Ecom Growth Insider · Podcast episode

Sebastian Valiente: Your Twenty Ads Are All Saying One Thing

With Sebastian Valiente · Published 2026-01-29 · 59 minutes · Hosted by Andrej Tumachowitsch

Sebastian Valiente: Your Twenty Ads Are All Saying One Thing

Sebastian Valiente quit software to start an agency with no experience and failed for months. His own read on why is unusually blunt: he was saying the right words with none of the confidence that comes from having done it, and people could tell. So he worked free for an ecommerce brand, launched ads, built pages, sent emails, and got a skill.

The idea he is best on is what actually breaks a plateau at eight figures. Not more creative, and not more variations. His distinction is between a sub-angle and a theme: twenty ads that look completely different can still sit under one umbrella, and once that umbrella is spent, more of the same will not buy you more spend. What does is a new umbrella entirely.

He is equally direct that at the smaller end the constraint usually is not creative at all. Plenty of accounts he has walked into could have doubled without touching an ad, because nobody was reading the numbers correctly.

What you will take away

• A sub-angle is not a new angle. His pet supplement example: benign lumps on an older dog is a theme. Positioning the product as an alternative to surgery, or twisting the knife on anaesthesia risk, or on the cost, are sub-angles under it. They find winners. What they do not do is unlock more spend, because they are all the same conversation.
• A new theme is a different problem entirely. Longevity, for the same dog and often the same buyer, is a separate umbrella. That is what opens a fresh pocket of spend. He says it is one of very few levers that can take an account from ten thousand a day to twenty inside a month.
• Personas and angles are not the same axis. A weak core spoken to a mother and to a college student is two versions of one message. They categorise by messaging theme rather than by persona, precisely because you can run many personas inside one theme and still hit the ceiling.
• At the smaller end the constraint is usually the reading, not the creative. Two failures he keeps meeting: not seeing that a lower efficiency with more contribution margin is more money, and reading Shopify in a silo while Meta spend drives Amazon revenue. Consolidate those and the answer is often that you can already double spend.
• Collecting research is easy now. Choosing what to test is not. Anyone can pull a hundred insights out of reviews, forums and comments in an afternoon. The skill he says is still unsolved is prioritising them, and he reaches for consumer psychology, Eugene Schwartz's mass desires among them, to decide which insight has the most weight behind it.
• They built the system around static ads on purpose. Statics let them put twenty-five genuinely different concepts under one messaging theme into the account inside a week, which is a message test rather than a creative test. On accounts spending eight to ten thousand a day they are launching thirty to forty concepts a week, in a single high-volume CBO where not every concept takes spend.
• Attribution is directional at best, so benchmark inside the platform. His method: establish what your true cost per acquisition is at a given in-platform number, then use that in-platform number as the benchmark. It is not a real value, but it is tethered to one, and it is the only data that exists at the ad level.
• Ugly usually wins, and not only because it is authentic. He agrees with the cliche but adds a second reason that applies to statics and landing pages: making something pretty often makes it slower to read. A plain static tells you what it says in a second. On AI creative his read is that the capability is already there, the giveaway is the audio, and anyone who knows how to replace and treat the voice can get past that today.

Jump to a chapter

• 00:00 - Introduction
• 00:52 - From software developer to performance creative
• 04:36 - Quitting the job on a cushion
• 06:30 - Why the first agency failed
• 08:20 - Working for free to get a skill
• 11:53 - Why brands plateau
• 12:22 - Messaging diversity at eight figures
• 14:53 - The smaller brand's problem is reading the data
• 17:12 - A lower ROAS with more profit
• 19:06 - Personas against angles
• 22:23 - Lumps and bumps, or longevity
• 25:04 - Why twenty ads can still be one message
• 26:25 - Is Andromeda actually a change?
• 29:40 - What creative strategy actually is
• 31:37 - Where the research comes from
• 33:43 - Volume against quality
• 36:37 - The quality floor
• 39:07 - Why they build around static ads
• 41:39 - How you know an angle is working
• 44:24 - An angle that surprised him
• 46:42 - The biggest misconception about performance creative
• 49:27 - Do only ugly ads work?
• 52:13 - AI-generated UGC
• 55:46 - The ethics, and where this goes
• 58:51 - Where to find Sebastian

Full transcript

Lightly edited from the episode captions for readability. The episode is the record - where they differ, the audio wins. Lightly edited from the episode captions, which carry timings but no speaker labels, so the transcript runs unattributed.

00:00 - Introduction

Welcome back to Ecom Growth Insider, the show for D2C founders, marketers and growth obsessed operators who want to scale profitably without the guesswork. Today's guest is Sebastian Valiente, one of the sharpest minds in performance creative right now. He's helped Brands scale to 300k, 400k to 1 million per month in revenue by systematically unlocking new messaging angles that break through plateaus and drop CPAs. I deeply respect the way he thinks about creative strategy, consumer psychology, and scalable ad systems. In this episode, we unpack the real reasons why brands get stuck at 500k or 1 million per month, and how to break past it with smart creative strategy and not just more content.

If you're running paid ads or trying to scale a D2C brand in 2026, this is one of the most valuable episodes I've recorded so far.

00:52 - From software developer to performance creative

Let's dive in. Sebastian, how are you doing? I'm good man. How are you? Yeah, doing great.

Do you want us to give us a quick rundown? Like how did you get into performance creative and what kind of brands are you helping to scale today? Yeah, sure. I guess I'll start kind of pre marketing career. So I was a software developer for a couple of years.

Throughout that time I'd always kind of been dabbling in various initiatives to try to start a business or do something I kind of always wanted to do a little bit more. And eventually I decided to quit my job and pursue starting an agency. Very, very ignorantly. I was kind of pulled in, I was, I was pulled in to that kind of like through like the YouTube SMMA rabbit hole. So I quit my job and tried to build an agency with absolutely zero prior experience or knowledge.

