Creative Diversity Explained: The Who/What/How Triangle Driving Our Top Scaling Ads

The Scaling Lab-Tried & True Media: Welcome back to another episode of the Scaling Lab. What an exciting episode. We've got John in the house, one of our top creative strategists to go over things like the triangle and creative diversity, what it really means and how you should be iterating creative to not tank performance in your account. Without further ado, my co-host with the most Kira Hug and I are jumping into this episode on creative strategy.

Let's do it. So before we dive deep into creative strategy, John, I just want to know like where you came from with this strategic mind and a little bit of history before we go deep into what you're doing today at the agency. So can you just go back in time a little bit and give us a couple of highlights of like where you've been to hone this ⁓ creative strategic mind? Like what's helped you get to this point so far?

Well, that's a good question. So. Probably the first starting place, copywriter by trade, right? That's where it comes from.

⁓ You know, there's the whole, the whole stick where everybody says to hand copy proven winners and stuff like that. Kind of build up your chops here on ending a copy. Yeah. So I did that and that was cool, but I was more into taking the proven winners and I would annotate them, right?

Pulling them apart, breaking them apart. A little bit less concerned with like... the language and the copy smithing and the stuff like that. And just asking why, why this and why this and why this and why in this order, that kind of stuff.

So it started there probably. And it's just kind of always been that way of like the pull stuff apart and kind of ask why. Not just be the guy to do plug and play stuff, but probably the, the biggest insights I've had with strategy, if you want to call them insights, has been working with AI, right? And so maybe a lot of stuff here will kind of always circle back to AI.

Right. ⁓ because we work in a, a fast paced business, creative volume is a big thing, right? ⁓ always trying to turn out more faster, ⁓ and speed is the enemy of good. Right.

And so it's always like, ⁓ I wish I had another me if I could clone myself. Right. And so with things like GBT or Claude, you can clone yourself, but you just have to teach it. And so in the process of trying to teach it, the different things that I do in my workflow, you have to really think deeply about it and pull it apart and like codify stuff to understand it.

Right. And so a lot of what I do today with writers in the agency, ⁓ spins out of that a lot, right? Like getting deeper insights on strategy from having to pull your brain apart and teach it deeply to someone else because you hope they can be your clone. So you were probably the kid, like the little kid who's asking your mom, like, why?

Like, why is this, you know, those little kids are just like, well, why this? Why is this this way? I can see that. the way it really started really was I'm a martial artist my whole life.

And when I got into that, it was the whole, like a really deep dive into it. It's like when I get into something, it's all or nothing in a sense. My wife knows this and she kind of gets like, oh, this look on her face. If I get into something new, cause she's like, he's gone, right?

He's not just going to like happen. He's gone. I'm not going to see him for a couple of months because he gets deep, deep, deep, deep, deep. I've always been that kind of person.

I like to pull things apart. Do you have a background in engineering? Like before copywriting, you were saying you like to take stuff apart as a kid and like, I'm sure there was something going on before you got into like full-time copywriter. What was your background outside of copy?

Yeah, that's a good question. I bounced around a lot in college, never knew what I wanted to do really. Engineering was not one of those things. But like I said, was anything I got into, I like to pull it apart.

It was in the screenwriting and story. It's like independent comic book stuff. And so I would deep dive into the writing there. It's big into psychology.

Maybe that's a big part of it. I was a psychology guy. Right? So even in college, it wasn't my major, but I took psychology courses because I'm like, I like this stuff.

Understanding how people think, what are motivations. Yeah. Yeah. Yeah.

I think anyone who's really into copywriting and marketing on the creative side is definitely into psychology or at least knows they need to value it. Awesome. Yeah, I asked because the whole like taking things apart, dissecting things, rebuilding them, understanding how they work. So you can kind of create a framework, all that stuff.

To me, like, ⁓ sounds like engineering. And I think I've asked that question a couple of times and people say, yeah, ⁓ yeah, how'd you know? It's well read between the lines, but cool. Well, one thing I'll say to that real quick too, is anything I've gotten into, I've always gotten into it with the mindset of like, the things I've gravitated to, I've always naturally wound up coaching or teaching in them.

And so my whole thing was, you you always get challenged by a student a lot of times. I'm like, I need to have an answer. Right. And so for me to have the answer, I have to really understand the thing.

Right. And so if I don't have that answer, I need to go find it. Yeah. So before you created this strategic triangle that we're going to talk about, can you talk about kind of what led to that?

What you were seeing across different offers and accounts at TNT and kind of what led to the creation of this? Like, hey, we actually need. to do something like this? Yeah.

That's a good question. So probably a bunch of things kind of coming together all at once. ⁓ know, meta shift with Andromeda is part of it. think like a lot of people you talk to about creative strategy, like Andromeda was a big thing where supposedly the creative is the targeting now, right?

And creative diversity matters more now, maybe more than before. Like smaller iterations aren't necessarily supposedly. the thing exclusively. And so then it kind of came from a question like, okay, well, what is creative diversity then?

Like what, what, what do you need to feed the algorithm to look different or to be different, right? So it thinks it's a different ad. So there was that kind of happening at the time, a while ago. And also, you know, a shift with how I interacted with the writers.

We're looking at things, you look at the data, you launch an ad. And if it performs, ⁓ you're not asking the same questions that you're asking if it doesn't perform, right? So if it's at target or if it's at 3X raw as whatever, it's like, it's great. And you just kind of kid like, what's up next?

You try to scale it, you keep going. When it doesn't work, you ask questions like, well, why did you do that? And not this other thing, right? Because it didn't work.

