0:00 In 2026 and 2027, creative is the lever 0:03 that sits within Meta ads. Media buying 0:05 fundamentally is not your advantage 0:07 right now. You need to become incredibly 0:09 good at creative strategy, at creative 0:11 design, production, then moving that 0:13 into the ad account, and analyzing the 0:14 creative to create this flywheel effect. 0:16 By the end of this video, you'll know 0:18 everything you need to know about 0:20 creative for 2026. We're going to be 0:22 running through why creative is the 0:24 growth lever, how Meta's algorithm 0:26 actually works in 2026, concept 0:28 architecture, hook strategies that 0:30 actually scale, all the formats you can 0:32 use and test, a testing framework to 0:34 take that creative and actually put it 0:36 into the account to get statistically 0:38 relevant results. We're then going to 0:39 move into creative volume, creative 0:41 diversity, and a financial model that 0:43 ties it all together. We're going to 0:44 talk about fatigue, scaling winners, 0:46 portfolio management, creative 0:48 production, so what do the teams, the 0:50 briefing, and the process look like? And 0:51 then lastly, we'll tie all of it 0:53 together and package it into the 2026 0:56 playbook. So with that being said, 0:57 diving into section number one, we have 1:00 why creative is the growth lever. I've 1:02 been media buying on Facebook since 1:03 2018. So I've been about eight coming up 1:06 on nine years in the Meta Ads Manager. 1:09 And I remember back in 2018, 2019, 2020, 1:11 there was so much leverage in media 1:13 buying. You actually didn't need to be 1:15 very good at creative, and as long as 1:17 you had a massive edge on media buying, 1:19 you could beat a lot of people. And the 1:20 reason for that is that there was a 1:22 massive skill gap. In fact, I would 1:24 argue that any arbitrage opportunity 1:26 that exists within a market is a product 1:29 of the skill gap that exists at that 1:30 point in time. So, back then, the 1:33 difference between someone that just 1:34 goes and opens up the ad manager and 1:36 someone that is a super advanced media 1:38 buyer that spends 12 hours a day doing 1:40 this and that's three years in, there 1:42 was such an enormous difference that 1:44 that difference allows for arbitrage and 1:46 allowed for the opportunity. Now, over 1:49 time, Meta has slowly compressed media 1:52 buying. The targeting is now fully 1:54 automated. I would recommend 99% of 1:56 people are just running broad only, so 1:57 there's no skill within interest 1:59 targeting. Andromeda now optimizes for 2:01 the distribution of the creative. 2:03 Consolidation is always favored across 2:05 all of these digital platforms now 2:07 because conversion data consolidates, 2:08 which allows for better machine learning 2:10 modeling. And all of the media buying 2:11 hacks that used to exist, like 2:13 duplicating ad sets or using bidding 2:16 models in a certain way, none of them 2:18 really work that well anymore. None of 2:20 them really give you a big arbitrage. 2:21 And so because of that, this arbitrage 2:23 opportunity has compressed. And instead, 2:26 where the big arbitrage opportunity 2:28 exists right now, is in creative. And it 2:30 is because most people, fortunately for 2:33 you, are terrible at creative. And so 2:36 because of that, there is this huge 2:37 opportunity that if you simply become 2:39 good at creative strategy, at 2:41 understanding what a good ad looks like, 2:44 and then being able to build a system 2:45 that can churn out a ton of high-quality 2:48 ads, this is where you now make all the 2:50 money on that. And by the end of this 2:51 video, you should unlock 2:54 this opportunity. So, the first thing I 2:55 want you thinking through here is the 2:57 portfolio model. So, let's say that you 3:00 have $1,000 a day in ad spend, and there 3:03 is two different scenarios here. 3:05 Scenario one is you have one single 3:07 creative that's holding that entire 3:09 $1,000 per day, and you're at a 4x ROAS. 3:13 And then you have another 3:16 pathway that you could choose. You've 3:18 got three ads holding $1,000 a day 3:20 together, and this is also at a 4x ROAS 3:23 blended across these three creatives. 3:25 Now, I would rather be in this situation 3:28 every single day of the week. And the 3:30 reason being is that this situation 3:32 decreases risk enormously within the 3:35 business, because any of these three 3:37 creative can fatigue, and the other ones 3:40 will pick up the slack, and you can 3:41 continue operating at this efficiency, 3:43 at this ad spend. 3:44 >> [snorts] 3:44 >> But over here, if this creative 3:46 fatigues, you're done. And I say this 3:49 out of PTSD, working with probably eight 3:52 to 10 clients now over the course of the 3:54 last six years who have come to us 40% 3:57 down year-on-year in new customer 3:59 acquisition. Their business is in a 4:00 terrible spot and they're going, "Our 4:03 business is declining. It's because we 4:05 built this eight-figure business off 4:07 three creatives that held a million 4:09 dollars in ad spend and now these three 4:11 creatives are all fatiguing and we're 4:13 trying to spin up variants and it's kind 4:14 of keeping us at some reasonable level, 4:16 but we just can't get back to the level 4:18 and we keep declining." And my response 4:20 these days is, "It's too late. We're not 4:23 taking on that account because you 4:24 should have fixed that six to eight 4:26 months ago." What people will do is they 4:28 will get one or two winning ads that can 4:31 hold a tremendous amount of ad spend. 4:33 And then what they do is they just stop 4:34 making more ads and they go, "Like, why 4:36 would we spend money on creative 4:37 production? That's useless. Let's just 4:39 double down and put all our ad spend 4:40 here and then focus on other areas of 4:42 the business." Yeah, amazing until these 4:44 creatives fatigue. And the only reality 4:46 of creatives in Meta ads is that every 4:49 single creative will fatigue. Every 4:50 single creative has a maximum spend 4:52 threshold that it will reach. And so you 4:55 want to be constantly thinking through 4:56 creative production ad account 4:58 management and creative through the lens 5:00 of this portfolio model, which is that 5:02 we want to get as many ads as possible 5:04 holding as much spend as possible. We do 5:06 not want to over leverage into one or 5:08 two ads because this is ultimately what 5:10 ends up destroying a lot of businesses. 5:12 Let me give you three common mistakes 5:13 that most teams are making when it comes 5:15 to creative. Number one is a lot of 5:17 people confuse activity for strategy. 5:19 I'll give you an example of this, which 5:20 is recently in the last few months I've 5:22 seen two ad accounts that are launching 5:24 2,000 plus new ads a month and they're 5:27 barely able to crack quarter million 5:29 dollars in ad spend per month. Now, for 5:30 those that work on large ad accounts, 5:32 you would know that that is insane. You 5:35 do not need two to 3,000 new ads a month 5:38 on a quarter million dollar budget. 5:40 That's so much volume. And they weren't 5:42 profitable. And so you're looking at the 5:43 fact that they're launching so much 5:45 volume, they're not profitable, They're 5:47 doing so much activity, but there's 5:50 clearly a gap in strategy. 5:52 Because if you were putting 2,000 great 5:55 creative into the account with a good 5:56 offer, with a good persona, with a good 5:58 angle, then you shouldn't need anywhere 6:00 near 2,000 ads a month. And so yes, I 6:02 think volume's good. I think you should 6:03 be launching as many ads as humanly 6:05 possible, but with a standard of quality 6:09 and with a strategy behind the creative 6:11 that's actually going into the account. 6:12 So do not mistake activity for strategy. 6:15 This is really common right now because 6:17 people think they can just launch more 6:18 ads and their brand will scale. In fact, 6:20 I hear this all the time, which is 6:21 people are like, "Okay, I need to double 6:23 my business. Do I just launch 40 more 6:24 ads a month?" 6:25 I was like, "If it's that simple, 6:26 everyone would do it. There's a little 6:28 bit more to just increasing activity." 6:30 The second one is misdiagnosing 6:32 bottlenecks. So with in creative, there 6:35 is a ton of different bottlenecks that 6:37 exist that could be holding you back. It 6:39 could be the quality of the creative. It 6:41 could be the format diversification. It 6:43 could be the quantity that's sitting at 6:45 the different stages of awareness or the 6:47 stages of the funnel. So you have 6:48 nowhere near enough top of funnel assets 6:50 to produce the scale that you actually 6:52 want. 6:53 This might be a bottleneck on not 6:55 actually introducing partnership ads or 6:57 employing different strategies into the 6:58 ad account to unlock more spend. There's 7:00 so many different bottlenecks that exist 7:02 just within creative that holds a 7:04 business back, and most people don't 7:06 understand all the bottlenecks that 7:07 exist. So they misdiagnose it. They 7:09 don't realize what the actual bottleneck 7:10 is, and so then they focus on activity 7:12 that doesn't actually move the needle. 7:14 For example, those two brands that I 7:15 spoke about before, they were not only 7:18 mistaking activity for strategy, but 7:20 they were also misdiagnosing the 7:21 bottleneck in the business, which is 7:23 that they thought just increasing sheer 7:25 volume would suddenly unlock growth, but 7:27 it's not doing anything for them. And 7:28 it's because they're trying to fix 7:29 something that actually doesn't need to 7:31 be fixed. And then the last one, which I 7:32 hope people are getting over by now, but 7:34 I still see it from time to time, 7:38 which is that creative is seen as a cost 7:41 center rather than a revenue driver. 7:44 And this is fundamentally because of the 7:46 shift away from media buying to 7:48 creative, which is that back in the day 7:50 when media buying had a lot of arbitrage 7:52 behind it, you didn't need a lot of 7:53 creative and a lot of volume of 7:54 creative. And so because of that, 7:56 business owners would often see the ad 7:58 spend in the platform, the spend on 8:00 Google, the spend on Meta, the spend on 8:01 TikTok, Pinterest, Microsoft as revenue 8:04 driving, but the actual creative that's 8:06 used to fuel that spend as a cost 8:09 center. We want to minimize that cost as 8:11 much as possible because we just want 8:12 the platforms to find us customers, but 8:14 that's the complete wrong way around. 8:16 The creative is actually what is driving 8:19 the revenue and Meta and these other 8:20 platforms are just driving the 8:22 distribution. Now, the fact that you can 8:23 actually just post this stuff on organic 8:25 and get free distribution as well means 8:27 that creative is also the revenue driver 8:29 organically. And so if you're thinking 8:30 that creative is a cost center, you've 8:32 got the completely wrong approach. We're 8:34 going to be diving into later in this 8:35 video exactly how much spend you should 8:38 have on creative production respective 8:41 to your ad spend levels so you can know 8:43 if you're currently under or over 8:44 investing in creative. Before diving 8:46 into creative strategy, you first need 8:48 to understand how Meta's algorithm works 8:50 because this is the distribution tool 8:52 for the assets that you're putting in 8:53 it. Now, everyone's probably heard the 8:55 word Andromeda. It's a massive buzzword. 8:56 I've talked about it in a lot of 8:57 content. And to put it simply, it's just 8:59 an algorithm change that Meta made 9:01 because they had a lot of GPUs for 9:03 training AI for their metaverse that 9:05 kind of fell flat, and so they 9:07 repurposed all of those training chips 9:09 towards a better ad serving platform. 9:12 With that, the mechanical shift that you 9:15 need to be across is that before 9:17 Andromeda, and to be honest, we weren't 9:19 even doing this pre-Andromeda, but it's 9:21 an easy way to simplify the concept and 9:23 then I'll go a little bit more nuanced, 9:24 which is that pre-Andromeda, you would 9:25 choose the audience. So you would say, 9:27 "Hey, 9:28 here is our interest. 9:30 Please target this interest." And you 9:32 would then go and load all the creatives 9:34 in. 9:35 And so you'd put creative one, creative 9:37 two, maybe you're putting three 9:38 creatives under this ad set. And you're 9:40 saying, "Here's the interest, here's the 9:41 ads, go and match them." Now, 9:43 post-Andromeda, what happened is you 9:45 actually do the opposite. So, you put 9:48 the ads into the account, and you don't 9:50 select any targeting. 