And then I failed miserably for several months. But eventually I decided to actually attempt to gain some real skills. And so I started offering to work for free for various e commerce brands. And one guy took me up and I I had basically worked for free for a couple of months and just doing everything I could to help grow his brand. So I was launching ads, launching landing pages, sending emails and I had been, prior to that I had been like consuming a ton of content and like studying like, so I had a lot of the theory on like how to run ads, how to make ads, good ones.

I just never done it. And so that was my first opportunity to like actually go in and, and learn through practice and was able to get him some results which was great. It wasn't anything massive but that doing it was enough to give me the confidence to like, when I spoke to people, they could actually tell, okay, like he's not a grifter, he knows what he's talking about. At least he has some real life experience. And so anyway from that free kind of gig I was able to land a freelance role and then eventually landed my way at an agency, as a, I guess as a media buyer.

And about eight to ten months after that I became a director at that agency. And then I quit that in 2024, September ish, and I went to go be a head of growth at a pet, supplement brand. I helped grow them to a million dollars a month and still working with them. And then that was my segue into. What I realized from there was like I really enjoyed working with multiple brands at the same time.

There was just so much more learnings that were happening. And so that was sort of my segue into I guess, starting, what is now Growthpod, which is still very young. When I first started, we were basically just like an all around boutique growth agency. Kind of just like we do everything to help the brands we work with grow. In the last four to six months, I would say I've kind of really niched down into, Performance Creative specifically.

I mean it's something we were always doing. But now wanting to sort of grow the agency a bit, I realized like, I need to focus and kind of, really hone in on one thing that we do well.

04:36 - Quitting the job on a cushion

And so today that's performance, creative. Got it. That makes sense. And initially, like, what gave you the confidence to just quit your job and start the agency? Yeah, so the real answer is, we, around that time, maybe like two months prior, my wife and I had sold, our condo.

And then we were going to, you know, at that time we were ready to start a family. And so the idea was that we were going to sell our condo and then buy, you know, a slightly bigger home so that we could, you know, start having kids. And this is around the time that the interest rate started going up, at least here in Canada. And long story short, we decided in light of everything, hey, you know, we're going to hold off, let's rent. And so now, we're renting and I have this massive chunk of cash in my bank account from the sale of my condo because it had gone up in some value and we had some equity and whatever.

So I had a massive stack of cash at the bank all of a sudden. And, it just kind of like dawned on me. Like up until that point I had tried a bunch of things, but my struggle was always I never feeling like I had enough bandwidth to really pursue other things. And that's probably not true, but that's just how I felt. And so I thought that was my opportunity to, you know, what, like, what if I gave a real attempt at something, quit and just did something like full time, like, could I make it work?

And that money was there as sort of as like, as a cushion. And so I was like, even if it fails, you know, we have some money to live on. I can go back and get a job. That was the idea. And and yeah, that's kind of what pushed me to to go and do it.

Obviously I had, I was consuming courses and content, so I, like I said I had a lot of the theory.

06:30 - Why the first agency failed

I didn't realize at the time that that was nowhere near enough, but that plus a little bit of that cushion was like, okay, cool, I'm gonna go and give this a try. Yeah. And now that you, that you have significantly more experience and launched a new agency, like looking back, what do you think were like the biggest gaps or the issues, why it didn't work out initially? Yeah, I mean it's pretty simple. It's just like, it's so obvious, which is just like, if you don't have legitimate value to offer people, like, no one's going to want to work with you.

Now there are some people that are so good at schmoozing and talking that they can get past that. I can't. I felt, I always felt it was so obvious. Like looking back, it must have been so obvious, people talking to me that like, like you're saying things that are right, but somehow I feel like you at the same time don't know what you're talking about. And it's true because I didn't have the actual experience, so it's like, that was a problem.

I didn't have anything to actually offer people. I had no legitimate skills, I had no real value to offer. And so I, I honestly ultimately came down to that. I, think. Yeah, Yeah, I imagine you were like talking to those, to those businesses, founders, whatever, and you, you didn't believe yourself that you're able to do it, but you tried to sell them on something.

That's what I'm, it's like, it's like you. I, I was saying a lot of the right words, but the confidence was missing. And I honestly believe that like that is what made it not work is like they, they didn't believe me in one sense. You know what I mean? Yeah.

And, and what gave me that confidence was just doing it. Like just actually acquiring the skill set, you know, and so looking back, it was so stupid.

08:20 - Working for free to get a skill

Like if I had just spent some time actually you know, just, just doing the work and learning, things would have turned out much differently. But is what it is. Yeah. And when you started working with the first, brands or first businesses or you mentioned you were working for free, like what kind of businesses were those? It was an E commerce business.

So this one was a, it was a, A shoe product. It was like a. So basically, like sneaker heads, they have their shoes, their Jordans and things like that. And they don't they obviously are trying to preserve them. And so it was a product that it was almost like a balloon that you inflate in the shoe that prevents the shoes from creasing.

So like when you're not wearing them, you inflate this balloon mechanism in the shoe and then it prevents creases. And so it was E commerce. Yeah. Nice. Yeah, I mean for me it was similar.

Like in the beginning I started working for free for businesses, but for me it was like local businesses. Because I had no idea how to get to E commerce clients and how to like get in contact with them. So I basically just reached out to like local businesses in my area, like gyms and stuff like that, and just offered to run their ads for free. And then later on, like got into E commerce. Yeah, you know what, for me it was a little bit of luck.

So I mentioned I was a developer, when I was doing software engineering, I dabbled a little bit in like coding NFT projects, if you remember, when NFTs were a thing. Yeah. And one of the dudes that I was kind of working with ended up starting an E commerce project. And so like, it just kind of worked out that way, that as soon as he needed someone, I was like, willing. I was like, dude, just let, just, I wouldn't, don't even pay me, like, let me do it type of thing.