And so the questions you ask are a little bit different and it's, you start getting answers like, well, I'm not really sure, right? Maybe. Sometimes, or, I did that because it worked over here in this other account. That tends to work in the agency.

So like we swipe those hooks, those are winning hooks. We swipe the winning format. And like, that's great. And that can take you far.

We spent a lot of money doing that. But like, when you start to go a little bit deeper in the context of we need more diversity, because that's what the algorithm says it wants. ⁓ And then also just doing a lot more post mortems and realizing, hey, you know, when we're in execution mode. trying to pump out a lot fast because we need volume, ⁓ you can often fall back on your intuition as a writer, right?

It's like you get a spidey sense for like, that I think that that feels good. ⁓ it makes sense to do, ⁓ and you don't really, you can't answer that why question, right? And so then when you are dealing with stakeholders and they want to know why and it's like, That's not really good answer, right? You're like, ⁓ it's because it's what, it's what worked before and we're doing the swipe thing.

I liked A more than B. Yeah, right. Right. Right.

And so like that's good, but it's like, what are you really testing here? Right. And it's like, so if you're, if you're running a hook test and you're testing proven frameworks for the hook, I've worked in the agency and one wins. It's like, great.

What have you learned there though that informs the next test? It's like, cause they, all of them were winners. They work across the agency. You know that.

Right. But how much data have you actually gotten to inform strategy moving forward? Right. As opposed to just let me do what has worked before, which is great.

And you like money, right? It'll make you money, but like there's more to it than that. ⁓ and so, so, you know, maybe I'm rambling here a bit to kind of circle back to, I think what your question was like, what were we seeing in the agency? was part that the data to what was happening in an ad, ⁓ and trying to figure out where did it go wrong?

⁓ what could we do, ⁓ as a next test to get it up to target, right. And finding that a lot of times the answers were just kind of like, I'm not really sure what I did in the first place. Right. there's that, but then also looking at accounts and saying like, Hey, we do need more diversity here.

And this is where it gets into some of the triangle stuff. Maybe it's like, well, what is diversity really? You know, a lot of the ads that we ran tend to fall into this one type of format. predominantly, or this one sort of topic if we're talking about like, let's say, nerve pain.

And there's a lot of ads about gabapentin here, right? And the majority have hits like, why are we still talking about gabapentin? There tends to be like a high cohort of gabapentin. There's so many other topics, which is an element in the triangle that you could talk about here.

So why aren't we doing that? And what could we do with that? Right? So just kind of all those things, I guess, at one time.

Yeah. And with creative diversity, can you just share where you think most writers or maybe even what you've seen with other brands out there where they go wrong with creative diversity? Like they think it's this, but it's, that's not actually what it is. Yeah.

So that's a good question. So creative diversity. So I've talked to a lot of different entrepreneurs and writers, different agency, different businesses. know, a lot of folks I talk to tend to lean on formats a lot.

It's like, I'm just going to test a new format. ⁓ And that's their diversification strategy and format testing is great. But if you're not changing who you're speaking to, or you're not changing your messaging, you're limiting your reach out there, right? You may even get ⁓ penalized eventually by meta for doing so.

You know, when I tend to look at a base level for diversity, kind of think of just general communications, right? It kind of boils down to three things and people that I've talked to at Metta, I've kind of said, yeah, it's kind of like how we look at it. It's who, what, how, right? Which sort of gets a little bit into what the triangle is, but it's very simple, very basic.

It's like nothing groundbreaking. Like who am I talking to? What am I saying to them and how am I saying it? Right?

And like with just within those three things, there are so many different things that you can do in terms of diversifying across each one of those sort of indexes. And a lot of people that I speak with ⁓ tend to fall on one or another, right? ⁓ They kind of sell themselves short, I guess we'll say, on diversity, right? Because this is the hot format and they get in this sort of ⁓ sprint of, just doing a lot of format testing and it's great, but you get stuck there and then you never test on the what or the who, right?

Well, there's a lot more diversity that lives in there. Yeah, one thing that to really quickly piggyback on your comment there, John, one thing that I see brands do, I was actually reviewing a brand last week that had probably somewhere in neighborhood of five to eight different people using the exact same script. And so it's like, they found the what, like what we're going to say, right? And now the who is, we're going to try and resonate with different who's using the same what by just, you know, leveraging different.

Avatar so to speak they were like real talent. They weren't like AI avatars, but you know It's like we've got our female who's middle-aged We've got you know ethnicity X X ethnicity Y and and that kind of a thing and and and I feel like that's kind of a failure like it can work But it's an oversimplification of I think what you're talking about of like who are we talking to it assumes that that what is gonna hit every who the same way and that's not true How do you feel about that?

Yeah, I know agree to 100 % so I think you know basically talent testing Avatar testing from the spokesperson, which you get that there. And, you know, we've had big wins, which is turning that dial, right? Like a script that's like on target, close to target, it's hovering around there for a while. ⁓ And you hit the right talent and the thing takes off, right?

It's like, maybe that like, that's the thing that you needed there for a particular reason. Like we've seen that you test regular Joe Schmoe type talent, man or woman. And then you just kind of put a doctor in that position. Same script, like you're saying, but it's the doctor delivering it and like the thing takes off.

there's that. credibility is inferred there through the doctor. And that's great. Then if that's all that you're doing, you're limiting yourself in terms of diversity, right?

So let's, let's say you hit with the doctor and the doctor takes off, the thing starts to scale. It's like, all right, great. So there's an indicator or a signal there that maybe the authority means something. So now you can change the voice, you change the script, right?