9:52 And then Meta goes and looks at the 9:54 creatives, and it goes, "Okay, I think 9:56 that these are going to resonate with 9:57 these people based on the transcript, 9:59 based on the words in the ads, based on 10:00 the copy, based on my understanding of 10:02 historical conversion data." 10:04 And then it will go 10:07 and find the people to target. 10:09 Now, this is fundamentally a way better 10:10 system because interest targeting has a 10:12 lot of downside. Okay, when you're doing 10:14 interest targeting, you're effectively 10:16 labeling people with an interest, let's 10:18 say pets, and then just arbitrarily 10:20 serving all your creative to people that 10:21 are interested in pets. But the reality 10:23 is is that it's not just people 10:25 interested in pets that want your 10:26 product. It's probably people that are 10:28 interested in pets within a specific age 10:29 range that also have other variables or 10:32 psychographic data points that are 10:34 similar between them. And so, there's so 10:36 much additional nuance that needs to be 10:38 captured. And this 10:40 serving mechanism ends up capturing that 10:42 nuance better than the old one. Now, the 10:44 reality is is that Meta was kind of 10:46 already doing this pre-Andromeda. We 10:48 haven't used interest targeting in most 10:50 accounts for like two years. But what 10:51 Andromeda really did is popularize this 10:54 new approach to structuring an ad 10:56 account. And so, there were still so 10:58 many agencies and so many people running 10:59 interest targeting, it was crazy, even 11:01 though we have a video that was put out 11:03 two years ago about this exact topic on 11:05 the Meta algorithm. And this was before 11:07 Andromeda. So, I was explaining this 11:08 concept before this whole update even 11:11 rolled out, and why we don't use 11:12 interest targeting. It's called How 11:14 Machine Learning Works in 2025 on Meta 11:18 Ads. But this has popularized this 11:19 approach, and now this is really how you 11:21 want to be thinking through the asset. 11:23 Now, why this becomes important is 11:24 because the creative that you loads in 11:26 that you load in is the targeting. So, 11:28 if you load a bunch of creative that's 11:29 all the the you end up just targeting 11:31 the same person. You don't reach new 11:33 people. If you end up loading in a bunch 11:34 of creative that isn't clearly 11:37 resonating with a specific target 11:38 demographic, then it's going to get 11:40 confused. For example, if you're calling 11:42 out everyone in your ads, if you're 11:43 like, "Hey, everyone, look at this 11:45 product." Meta is going to be very 11:46 confused cuz it's going to go, "What 11:47 should we target the entire population? 11:49 Is there a subset? Who should we 11:50 actually go after?" You also need to 11:52 think through this targeting mechanism 11:54 at an ad set level in terms of 11:55 structuring because if you just put a 11:57 bunch of ads together that are all 11:58 targeting different people under the one 12:00 ad set, the ad set will also get 12:01 confused. Who are we targeting? So, 12:03 there's a lot of of additional nuance 12:04 that we'll break down here when we get 12:06 into the structure and testing section 12:07 of this video. But, this is 12:08 fundamentally the key concept that you 12:10 need to understand here. There's also 12:11 the creative similarity score that 12:13 rolled out of Meta recently. Now, most 12:15 people still can't access this in their 12:16 ad accounts, but if you reach out to 12:17 your dedicated Meta reps, they can get 12:19 you this score. So, you can have an idea 12:22 of what your similarity score is in the 12:24 business. And as it currently works, the 12:26 higher the number, the worse it is 12:29 because the more similar your creatives. 12:31 Now, why this matters so much is because 12:33 of Andromeda bundling. 12:35 And so, if you have, as I said before, 12:37 multiple different creatives here, let's 12:39 say one, two, three, but they're all 12:41 slight variants. Let's say they're the 12:43 same campaign shoot, but the model's 12:45 just standing in a slightly different 12:46 position. Or let's say it's UGC, but 12:49 maybe there's a slight variation in the 12:52 script halfway through the video. Or 12:53 let's say these are actually very 12:54 different videos in terms of formatting, 12:57 but they're all speaking to the exact 12:58 same audience. Well, what ends up 13:00 happening is Andromeda will bundle them 13:03 together and serve them all to the same 13:06 pool of audience. Now, because of this, 13:07 because they're similar creatives going 13:09 to the same audience, what happens in 13:10 the ad account is frequency goes up. So, 13:13 frequency is how many times users are 13:15 seeing your ads. 13:16 Obviously, these ads are just going to 13:17 serve to the same people over and over 13:18 again, so frequency goes up. Your reach 13:21 will go down, so you'll stop reaching 13:22 that new users. 13:24 And as a function of this over time, 13:27 your return on ad spend will go down. 13:29 And so you will start seeing diminishing 13:31 returns in the platform. 13:32 So instead, the position that you 13:34 actually want to be in here is launching 13:35 three creatives into the account, but 13:37 these three creatives talk to different 13:39 people, 13:40 and therefore 13:42 reach their own audiences. 13:45 And with that, frequency declines 13:47 because you're reaching different people 13:48 with these ads, it allows you to unlock 13:50 more spend because as they scale, you 13:51 can almost think about this circle 13:53 expanding as it goes to cold on cold 13:54 audiences. 13:56 But because these are talking to 13:57 different people, you can see there's 13:58 not much audience overlap. 14:00 And as a function of that, you're going 14:02 to be able to scale to new audiences 14:04 effectively without just increasing 14:05 frequency and saying efficiency decline. 14:08 Another incredibly important concept 14:10 here about how the platform algorithm 14:12 actually works is in sequencing of 14:14 creative and the reliability of return 14:16 on ad spend as you decline through the 14:18 account structure. So when it comes to 14:20 sequencing creative, the reality is, and 14:23 I think everyone watching this will 14:24 agree with this, which is that if you 14:25 just serve one ad to a user once, will 14:28 they buy? And the answer 99.99% 14:31 of the time is no. 14:33 Generally speaking, you need to see an 14:35 ad more than once to be confident in 14:37 purchasing. Now, yes, there are some 14:39 people out there that are getting one 14:41 hit by ads and suddenly buy, right? That 14:42 does happen. In fact, it's probably 14:44 happened to me at one point in my life, 14:46 but it's very, very rare and you 14:47 definitely cannot build an efficient 14:49 business off of those people. And so 14:51 because of that, you need to serve them 14:53 generally more than one ad. Now, you 14:55 could serve them this ad again, a second 14:57 time. However, 14:59 Meta put out a article ages ago, it was 15:01 like four or five years ago, which said 15:03 that after two impressions of an 15:05 identical creative, probability of 15:07 purchase declines precipitously. And so 15:10 if you look at the probability of 15:13 someone buying, let's say this is a 15:14 percentage, and then you look at the 15:16 amount of ads that they are seeing. So 15:18 this is frequency of an identical ad. So 15:21 we're talking about the same ad here. 15:22 You're not showing them different ads, 15:23 just the same ad. 15:24 When you show them the ad once, there 15:26 might be a certain percentage uh chance 15:29 of them buying. Then, when you show them 15:30 the ad twice, it might be the same. It 15:32 actually might even be slightly higher. 15:34 Because on second viewing, now they're 15:36 warmer and they're more likely to buy. 15:38 But then on third viewing, it declines 15:40 massively. 15:42 And then on fourth, massively. And then 15:43 it pretty much declines to zero. And so, 15:46 there is very little purpose at all in 15:48 showing a user the same identical ad 15:51 more than twice. And the way to actually 15:53 stress test this idea that I'm telling 15:55 you right now in your own ad account is 15:56 you can pull your ad account up, you can 15:58 go down to the ad level, look at the 16:00 last 30 days, 60 days, and look at 16:02 frequency at an ad level on cold 16:04 campaigns. And what you will see is that 16:06 cold ads almost never have a frequency 16:09 above a two. And if they do, you have a 16:12 creative fatigue issue. You do not have 16:14 enough diversification and volume in 16:15 your account, and that is a red flag. 16:17 Any good account that we're running, 16:19 which is all of them, um no one has a 16:21 frequency above a two on a cold ad. Just 16:23 doesn't happen. Because if we see this, 16:25 immediate red flag. We need to fix it. 16:27 We need more diversity. We need more 16:28 volume. Now, the big kicker here is that 16:32 what Meta are published in that 16:34 statement, and if I can go and find that 16:36 article, I'll put it in the description 16:37 below, is that if you then go and 16:39 introduce a novel creative, the 16:42 probability of purchase stays high. And 16:45 so, after seeing the same ad once and 16:47 then twice, if we go and serve a 16:48 different ad in as the third impression, 16:52 probability stays up here. 16:54 And then if we go and serve a different 16:56 ad as the fourth impression, 16:57 probability stays up here. 17:00 And so, we can actually keep probability 17:01 of purchase up by rotating in new novel 17:04 creative. Now, the question becomes, how 17:07 does the algorithm do this and how does 17:08 it decide the sequencing of creatives to 17:11 these users to try to get them to 17:12 convert? And this is where sequencing 17:14 comes into the ad account structure. 17:16 When Meta is optimizing across serving 17:18 users ads, they're optimizing across 17:22 understanding what sequence of creative 17:24 actually gets the purchase. So, is 17:26 showing users ad one, then ad two, then 17:29 ad three going to get the click and 17:30 purchase? Or is it showing users ad 17:32 three as the opening creative, then ad 17:35 two, then ad one going to get the 17:37 purchase? Or is it some other kind of 17:38 combination? Are we coming in here and 17:39 then going back here? There's all of 17:41 these different combinations of serving 17:43 these ads in a particular order that 17:45 might yield a click and might yield a 17:46 conversion. Now, the kicker here is that 17:50 Meta only attributes based on last click 17:52 attribution. So, if we assume that the 17:55 first sequencing example is what is 17:57 actually occurring. They're going 1 2 3. 18:00 If the click occurs here, all of the 18:03 return on ad spend, all of the revenue 18:05 gets attributed to this ad and it looks 18:07 really good. Let's say it's at a 6X 18:08 ROAS. 18:10 Ad two, maybe some people buy at this 18:11 stage in the sequence, but not a lot. 18:13 And two at a 2X. And then on this top 18:16 ad, we're down at like a 1.4X. When you 18:18 look at this and you look at spend 18:20 distribution, let's say spend is equal 18:22 across all three creative. Naturally, 18:24 you go in here and you go, "Oh, this is 18:27 obvious. Turn ad one, turn ad two off." 18:29 Cuz ad three's getting all the 18:30 performance. But that would be a bad 18:32 idea. And this is the concept in a lot 18:34 of our content about do not turn a 18:35 creative off when you are hitting KPI. 18:38 If this is hitting your performance goal 18:40 overall as an ad set, don't touch 18:42 anything. Because Meta is sequencing 18:44 between these creatives on purpose and 18:46 it is creating an outcome. Meta's 18:48 optimization is not based on last click. 18:51 Really important concept to understand. 18:53 Meta's optimization is actually based on 18:56 multi-click and view. And so, Meta sees 18:58 that people are viewing this ad, viewing 19:00 this ad, then clicking on this ad and 19:01 buying, and therefore it distributes 19:03 spend here, knowing that it played a 19:05 role in the purchasing journey. But you 19:07 don't see that. Unfortunately, we don't 19:09 get any visibility into this at all 19:11 within the ads manager, which is 19:12 actually really annoying. And so, as a 19:13 product of that, we just need to trust 19:15 the system. We also need to think about 19:18 the creatives that we're putting into 19:19 the account. How are they working 19:21 through this purchasing journey? 