnings container as a framework for thinking about doing more.

[00:18:34] Joy: the learning container is, is marking calendar, and the reason it is learning container is because it has the highest hit rate, but also it forces me to test everything and it tested in the

[00:18:43] most cost effective way because you get to test everything together. I have the highest hit rate, but also. Doing individual offer testing and let's play the scenario is like, I want to test if given a gift card versus giving 25% off versus a sample kit versus a quiz kit.

[00:19:00] Like I want to test all these things in Evergreen. What is to stop? Like if you play this journey out, what's going to happen is you're going to test this Evergreen and there's gonna be prime day, and then you're gonna take your best offer and you're gonna use it in prime day. And the only exception to this rule is it's Black Friday.

[00:19:15] I'm gonna use a site discount because that works well. For all the other smaller moments, you're going to take the best evergreen offer. I'm saying Then just test the Evergreen offers in the marketing moment, and if it wins in the marketing moment, then you can use it in Evergreen. Just just flip that script because the chances of success of having a bad offer working in market moment is still hard,

[00:19:37] so you would actually never lose money, like you can never lose. If you use this, learn this ilog of a marketing calendar to test evergreen stuff across the board and then transfer things from Marketing calendar into Evergreen instead of the other way around. Because the other way around is the most cost heavy, worst use of everything's like setting up to win.

[00:20:00] Richard: You find like a specific marketing moment, as in sense excuse for test, for offer testing basically, and then pull, yeah, offer creative. Actually, I think you have like sort of like a hit list here, landing page creative hitting on certain trends and then take those and then translate it over to your evergreen.

[00:20:17] Just like we often say sort of, I don't know if this is an irony or not, but like over Black Friday we say like Evergreen performs best over Black Friday, so there's this sort of like reverse of that. Then over across the rest of the year, it sounds like basically what you wanna be doing is producing marketing calendar moments, specific creative, and then taking that, turning it into Evergreen.

[00:20:37] Then that's the stuff that ends up performing best in Evergreen, which then later down the line in Black Friday will be the thing that performs best. So it's creates, like you were saying, this sort of virtuous cycle of offer testing in moments, pulling it into Evergreen, and then having that all come together over holiday.

[00:20:52] Joy: You gonna have this like as a, as a mind image. It's like success is on like five pillars in a marketing calendar and in an evergreen period. It's just like on one. So like

[00:21:01] success might be on five pillars of. You can just make a good landing page and the market moment will work. You can just find the right trend and it'll work, or you can find just the right offering will work. Meaning there you get to take five shots at winning the same day. The only thing that's constant in all of world is time. So if it's the same day and I get to go and take five shots to be successful at that day versus take one shot in an evergreen day, like I will just take the five and I, the only thing I can't get back is that day ever. So if I win on that day. That's all I care. And if I

[00:21:33] get to take five shots with a marking calendar, and that's why I'm saying it's a learning container. The reason it is a container is because it'll contain the information of what won. And you wouldn't know this. You would know this. Did my click through it actually improve?

[00:21:45] Really good. Then, you know, the creative work or the offer work and you can just basically make subsets of things of, okay, what is the learning I need to go and carry forward from this marketing calendar? Okay. I found a landing good format that worked. The next time I'm gonna go and test something, I would probably hold that constant because I care more about testing the offer.

[00:22:00] So I would hold everything else constant to the last winner I have found, and then I'll change. I will take big swings and small swings across the things I want to test. And then again, I'll play the same game all the time and I will just win more than if I were to not do anything.

[00:22:17] Richard: Right. So the idea there is like, just to make sure that I'm clear on this and that it's kinda laid out, is that the, the five pillars you say during the marketing moment are you have the opportunity in a moment to test five things at the same time. Is that essentially what you're saying? Like you're trying a landing page, you're trying a new offer, you're trying a new creative, whatever the case is.

[00:22:35] And then you could pull those into Evergreen and then test them one by one to isolate those variables and like get a better sense of, of what's

[00:22:42] Joy: You, you, you don't need to, I don't think you need to isolate them out. You can just like, you can just see what one, based on the marketing counter moment,

[00:22:49] looking at the metrics, like what improved over baseline is basically what you can do if, if click through it, improving your creative one. The idea of the five shots is basically if I were to test just an offer on Evergreen Day, I only get one shot at it.

[00:23:02] I either

[00:23:02] find a good offer, I or I lose. And here I'm saying. Either everything can work, but I'm saying even if nothing works and only one thing works, I still win on the day. I still

[00:23:12] take more cash home and I get the learnings. If you argue, and people do this all the time, it's like I want to do individual.

[00:23:19] I don't wanna do multi variant testing, I wanna do single wear testing. And the reason for that is because at the cost of some money, I get to go and take clean learnings. And as sim for your biggest business, you don't actually have that luxury. And you still, so this is the best in the middle. Like you get to go and take the learnings and the cash home.

[00:23:43] Richard: I lost you at the end there. You say what The last thing you said

[00:23:46] Joy: Basically this is like, this is the best way to do it, which is instead of sacrificing cash for cleaner learning,

[00:23:54] I get to go and take the cash home and learning home by testing it in a learning container of a marketing calendar. And I just keep ya it forward.

[00:24:03] Richard: Yeah. Makes sense. All right, cool. Well, I think that kind of covers the basis in terms of like what we're gonna talk about today. Anything else that you wanna hit? Any like final piece of advice that you want to give for seven figure brands here?

[00:24:13] Joy: I think you

[00:24:14] should we do this like I think Tara posted about this, how we do this internally, getting Claude to look through every conversation you have with your team members. Across Slack, zoom, like standardized communication. Track it every day, get cloud to be the management layer, and

[00:24:32] then having a dashboard or a leadership board on how, how many marketing calendar moments you were able to pull off in a month and try to force to have no blank days. And then tracking and asking where I'm missing, like where are the inefficiencies from Claude, because it has view into everything. Is it probably the best hack you can use to build a sculpture and you should

[00:24:50] thrive for it? Because there would be a day, probably by the end of this year, where if you don't, you will probably not be able to grow because the competition will be so much.

[00:25:00] Richard: Yeah, that's right. No, I think that that's a good point. That like work, working on the infrastructure of your culture, that's maybe one of the places where if you think about, okay, so if your time isn't going into spending 60 days on a landing page anymore, where is your time going? And it feels like.

[00:25:16] Where your time should be going is essentially developing the infrastructure, including things like ai, whatever, to build the culture that then creates tons of output, right? So the input work should be going into the system rather than the actual execution of the thing itself. Which is maybe the takeaway here.

[00:25:31] So.

[00:25:32] Joy: I think. We, we launched P six and this is like, actually this is applicable for P six and PE seminars. Like what happens in P six. Like we, we can tell you the best thing to go and do and we can give you all the inspiration like on how you should do it. And like we can tell you've been there, done that this is the way to win. But again, your execution speed matters there also.

[00:25:48] So like across P, all the PEs, like I think this is the most valuable thing you can build internally as a brand.

[00:25:55] Richard: that's right. All right folks, well I think that's gonna do it for us for this session. Well, obviously we're gonna have joy back, I wanna say next month maybe. I can't remember the frequency that we do this, but we'll have joy back to talk to the specific issues of the seven figure brand. But until next time, folks, it's great talking and we'll see you next time.

[00:26:10] Bye. Pow.