Which is another dial on the triangle. And so you can start to, if you're not thinking those in those terms, you really limit yourself in terms of, of your diversity. In terms of ⁓ the types of tests you can run possibly the scale you can get let's talk through the triangle because you've talked about parts of it now like who what how but I think you know for someone who's not familiar with it they may not be able to grasp the full triangle so let's actually talk about the triangle and the components of it ⁓ and Then we can kind of zoom in on different parts of it afterwards.

All right, cool All right. So so then the question is ⁓ like what is the triangle right? What does it look like? Why is it a triangle?

What are the points in the triangle? Why those points? Why are you missing other points? Potentially right.

So the triangle really at the base level is the who, the what, and the how. And that's nothing groundbreaking or ground-shattering if you look at a lot of creative strategy talk out there. People may have similar models, right? I think you've seen the who, the what, the how, the when, all this kind of stuff.

So there's an issue of semantics, like how I define it. But it's who, it's what, and it's how. Who am I talking to? What am I saying to them?

And how am I saying it? Right. And that's a simplistic way of putting it, but it's also, that's... There's a lot more to it than that.

And so it's a triangle because, you you pull things apart to codify them and study them, but the reality is there's bleed over. Right. So without getting into what's around the other edges, the triangle just yet, like if I change, ⁓ if I change ⁓ who I'm talking to, like what I'm going to say is going to change as well, probably. Right.

And so like, you have to understand that as well, right. Like as you change. In an ideal world, when you're doing testing, you isolate one variable, everything else is held in control. Yeah, that's great.

Right. ⁓ But the reality is it's not always going to be the case. Right. And so, what's around the other edges of the triangle?

Right. So in between the who and the what on that edge would be the topic. Right. So it's like, ⁓ who am I choosing to talk to?

What do want to say to them? But what am I saying to them with regards to what topic? So for example, if ⁓ you're talking to people with nerve pain, right. say, okay, I'm talking to someone with nerve pain, very superficial, generalized segments, old people with nerve pain.

⁓ What would be the insight? Like what's the new insight that I want to deliver to them? Right? The new thing, new piece of value or information I'm to give to them that's going to ideally ⁓ persuade them to click the ad, right?

Ultimately buy the product. What is that new insight? Well, I need to know, you know, what are we actually talking about though? When it comes to nerve pain, you can be talking about...

the cessation of nerve pain, can be talking about the solutions they're using. We talk about gabapentin, you can talk about capsaicin creams, you can talk about dismissive doctors, so experiences, right, or micro moments. So if I know the topic, I have a better idea of the what, right? So that's between the who and the what.

⁓ When it comes to the who and the how on that edge, it's how you say a thing, right? So what lives there is the speaker who's delivering. The actual script, right? And you can use a couple of different terms to define that.

like to use the voice would be sort of my keyword there. What's the voice at the point of view or the person's posture, right? ⁓ and actually give an example of that in a second. So then on the how edge, what we look at on the how edge is the format, right?

So here at TNT, we've got a bunch of like formats that we've proven across multiple accounts to be like, you know, ⁓ a solid chance or a higher chance at a strike. And so things like expert debunk, hidden anatomy, nope cast, those kinds of things. Right. And so, ⁓ as an example, expert debunked the default speaker there is just a neutral voice, a neutral tone, right?

They have no real, ⁓ point of view, let's say in the way that, ⁓ if you said, ⁓ I want to have a nurse, right. Who got fired from her job for revealing the secret about nerve pain, right? That's a very specific point of view. very specific voice.

So I want to be aware of who is speaking sort of like what's their tone, their voice, their point of view. It's good to know because I know if it's neutral and I get a win there, I could say, well, how can I turn that dial, right, and change the voice there, right, and iterate within the format by changing the voice. All right. And so then the last edge would be the what and the how, like on that high edge.

And that's the frame. That's the one that's kind of the most confusing ⁓ for people. When I discuss it, because it becomes a matter of semantics, they say like, oh, it's just the hook. It's the frame.

Which is, I consider the frame, the way in which you're positioning your topic in the context of the problem. So hooks spin out of that. But when it comes to the triangle, I'm not concerned with copywriting at the triangle, right? I'm thinking at a higher level, right?

Just in terms of communication. It's like, what are we talking? And so, so let me stop there. I don't want to keep rambling.

Is that clear? Could you give some examples of frames? So when it comes to frames, what you're thinking about frames is a psychological frame. Warning signs would be one frame.

Right? So let me give you a specific example. If our problem is nerve pain, right? Let's say the problem always sits at the center of the triangle, right?

So it's in the middle. Like everything is in the context of the problem. We're talking about nerve pain. So who do I want to speak to, right?

Well, I'm going to choose to speak to people who are dealing with sciatic nerve pain. Right. I'm going to choose that segment of the marketplace for reasons, right? My research bears out like sciatica is actually a hot trigger.

You should talk to these people specifically in an ad. So I'm going to choose to talk to people about sciatic nerve pain. Well, what is my topic? What are we talking about?

Well, my research shows that like, you know, especially on YouTube, stretching and exercises for sciatica are a huge thing. So I to talk about stretching, right? So I'm talking to people with sciatic nerve pain about stretching and then what is it I want to reveal? reveal that actually stretching can make your sciatica worse, right?

And that is true for quite a number of people. And so then if you're looking at the frame edge and say, how would you frame that? And say, well, from a warning sign standpoint, I could say something like, you know, if stretching only makes your sciatica worse, it's a warning sign of this hidden circulation issue, right? Now say hidden circulation issue, I know what my product is.

I know what the unique mechanism of the problem is over there, the unique position in there. So I'm trying to tie the two things together. Right. But that would be a warning sign frame.