19:22 Normally, if you're good at creative, if 19:24 you're good at creative strategy, you 19:26 can look at these three creatives, and 19:27 you can go, "Oh, it's pretty obvious to 19:29 me that this is actually the opening ad. 19:31 This is then probably the second ad, and 19:32 then this is the converting ad." Because 19:34 this is a super unaware top of funnel 19:36 creative. This is then a product aware 19:39 creative, and then this is a bottom of 19:41 funnel very aware creative. And so, it's 19:43 super obvious to me just looking at the 19:45 three ads that that's what the sequence 19:46 looks like. But, if you don't understand 19:48 creative strategy, you're going to look 19:50 at this and go, "I don't understand 19:52 what's going on. Why are the ROAS like 19:53 this? Turn off, turn off, turn off, 19:55 destroy performance." We're going to go 19:56 into this in a lot more detail later on 19:58 once we start going through the testing 20:00 structure and actually KPI'ing and 20:01 understand what is a good ad, what is a 20:03 bad ad, but this is a really important 20:05 concept to be across. 20:06 Second concept to be across is that the 20:09 reliability of ROAS declines as you go 20:12 down in the account structure. So, 20:14 everyone who's opened up a Meta ad 20:16 account knows the basic structure, which 20:18 is you have campaigns, you have ad sets, 20:20 then you have ads. Right? The campaign, 20:23 you set up the budget typically, and you 20:24 set up the overall uh optimization of 20:26 it. So, is it optimizing for sales, 20:28 reach, etc. 20:29 Ad set, this is where all the targeting 20:32 occurs. Now, people forget that these 20:33 days because people go, "Oh, targeting 20:35 is just broad, so I don't even really 20:37 know what's happening at the ad set 20:38 level." But, it's really important to 20:40 reinforce targeting is occurring at the 20:42 ad set level. And then down here, the 20:43 ads, this is obviously the creative that 20:45 gets served to the end user. Now, when 20:47 we look down at the ad level, and we 20:49 start assessing creative, the issue here 20:51 is everything I just went through. We 20:53 can look at the ROAS numbers, 20:56 but they're kind of meaningless because 20:57 they're last click attribution, and 21:00 there's a multi-click or multi-view 21:01 through journey occurring here. And so, 21:03 I can go in and say, "Yeah, this has a 21:05 6x ROAS, is Let's like try to scale it 21:07 or make more ads like that. But it's 21:09 often a bad decision. So instead, when 21:11 we move up to the ad set level right 21:14 here, now we can look at all of these 21:16 returns as a group. And we can go, let's 21:18 not go into the individual ads because 21:20 the reliability of uh the attribution 21:23 here is terrible because there's so much 21:24 happening across these different 21:25 sequences. So let's just go up to the ad 21:27 set and look, okay, at the ad set level, 21:29 what's our ROAS here? And maybe overall 21:31 it's at a 3x. 21:32 Now this is a lot more reliable. It is 21:34 worth noting that cross-serving of ads 21:37 still occurs at the ad set level. So you 21:39 can have uh multiple different ad sets. 21:42 So if I just throw another ad set in up 21:44 the top here, 21:45 we'll call this ad set two. And let's 21:47 just put a couple creatives under here. 21:51 A sequence that can occur is that people 21:53 see this ad in ad set one. 21:55 They then see this ad. 21:57 And then they get served another ad set. 21:58 Now it's not prioritized by the 22:00 algorithm. The algorithm will always 22:02 prioritize serving adjacent ads under an 22:04 ad set due to the way that targeting is 22:06 siloed, but there is cross-ad set 22:09 targeting. And so because of that, we 22:11 might have a 3x here. We might have a 4x 22:14 up here. This ad set seems to be doing 22:16 better, but it's actually just because 22:19 this ad set is picking up some 22:20 bottom-of-funnel conversions being 22:22 generated from this ad set. That can be 22:23 completely avoided as long as your 22:25 structure is good and you're 22:26 diversifying concepts across the ad set 22:28 level, you shouldn't have much cross 22:29 targeting. And you can measure all this 22:31 and we'll go into it later on, but the 22:32 whole idea is ad set is better. 22:35 We can trust this number. We can trust 22:36 it more than the ad level. Then once you 22:38 go all the way up to the campaign level, 22:41 the campaign ROAS number, as long as 22:43 it's on 7-day click attribution, is 22:46 super reliable. This is generally very 22:48 congruent with the actual P&L as long as 22:51 it's on 7-day click and as long as 22:53 existing customers are excluded. And so 22:55 because of that, when we're making 22:56 decisions within Meta, when we're 22:57 thinking about structure, when we're 22:59 thinking about introducing creative, we 23:01 want to be thinking about measuring at 23:03 the ad set or the campaign level, not 23:05 the ad level, and we want to be thinking 23:06 through this idea of sequencing. So, the 23:08 practical implications of Andromeda and 23:10 the way that the algorithm and ad 23:12 platform currently works is fourfold. 23:14 Number one, you need concept diversity. 23:17 We're obviously going to go into what a 23:19 concept is, how to break it down, how to 23:21 come up with concepts, what diversity 23:22 looks like later on, but number one is 23:24 you need diversity within the creative 23:26 going into the account. You also 23:28 obviously need that introduced into the 23:29 structure so that you're having a 23:31 concept per ad set so you don't have any 23:33 of this cross targeting going on. 23:35 Number two is you need format variety. 23:37 Okay, you can't unfortunately just have 23:39 the same format of all of your creative. 23:42 And the reason being is that Meta will 23:44 prioritize formats that people resonate 23:46 with. I'll give you an example, which is 23:48 that I never ever, and I could scroll my 23:50 Instagram feed for the next 3 hours and 23:52 show you, I will never get a piece of 23:54 UGC under my profile. Just doesn't 23:56 happen. And it's because I never click 23:58 on them, I never resonate with them, and 24:00 so Meta inherently knows that low-fi UGC 24:03 content should not be served to me. And 24:05 so, if your ad account is just low-fi 24:07 UGC, I'll never see it. Now, same thing 24:11 actually for DPA ads. Now, DPA are very 24:14 bottom of funnel. This isn't really 24:15 cold, so I'd have to be already very, 24:17 very product aware and on the website, 24:19 but I don't think I have ever been 24:22 served a DPA almost in the history of 24:25 the ads that I get served on my personal 24:26 profile. Now, that is likely a product 24:29 of the fact that DPAs are not a creative 24:32 type that I resonate with. I never click 24:34 on them. And so, if you just had UGC and 24:36 DPAs, it's very unlikely you'd ever 24:38 reach me. So, you need to be making sure 24:40 you have as much diversity in format 24:41 type, and I'm going to be giving you all 24:42 the formats and how you should be mixing 24:44 and matching them later on so that you 24:45 can reach as many people as possible. We 24:47 then have hook quality. So, this is 24:49 going to be a big section of this video, 24:51 but with Andromeda, any kind of persona 24:54 call out that occurs within the actual 24:56 creative is going to go and prioritize 24:58 that persona. So, if I make an ad and at 25:00 the start I say, "If you're a business 25:02 owner doing over $5 million a year," 25:04 Andromeda is going to read the 25:05 transcript, it auto transcribes every 25:07 single ad, and it's going to go, "Okay, 25:08 let's try to find a similar audience to 25:10 what he just said." What if my ad or my 25:12 content has nothing to do with that 25:13 audience? Well, then we've just pushed 25:14 the creative to an audience that's not 25:16 actually relevant. And people do this a 25:17 lot. People's hook actually isn't as 25:20 good as they think it is. People call 25:22 out a persona without problem agitating. 25:24 There's the so many different issues 25:25 that actually lie in hooks, which is why 25:27 we're spending so much time on it, cuz I 25:29 think people are terrible at hooks. Most 25:30 ads I see, they're just not good. And 25:32 then lastly, is that with Andromeda, 25:34 volume is absolutely critical. The spend 25:37 capacity of most ads isn't very high. 25:40 Most ads that you went to the into the 25:41 account, if we look at mean ad spend per 25:44 ad unit, sits at about $700 to $1.5 25:47 thousand depending on the ad account. 25:49 And so, generally, you can't get much 25:51 spend through most ads. So, you need 25:52 volume to compensate, but you also need 25:54 strategy at the same time. Because if 25:56 you're just putting sheer volume into 25:58 the account, which I've seen so many 26:00 people do, with no actual strategy 26:02 behind the volume, it doesn't actually 26:03 work. So, let's run through concept 26:05 architecture. This is how you need to be 26:07 thinking through the segmentation of 26:08 different creative. This is how your 26:10 naming convention should be set up on 26:11 Meta as well. This is how you should be 26:13 ideating in a Google Sheet or whatever 26:15 format you want to think through this. 26:16 And we'll also talk about the AI 26:17 implications into this ideation process. 26:20 So, what actually is a concept? A 26:22 concept is a persona, angle, and offer. 26:25 And so, to have diversity within the 26:26 creative that's going into the account, 26:28 you need one of these three variables to 26:30 change. If the offer changes, it will 26:32 resonate with a different audience, with 26:34 a different pool of people, because it's 26:36 either a different product, or if it's 26:38 it's a different value proposition or 26:39 packaging of the product. If the angle 26:40 changes, obviously, it's going to 26:42 resonate with with someone different, 26:43 because you're problem agitating on 26:44 something else. And so, if you sold skin 26:46 care, you might start by problem 26:48 agitating on wrinkles, but then you 26:50 might have another ad that problem 26:51 agitates on that. They're going to reach 26:53 two very different audiences even though 26:54 it's the same product because the angle 26:56 changed. And then lastly we have a 26:58 persona change. Now the easiest way to 27:00 think through a persona change is 27:02 obviously who are we talking to, but 27:03 also who is actually in the ad. And so 27:05 if there's a 60-year-old woman in the 27:07 ad, it's obviously going to resonate 27:09 with a very different target demographic 27:10 than if there is a 20-year-old woman in 27:12 the ad. And so because of that we can 27:14 adjust the persona and that will 27:16 inherently adjust who resonates with the 27:18 creative and who it will target on the 27:19 platform. And so one of the easiest ways 27:21 to ideate here and come up with almost 27:23 infinite creative concepts is to start 27:26 ideating on who are all the different 27:27 personas that we want to reach. What are 27:29 the core personas in the business? And 27:30 you can go through and you can probably 27:32 come up with two or three or four. And 27:34 then if you really drill deeper, you can 27:36 come up with tons more. So yeah, we 27:38 could go like females age between 40 to 27:40 45, etc. But then we can start digging 27:43 into like what are the actual pain 27:44 points. And so females age 20 to 20 five 27:47 dealing with acne. Females age between 27:49 30 to 35 dealing with acne which is a 27:51 slightly different problem because acne 27:52 shouldn't be occurring at that age so 27:53 maybe it's hormonal acne. And then we 27:55 can go women age between 40 to 45 with 27:59 acne as well. That's also a completely 28:00 different problem that should be fixed 28:02 by now and so what are the different 28:04 underlying causes that we can then 28:05 resonate within that persona. And so you 28:07 can build personas out in ridiculous 28:10 detail based on all of the problems that 28:12 the product solves. By the way, this is 28:14 the case in every industry as well as 28:17 fashion. So in fashion people really 28:19 struggle to get their head around 28:20 concept architecture, but with personas, 28:23 it's the same thing. Most large retail 28:25 fashion brands will break their target 28:27 demographic down into like four core 28:28 personas and they'll give them names. 