And so it worked out that way. I, at the same time I was like, for months I had been like dabbling in the E commerce community, connecting with people on LinkedIn, you know, consuming the content on Twitter, etc. So like I was in there, and as soon as I was attending events, but but yeah, I also tried some local stuff, but like I was, I sought out, That's so funny. I was so desperate. I sought out like, yeah, like not chiropractors, what are the guys called?

Ortho, Orthopediatricians. And this one guy responded to me and he was kind of skeptical and I'm like, dude, I'll pay for the ads. Just pay me if like people book or whatever. Right. And I had no money at the time, but I did it anyway.

Just like, I need, I need to do something. You know what I mean? And that was that, that was, that didn't last long. But yeah, like I said, I think I, I had always had my sights set on E comm. I.

Even when I was like years prior, like learning meta ads on the side and stuff like that, I was always kind of doing it within the E commerce realm and that was always kind of my interest. So yeah, I kind of positioned myself in a bunch of communities and so it ended up working out that I was able to like kind of like break in if that makes sense. Yeah, yeah, yeah. Nice. And now going a little bit more to the, to the E commerce side of things, like what do you think are the most common reasons that you, that you see brands plateau at like 200k, 500k or even like 1 million per month?

11:53 - Why brands plateau

Yeah, the thing is there, yeah, I mean all of those. It's like, it's so, it's so nuanced. Right? A 200k a month brand versus a 500k versus a 1 million. They could all have various reasons as to why they plateau.

I can start towards the let's say the $1 million a month because this is something I deal with with some of the brands that we're working with now and obviously still varies brand to brand.

12:22 - Messaging diversity at eight figures

One of the biggest levers that we're trying to pull right now for eight figure brands is is, is messaging diversity. So if you have a. Let's. So, so let's say we have a brand that has been able to scale to a million dollars a month and it's really based on let's say one, main angles slash Personas, slash messaging points. And that's worked.

What we're trying to do now to unlock ultimately more spend at the same CPA is essentially trying to unlock like net new angles entirely that we can spend into. This is something that even honestly it's, it's, it's kind of something that I've been learning even in like the last couple of Months, which is like how I understand messaging diversity has even changed because what, what we've done for a while has been like okay, you have like this like main angle and like let's try to find new sub angles within that that are different to the main talking point but still within this umbrella, or like theme of messaging. And we would find winning ads, let's say doing that. But, and they would take spend. So in one sense they would be winning but they wouldn't allow us to spend more.

So typically so we, we operate within a CBO structure. Everything just goes straight into their one campaign. Typically I, And so we would have ads that do take spend but they weren't able to necessarily enable more spend on the, at the campaign level. Overall, And so it's like, oh, we are doing messaging diversity, we are doing new angles. And that was true, but basically they just weren't diverse enough.

Whereas now what we're trying to do is like unlock completely unique net new angles unrelated to what the main thing was to try to essentially open up new pockets of spend, if that makes sense. So that's, that's a massive one. And that's the most sophisticated one I would say. What I find the lower, the smaller that you get, the problems become a little bit more from what I've seen at least it's a lot less like there. It's almost like a lot of times it's not even real constraints, it's just a lack of understanding, it's a lack of data interpretation.

14:53 - The smaller brand's problem is reading the data

So, I talk about this a lot. I've written about this quite a bit where like I have operated within businesses where we were able to like basically double the business without doing anything. It was like the problem was that the brand owners didn't know how to interpret the data. They didn't know that. Oh, wait a second.

So if I look at as if I'm just profitable, on you know, first order and I have margin on that first order, I can just continue to spend and my contribution margin will grow. Even even though my em might come down, let's say I still have contribution margin on the first order so I'll still make more money. Or another one that that has come up has been them not realizing that their meta spend is driving revenue and profit on Amazon. And so they interpret Shopify data in a silo and they're like, oh, well My every time I increase spend my Shopify sales aren't increasing with it. But it's like you don't see like, oh, it's actually happening over here.

And then when you look at it all consolidated, it's like, oh, actually we can like double spend and we'll be fine and we'll make more money. That a lot of the issues in the lower range and to it's I find tend to be like data interpretation issues less than like we need to do something different if that makes sense. I know it was a lot, but yeah, that definitely makes sense. I mean we had that happen with with a few clients already that we for example doubled the budget on Meta and on Shopify. Like sales obviously go, go up, but then the client reaches out to us and is like, damn, like Amazon sales are going crazy.

Like, did you change something? Are we sending traffic to Amazon? But it's just like the byproduct of, of, like, spending more on meta, because obviously you reach significantly more people, you get way more awareness. And a lot of people just search for the products on Amazon. Yep, 100%.

Yeah. And then if you consolidate the data, you realize, oh, we're actually doing. We're killing it. If you just look at Shopify, it's like, okay, this looks bleak. Turn it down.

But when you consolidate, it's like, oh, we can actually sustain this and maybe more. Right? So, yeah, yeah.

17:12 - A lower ROAS with more profit

And the other thing with the, with the MER going down, but the contribution margin going up, that's. In the end, it comes down to, balancing, like, efficiency and absolute returns. Because obviously, like, the efficiency of the ads might go down, but, the absolute returns and the profit that you're making is going up. So it still makes sense to scale. Yeah, yeah.

It's crazy because to us it seems really simple, but I've, like, watched time and time again, people, like, it's very difficult to grasp for some reason. And so it almost feels kind of like, not unfair. What's the. I don't really know, but it's like you come into a business and they pay you money and you're like, oh, actually we could just turn this up. And that's it.