⁓ Another frame would be, what do they want to know frame? Right. Probably the simplest frame, a very powerful frame. Right.

It's the whole Robert Collier ⁓ enter the conversation already in their head. Right. So my research, all this always goes back to research. Research bears out that one of the questions they're asking is, why is stretching making my sciatica worse?

So that literally, right? So I have a, what do they want to know? That's the category for the frame. The hope that I could deliver there, a simple basic hook is just literally recite back the question they're asking.

Why is stretching making your sciatica worse? You know, here's why, watch this. And so you can kind of go on and on that how many frames exist, right? It's like almost unlimited.

It depends on. how you want to split out your frames depending on what kind of psychology you subscribe to. But those would be two types of, two ways to frame ⁓ your insight on the same sort of conversation. So the psychological aspect for the warning sign is fear, right?

And then would you say it's driven by fear and then the frame where it's like, well, what do they want to know? What are they already chatting about? posting about on Reddit or forums, that psychological frame would be more curiosity or what would that be? Yeah, right.

And so this is where things start to get sticky, right? And so there's overlap. You separate it out to codify. You can say, it's fear based, right?

So you can base your frames. Say, I'm going to create a set of frames based on emotions, Fear, greed, hope, all that kind of stuff, insecurity. And you can develop frames around that. The reality is when you...

Execute on a frame and turn it into the hook. Oftentimes you will mix in elements across both, right? Like you can have a pure fear-based frame, but you may have one that brings in authority as well. Right?

So like, yes, that's one way to divide up your frames is through emotion. You know, there are different, many different schools of thought on psychological frames. Like I'm a fan of Sally Hogg's head's ⁓ fascination, right? The original iteration in the book.

And like she divides up her fascinations into like mystique. What is it? Taboo, vice, right? There's prestige, authority, and so like you can kind of cut your frames that way as well.

So, you know, one thing with the frame, I may have mentioned earlier that, you know, the way that I look at the triangle is that I separate things out to codify it, but reality is that they cross over. When you're defining an edge of the triangle, oftentimes it will start to self-select other edges. Right. Or it'll narrow the field.

Right. And so we know if we were looking on like the how edge and you're picking your format and saying that like I'm doing a note cast, right. Which is like sort of known for the note cast around here. It comes pre-packaged with that sort of demolition frame.

Right. Fiber helps you poop, right? Nope. Right.

Like that's self-selected for you as like you need to test that frame. Right. ⁓ With that sort of type of format. Same thing with an expert deep on, like a hook that is worked continually in that format would be something like, you know, the fastest way to clear out stuck poop, right?

A lot of experts get this wrong, let's see what you think, right? And that's sort of is the hook, right? Which is carrying that sort of frame. And so, that's...

A key thing to keep in mind when you're going through the triangle is that as you're picking pieces around the edges, they oftentimes will self-select up. Okay, so did we cover everything in the triangle? Is there any part of it that we missed? ⁓ No, we covered the whole thing.

So where is the biggest opportunity with the triangle for brands where, you know, it's like we're all, like you said, we're focused on testing maybe different... personas ⁓ or just avatars, but we're missing out on this opportunity over here, which could be the frame or could be something else. Where do you feel like it's worth focusing more attention? Yeah.

So I'll tell you first how we use it. I tend to use it with writers ⁓ because you can use it a number of different ways. You can use it just as a brainstorming tool, right? Becoming up with angles just in general.

You can use it as a strategic tool. You can use it as a diagnostic tool. We use it postmortem. ⁓ to ⁓ give you ideas on how to iterate.

Right? You know, I'd say that one of the ways that we've been using, at least I've been using it with my writers, is for strategy. A lot of times when you launch an ad, for a lot of people, it's like, ⁓ this has got to be a hit in the home run, right? Otherwise it's a failure.

Versus, well, you know, looking at it from the stamp on the triangle, it's a minimum viable creative. And I know I'm going to build this thing out over a couple iterations ideally, right? budgeting and account management in the context. And that's always a thing, but like, you know, I'd like more at bats at this ad and I just need to get data.

Right. Cause if I look at something like the triangle, you know, when I'm picking different points on the triangle or assessing what my options are within those different points in the triangle, I'm not just saying, ⁓ it's an idea that I want to test. Like I'm placing bets. Right.

And so you shift the mindset to like, I'm placing bets. Then it's like, well, you're betting on something, but you're also betting against something else. Right. So first and foremost, it's like, well, what are my options here?

Right. And so for a lot of people that I work with with the triangle for the first time, ⁓ you know, it can seem overwhelming because, you know, it's a lot of words and semantics and it's like, ⁓ you talk about frame, but I think about it differently. Right. Whatever.

⁓ but one of the things that people respond to a lot is like, ⁓ like I didn't even realize that that was an option to test. I typically don't think that way. Right. So like, I'm going to run a hook test here and, ⁓ But I don't know what I'm actually testing in the hooks other than like I swipe these hooks because they're proven hooks.

Right. So if you look at something like, I'll give you an example of what we've kind of put in the motion recently. Right. So we're going to run a warning sign frame ⁓ for a nerve pain offer.

Right. And what we want to find out is like, which, which symptom experiences the strongest pull. Right. So like, is it, is it the feet feel ⁓ burning?

Like you're walking on hot coals, is it pins and needles? Is it stabbing, shooting pains? And so we're to use the same framework, the same frame, which is like, if your feet burn at night, ⁓ you know, it's not bad nerves. It's a warning sign of this hidden circulation issue.

Right. And so we're going to say if your feet burn at night, if your feet feel like pins and needles at night, right. If whatever, however many symptoms we're going to test. And so if you know that going into a test, it's not just coming out of the test saying like, hook one, one, now what do we do?