28:30 But the reality is you can go way deeper 28:31 than this because a persona could be a 28:34 particular problem that a particular 28:37 person is facing. And so there's this 28:39 particular person who's a mom age 28:41 between 30 to 35 with young kids and the 28:44 problem that they face is they need an 28:46 affordable dress that can go from a work 28:49 dress into a night dress on a Friday 28:50 afternoon. That is a persona. That's a 28:51 persona that you could make infinite 28:53 creative for for that exact particular 28:55 pain point and circumstance. And you 28:57 could probably think of 400 examples of 28:59 that. When we move into angle, it's the 29:01 presentation of how we're actually going 29:03 after the problem. So, what is the 29:05 angle? What is the value proposition? 29:07 Why would you purchase the product? And 29:09 then the offer is obviously really 29:10 straightforward, which is what are we 29:12 actually offering? The key here is 29:13 personas 29:14 >> and demographics. They're archetypes 29:17 that are defined by problems, desires, 29:19 and buying triggers. So, we want to be 29:20 thinking about who are our avatars. And 29:23 you want to think hard on that. Spend 20 29:25 minutes on it. Then, what problems do we 29:27 actually solve for these people? 29:30 Then, what are the differences in pain 29:33 points, language as to how we talk to 29:36 this person, and the proof that is 29:37 required for this person? And once we 29:40 understand these three, we should be 29:42 able to build out a whole lot of 29:43 personas, which you can then, once that 29:46 is done manually, go to an LLM like 29:49 Claude or ChatGPT or Gemini, 29:51 give it the entire persona list, and 29:53 tell it come up with 20 to 30 more, and 29:55 give it as much context as possible, 29:57 which you should already have loaded in 29:58 there in some capacity about the 29:59 business, about the target demographic, 30:01 etc. Claude, OpenAI LLMs, they're only 30:04 going to give you averages. And so, if 30:07 you just give it three or four personas, 30:09 and you give it your website, and you 30:09 say, "Hey, come up with some more 30:10 personas," you're going to get an 30:11 average outcome. And everyone's going to 30:13 be doing it, and you're going to get the 30:14 same as them, and you're going to revert 30:15 to the mean. So, what you want to do is 30:16 if you are going to do this, you need to 30:17 put all the manual effort into actually 30:19 building out proper 30 to 40 personas 30:22 that are actually good, and then you 30:23 need to give it an unbelievably wide 30:25 amount of context as to the business and 30:26 the target demographic and everything 30:27 you can, so that it can actually ideate 30:29 properly on another 20. And of the 20 30:31 that it produces, you shouldn't just be 30:33 putting that in there. You should be 30:34 refining down, probably cut down to four 30:36 to five, give that back in, say these 30:38 four to five are good, ideate on another 30:40 10 to 20. And then that way you can beef 30:42 this out from 20 to 30 personas or 30:44 whatever you get to and double the 30:46 number. One really important note here 30:48 is that people die in persona work by 30:51 keeping the personas too broad. Broad 30:53 personas end up producing forgettable 30:56 creative that is nowhere near specific 30:59 enough for the target demographic that 31:01 you want to resonate with. And so if you 31:02 come away with a persona while you're 31:04 building this out, which is let's say 31:06 women age between 25 to 45, this is a 31:11 terrible persona. The creative that 31:13 you're going to make for this audience 31:14 is nowhere near specific enough that 31:16 it's going to resonate with enough 31:18 people to be able to hold a good amount 31:19 of spend. As an example, instead of 31:21 going for the persona busy professional, 31:23 which I actually see so much when when I 31:25 tell people to do this concept build 31:26 out, and instead you would want to 31:28 delineate this into a consultant who 31:31 flies twice a week and needs a carry-on 31:33 wardrobe that works for boardrooms and 31:35 bars. Way more specific. And now with 31:36 that persona call out, you're already 31:38 thinking about how to talk to this 31:39 person. You can almost visualize who 31:40 this is, what their problem is, what we 31:42 need to solve, what the creative should 31:44 look like. It gives so much more 31:45 specificity into the actual briefing and 31:47 production process. And then guess what, 31:49 we actually reach this audience 31:50 properly. But when we say busy 31:52 professional, what is a busy 31:54 professional? That is so wide, there are 31:55 so like the age is all over the place, 31:57 their gender is all over the place. What 31:59 do they actually do? What do they get up 32:01 to? Why do they even need this product? 32:02 This is so broad that you will you will 32:05 end up with mediocre creative. You are 32:07 better off making five ads for this 32:09 target demographic than five for this. 32:11 Now five for this will scale harder. If 32:13 you can actually do a proper unaware ad 32:15 that is going to resonate with this 32:17 entire pool of people, yes it's a bigger 32:18 total addressable market, so yes you'll 32:20 be able to hold more spend. But the 32:21 reality of you making a good ad for an 32:23 audience this broad is actually very 32:25 unlikely. And so you are way better off 32:27 making more ads for very specific 32:29 personas, and then as you start to 32:31 become a eight mid-eight figure brand, 32:34 you can start to play around in trying 32:36 to get ads to work for very broad 32:38 audiences, but the skill level required 32:40 here is so high and it is just not 32:42 required for most ad accounts. Most ad 32:45 accounts you can just go hyper specific 32:46 and get way better results than needing 32:48 to go and try to target the entire of a 32:50 population. Another example here is 32:51 instead of doing, let's say, health 32:53 conscious parent, you could target a mom 32:56 who's tried six different supplements 32:58 because her toddler won't eat 33:00 vegetables. Once again, way more 33:02 specific. Now, you could go away and you 33:03 could write a script for that. You know 33:04 exactly what that ad looks like. You 33:05 know who who should be in it. You know 33:06 who you're talking to. You know what the 33:08 proof needs to be for that particular 33:10 target demographic. You know what the 33:11 language needs to be and you know the 33:12 pain points of that particular persona. 33:14 Now, an angle is the specific argument 33:17 or perspective that you take when you're 33:19 speaking to a persona. This is where a 33:21 lot of creative strategy actually ends 33:23 up failing because brands default to 33:25 product features rather than crafting 33:28 genuine angles that resonate with the 33:30 persona. So, I'm going to give you a lot 33:31 of tactical application here. Number one 33:33 is you want to problem agitate in some 33:35 capacity. You don't need to do this in 33:37 all of your creatives. In fact, 33:39 especially if people aren't even problem 33:40 aware, then you can't even problem 33:41 agitate, right? But you want to be 33:42 leading with pain as much as possible 33:44 and you want to make the viewer feel the 33:46 pain, feel like they have been heard 33:48 before you go and introduce the 33:50 solution. Another great angle here is 33:52 approaching with a contrarian truth. So, 33:54 you want to challenge a common belief 33:56 that people have which allows them to 33:57 reframe. An example of this, 33:59 particularly just as a hook, is 34:01 everything you've been told about X is 34:03 wrong. Everything you've been told about 34:05 the protein industry is wrong. And then 34:07 you go into obviously an educational 34:09 piece that then might problem agitate 34:10 that then obviously loops into a 34:12 solution. You want to typically always 34:13 have social proofing in the angle. You 34:15 want to build some kind of authority or 34:17 expertise in 34:19 the domain. So, this is more 34:20 specifically domain expertise. So, you 34:22 don't need, for example, a doctor in a 34:24 lab coat showing up in every single ad 34:26 for every industry. Right? If you're 34:27 selling a running product, you want a 34:29 marathon runner showing up. You want 34:30 someone with authority in the domain 34:32 that you're trying to sell it. Curiosity 34:34 is great and is commonly in most angles 34:36 that do well, and this is specifically 34:38 creating what's called a curiosity gap. 34:40 And so, when you open up a video asset, 34:42 and you can actually do this in image 34:43 assets, too, but when you open up the 34:45 video, you open up and create a 34:47 curiosity gap or an information gap. As 34:49 an example, you'd say, "Here's what the 34:52 top 1% of millionaires know about saving 34:55 money that they don't want you to know 34:57 about." And then you don't answer it 34:58 straight away. And you go into a story, 35:00 and you go into something else. You go 35:01 into any of these other aspects. You 35:02 problem agitate. You have a contrarian 35:04 truth. You go into some degree of social 35:06 proofing, although that's probably 35:07 stacked towards the back of the ad. And 35:09 then eventually, the ad goes on and on 35:10 and on. You pull people through this 35:12 journey. You educate them on everything. 35:13 And then you end up actually answering 35:15 the curiosity gap. And so, you didn't 35:18 have any information. You want to get to 35:19 the information, and eventually the 35:20 information's provided to you. And so, 35:22 that's how you hold people through an 35:24 ad. All good angles also have some 35:26 degree of comparison or objection 35:28 handling. And so, often people who are 35:31 watching your ad have tried to solve 35:33 their problem through other solutions. 35:35 And so, you want to tell them why your 35:37 solution will work versus all of the 35:38 competitors. You've probably tried this. 35:40 You've probably tried this. You've 35:41 probably tried this. If it's a piece of 35:42 user-generated content, you'll commonly 35:44 see the actual person in the script 35:45 going, "I tried this, just like you 35:47 would have, and it didn't work for me. I 35:48 then tried this. I thought it was 35:49 overhyped. It didn't work. And then 35:51 finally, I was like, okay, I'm going to 35:52 give one more thing a shot, and it 35:53 actually worked." And then this one is 35:55 the least compliant, but it is the best 35:57 performing component to add into an 36:00 angle, so I'm going to throw it in, 36:01 which is a transformation, aka a before 36:05 versus after. Uh before versus after's 36:08 Meta does not like them. You have to 36:09 sneak these into creatives in a way that 36:12 the algorithm doesn't flag it. But these 36:14 are some of the best performing 36:15 creatives ever, because people want to 36:16 see visually a before and after. I think 36:18 this is basic human psychology. Before 36:20 and after's will always crush for as 36:22 long as they exist. And because of that, 36:24 Meta doesn't like you uh running them. 36:25 So, when we're thinking about the angle, 36:28 we're starting at the persona and then 36:30 we're thinking about what angle and 36:31 approach do we need to be able to 36:33 resonate with this target persona. And 36:34 what you will notice is I've given you a 36:36 lot of tactical application that can be 36:39 included or built together into a script 36:42 that creates the angle. So, the real key 36:44 with the angle is actually you can 36:46 simplify it into a one sentence, which 36:47 is if our persona is women with kids age 36:50 35 to 40 that are having energy issues, 36:53 we're going to then sell them creatine 36:56 powder and the angle is going to be that 36:59 it solves energy issues based on 37:01 particular studies and then we can 37:03 continue training it out. The thing is 37:05 the angle realistically, if you're going 37:06 to write this out properly, we're just 37:07 talking about the script, okay? And so, 37:10 you just want to write the whole script. 