And, it's almost like too simple, too easy sometimes. But, but, but then, but then you look at how much time it takes to explain this to founders and like, oh, okay, you really, really don't get it. Like, if I was in here to show you this, you would not have gotten this. But yeah, it's. It's surprising how hard it can be sometimes for founders to grasp that.

But, yeah, yeah, I mean, I get on calls with founders on a weekly basis where I ask them, like, what, the goals are, and they're like, yeah, we want to get a 4x return ad spend, we want to get a 5x return ad spend. But they don't even know really why. They just estimate that that is, like, a good number that feels comfortable for them. And that they're like, profitable if they get that. But if you dig into the numbers, oftentimes you can go way lower in terms of return ad spend and you will maximize the.

The profit. Yeah, exactly, exactly. And I found it really interesting what you shared about the creative diversity, like, especially for bigger brands, 1 million a month.

19:06 - Personas against angles

And plus, does, that kind of go hand in Hand with unlocking different, demographics, different target, like, different buyer Personas, because I know that when we look at a brand, we oftentimes come up with, like, three, four different buyer Personas, that we want to target. And then for each of those Personas, we want to have different messaging, different angles, and they have, like, different problems that we need to address. Does it go hand in hand with that, or is it a different approach? I would say it's definitely related. And these terms, everyone has different ways that they define these terms.

So when I hear Persona, I obviously think of, like, a profile of a person, and, an angle, can be attributed to technically multiple Personas. Like, you could be speaking to a pain point in different ways to different people. Right. I'm trying to think of an example that I can use without, without outing one of my clients. So let's say.

Okay, so here's a good one. So let's say, like, the pain point is, a weak core. Right. Obviously the way that you speak, to a solution for that pain point to a mom or to a college student will be slightly different. They have different contexts.

Right. What we're finding, though, is the. So let's say if I. If I'm positioning a fitness product, as a solution to, let's say, the pain point, which is a weak core, I could speak to that different. Like I said, you speak to a mom differently than you would a college student.

To me, those are what I would call kind of like sub angles. Right. So it's like, problem is, we core, solve a weak core, for whatever. Like when you're, so that you can be, a better mom at home with the kids, or, like, so that you can, whatever, look good if you're. If you're a young guy, college student.

But, yeah, I'm trying to think here. Hold on. I want to make sure that I say this correctly. So those would be potential sub angles under an overarching pain point, which works well to find winning ads. What I'm finding, though, is when we plateau, the sub angles basically don't cut it anymore because they all kind of fall under one overarching theme.

And. And because a Persona is, tethered to, like, the we. The way that we categorize our ads is not by Persona. That's the thing. We categorize them by angles, Slash messaging themes, if you will.

And so, you could be doing multiple Personas under the same messaging theme. I think that you'll still hit that plateau in certain cases. What we're trying to do is unlock completely new themes. You might still be speaking to the same Personas technically, like the same type of people, but you're addressing a different problem or desire entirely from, in this case, let's say, a weak core.

22:23 - Lumps and bumps, or longevity

Does that make sense? It does, yeah, definitely. But can you give a good example of, like, what, for you, a theme is and then what the sub angles are? Yeah. So, okay, so I'll give you an example.

So, in the pet niche, for example, certain, dogs. Dogs get, have this. Or senior dogs get this problem called, benign lumps. So sometimes dogs will develop lumps on their bodies that are, benign. They're not gonna kill them, but owners don't like it.

They want to get rid of them. And so, there are various sub angles under that. Well, okay, you could speak to the fact that, you know, let's say you have a product that solves this, a supplement. You could position the product as an alternative to, let's say, surgery, which is usually what dog owners would. Would have to get into to be able to get those removed.

You could, speak to. You could position as an alternative to that, you could, or you could even just, inversely or similarly, kind of like twist the knife on the pain points around surgery. And so you could be like, not just, hey, do this instead of surgery, but hey, don't do surgery. Because anesthesia, can be harmful to senior dogs, or it's super costly, et cetera, et cetera. To me, those are all sub angles under the overarching messaging theme, which is lumps and bumps, benign lumps and bumps as a pain point.

So even though the messaging and what you're saying is different across the ads, there's still under, to me, this one big category of lumps and bumps. And so what I'm saying is a, completely new messaging theme would be like, longevity. Here are ingredients that whatever, that can help your dog live longer. And now that is the new desire and pain point. Those are completely separate categories of messaging.

And you might have Personas that overlap across both. But what you're. What you're. Yeah, the overarching theme of what you're talking about is completely different. So that's what I mean.

So it's like within an overarching category, you can be diverse in what you're saying, yet still be under this really big umbrella. And so it's like what we're Trying to do now to kind of break plateaus for the bigger brands is like open up new umbrellas entirely of messaging. Okay, so would you say the overarching angle is always about one specific pain point, one specific dream outcome, and then basically the sub angles are digging deeper into that pain? Exactly. Yeah, exactly.

Yes.

25:04 - Why twenty ads can still be one message

Yes. Okay. Okay. Interesting. Yeah, because I definitely noticed that there are a lot of brands that have one main angle and then they attack that from 10 different approaches, 10 different sides.

And then at some point they just can't scale anymore because that's that angle is that or the, the audience pool that they're like getting with that is just too small to, to scale. Right. Yeah. Or like, I mean like the way that it usually we see it comes out is like you just can't spend more without, things breaking. And the thing is there are other, there are other levers that you can pull.

Right, but are there other levers that will allow you to, you know, increase spend by 50%? I feel like this is one of the few that like, will allow you to do that. There aren't many things that you can do to like, go. There aren't many things that will allow you to go from, you know, 10k a day to 20k a day within like a month. Like unlocking completely new buckets of messaging to spend into.

And the nuance is tricky because like I said, you could be, you could have 20 different ads that technically say different things. And so it feels diverse, but then when you kind of zoom out, it's like, oh, these are all under the same overarching theme.