It's like, no, I know I'm testing symptoms first. So whichever is the strongest signal, I then know in my next round of testing what I'm going to change in the body, say. Let's say, cause I wrote a body, it's a little bit more generalized cause I don't know which symptoms gonna ping or is gonna hit. So I'm go back in and edit the body specifically for pins and needles cause pins and needles work, right?

Or I'm gonna build out a pre-funnel or a hook stack based on pins and needles, right? and do like a topper, which is ⁓ a UGC sort of a transformation story around pins and needles to run in front of this thing. Right? So it starts to, it helps you to think a little bit more forward thinking down the road and plan strategy.

Now things don't always work out as planned, right? There's always bumps in the road, but you're not just kind of, for lack of better way to say it, throwing shit against the wall and seeing what sticks. Right? That can work great.

It works great for lot of people. You're working based on your intuition, like what you know, based on your skillset. Yeah, this feels right. ⁓ But oftentimes what you think is going to win doesn't win.

Right. And so when you, when you're working and understanding, ⁓ what we feel are key performance drivers around the triangle, ⁓ and understanding the options you have available within there, you could say, well, I'm making a bet on A for this round. Right. as opposed to B.

And so then when you come out of that, if something hasn't performed, let's say, it's like, well, I know what I bet on and I know what I bet against. So let me look at the things I bet against. Right. Or another way that we've used it ⁓ recently was in a post-mortem looking at an ad that we felt really good about ⁓ that underperformed.

And so we compared, we broke out the triangle for that ad and compared it to one for the top performing ad in the account. and say, well, where is it different? It's like, ⁓ it talks to the same person. All right, cool.

Who has the same problem. Okay, cool. ⁓ The topic is actually the same, right? But the insight that we're delivering ⁓ is different, but so is the frame.

And it's interesting that the frame that we went with is a contrarian frame, right? Whereas the one that we use in the top performing ad is just answering the question of what is it they want to know, right? And when we look at the data and the data shows where the average view time is, and you look at that point in the video, it's when they start talking about that sort of framing, right? And so then you have theory, well, maybe they're calling bullshit on that contrarian frame, right?

We know in the account, they respond to the one that simply just answers the question that they have. So what would happen to this ad if we run a test and put that way of framing the conversation, the topic in there, right? And so that's something that we're rolling out now. So it just helps you think, helps you think strategically, helps you think differently.

My biggest thing is, getting feedback like, oh, I didn't know that was an option. I never thought about it like that before. Right. And when you start doing that, you see just how much diversity you can sort of create in the account.

You can do it strategically. You're not just throwing shit against the wall. So for anyone listening who wants to roll out something like this, you've given that great example of using it in a post-mortem session with your creative team where you're looking at ads that have performed or not performed. What are some other ways they could start rolling out this triangle or something similar?

Maybe even it's like a simplified triangle just to start testing in a more strategic way and thinking in a more strategic way. Like how else could we roll this out? Yeah. Well, I would start with just the three points on it.

The who, the what, now. You know, again, it sounds so simplistic and it sounds so obvious, but I've talked to a lot of people and it's just like, oh, I've never stopped to think about it like that. Right. And so.

You know, cut up who you're talking to, like what, what, who you're, you're total accessible market, right? Like who, who, who could you talk to that would make sense for this product, right? Instead of make a list. maybe as a simple way to do it, I would say is, ⁓ cut up your who category by symptoms, the way in which they experienced the pain.

You know, there's a lot of talk out there about micro moments. You can kind of go that deep, but just like, are the different ways they experienced the pain, right? And cut up your segments that way. Cut up your segments by ⁓ causal belief.

Right? That's how I want to divide my segments sometimes. Right? So an example would be a nerve pain offer.

I know that there's a good portion of the market who have ⁓ mystery nerve pain, idiopathic nerve pain. They're told by their doctor, we don't know what it is. We've ran all the tests, it's idiopathic nerve pain. Another big segment is diabetics, pre-diabetics.

So blood sugar. Right? So people who believe that there is no answers to their nerve pain, people who believe that blood sugar is causing their nerve pain. ⁓ post-chemo people who believe chemo cause their nerve pain, right?

And run a test that way. ⁓ as you change who you're talking to, it also changes the conversations available to you. Right? So if I'm going to talk to someone who's a diabetic, so on that what edge, what am I going to say, or what am I talking about with them?

You're talking about blood sugar, right? In a way that you wouldn't talk about blood sugar if you're talking to the general market. I'm going to talk to everybody who has nerve pain. Right?

You're probably going to talk to nobody. In a sense, if you do that, right? And so, you know, take those three ways of cutting up your segments and then say, well, what conversations are, so the loudest conversations based on my research for each one of these people. if you look at, again, your nerve pain people who believe it's because of their blood sugar, it's going to be, you know, blood sugar itself was a topic.

Cutting out sweets and carbohydrates, changing their diet, taking their metformin, they get the blood sugar down, like the pain is still there. pain in the fact is actually spread. Right. And so you could say, well, okay, if I'm going to cut the market that way, I can have these three conversations.

What does that want to say? That is insightful enough, right? That it sets up my product is almost like the obvious answer to solve their problem. Right.

And then the how edge it is, what are all the different ⁓ formats that I can deliver this in? Right. And so, so how can I communicate? ⁓ that message to that person, right?

To the person who has diabetic, the diabetic nerve pain is because of blood sugar. And you want to tell them, that blood sugar is actually your blood flow? What are some, some very different ways that I can actually communicate that message to them? Right.