37:11 If it's a static, this is obviously 37:12 going to be a bit tighter and you could 37:13 probably encapsulate this in a few 37:15 sentences, but the idea is we want to 37:17 figure out what angle are we approaching 37:19 selling this person on with a myriad of 37:22 different strategies that's going to 37:24 resonate with them. And so, if we take 37:25 for example that, uh, woman age 40 to 37:27 45, uh, with creatine and energy issues, 37:30 well, we're going to want social 37:32 proofing, but we want to we're going to 37:33 want social proofing for that particular 37:35 persona. So, we're going to want either 37:36 case studies around persons people like 37:38 that or the clinical studies that we're 37:40 going to reference, which is going to 37:41 give us expertise and authority, needs 37:44 to reference their particular age and 37:45 gender. We want to think about problem 37:47 agitation on that particular persona. 37:49 So, how are we going to agitate to that 37:51 problem? Are we going to pull in some 37:52 kind of contrarian truth or curiosity 37:55 gap that's going to hold them into the 37:56 asset? Can we do any kind of comparisons 37:59 to stuff that they've tried in the past? 38:01 Has this persona tried things or is this 38:02 very new to them? And then, can we pull 38:04 in any kind of transformation at the end 38:07 as a piece of social proof to back into 38:09 the creative? And so, we choose the 38:11 elements that we want, we think about 38:12 how we're going to slice and dice these 38:14 into a script and that becomes the 38:15 angle. And we're going to build out some 38:17 actual examples of this in a moment, but 38:19 before we do, I want to talk about the 38:21 testing priority order. Which of these 38:24 three components has the largest impact? 38:27 What should you be thinking about 38:28 testing the most of or the least of? 38:30 Number one at the top is angles. So, 38:32 this is exactly what we're just 38:33 discussing. This is the message or the 38:35 argument. The reason why this is such a 38:37 high priority is because we can sell to 38:39 any persona that is relevant to the 38:40 product. You should be able to 38:41 effectively look at who has bought from 38:43 you in the past. What persona do they 38:44 sit in? Why did they buy from you? And 38:47 then, we just need to craft an angle to 38:48 that particular persona. If that persona 38:50 isn't purchasing from us, we know that 38:52 it's actually really an angle or an 38:54 offer issue. It's not an issue with the 38:55 persona cuz we've already proven that 38:57 these people buy from us, okay? And so, 38:59 when something's not working, we 39:00 typically don't want to look at the 39:01 persona as long as you did it well. 39:03 Like, let's not do broad 25 to 39:06 45-year-old females. Let's get hyper 39:08 specific into a very specific person 39:10 that has purchased from us before, and 39:12 let's target them. But then ultimately, 39:14 it's going to come down to an angle or 39:15 offer issue, which is why angle is the 39:17 first thing you should test if things 39:18 aren't working. You need to rotate that 39:20 and improve the angle. And then the 39:21 second is the offer. Now, fundamentally, 39:24 the offer actually might be the number 39:25 one most important thing out of all of 39:27 this, but the issue is is that you 39:29 actually don't have much flexibility in 39:30 changing the offer. What do I mean by 39:32 that? The offer is two things. The offer 39:34 is the product, and then the offer is 39:36 the value proposition or the way that 39:38 we're framing value. Now, when it comes 39:40 to the product, if you have a single 39:42 product and ads aren't working, well, I 39:44 can't really tell you to change the 39:45 product, right? Now, you could. You 39:46 could change the whole business, but the 39:48 reality is is that the product is the 39:50 product. We need to change the angle or 39:52 the persona, right? So, even though this 39:54 is the most important thing, the product 39:55 ultimately is king out of anything, we 39:57 don't have much flexibility here. Now, 39:58 if you have a very large skew count, and 40:00 maybe this is a new product that's being 40:02 launched into an existing 20-product 40:04 portfolio, then sure. Then we can go and 40:06 say, "Is this an offer issue? Is this an 40:08 issue with the product?" And we can 40:09 start to troubleshoot accordingly. But 40:10 for most people, you're pretty stuck on 40:12 the product. Then when it comes to the 40:14 value proposition, the value proposition 40:16 we've got a couple options, right? We 40:17 can go and try testing bundles, we can 40:20 go and try testing like a gift with 40:22 purchase, we can go and try all of these 40:24 different repackaging of the offer, 40:27 right? We can do like tid as well, so 40:29 you get percentage off when you buy two 40:30 or three. We can go and throw in like a 40:31 subscription offer. So, there's a lot of 40:33 testing that we can do here. And we do 40:34 this with a lot of clients is we'll come 40:36 up with a bunch of different offers and 40:38 how we should reframe value to try and 40:39 get the funnel working more effectively. 40:41 But ultimately, you are limited here. 40:43 There's only so many offers you can 40:45 test. And we run into this issue, which 40:46 is like, "Ah, we've tested like seven 40:48 variations of how we can spin these two 40:50 to three products. And now we're 40:51 relatively stuck. And the only way that 40:53 we can increase the value perception on 40:55 the offer is to just do harsh deep 40:57 discount. And we don't want to do that. 40:59 We don't want to go and just eat 40% 41:01 into margin and build a brand off a 41:03 heavy discount reliance." And so, 41:04 although the offer is unbelievably 41:06 important, this is a huge variable that 41:08 we do a lot of testing on. And we do a 41:09 lot of offer testing for clients. It's 41:12 something that eventually you're going 41:13 to hit a wall on. There's only so much 41:15 offer testing you can actually do. And 41:16 once you've found the best performing 41:18 offer, you're stuck on it. And so, that 41:19 is where then we want to be doing a lot 41:21 of switch ups in personas. And personas 41:24 is ultimately what's actually going to 41:25 give us breadth and volume in the 41:26 account. And then I have number four in 41:28 here. There is actually a fourth 41:29 component of concept design. Now, this 41:32 didn't really used to exist 41:34 pre-Andromeda, but it's an additional 41:36 component that we've added in recently. 41:38 And it's because of the way it then 41:40 Andromeda will take different formats 41:42 and serve it to different people. So, as 41:43 I said before, I don't get UGC content. 41:45 I don't get dynamic product ads. And so, 41:47 you could do all the mix and matching 41:49 that you want of personas, of angles, of 41:51 offers to try to hit me, but if the only 41:53 ads in the account is UGC and DPAs. And 41:56 so, there is a fourth component to be 41:58 able to reach more people, which is the 41:59 format. You can have the same persona, 42:01 same angle, same offer, and you can 42:02 change the format and you will reach a 42:04 different audience. However, this is 42:06 nowhere near as effective as these other 42:09 components. If you do this and then you 42:10 change the format, the reality is you're 42:12 probably going to hit a lot of the same 42:14 people cuz you are talking to the same 42:15 people, you are going with the same 42:16 angle, and you are offering them them 42:18 the same thing. You'll get a bit of 42:19 novel reach, you'll obviously reach 42:20 people that don't resonate with the 42:21 current format, but this isn't like a 42:23 huge unlock great variable in the 42:25 account. And that's why it is down at 42:27 number four in the testing priority. 42:29 Now, the big issue here and the reason 42:30 why I spent so much time going through 42:31 the testing priority order is everyone 42:33 gets this completely the wrong way 42:34 round. People spend all of their time 42:37 testing formats. They go, "Oh, let's 42:39 test some UGC cuz UGC will work." UGC as 42:42 a concept doesn't work. It is the 42:44 script, it is the persona, it is the 42:46 angle, it is the offer that is contained 42:48 within this type of content that is what 42:50 works. Just like a professional photo 42:53 shoot doesn't work. The format isn't 42:55 what works, it's what's contained within 42:57 the format. Then people move up to the 42:58 persona and they go, "We need to reach 43:00 different people." Okay, we understand 43:01 that in Andromeda we need to talk to 43:03 different people, so let's talk to 43:04 different people. This is the second 43:05 priority. It is not because if your ads 43:08 aren't working, fundamentally, you have 43:10 bad messaging and you have bad arguments 43:12 for why they should buy. The creative is 43:14 just bad. And creative being bad is 43:16 normally a function of the angle and the 43:17 scripting. And so this is what you 43:18 should be fixing. Changing up the 43:20 persona is just you trying to target 43:22 different people with terrible ads. Then 43:23 they go to the offer and they say, 43:24 "Well, all the big influencers, Hormozi, 43:26 says offer is king. That's the thing 43:28 that matters the most, so we need to go 43:29 to the offer." Sure, but you're going to 43:31 once again fatigue offer testing pretty 43:33 quickly. There's only so much you can do 43:34 with your one or two product. And then 43:36 they end up at the angle, which is their 43:37 lowest priority, but it should be the 43:38 highest. So this is how you need to be 43:40 thinking through testing priority. Now, 43:41 the practical application of thinking 43:43 through this format type actually 43:45 translates directly into the account 43:47 structure, 43:48 >> [snorts] 43:48 >> which is that when you have a campaign, 43:50 you want your ad sets to be segmented 43:53 out based on concept. And the reason 43:55 being is that the targeting exists at 43:57 the ad set level. And so if you go and 44:00 put in, as I said before, a bunch of 44:02 different ads that are a bunch of 44:03 different concepts resonating with 44:05 different people, well then the ad set 44:07 that's trying to figure out who to 44:09 target is going to get confused. It's 44:11 going to go, "Wait a second, this is 44:12 resonating with men, old men, this is 44:14 resonating with young women. Like, why 44:15 are there all these different angles in 44:17 here? Who do we actually target?" Now, 44:18 yes, there is some siloing of target 44:21 demographic targeting at an individual 44:22 ad level, but a lot of it still sits at 44:24 the ad set level. And so, you're going 44:25 to get worse performance when you go and 44:27 bundle all different concepts together. 44:28 Not only that, because you can't trust 44:30 the reliability of ROAS at an ad level 44:33 very well, when you go and put all these 44:34 together and you don't hit KPI, you do 44:36 hit KPI, let's say you maybe you got a 44:38 3x ROAS here, well, then you're going to 44:40 look at these ads 44:42 and not actually understand what concept 44:44 worked because there's three concepts 44:46 here. So, what is the learning from this 44:47 ad set? Well, if it was all under one 44:49 concept, one persona, one angle, one 44:50 offer, we could go, "This concept works 44:53 really well. Let's go and make more like 44:54 this. Let's double down in this 44:56 direction. Let's not over leverage. 44:57 Let's think about portfolio management, 44:59 but we should definitely put a little 45:00 bit more resourcing here cuz there's a 45:02 squeeze available." If instead you have 45:04 three different concepts in here and 45:05 it's hitting KPI, what what's the 45:06 conclusion? Is Is all the concepts 45:08 working, but this one doesn't have a 45:09 good ROAS? So, we know that ROAS at the 45:11 ad level isn't reliable. So, like, what 45:12 is it How do we draw conclusions out of 45:14 a campaign structure with mixed-up 45:16 concepts at the ad set level? You can't. 45:17 It's very difficult to do. You just got 45:19 to hope your ads are good and that you 45:20 get a good return. But, if you want to 45:21 actually have any kind of learning 45:23 feedback loop, you want to be 45:24 structuring 45:26 the account based on concepts. All 45:28 right, so here's a full buildout that 45:30 I've just put together of a persona, 45:31 angle, and offer that I think would 45:33 actually crush. The formats we can add 45:36 as we go along here. So, and you're 45:38 going to have to excuse my writing. 