26:25 - Is Andromeda actually a change?

Maybe that's why we're not able to actually kind of break past where we're at now. Yeah, yeah, yeah. And have you, have you seen that it's now even more important with like Meta's updates, Andromeda and all the diversity that they're pushing? No, not, I don't, not, not particularly to be honest with you. I don't, I don't super buy into the whole Andromeda thing in the sense that like, like, I don't think anything actually changed in the last three to four, five months.

I think, I think messaging, like creative volume, creative diversity. Messaging diversity has always been a thing like for the last like two years. I think LinkedIn and like Twitter just caught up like seven months ago. Not to say that nobody else was already on that, but I'm seeing like the, the zeitgeist of like, you know, DTC meta marketing is now kind of like catching up to that. But, I.

It's always kind of been a thing and I think, yeah, so I don't, I don't have, I don't attribute any particular significance to like the Andromeda stuff. But in general, like it's. Yes, I think it's just, it's another way of saying you need, it's another type of diversity that you need in the ad account. It's also one that not a lot of people talk about. Everyone talks about volume and creative diversity, but not a lot of people talk about messaging diversity.

Because you could have again, 50 to 100 different ads that all look visually unique but are all ultimately saying the same thing. That's not, I'm not saying that's not valuable, but I think as a lever it runs out in terms of contribution, at some point. Right. Like it's not like endless, I don't think if that makes sense. Yeah, yeah, it does.

And interesting that what you're, what you're saying about, about Andromeda because I definitely think everyone is, is going crazy and putting too much weight on it. But on the other side, we noticed a change like I would say between six, seven months ago or like four. Yeah, four to six months ago that definitely the brands that had like less creative diversity also like messaging diversity and like more just the same. Yeah. The same angle, the same creative from like different sites that they had a drop in performance and that the brands that already had more diversity, that those brands didn't have that drop.

And I think it was always important, but from, from the data that we are seeing, it definitely is now even more important than it than it was before. Sure, yeah. Now for you, like, what does creative strategy mean? And how do you come up with a creative strategy for a new brand that you work with? Man, that's a big question.

You might as well ask me what is the meaning of life? No. So what is, what does creative strategy mean? What does creative strategy mean? Well, I think that, I mean, I don't know if I could define it in a one liner ultimately, but for us it ultimately comes down to, I would say there's kind of like three stages, right?

29:40 - What creative strategy actually is

So there's there's the research phase and trying to understand, trying to understand the consumer that you're targeting. What are their, you know, real life experiences, real life pain points, real life desires. Transforming that data into like structured formatted insights and then converting that into tests that ultimately run in the ad account and then obviously having a feedback loop. So like, so research, you structure insights, test, you know, find a way to prioritize those insights. How, however you want to do that, start testing and then feed that data back in the performance data back into that cycle, to then inform your next tests.

At a high level, that's ultimately what it comes down to. Now you could obviously you could, you could like, zoom into any one of those stages and there's a whole other mat, like branches and branches of various skill sets and things to do. Like within, you know, deploying ads is like, you got copywriting and video editing and static ads design. And those are all conversations of their own. But at a high level, that is how I would describe the creative strategy process.

And that is, that's our process. I'm at a high level. It's like raw research format insights based on what, we know as creative strategists, figure out a way to sort of prioritize that, those insights. By that I mean, like, you could, you could come up with a hundred different insights from your research, but which ones are you going to test first and why? Right.

And then obviously once that gets ready, that goes into a brief and then the ad production process starts, which is a whole other thing.

31:37 - Where the research comes from

But, yeah, that's what I would say. Yeah. And what would you say are some, some overlooked sources in terms of getting those creative insights? Especially in the beginning when you, when you do like the research. Overlooked?

Well, I think everyone already kind of knows that, you know, Reddit reviews, are, are some of the biggest ones, forums and things like that. I think, social comments is obviously a big one. It's a lot harder to find those. Excuse me. To find those.

I don't know that the, it's, it's funny. This is something that even that we're kind of, thinking through more right now, which is like it, it. No one has a problem with pulling data. I think everyone but we, But I think we're all still trying to figure out, like I said, it's the second part after that's like, how do you actually prioritize that data? Strategically, meaningfully.

Meaning like anyone can like pull up cloud code and be like, hey, go find me information about this. That's going to pull you 100 different sources and this, that or the other. But then the question goes, okay, so what are you going to do? Just test 150 insights across 4,000 ads? It's like you could, that's super.

Inefficient. But you could, I guess cost cap guys might argue that. But I think the, the real skill is like taking all that data and then figuring, okay, what, what do I have the most confidence behind and why? What's his name? Eugene Schwartz, talks about mass, desires.

So like that's an example is like how, you know, understanding psychology, consumer psychology, for example, to be able to be like, hey, okay, we have 100 these insights. Which one of these, associate the most with some mass desires that I'm aware of? As an example, I think so that process is probably the most overlooked one. But yeah, I would like, no one has a, we don't have a problem these days.

33:43 - Volume against quality

I don't think anyone has a problem these days with pulling data, I would say. Does that make sense? Yeah, yeah, definitely. And how would you say, how do you balance like creative volume with like quality, like really prioritizing and figuring out what works. And like, if a client, if a, if a brand asks you, like, how many creators should we test?

Like, how do you approach that? Yeah. Oh, those are two different questions. So, so how do we balance creative volume? And then if a brand asks me, so I guess I'll answer the second one which is like, I guess it depends on.

Obviously this assumes that creative is your problem. Right? Because creative isn't your bottleneck. Meaning like, and I don't just mean like the ability to produce creative, but like some brands, like, the reason they can't grow has nothing to do with creative. Like we said earlier, like, you have, your ads are fine.