And so a lot of people will split it up by just like UGC car confessional might be one. ⁓ our expert debunk right where it's a, it's an authority in the studio reacting to a video on screen is another one. ⁓ two person podcast is going to two people. talking about their nerve pain and one of them is someone who's had the nerve pain, used to believe it was their blood sugar, but then found this new thing out was actually the blood flow and it solved everything for them.

So I would look across those three categories, ⁓ starting with the who. If you start with the who, it'll start to inform everything else. Yeah, so you mentioned earlier AI and how you want to clone yourself and ⁓ can you kind of go deeper into how you've been using AI to help you with the triangle and strategic thinking and how you've shared that with the team of writers as well. Yeah.

So I've always been a big AI guy since like chat GPT was a baby. so, you know, it started out with prompt engineering and developing prompt flows to perform a task that I wanted to sort of offload so I can get more done faster. You know, it migrated into I'm a Claude guy, right? And so when Claude got Claude projects and you could do system prompts, you can kind of create like bots, right?

15,000 word system prompts, but it really popped off when Claude got memory and skills. And so in that process, like I've probably created like 75 skills by now. 50 is your limit in Claude, right? I learned it out when I got to 51.

So I've had the sunset a few, but like in that... John, can you also define what a skill is? Like define some of these terms. Oh yeah, yeah, sorry.

So a skill is, like the scene in The Matrix when Neo's pulled out of The Matrix, right? And he gets uploaded with the Kung Fu program and he comes out and says, I know Kung Fu. It's like that. You're just loading up programs into Cloud that teach you how to do a specific thing, right?

And so like in that process, I've got to pull out of my head, what is my workflow? What are the things that I actually do? You know, cause you can do a particular thing, but if you, you can do it a couple of different ways. And sometimes you want to do it a couple of different ways for different reasons.

So you're teaching all this stuff to Claude so it can be a second you. And in that process, you get to a point where it's like, ⁓ there's a lot of execution skills here, scripting skills, copy, chiefering skills, research skills, ideating skills. It's like, but I really need some strategic level skills, like as a hive mind that can kind of manage all this stuff. Right.

So that's the triangle is one of those things ⁓ that sort of comes out of that. And so. ⁓ how we've been using those lately in the agency, at least with the writers that I'm working with, know you use some of these as well. ⁓ We'll use them in creative jams for ⁓ quick research, ⁓ quick ideation.

So a ⁓ good example would be on the triangle, the how, the format, right? It's like there's a million different formats that you could run. You'd like to test the ones that are working the best, like now currently in organic or in paid. And so there's a skill called format explorer.

Right? That has all that stuff loaded up in its brain. You just feed it an angle or an idea and say, run format explorer. Right?

And then it gives you a look, you know, this nice, pretty looking document with maybe 50 different looks at how you could take the angle and spin it out into the different formats. Right? And oftentimes, you know, we'll pull out four or five, we're like, oh, right. like, oh, I can write that right now.

Like I can totally see that. I get the vision, I get the voice. It totally makes sense. Go write it.

Right. We'll also skill stack because we have scripting skills for every skill, every sort of TNT special. ⁓ We teach Claude how to write it, right? And so you can get a quick look with format explorer, what an angle looks like across different formats.

You pick the one you like, you run your scripting skill and it'll split it out for you, right? So like in quick order, ⁓ you can have quite a number of net new scripts that will get you to... ⁓ gets you to a good first draft very fast. I know plenty of people have prompts that'll run like hundred different scripts, whatever, in 20 minutes.

You know, some of those are just kind of throwing stuff against the wall. The way we like to build them here is a little bit more strategic and formed by our data. But oftentimes now in creative jams, it's not just coming to the jam with ideas, but it's coming to it like with scripts, right? That are like fully fleshed out and developed, right?

And we're kind of like expediting the process because we can get to that point so fast. Totally makes sense. I'm just wondering how someone listening, if they're not even familiar with skills and the creation of skills, like this might feel overwhelming to them. So I don't think this needs to turn into an AI training, but could you just kind of talk through maybe how you think about the questions you're asking when creating a skill, just as a starting point or if there's a better starting point.

Yeah, no, no, no, sure, sure. So like I said, I'm a cloud guy and you can kind of create skills through other AIs, right? With Gemini has gems. GPT has GPTs, this has skills.

so, know, Claude has its own built-in skill on how to create skills. So like that part is kind of taken care of for you. You just let it know, hey, I want to build out a skill here. The biggest tip I can give is just have a conversation with it.

Tell it what you want to do. Right. And then walk it through the process. So like a scripting skill, for example, if we take a format that's working for us and we want to be able to get quick looks from Claude really fast, you teach it how to write that format.

So you're going to take your best performing ad with that format, feed it to Claude, have it break it down. You have a conversation with it to make sure it broke it down the way you want it to break it down. Then you need to pressure test its thinking. Right?

So you want to have a conversation with Claude where you get it to the point where he understands what to do and how to do it. But like why it's doing what it's doing at every turn in that particular script. Right? It needs to understand how to think.

Cause where the skills can easily fail is when you open up a fresh new chat and have it run the skill. Can the skill perform as well as it did in the chat that you built the skill? Right? Can it transfer that knowledge?

Right. But a lot of that goes back to the strategy. You have to teach it the strategy behind what you're doing. It's not just like, hey, plug and play on this particular script.

We're just going to swipe it. Cause a lot of times you wind up trying to force a round peg into a square hole. So it has to understand strategically how to think about that and how to adapt it. Um, and yeah, you just have a conversation with it.