45:41 Persona: casual runner training for a 45:43 sub 4-hour marathon. It's a male. 45:46 They're a gym-goer. So, they're not just 45:47 a runner, but they've gone to the gym. 45:49 They follow XYZ influencers. So, there's 45:52 particular influencers that we've called 45:53 out and we know that they're probably 45:55 following them in some capacity. And 45:56 they make about 70k a year per annum. 45:59 This is important in terms of framing 46:01 value around what we're actually going 46:03 to try to sell to them. So, where do 46:04 they actually sit in terms of income 46:06 level? In terms of the angle, we're 46:08 going to come in with science backing 46:10 because we know that this person 46:11 probably follows someone like Andrew 46:13 Huberman, so we can come in with at 46:15 least an angle around him. We're going 46:16 to go through how creatine improves 46:18 recovery, but it doesn't improve 46:20 recovery in general cuz we want to talk 46:22 to this persona. So, it improves 46:24 recovery for the long run. Now, people 46:26 are only going to be doing a long run if 46:28 they're training for a marathon. That 46:29 enables a 5-minute improvement on the 46:32 marathon time. So, we're directly 46:34 connecting the entire angle to the 46:36 persona, and we're giving them an 46:37 expected outcome. Then, the offer is 46:39 relatively straightforward. Creatine 46:41 gummies, buy two, get one free. Now, 46:43 this offer was selected by myself just 46:45 due to the income level here. So, we 46:47 probably won't increased perceived value 46:49 cuz creatine gummies are generally quite 46:50 expensive, and it's going to be hard to 46:52 push it onto someone and to reframe them 46:55 against creatine for running. The actual 46:57 format here, this is where we can chop 46:58 and change this, and we can have like 10 47:00 different ads, right? We could go into a 47:01 VSL format. Probably not going to work 47:03 as well on this age demographic, but it 47:07 is a potential option. We could go into 47:08 UGC, which I think is going to be great. 47:10 We could go into founder talking head. 47:13 We could go into a more educational 47:14 high-fi style piece. 47:17 You could also actually translate this 47:18 into a carousel. Formats go on and on, 47:20 but they're probably the starting ones 47:22 that I'd go for. VSL, I might just leave 47:24 out for now because the age is too low 47:26 for a VSL to work on. And so, that is 47:28 one concept. We're now we can come into 47:30 here, and we can start changing one of 47:32 these variables to be able to mix and 47:34 match it to different target 47:35 demographics. And so, the reality is is 47:37 that this angle will also work for women 47:42 that are training for a marathon. So, we 47:43 can change over that persona, and we 47:44 have now a whole different concept that 47:47 we can start making an ads for. Now, we 47:48 can tweak the angle a little bit, and 47:50 rather than making it for long days, we 47:52 could make it for people that are trying 47:54 to get up to a 5K run. And if you're 47:56 just started running and you're trying 47:58 to get to a 5K run, creatine is a great 48:00 way to build up initial improvements in 48:01 recovery, and therefore we can now 48:03 resonate with a different persona. So, 48:04 you can start to mix and match this 48:06 accordingly to reach way more people. 48:08 Now, on hook strategy, everyone thinks 48:10 that the hook of an ad is a trick to try 48:12 to get someone to watch the video. But, 48:14 the hook of an ad is a play towards 48:16 relevancy. What you are doing is 48:18 promising the user that what they're 48:19 about to watch is relevant and 48:21 contextual to them. And that distinction 48:23 matters enormously because a lot of 48:26 people think that they should just be 48:27 baiting people into watching. They 48:28 should be trying to hook them through 48:30 some kind of meme at the start. And yes, 48:32 that'll give you really high hook rates 48:33 if that's what you want to optimize for, 48:35 but it's not actually going to give you 48:36 conversions. Now, there's actually three 48:38 different layers to the hook. You have 48:40 the visual. This is what's actually on 48:42 the screen. 48:43 You then have audio. 48:45 This is obviously the underlying audio. 48:46 And then you actually have the copy that 48:49 overlays the visual. This could also be 48:51 the primary text that sits above the ad. 48:53 Now, the reason why I'm even putting 48:54 this as a section in the video is 48:56 because the hook is so important in 48:59 video assets because 80% of people never 49:02 watch past the hook. And so, if you're 49:03 going and serving ads into the 49:05 marketplace and 80% of people never even 49:08 watch your ad cuz they don't like the 49:09 hook, so they just skip straight past 49:11 it, well, it's pretty critical then for 49:13 us to try to optimize the hook, okay? If 49:14 most people are never going to see past 49:16 this point in the video, then let's make 49:18 sure this video is relevant, providing 49:19 value in some capacity. You're trying to 49:21 pull people through the video that are 49:22 relevant and need to watch. An example 49:24 of this in terms of resource allocation 49:26 is that you can spend an hour going and 49:28 making two hooks and then five different 49:31 bodies for this ad, or you can go and 49:33 spend the same amount of time, but just 49:35 make two bodies and then 15 hooks, and 49:39 you will probably always get better 49:40 performance here given that the actual 49:42 scripting is good on the bodies. And the 49:44 reason being is that the delta available 49:46 in performance increase is normally 49:48 higher in the hook than it is in the 49:49 body. Once again, assuming scripting's 49:51 good. If you're not good at scripting, 49:53 well then like one of these bodies might 49:54 actually be a good script and then it 49:56 will hit and you'll get a 49:56 misrepresentation of the actual impact 49:58 here of the different components of the 50:00 ad. Now, in terms of rating whether a 50:02 hook is good, cuz I do this a lot. I do 50:04 a lot of reaction videos on Instagram 50:05 saying this is a bad hook, this is a 50:07 good hook. And I'm rating how well 50:09 people have done in putting a hook 50:11 together. Well, there is actually a 50:13 grading system that you can think 50:14 through as to how good your hook 50:16 actually is. Number one is clarity. Can 50:19 a stranger understand what you are 50:21 actually talking about within the first 50:23 3 seconds or are you just trying to 50:25 confuse them as a way to hook them in? 50:26 Number two is relevance. Now, people 50:28 mistake this for a persona call out. 50:30 They mistake this for me taking the 50:32 persona that I was talking about before 50:34 and just calling them out at the start. 50:36 Hey, if you're a mom that has kids and 50:38 you're age 30 to 35 and you have this 50:40 problem, well listen up. Persona call 50:41 outs are super overrated and it's 50:43 because no one wants to hear themselves 50:45 called out. People want to hear the 50:47 problem that they're actually dealing 50:49 with and so you'll always see much 50:50 better performance agitating on a 50:52 problem than you will on any kind of 50:55 persona call out. And I have this from 50:56 personal experience. Anytime I get an ad 50:58 saying agency owners who need XYZ, 51:01 listen up. I never watch the ad. I move 51:03 on. But if they agitate on a specific 51:05 problem, it feels much more genuine. I 51:07 actually wants to I actually want to 51:08 watch through and find out what the 51:09 solution is to the problem that I am 51:11 actually currently facing. So, you want 51:13 it to be relevant but persona call outs 51:15 are nowhere near as effective as just 51:18 making the problem agitation relevant to 51:19 the person. We then have novelty. This 51:21 isn't a requirement but this is a really 51:23 good way to create white space within 51:26 the ad that you're making. And so this 51:28 concept of white space is also this 51:30 concept of purple ocean theory, which is 51:32 that you have a blue ocean over here. 51:35 You then have a red ocean over here. 51:37 Now, this is actually one of my favorite 51:38 books that I've ever read. I actually 51:39 reference it in a lot of content, blue 51:41 ocean theory. This is that you want to 51:42 reposition your product into a market 51:45 that differentiates you from everyone 51:46 else so that you're not competing on 51:48 pricing, you're not even getting 51:49 compared, and you're seen as a 51:50 completely novel product or service 51:52 within the market. Red ocean means that 51:53 you're selling the same stuff as 51:55 everyone else, and then you're just 51:56 getting priced down to zero. These are 51:57 terrible businesses. These are 51:58 incredible businesses. The issue with 52:00 blue oceans is they don't really exist 52:03 because if you're going truly blue, 52:04 you're selling something that's never 52:05 been sold before, and that's kind of 52:08 risky cuz there's no actual market 52:10 knowledge. You need to do tons of 52:11 education, and there's no proven product 52:13 market fit, either. So, yeah, you could 52:15 reposition into something that you think 52:16 people might want, but they might not 52:17 even want it. And so, you realistically 52:19 want to sit like somewhere in the middle 52:21 here where you're repositioning into 52:22 something novel, but it still has proven 52:25 product market fit. It's still proven 52:26 that it is going to work. It's just not 52:28 what everyone else is doing. It's 52:29 slightly adjacent. And so, we do this a 52:31 lot in our own content in terms of 52:33 format testing. So, we'll rotate in 52:34 different formats cuz we want to test 52:36 new things. We'll take ideas from other 52:37 industries, from other people. We'll 52:39 come up with our own ideas to try to cut 52:41 through and have novelty that people 52:43 haven't seen before. I would say this 52:44 YouTube video is actually an example of 52:46 that, which is that there isn't many 52:47 creative courses on YouTube for meta ads 52:50 that are 2 plus hours long. So, you want 52:51 to be thinking through that concept, 52:53 too, in the hook. Now, once again, it's 52:54 not a requirement, but it will help in 52:57 standing out without having some kind of 52:59 gimmicky tactic to hold people through 53:01 the ad that's not actually unique. The 53:03 question I would be asking yourself here 53:05 is has this hook been done a million 53:08 times? And if the hook's been done a 53:09 million times, probably don't do it. The 53:11 next is specificity. So, is there a way, 53:13 and you're not going to be able to do 53:14 this in a lot of hooks, but in our 53:15 specific hook, is there a way we can use 53:18 numbers, maybe quantify names, or maybe 53:21 even quantify outcomes right out of the 53:23 gates 53:24 that's going to build authority into the 53:27 hook. Anytime you can start a hook by 53:29 name-dropping someone that's super 53:31 famous that's relevant to the product in 53:32 some capacity, it always does better. If 53:34 you can name-drop some kind of number, 53:36 67% of people got this result, then once 53:39 again, it's going to build authority. If 53:41 you can speak straight to a specific 53:42 outcomes, it's going to build authority 53:44 and it's going to be more specific to 53:45 your target demographic that you want to 53:46 resonate with. And then lastly, we have 53:48 credibility, which is fairly obvious. 53:49 Now, with credibility, you're usually 53:51 not going to be able to do this through 53:52 the copy or through the audio. It's 53:54 typically going to have to be done 53:56 through the visual. And so you were 53:57 going to need either like someone famous 53:59 that pops up at the start, you're going 54:01 to need someone that is uh contextually 54:03 relevant. So if you're selling 54:05 supplements, you have a doctor who pops 54:06 up. If you're selling running 54:08 supplements, it's obviously like a 54:10 runner that pops up. You want to be 54:11 thinking about how you can build 54:12 credibility. Now, you don't need 10 out 54:13 of 10 scores on all of these to be able 54:15 to have a good hook. You really only 54:16 need a couple of them done well. This is 54:18 how you should be thinking through hook 54:19 creation. Now, let me give you a bunch 54:21 of hooks that you can go and start 54:23 using. Number one is problem agitation. 