You just don't know how to interpret your unit economics, you know? Right. Assuming that creative is your problem, how many ads should you test? I would say, I mean, there's probably a point. There's probably is too much, I don't know what the number is, but like as much as you can.

I would say I, I, I have no problem. And obviously it also depends on your, your, your account structure. Like if you're running aboard. So yeah, so like I would say when I say as much as you can, I'm operating within like a highest volume CBO mentality. If you're doing ABO and you're allocating budget to each individual test.

I mean that, I don't know, I don't, that's harder. I'd have to think about that one more. But generally speaking, as much as you can. I mean, we're, for some of our brands we're launching. Let me think anywhere between, 30 to 40 concepts a week.

So multiply that by three, whatever, 100 to 120 a week. And those are, those are accounts that are spending, let's say between 8 to 10k a day type of thing. So it's, it's a lot for the. They're not spending that much. And I think most people say that we're, we're, we're testing a lot, but it's not like every single individual concept is getting spend.

Right. Yeah, so it's, it's a tough question. I don't, I don't know that there's a, there's an answer. Like, how much? Exactly.

I don't have one. At least. And then your other question was, how do you b. How do you, how do you balance volume with quality? Yeah, that's, that is, How do we balance volume with quality?

The thing is, we built like for the system that we've built for producing our ads.

36:37 - The quality floor

Quality kind of is a constraint within it, meaning you could only produce ads that would pass our quote unquote, quality check, if that makes sense. So we have a certain standard, like, okay, and this is just us especially because like, you know, ultimately, like, the brands are our clients. And like, if brands were like, if, like, if every brand was okay with AI slope, then sure, we could probably get out a ton of ads, but like, they're not and I get that. Right. So, we, we have like this minimum standard for.

And this is purely like, aesthetic as well, that I'm referring to right now. Like, I would say we have a, somewhat of a minimum standard for like, hey, like our ads need to pass a third, a certain threshold and if they don't, they're just not getting approved basically. So yeah, I don't know if that answers the question, like how we balance volume and quality, but like, we kind of built it in as a check into our system. So I don't want ads going out that we're not going to be proud of. In one sense, that doesn't mean that they're all beautiful, but it just means that they're not all.

Like, it's not like, I honestly think like, the big, the new low is just like, just like pure AI slob. If it just looks super, obviously AI slop, it's like, yeah, for us it's just not going to pass. That doesn't mean that won't work, performance wise, but it just doesn't pass. So does that make sense? Like we kind of implemented like a minimum quality that you have to pass.

Be able to, get it out. Yeah, yeah, it definitely makes sense. And when you come up with like completely new angles and you're testing, do you still like, even though you have like the quality threshold that they have to pass, do you test them like more dirty and quick just to see like whether it works or not, or do you still invest like quite a lot of time into, into making sure that the quality is, is right? So we do, I would say we do everything pretty quickly still. So This is what I'm saying.

So it's like, it's all, it's like we're in one sense we're maxed out in terms like we work as fast as we possibly can already, so every and everything flows through that, if that makes sense. So it's like. No, so, we don't, we wouldn't, we don't necessarily drop the standard for like, stuff like newer tests, but that's okay because we're already extremely fast. Does that make sense? Yeah, yeah, yeah.

And is there a specific type of creative that you, that you like to start with when it comes to like testing new stuff?

39:07 - Why they build around static ads

Definitely. So, so we're, as an agency, we're predominantly focused. This is a good context to give. So I mentioned that we're performance creative, but we're, we're not only performance creative, but we've kind of, our huge focus. And all of our systems on the ad production side are built around static ads primarily.

And so the, a big part of the reason for that is because we were able to like build a system that kind of does everything that I just said, ensures quality, but volume as well. And the biggest bonus is that to answer your question is that it enables super fast message testing because we can pump out, you know, 25 net new unique concepts, not variations Net new unique concepts under let's say one messaging theme, maybe even write a listicle. And then boom, we have this super full test for this messaging theme that just, that just went out within a week. So, so to answer your question, and like we just, we have not figured out how to do that on the video side yet. It would just take a lot longer to do that much, and maintain a certain standard of quality.

So, so yeah, so statics, to answer your question. Yeah, definitely. And, and when you figure out like an angle or static ads that, that work the best, I assume on the one side you make just a bunch more of, of like Ads that are like creatives that are targeting that, that angle and hitting it. But at the same time are you also then doing like other types of creators for, for that. Oh definitely.

Definitely. Exactly. So and this is what I try to tell new brands that were, you know, coming to work with us, which is like we're not necessarily designed to be your your full creative team. We're more so there to provide volume and help you double down, but really to unlock messaging and then double down on that messaging. But once you have that insight and you know what's working, it's super prudent to.

Okay, now let's start investing in creators high production content, like podcast ads or, or VSLs and this out of the other. So we and so we're not the ones that go and do all of that but it's definitely, that's the right move. It's like once you unlock. Okay, cool. We know this is a whole new segment of, of messaging that we can tap into.

Then it's like drop everything else. We're just doing this and put all resources here type of thing.

41:39 - How you know an angle is working

Yeah, yeah. How do you really know that an angle is working? Like obviously positive like contribution margin. And I assume that you're able to scale it to some degree. Well this is where it gets tricky, right?

Because so let's, so let's say if you're an eight figure brand and you're spending 10 to 20k a day on one angle and again let's assume for the hypothetical scenario you're running a cbo. The problem is this is a kind of a question of attribution. Like how do you know that the angle is working? Well you said contribution margin dollars, but that assumes that you can track that at the ad level. Slash ad set level.

Debatable as you know. Yeah, I, so here's my approach. I and again, this is just how I look at it. I for the most like attribution to me for like for the most part is just not real. Even with tools like triple wheel, it's directional at best.