I know that sounds super simple. But it really is the best prompting advice I could give right now is like talk to Claude the way you would talk to your bud. ⁓ it makes me think of a recent episode of open residency that, ⁓ had Dan Martell on it. ⁓ and he's super big AI guy, like big proponent of Claude and Manus.

And he was saying like, if you don't know what to do with your AI, like you don't have to know because you can ask it and start from there. of like, this is my outcome. This is what I want to do. Ask it.

And then it's going to go ahead and walk you through the steps. And I've done this a couple of times in building skills on Claude. And, I love how it does the multiple choice questions and it walks you through it. Okay.

Am I understanding this? Do I get that? I need clarity on this. And then you give it all the feedback that it needs and then set it off and running.

So it's a really great question, Kira. And for anyone who does feel intimidated by building skills or what is into the more like intermediate and then advanced level uses of AI and not just treating it like a super intelligent search engine that like, just ask it. Don't be scared by it. Show up and actually say, I'm a complete noob.

I don't know what I'm doing, but this is where I want to get and I want to build a skill for it. And then it will walk you through the process. It's great. honest with it.

One of the best questions you can ask too is as it starts to build it out, it's like, well, how do you think, how do you feel about this skill? And it'll tell you, it'll tell you what it likes. it's like, yeah, it's 80 % there, but actually, you know, it doesn't do these other things well. So it's like, okay, so then just do that.

that gap in and just keep doing that until it's like, yeah, it's like 95 % there. You just need to test it. It's like, okay, cool. there.

We're good. You know, but you need to know how you think about the skill that you're trying to teach it. Right? So like, if you don't really understand how you do what you do, why you do it, you're have a hard time creating the skill.

So much of this triangle ⁓ is just about the thinking process, which is great news for all of us that we can strengthen our strategic thinking, especially when AI seems to be taking some of the thinking out of our tasks and day-to-day projects. It's good to really strengthen that muscle. So I'm curious how creating the triangle has helped you think differently or expand the way you think if it has. And like what that looks like for you as you've been using this more and more.

That's a very good question. What I will say is, you know, and again, everything kind of comes back to AI sort of unfortunately, but it's just in that process of, getting it to be a second you, it forces you to go deeper and understand like the things that you do intuitively. Right. And it's really good at helping to clarify things for you and put things into words that you couldn't put before.

They'll just sort of like turn a little dial in your head and make you think a little bit differently. I'd say ⁓ one of the biggest things ⁓ that has changed in my thinking is how deeply I question things, I guess. Right. So like one good thing that the AI is really good at is it feels very confident in its responses to you.

They give a very binary answer, black and white. like, ⁓ yeah, this because of this. I was like, yeah, but you don't have enough data to come to that conclusion. Right?

Like that sounds really good. Right? And like, just to hear yourself talk that out loud, I like, yeah, I sound really smart. Like that sounds really good.

And it's like, But you don't have enough data to stand on that conclusion. Right? So then you always have to caveat. So then you have to ask yourself, then what data am I missing and how can I get it?

What do I need to do to get it? Right? And so I think what it's changed for me is like, I always have a big BS meter when it comes to AI anyway. ⁓ But just questioning what I think I know a lot more about things, right?

Even stuff that I learned years ago from copywriting or whatever, right? It's like, ⁓ like it now makes sense to me maybe in a different way because I've had to question it deeper and I've gotten answers to things that never really made sense to me before, but I just accepted them as true because X, Y, and Z guru was saying it. Right. And so it sort of, ⁓ it puts you on that path if you're really mindful of it.

of, me find out the things that I really don't know with full confidence. And then like, how can I know them? Right. And when it comes to what we do here, it's always about opening up, winds up opening up new testing opportunities to find that answer.

like, no, I haven't run that kind of test before. That's why I don't have that particular type of data. Right. Or I don't have that particular learning, or I can run a test that we've run.

a hundred times before, but in a slightly different way, because now I'm looking for a specific type of learning, right? And it then always sort of advances your thinking. love that you mentioned multiple times ⁓ cloning yourself essentially, because that's something that, ⁓ whether it's AI or not, the thing that came to mind is some content that I saw from Callaway, who is a big like content educator and... Yeah, Callaway's going...

Yeah, Calaway's dope. And so what he was talking about is spending essentially three months training someone how to think like him. And so in this case, it wasn't AI, it was a person, but you know, somebody at the highest level of content production and you know, he's just gotten amazing YouTube channel and does long form content. We're not talking about ads in that case, but the value of being able to clone yourself so that the thing that you need, the output that you need is done exactly the way that you would do it without you having to do it.

Yeah. Super valuable. 100%. Yeah.

Now with things like cowork and code and stuff like that, you can just start to like, you can run a dispatch from your phone, right? And you can get home and Claude's got a Google doc for you with 15 scripts. If you know what you're doing, 100 scripts, right? I wrote a hundred ads while we were in this podcast.

Yeah. No, I'm kidding. This leads to, I guess my final question is just with... AI making it easier than ever for anyone to produce that volume, like both of you were just talking about, where does that leave teams that don't have the strategic thinking or a strategic system underneath it?

That's a good question. I mean, cause even, even, you know, you can produce it high volume and at speed with the AI, but if you're not producing it in a way where there's a strategic system underneath that, you know. You can still run into problems. Like obviously like budget and how you manage the account are a big part of that.

You know, we could run a skill now and create a hundred scripts. Are we testing those hundred scripts this weekend? Like, no, not in the types of ads that we do, right? If we're doing 30 second e-com ads, it's a little bit of a different story, right?

You can code up some things so AI can develop the actual video assets for you. It's a different story, but like we've got scripts that's in millions of dollars that are six minutes plus. Right. And so it's a little bit of a different beast there in terms of production and the production quality that we're looking for.