54:25 Do you want to lead with the pain that 54:27 the viewer is already feeling and you 54:29 want to make them feel seen. An example 54:31 of this is "Are you struggling with back 54:33 pain from desk work?" This will only 54:36 resonate with problem aware audiences. 54:38 So you can get really good performance, 54:39 but it isn't going to be scalable above 54:41 that stage of awareness. Number two is a 54:43 contrarian truth. So you challenge a 54:44 common assumption, you create cognitive 54:47 friction that demands a resolution. An 54:49 example here is everything you've been 54:50 told about protein timing is wrong. And 54:52 this works across all levels of 54:54 awareness and it works strong in organic 54:56 and paid. Number three is specific 54:57 proof. So you want to lead with a 54:59 measurable outcome, numbers, time 55:00 frames, specificity. An example of this 55:03 is "I lost 12 kilos in 90 days without 55:06 giving up pasta." So you build 55:07 credibility throughout the hook. The 55:09 next one is a curiosity gap. We spoke 55:12 about this before, which is that you 55:14 create an information gap that can only 55:15 be closed by watching. An example is 55:18 "What the top 1% of brands know about 55:19 meta ads that you don't." And a funny 55:21 side note is that my average view time 55:23 when I use that hook or some variation 55:25 of that hook is over 45 seconds. And so 55:28 that hook actually works incredibly well 55:29 because it creates an information gap in 55:31 my own organic content. Truth bomb, 55:33 there's actually one specific client 55:34 where this is just unbelievably well 55:36 for. So that's why I threw it in here 55:38 cuz I don't see many people doing it, 55:39 which is that you lead with an 55:40 uncomfortable truth about the product's 55:42 price, ingredients, or industry. The 55:43 honesty itself becomes the hook. And so, 55:46 the hook in this instance is something 55:47 along the lines of, "This product costs 55:49 $120. Yeah, it's super expensive, but 55:53 it's our best seller and everyone buys 55:54 it. Here's why." It ends up working 55:56 exceptionally well for premium or higher 55:58 average order value products where the 56:00 price ends up being the main objection 56:02 that people have. So, you confront it 56:04 way up front, and it actually works 56:06 really well at cutting through on cold 56:07 audiences. Cuz even if I've never seen 56:08 the product before, if someone's opening 56:10 with, "This thing is expensive." I'm 56:12 like, "Okay, what is that thing? Teach 56:13 me more." And then it can go into a very 56:15 very educational piece. Obviously, 56:17 that's only going to be relevant to very 56:19 specific industries. The next is 56:22 psychological confrontation. And so, 56:24 this is being direct and pattern 56:25 interrupting with a statement that feels 56:27 like a friend calling you out. An 56:29 example here would be, "You know that 56:31 drawer full of products you never use? 56:33 That's the problem." Or, "Stop buying 56:35 things that make you feel guilty for not 56:36 using them." It's not clickbait, but 56:38 it's playing on an emotional pain point. 56:40 And so, the viewer ends up feeling 56:42 personally addressed, which ends up 56:43 acting as a scroll stopper. Once again, 56:45 all of these things also apply into 56:47 organic, which is where I think getting 56:48 good at organic content is actually a 56:50 very critical skill for getting good at 56:52 creative strategy because it just 56:53 translates into paid. We'll keep going 56:55 here, which is you can also do sensory 56:59 I'm going to call this a texture hook as 57:00 well. This does unbelievably well in 57:02 fashion right now. This is just an ASMR 57:04 hook. So, you don't need any words, 2 to 57:06 3 seconds, close-up of the fabric being 57:08 handled, a product being squeezed, 57:10 liquid being poured, package being 57:12 opened, something ASMR related. These do 57:14 incredibly well. Uh we have a lot of ads 57:16 right now that are high performers on 57:18 this hook. You've then got a founder's 57:20 letter. You can go and apply all of 57:21 these into most brands. So, you want to 57:23 open with the founder and their origin 57:25 story. So, we started this because {dot} 57:27 {dot} {dot} Or, "Here is why I left my 57:29 career to go and build this." The more 57:31 risk you can build within the hook as 57:33 well, the better. And so, you want to 57:34 build tension. And this is once again a 57:36 big organic strategy that you just want 57:38 to translate into paid, which is that 57:40 you want to have high stakes within the 57:42 hook if you're going to have a founder's 57:43 letter style hook, so that people have a 57:45 reason to continue watching. I 57:47 remortgaged my house and put $300,000 57:50 down to be able to order my first order 57:53 of this product. And here's what's 57:54 happened since. Like that hook, I could 57:57 almost guarantee it would do well. Now, 57:58 obviously, that has to be a real 58:00 circumstance, which is why that hook 58:01 does well. Right? It's things that you 58:03 can't fake. If you actually did put so 58:06 much risk on the line, it's a great 58:08 angle to be able to talk about that 58:09 risk. The last one here is social proof. 58:13 So, you want to lead with some kind of 58:14 credibility marker. 90% of customers saw 58:16 results in 30 days. Now, this is nine 58:18 different hook types that you can go and 58:20 take. You can probably apply at least 58:21 seven of these into your particular 58:23 brand right now. There are infinite hook 58:25 strategies. Right? I could write hooks 58:26 here all day long. I could fill the 58:28 whole board with 50 different hook 58:29 types. You can go to Claude or ChatGPT 58:31 right now and say, "Give me all the 58:32 different hook types that I can run." 58:33 The idea is that you want to take hook 58:36 strategies and you want to translate it 58:37 into your brand, but the scripting is 58:39 going to be critical. And where you're 58:40 going to really see uh a standout from 58:43 competitors is over here in this novelty 58:45 piece on whitespace. So, you can become 58:47 an eight-figure brand by using these 58:49 hook strategies and having relatively 58:50 good scores over here, but if you want 58:53 an ad that's going to hold half a 58:54 million dollars in spend, you do 58:55 probably want to generate something 58:57 novel at the beginning of the creative 58:59 that's going to stand out from 59:00 competitors. And then what's going to 59:01 happen is all your competitors are just 59:02 going to go and copy it. And so, you 59:04 will set the trend. Now, I've spoken 59:05 about the hook a lot, but an an 59:07 important subcomponent of the hook is 59:09 the bridge, which is where you move 59:11 between specific components within the 59:12 video. Now, the two big bridges that 59:14 often occur is when you go from hook to 59:17 body, and then when you go from body to 59:19 CTA. These two bridges often don't end 59:22 up being done well. And so, what ends up 59:24 happening is if you think about product 59:26 awareness on the Y axis and then you 59:28 think about time going through the video 59:30 asset on the X axis, what ends up 59:32 happening is people hook you in and 59:34 you'll become aware that it's an ad and 59:36 then suddenly as you bridge to the body, 59:38 some people will just instantly 59:39 introduce the product. So they'll hook 59:41 you with something like, are you trying 59:43 to train for a marathon and you want 59:44 better recovery? Well, creatine from 59:47 Blue Zones Digital is your solution. 59:49 That ad won't do well. Hook was good, 59:52 bridge ruined the whole thing. The whole 59:54 body could be good. Like the body 59:55 scripting could be perfect, hook 59:56 scripting could be perfect, call to 59:58 action could be great, but the bridge 1:00:00 ruins the whole asset. And so you don't 1:00:03 want your asset to look like this where 1:00:05 there's this massive step change in 1:00:07 product awareness as a user goes through 1:00:09 the creative. Instead, you want to blend 1:00:11 these out so that people are slowly 1:00:13 becoming aware that they're being sold 1:00:15 to. And so instead, you're looking like 1:00:17 this. And so I don't really even need to 1:00:19 give you that much information on how to 1:00:20 do this. To me, it's pretty intuitive 1:00:23 which is just blend in the introduction 1:00:25 of the product. Don't just suddenly 1:00:27 introduce it or else people go quickly 1:00:29 from oh, I was watching something that 1:00:30 was educational or that was teaching me 1:00:32 to oh, I'm actually being sold a 1:00:33 product. And people don't intuitively 1:00:35 like that feeling. This also, this jump 1:00:37 in bridge will typically ruin the flow 1:00:40 of scripting in most very good top of 1:00:42 funnel unaware ads. And so when it comes 1:00:45 to really good top of funnel unaware 1:00:47 ads, what they do is they move people 1:00:49 through the stages of awareness as you 1:00:51 progress through the ad. This is 1:00:52 actually something that VSLs, video 1:00:54 sales letter ads do unbelievably well 1:00:57 because they get the capacity to teach 1:00:59 through the asset. And so you start with 1:01:01 unaware. So you're hooking people in 1:01:03 with something that they might be 1:01:04 interested in, usually some kind of 1:01:06 curiosity gap. Then you're making them 1:01:07 problem aware. Okay, now I understand 1:01:09 that there's a problem with drinking 1:01:11 water. If I just drink water every day, 1:01:13 not only does it not have electrolytes 1:01:15 which apparently I need, I didn't even 1:01:17 realize this was a problem, but also tap 1:01:19 water's incredibly bad for me. Okay, 1:01:20 that's interesting. What's the solution? 1:01:22 Oh, the solution is is that I need this 1:01:23 amount of salt in this kind of ratio and 1:01:25 then I need to stop using tap water. 1:01:27 Okay. Ah, there's actually a product 1:01:30 that solves this problem. It's canned 1:01:32 mineral water with electrolytes in it. 1:01:34 Ah, okay. I was unaware of anything. I 1:01:36 became problem aware. I became solution 1:01:38 aware. I'm now product aware. Okay, this 1:01:40 is going to fix all of these problems 1:01:41 that I didn't realize I had and now I 1:01:43 move down into the aware stage of 1:01:45 awareness where maybe I convert on a 1:01:46 retargeting ad. Now, this is generally 1:01:48 how you want to be thinking through 1:01:49 scripting on really good top of funnel 1:01:50 ads, but what most people do is that 1:01:52 they'll start really unaware. They'll 1:01:54 then go problem aware and then like in 1:01:56 here they'll just introduce the product. 1:01:58 And you're like, "What are we doing? We 1:01:59 haven't even moved them to the next 1:02:00 stage of awareness. We can't introduce 1:02:02 the product yet." And people do this cuz 1:02:04 they think in retention curves. And so 1:02:06 they go, "Well, if we look at the 1:02:07 retention curve on an ad, which is how 1:02:09 many people are still watching on 1:02:10 average as we move through a creative, 1:02:12 the retention curve will look like 1:02:13 this." And so intuitively you go, "Well, 1:02:16 80% of people are going to drop off by 1:02:19 the 90-second mark. So we need to 1:02:20 introduce our product before the 1:02:22 90-second mark cuz we don't want most of 1:02:25 the users to not know they're being sold 1:02:26 our product. It's the wrong way to think 1:02:28 through it. Okay, you actually want to 1:02:30 introduce your product as late as 1:02:31 possible in the creative because you 1:02:33 want people to be problem aware and 1:02:34 solution aware and really pre-framed and 1:02:37 engaged, ready to buy once the product 1:02:39 is actually introduced. You don't want 1:02:41 to introduce the product too early 1:02:42 before they're actually aware of what 1:02:44 the problem and solution is. And so you 1:02:46 often end up getting way better 1:02:47 conversion rates out of an asset by 1:02:49 moving the product introduction and call 1:02:51 to action way later because these people 1:02:54 are now way more qualified to actually 1:02:56 buy and are way more higher intent. And 1:02:58 so yes, you can introduce it sooner, but 1:03:00 they're not high intent enough that you 1:03:02 will actually have worse conversion 1:03:03 rates and even though it's more total 1:03:05 people, you will ultimately get worse 1:03:07 total purchase volume out of the asset. 