But then if that's the case, like I'm like so it's directional. So basically you can't really know. And so the only other data that you have at your disposal is basically the ad account data. And so the way that we do it is we know for example that okay at this in platform cpa, we know that our true CPA is X. Cool.

So now Y being our in platform CPA is the benchmark for in ad account numbers, Y isn't necessarily real or true, but it's tethered to a true value. And so it is now a benchmark in the ad account. Right. So to answer your question, so if our in platform CPA benchmark is $80, if new ads going to new Lander PDP are getting significant spend and matching or beating that in platform cpa, then it's a winner is how we would do that. But are you still looking at the in platform CPA on a whole account level or on an ad set or ad level?

What do you mean? Because sometimes, like you mentioned, you have this benchmark for what the in platform CPA should be like. Yes. And then you launch a new set of ads and they get spent from meta and for the new ads, the CPA is not hitting the benchmark, but you notice that the whole account is still hitting the benchmark. That's interesting.

That would be an interesting. If that were to happen, I would, I would not write off those ads. If that were, if that was like demonstrable, that you could very clearly see that that was happening, then I would be like, okay, cool, this, this is doing something positive.

44:24 - An angle that surprised him

Let's keep it. I haven't had that happen, but I, but yeah, I would, I would. If that did happen, I would agree. Like, no, that's. That's probably a win too, for sure.

Yeah. Yeah. Do, do you have an example of an, of an angle that like crushed it and like surprised you? Like you didn't expect it to work, but it just worked. Let me think.

I, I think that there are. Yeah, there are, there are definitely some. Like, for example, for, for very. For a fitness brand that we work with, we test. So I.

Generally when you think about fitness, obviously the most like, the strongest desire that you can tap into, generally speaking is going to be something related to weight loss. And so that's something that we kind of have been testing more around without making crazy claims. But. But it's interesting when even testing things like speaking to like people that have like, lack of commitment or kind of like the, the typical scenario where someone commits to some type of workout and then gives up on it in whatever, in a couple of days, stuff like that. That seems to me at least, yes, it's obviously things that people go through.

Is it a strong enough problem that will provoke, someone to purchase? My initial thought would have been like, no, but there are angles like that that end up, popping up like, oh, wow, that's actually, that's actually working. Cool. Interesting. That's one I could think of, I guess off the top of my head.

Yeah. Do you ever get brands that they, that they don't want to test a certain angle that you, that you propose? Not yet. Not yet. I don't think that this happened.

We definitely, we obviously get brands that you know, will give us guidelines like hey, like you can't, for example, you can't mention other brands. And I understand that it's like, okay, cool, we're not gonna. But no, that hasn't for the most part been an issue on our end, I would say.

46:42 - The biggest misconception about performance creative

Okay, what would you say is the biggest misconception that brands have about performance, creative or what do you wish they like brands and founders would understand more about it? I think people, it's not even just brands and founders. I think it's everyone that's not a performance marketer. They just, people just don't understand the difference between direct response content that converts and great looking ads ultimately. And I think everybody knows that in the performance world, but like People just outside of, you know, this, this world don't just don't get that part as much, which I understand.

It's, it's a little weird. It can be weird. It's like you look at a winning piece of content and most people be like really? Like that that's driving you millions of dollars. It's like, yeah, that's probably the biggest hurdle that most brands need to kind of get over, at least initially.

Yeah. And how do you balance that when you talk to founders? Like are you 100% direct response approach? So we're like, yeah, for the most part, yes and no. So, okay, so the more recent systems that we've developed, on the ad production side allow us to produce stuff that it's actually very interesting.

Like what AI has allowed us to do is like produce visuals that otherwise would have taken hours to do. But now it takes us seconds. So like we can, it's kind of cool. We can. Lately we've been able to kind of still do very direct response messaging and copy, but with like very intriguing visuals.

And so lately we've kind of been, been able to make ads that like, I'm super proud to like put up on a website because it looks great, but it's still super direct response. But, but, but generally speaking, yeah, no, we're, we're, we're. I'll Never. I'll never sell that in that way to a brand. Like, hey, like, we're here to like, perform and like, we're not here to make stuff that's pretty.

And. And usually the way that I'll. And this is with anything explain that to a brand is like, it's just with data. It's like. Because once you see the numbers, it like, you can't really argue with it.

It's like, hey, here's an ad. Do you like this? No cool, dude. Like, this is performing the best. Do you like this ad?

Yeah. This is losing you money and then they get it. You know what I mean? And so that's usually how, we go about that conversation. Yeah.

Yeah. What do you think about people that say that only ugly ads work?

49:27 - Do only ugly ads work?

Because I've heard that a few times. Yeah, it. I'll be. I mean, I tend to agree. It's not.

I mean, only is obviously a very, a bit too dogmatic. But generally speaking, like, if you took all the ads in the world on meta and you categorize them like all the winning ads in the world on meta, you categorize them by ugly and not ugly, there'd probably be more ugly ones than not. And why do you think that is so? I mean, I think it gets nuanced because you could have a different conversation about statics than you could about video. Let's say on the video side for sure.

On the video side it's easier. And, and I think al. Like, I'll give you an example. Like, we have an account where we've got fake podcasts that look great, direct response scripts still. We've got VSLs, and all these are working and taking spend.

Don't get me wrong. Fake podcast, VSLs, founder ads, static and all that's great and working. And then this girl pulls up in her car, sets up the phone, talks for like a minute, zero editing, no B roll, nothing. And it just boop straight to the top of the ad account. And, and, and like, why is that?

I don't know. I think. I think there's an element of like, I think it's probably just authenticity. It's like you. I think we just easily.

There's like a filter. And you know, Barry Hot talks about this like there is a filter in our brains that like, it's like an ad filter. It's like if you get the sense that you're being advertised to, your initial instinct is to like, swipe right. And obviously if you can, you know, speak super well to a pressing knee, then you can get past that, but you have that working against you. But a super ugly video of like just someone just talking in their car, it's like that passes that filter.