And so, you know, I think it's easy to look at how much AI can do and how fast it can move. And if you're someone who's a copywriter, even a video editor, that can be kind of scary. But the AI still by default sucks at context. It's really on its own, not great at writing scripts unless you've taught it well.

And even then... You know, if you, if you sign up for ⁓ error notifications from Claude, you'll see how often it's erroring like all day long. And so you can run a skill this morning and run it in the afternoon and get entirely different results because it's just depreciated output content or quality. ⁓ And so I would say as long as you're, if you haven't already shifted towards just more strategic thinking as a team, like if you're doing that, you're doing the thing that AI sucks at.

now. so like volume is great, but volume for the sake of volume, isn't strategy. Some people have that as a strategy is like, I'm just going to throw it all against the wall. I'm going to hit something eventually like more at bats.

Right. ⁓ but that's not necessarily how we do things around here. It's like thinking about that earlier when you were talking about iterating. ⁓ so just to jump in on that, like, thinking you're achieving creative volume because you're able to do a large number of ads, whereas like really what you should be trying to achieve is like what the necessary volume is for the volume of ad spend that you have.

knowing how much your average ad spend per ad unit is and not making more than that and then being strategic within that number instead of saying, we're going to do 2000 ads this month because we need a lot of ads. And anyway, so I just wanted to call that out because Volume for volume sake is just more throwing shit at the wall like you said. like stuff can stick, but to your point, like the more strategic you are, the more you understand why you're doing something, then the better results you're gonna have.

And so if you limit yourself to achieving that maximum volume, you do wanna hit the maximum volume, but knowing where that ceiling is so that you can have more intention. You're maximizing the amount of effort and intention that goes into each ad unit. And you're probably going to get to wins faster than if you're just like, Hey, let's make 2000 ads. And then some of them stuck 80 % of them only spent $20, but who cares?

Because you've got some winners and then you're just blindly like pushing forward. you know, it's like a, on a hamster wheel kind of ⁓ effort towards volume and diversity. It's not the intention ⁓ and like real strategy of diversity and volume. Yeah, because it's also the question then too, like ⁓ of learning.

Like what are you learning from that? How is it impacting your variations if you're looking at the scale? Right. And so, you know, to each their own, to some degree, depending on the type of business you're operating in, some strategies make more sense than the others.

Maybe throwing shit against the wall with a hundred variations, taking a modular approach is the way to go for you there. Okay. My final, final question is maybe tricky because I want you to look into the future and What, where is strategy going with ads in the near future? Like what do we need to do as creative teams, as CMOs, as founders to get ahead over the next year or so?

Beyond implementing your system, like, but what else should we be looking for? The contingent there is always like, you know, what's the algorithm going to do, right? That's a tough question. We're asking you to hit a moving target, John.

Hit a moving target, right? Like AI is evolving so fast and the algorithm over the past year and a half has changed so much that those two things combined basically like make it impossible to look into the crystal ball. But I guess, you know, if, if you have to, we're going to force you to, then let us have it. Yeah.

Right. So, so here's one thing I've been thinking. I say it's off the wall. not even really off the wall because it's happening in certain places.

I was talking to a friend earlier. you know, who has their own business and is a little bit frustrated. like, yeah, you know, my business already is starting to be like, we're developing AI bots that talk to AI bots on the other side and there's no human interaction. It's just bot to bot, right?

You already kind of see marketplaces where like AI is hiring human people or AI bots are doing purchases for you. And so like, you know, one thing I find interesting to think about is at some point, it going to be The ads that you're using AI to develop are going to be served to AI bots that are deciding on whether or not they're going to click and buy it for the actual real person. And like, what does that look like? You know, because it's happening in spaces already.

New metric bot hook rate, bot hold rate. Yeah, exactly. And like we're laughing at it now. I'm curious to see next year come back at this point and like see how much that sort of has encroached on what we do.

Cause you know. It's the AI algorithm scrutinizing and deciding like who it thinks it should be showing it to. ⁓ AI is integrated more and more in our devices, right? Like ⁓ it's going to be AI ad bots writing ads for bots that are making the purchases for the people and doing the clicks and stuff like that.

So like how would you think strategically about that, I guess, is sort of the question. You asked the goes back to the approach to how to build skills. You asked the bot. This is what we need to do.

Go do it. How do I do it? Yeah. Exactly.

Your who, your avatar soon is going to be a bot. Right? And you're like, how do I write for the bots themselves? Not just for the algorithm, but for the bots that are going to watch my ad.

Okay. This will be a future episode. We'll bring you back for that one. Okay.

Absolutely. All right. Well, John. Thank you so much for going deep into the strategic triangle and outlining that for us and giving us this tool that we can use and brands can use and creatives can use, teams can use to really think differently about how they're scaling and iterating on ads.

So this has been really helpful. Thank you. Cool. John also wanted to throw a thank you to you for coming on, elaborating on this.

know it's something that you've Obviously mentioned working with writers and people saying, ⁓ I never thought about it that way. I've had multiple people bring that up in conversations that you weren't a part of. ⁓ so it's definitely shifting the way people are thinking. ⁓ so yeah, rolling right into the outro.

If you've got any questions about the triangle, ⁓ if you're not really sure on something that John was describing, or you want more examples, leave us a comment or shoot us an email. ⁓ all that stuff is in the show notes. And if you liked the episode, give us a thumbs up. Follow along, make sure to hit the bell icon so you get notified anytime we drop episodes.

If you're listening on Spotify or Apple Podcasts, make sure to follow along. And other than that, thanks for being with us and we'll see you in the next one.