Eventually you might realize it's an ad, but you've already made it past the initial filter. You're already five, 10 seconds in, you know what I mean? So, so yeah, I read something recently too, which I think is true, which is that a lot of the time because like this even extends to like landing pages where it's like a lot of the time it's not even that, it's, it's that it's ugly. It's that when you're trying to make it pretty, it's sometimes even just makes it harder to read. Or it's much less like scannable.

This is more for statics and landing pages. Like if the, if the static ad is just plain, it's like, yeah, no one loves to look at it but like I know exactly what it's saying in a second versus if it's like super nicely designed.

52:13 - AI-generated UGC

You've got these nice fonts and it's like three different fonts and like they're all spaced differently. It's like it looks great but it probably took me a couple extra seconds to read it. And I bet that plays into it as well, you know? Yeah, yeah. And talking about the authenticity, what are your thoughts on AI generated ads and more specifically AI generated UGC content and video ads?

Well, I mean it's already like you could like. Yeah, so it's so close. I mean it's already there in many ways. It's obviously not perfect. But if, and this is kind of just kind of goes with AI.

In general, it's like the capabilities are actually already there. It's just that people don't know yet on mass how to leverage them, how to extract them. Meaning like Sora or Gemini can create pretty realistic looking videos. By realistic, I mean like UGC looking type videos. Now a couple of things.

Do you know how to prompt it? Well, do you know how to give it a right reference, starting image? That's like level one. Can you get past that? Cool.

But then the next problem that you'll usually run into with like Gemini or Sora is and right now the biggest giveaway on any AI video is not the visual, it's the audio. It's like when they're talking, it's funny that they haven't been able to fix this yet. I'm Sure. I don't think we're far from them fixing this though. It's like you can very clearly tell like there was something weird going on with the audio and it usually comes out in the voice.

But and I've seen this, if you're sophisticated enough and you know how to first replace the dialogue with something from 11 labs, now you have something a little better. The problem though, and I didn't mention this, but prior to being a software, I was an audio engineer. So a lot of the things that we would do as an audio engineer was like if you worked on a film, let's say, and it was a scene of two people talking in a room. Sometimes what will happen is an actor will get lines wrong or the, whatever they, to change the lines. And so they'll go back into a studio afterwards, re record their lines and then the engineers, the audio engineers will place it back into the scene as if they're saying it then and there.

Part of what they need to do to make it feel real is because if you just take a professionally recorded, you know, script and just plop it into a dialogue in a room, you're going to know it's off because like this guy's voice sounds like he's in the room. The other guy sounds like he was recorded in a professional studio. So, so what they do things like they will muddy up the voice a little bit. They'll add like reverb, really short reverb to the voice to make it sound like he's in the room with the other person. I say all that to say if you're sophisticated enough and people do this, you could do the exact same thing on ugc.

Like replace the voice with a voice that doesn't sound odd, add room tone to it and now you have something that feels like incredibly real. So what are my thoughts? Like it's already there. It's extremely usable.

55:46 - The ethics, and where this goes

It's not perfect, but it's pretty close. And that's both cool and obviously kind of scary. But Yeah, yeah. How do you see that from like a moral standpoint and also like long term if we imagine like, I mean I don't think it will take, take a long time but like in three to six months. But every brand can pump out like realistic looking UGC content that is just completely fake.

Yeah. So what's the question? How do you see that from a, from a moral standpoint first of all? And then like where do you see the, the industry go over the next like three to six months. Yeah, yeah, yeah, that's a, I mean that's a great.

I think, I think, I think, like, ethically, like, as it depends on what's being said for sure. It depends on what's being said ultimately. Like, if you're, There are probably certain scenarios where you're. If you're like straight up fabricating like. Okay, I'll give you an example.

Like if you create a ugc, a video of. And the thing is all the same things that would probably be wrong in, you know, normal UGC would be wrong with AI. Like, so, like if you create a video and it's like a normal dude, but you give him a lab coat and you're like, hey, My name is Dr. Mike. My patients.

Here's a, Here's a study that I did on my patient and this is what we found. It's like all fake. It's like whether it's AI or not, like, you probably shouldn't do that. Right. Outside of that though, it gets grayer, I would say.

I mean, because ultimately it's no different than normal ugc. It's no different than a normal person making claims, an actor making planes. It's just that it's not a real actor. So I would say any qualms you have with let's say using AI UGC to write ads, you, to be consistent, you'd probably have to apply it most times to just normal ugc, if that makes sense. Yeah, I will say, on LinkedIn, I.

I made a post on LinkedIn and I took a picture of myself sitting at my desk and then I regenerated it to put like the Flash costume on me. And and then when I posted it, LinkedIn added a little badge to the top left corner saying that this was AI generated content. And I honestly think that's where like that's what we're going to have to end up doing. Because otherwise, like, and I don't even know if that's going to be a true fix, but like, otherwise how, like it's just going to get too crazy. Literally you just have no way of knowing anything that you're looking at is real.

Yeah, definitely. I think Meta will implement something like that. Obviously there will probably be ways to circumvent it and to go around whatever. But yeah, otherwise it will get messy. Yeah, for sure.

Awesome.

58:51 - Where to find Sebastian

That was cool. Where can people follow you, your work or get in touch if they just want to scale with better creative? Yeah, yeah. So we're at GrowthPod CA. You could find me on LinkedIn.

Sebastian Valiente. I'm mainly active there. I do have Twitter, but I'm not as active there. But, can hit me up on LinkedIn or you can go straight to GrowthPod CA and book a call. Awesome.

It was great to have you on. Thanks, man. I appreciate you having me. Yeah, then talk soon. Okay, bye.

Bye.