1:03:09 And so that is something to be really 1:03:11 careful of when you're thinking about 1:03:12 bridging. Now, to wrap this section up, 1:03:14 one final question that I get a lot is 1:03:17 does hook testing still work post 1:03:19 Andromeda? Because isn't the whole point 1:03:21 in Andromeda that we need diversity? And 1:03:23 if I have a 90-second Facebook ad, me 1:03:27 just changing the first 3 seconds at the 1:03:29 start, that's not diversity. 1:03:31 Right? That's the same ad. We just 1:03:33 changed the first 3 seconds. 1:03:35 Meta will still recognize a different 1:03:37 hook as a different creative. It will 1:03:39 get its own creative ID. It will not get 1:03:42 pulled together with this asset. It will 1:03:44 go and reach a novel unique audience. 1:03:46 Now, obviously, there is going to be 1:03:48 some degree of overlap because 87 1:03:50 seconds of this asset is going to be 1:03:52 talking to the same person with the same 1:03:54 problem with the same offer. So, if we 1:03:56 actually look at the concept framework, 1:03:58 it is the same concept. It's just the 1:04:00 hook chain. So, the question becomes is 1:04:01 hook changes worth it? Yes. Why is it 1:04:04 worth it? Because it is so low effort. 1:04:06 Rotating in new hooks is almost free. 1:04:09 Okay, when you record the ad, you just 1:04:11 record the hooks six times and it takes 1:04:13 an extra few minutes and you've got six 1:04:15 different hook variations. Or you can 1:04:17 take existing high performers and you 1:04:18 throw new hooks on them. We have had ads 1:04:20 that have spent 50 100 thousand dollars 1:04:22 fatigued, then we just rotate 50 new 1:04:25 hooks on and we get another 100 thousand 1:04:27 dollars to spend at the same efficiency. 1:04:29 And those hooks took like 1 hour to 1:04:32 shoot. And so, you can very much so keep 1:04:34 winners alive through changing hooks. 1:04:36 You can also take ads that didn't work 1:04:38 and then suddenly make them work by 1:04:40 fixing the hook that had a really bad 1:04:42 hook rate. So, yes, absolutely hook 1:04:44 testing still works post Andromeda. You 1:04:46 should be doing it. It is still the 1:04:47 highest leverage point in videos, 1:04:50 organic, ads, YouTube, everything. All 1:04:52 right. So, when it comes to formats, 1:04:54 there's five questions that you need to 1:04:55 be asking yourself when you're thinking 1:04:57 on a format. Number one, what is the 1:04:58 concept? We need to define the concept 1:05:00 first before we even think about 1:05:01 formats. You want to go out, build your 1:05:03 persona angle offer, then come back to 1:05:05 thinking about the format. You don't 1:05:07 start with the format and work 1:05:08 backwards. This is the last step. 1:05:10 Secondly, does this concept require 1:05:12 education? Now, if it requires 1:05:14 education, you likely need to go into 1:05:15 some longer form asset that's going to 1:05:18 educate. Number three, can the value 1:05:20 prop be communicated in a single frame? 1:05:22 This is obviously going to lean yourself 1:05:24 towards images, carousels, something 1:05:26 that's easier to produce. Can obviously 1:05:28 still be in a video, but you do have an 1:05:30 advantage with images that they're very 1:05:32 easy and cheap to produce and you can 1:05:33 produce a lot of them. So, anytime you 1:05:35 can produce statics over videos, you 1:05:38 generally want to do it. Now, there's 1:05:39 two different approaches to statics, if 1:05:41 we just leave these out for the moment, 1:05:42 which is number one, you can use statics 1:05:44 to quickly test different angles and 1:05:46 different concepts, get fast feedback, 1:05:48 and then develop that into a video 1:05:50 format where you can go a lot deeper and 1:05:52 create a more scalable asset. And then 1:05:54 number two is that images are very cheap 1:05:57 to make and so they often end up 1:05:59 outweighing videos in the cost of 1:06:01 production. So, if you look at the 1:06:02 average spend of an image versus a video 1:06:04 in your account, it might actually make 1:06:05 a lot of sense to just go way harder on 1:06:07 images because the cost of production is 1:06:08 super low and the average revenue 1:06:10 generation is decent in proportion. And 1:06:12 so, this is what is called cost of 1:06:14 production against the average revenue 1:06:18 per ad. This is something that we 1:06:19 calculate for everyone during the audit 1:06:21 process because we want to understand 1:06:23 what is the average revenue per ad that 1:06:25 you put into the account. For most 1:06:27 people, it's about like $2,000-ish. 1:06:29 And then what is your cost of production 1:06:31 against this delineated down to 1:06:33 contribution margin. And so, we minus 1:06:36 off gross profit, we minus off the cost 1:06:38 of serving the ad, we might find that 1:06:40 average contribution margin per ad unit 1:06:42 is, let's say, $350. 1:06:45 How much is it costing us to make an ad? 1:06:46 It might be costing us $400. Okay, this 1:06:49 equation doesn't work. We're currently 1:06:50 paying more to produce an ad than we're 1:06:52 getting contribution margin out of each 1:06:54 ad. We can then go and delineate this 1:06:56 down to images specifically. We might 1:06:58 find out images cost us about $50, but 1:07:01 they're producing us $300. 1:07:04 Videos are costing us 500 and we're only 1:07:07 getting 400 here. And so we actually 1:07:09 find out even though videos do perform 1:07:10 better, the actual cost of production 1:07:13 versus expected contribution margin is 1:07:14 worse than image assets and therefore we 1:07:16 should actually just double down on 1:07:18 images. And I know people doing 1:07:20 $200,000 a day in revenue on just AI 1:07:23 images. And so there's absolutely no 1:07:25 reason as to why you can't do hundreds 1:07:26 of thousands of dollars a day on images 1:07:29 only. And if you can't, it just ends up 1:07:30 being a skill issue. Number four is does 1:07:32 the product need a demo? This is just 1:07:34 going to be a general question that 1:07:35 applies to the business as a whole, 1:07:37 which is how much education do we need 1:07:39 to force people through because 1:07:41 obviously you're going to be relatively 1:07:42 restricted with images then unless you 1:07:44 get really fancy and you you have a good 1:07:46 designer. And then lastly, which ties 1:07:47 into both of these points, is what stage 1:07:49 of awareness are we sitting in? Okay, if 1:07:51 we're sitting in a very high stage of 1:07:53 awareness, we need a creative that is 1:07:55 going to be a longer format that can 1:07:56 educate, typically a video. If we are 1:07:58 sitting at a very low stage of 1:08:00 awareness, like the product aware or 1:08:02 aware stages, uh we can just go for DPAs 1:08:05 and images and get away with very bottom 1:08:07 of funnel ads. I want to reiterate that 1:08:08 the belief that image ads are bottom of 1:08:11 funnel and video ads are top of funnel 1:08:13 is fundamentally a skill issue. Once 1:08:15 again, I have seen multiple ad accounts, 1:08:17 I've talked to multiple people that do 1:08:19 hundreds of thousands of dollars per day 1:08:21 in revenue off just that. And so if 1:08:24 you're saying, "Oh, statics don't work, 1:08:25 they're bottom of funnel assets." It's 1:08:27 because yeah, the way that you design 1:08:28 them and the way that you approach them 1:08:29 and concept them out is bottom of 1:08:31 funnel. But you can make top of funnel 1:08:32 statics. We're now going to cut out two 1:08:35 three ad breakdowns that I have done on 1:08:38 static images. We're going to go through 1:08:39 an organic tweet format, which is an 1:08:41 IM-8 health ad. We're going to go 1:08:43 through a creatine gummies ad. And then 1:08:45 we're going to go through a GLP-1 hair 1:08:47 loss ad and you will see three statics 1:08:49 that are holding tons of spend on top of 1:08:51 funnel. This is an incredible static ad. 1:08:53 4.7 thousand likes down here. It's by 1:08:56 IM-8 and let's break down exactly why 1:08:58 it's so good. Number one, it's a new 1:08:59 format. They're actually using a tweet 1:09:01 here and they're re-posting it, so it 1:09:03 feels organic to the platform, but it's 1:09:06 actually an ad. Now, further to this, 1:09:07 they're reinforcing it by using not a 1:09:10 partnership ad, so they're not 1:09:11 collaborating with IMA and Dr. James, 1:09:14 but it's just Dr. James. So, they're 1:09:16 running it as effectively a whitelisting 1:09:18 ad through this profile, so you don't 1:09:20 know that this is an ad. In fact, you 1:09:22 would look at this for a very long time, 1:09:23 and it's only until you start getting 1:09:25 through the copy that sits under the 1:09:27 creative over here, that you actually 1:09:29 realize that you're getting sold 1:09:30 something. So, what's so good about the 1:09:31 static over here, other than it being a 1:09:32 common format and that you're not really 1:09:34 knowing that it's an ad, so it feels 1:09:35 organic? Well, symptoms of vitamin 1:09:37 deficiencies. So, right away, we're 1:09:38 problem agitating into okay, here's some 1:09:42 education around how vitamin 1:09:43 deficiencies could be impacting you. 1:09:45 Then, this is an unbelievably wide set 1:09:49 of symptoms. So, if you have dry eyes, 1:09:51 if you have poor vision, if you have low 1:09:52 energy, if you have bad skin, if you 1:09:54 have cracks on your mouth, if you have 1:09:56 fatigue, if you have bleeding gums, if 1:09:57 you have painful muscles, if you have 1:09:59 excessive bruising, all of these things, 1:10:02 very, very wide total addressable 1:10:03 market. Well, then, it's an issue with a 1:10:06 vitamin deficiency, and it tells you 1:10:07 exactly what the vitamin deficiency is. 1:10:09 So, you get education around okay, 1:10:11 here's my problem, here's the solution. 1:10:13 Then, you start probably reading the 1:10:15 copy, and you go, "The best multivitamin 1:10:17 equals IMA, daily ultimate, 10% off. 1:10:20 Here's the thing." Then, you get more 1:10:22 educational pieces here about how the 1:10:24 product actually works, and obviously, 1:10:25 because this is an ad, a big call to 1:10:27 action will appear at the bottom here 1:10:29 that it will allow you to click straight 1:10:30 off the site. So, it problem agitates to 1:10:32 a very wide target demographic, it then 1:10:35 educates about exactly what the solution 1:10:38 is in terms of vitamin to solve your 1:10:39 particular problem. It does it in a 1:10:41 format that's super organic. It then 1:10:43 maintains that organic feeling by not 1:10:45 having the brand and company name 1:10:47 attached with the post, and then this is 1:10:49 obviously driving off to cold audiences, 1:10:51 probably driving off to this guy's 1:10:52 audience, which is even better up 1:10:54 because we're obviously piggybacking off 1:10:56 his organic. This is an advertorial 1:10:58 static card. These are really good. They 1:11:00 perform incredibly well. They only 1:11:02 perform well though if you go to an 1:11:03 education-based landing page or you use 1:11:05 the primary text to educate heavily. So, 1:11:07 let's break it down. The best hair 1:11:09 growth products of 2025 and the ones to 1:11:11 avoid. Here's what the big brands are 1:11:13 not telling you. And then we have a 1:11:14 bunch of different products here. So, we 1:11:16 know or we at least assume, okay, this 1:11:18 is some kind of blog. This is some kind 1:11:20 of article that's going to educate us 1:11:21 around what hair growth products we 1:11:23 should actually be buying. Now, even 1:11:24 better, it's running through a doctor's 1:11:26 page. Okay, so there's a bit of 1:11:28 authority here. As a dermatologist, I 1:11:30 set out on a mission to find the most 1:11:31 effective hair growth products. Not very 1:11:33 long copy, so they're not selling us on 1:11:34 anything here. Okay, so we want to find 1:11:36 out more. So, what do we do? We go and 1:11:38 we go and click learn more. Now, from 1:11:39 here we end up on a website called 1:11:41 dermatologist reviews. Now, this website 1:11:43 is owned by the brand that's trying to 1:11:45 sell you on something. This is a really 1:11:47 good funnel. So, hi, who is the person? 1:11:50 Why do they have authority? This is the 1:11:51 blog posted. I set out on a mission and 1:11:54 here are just the products. Right. And 1:11:55 then it goes into all of the reasons as 1:11:58 to why and it goes into selling that 1:11:59 product. And then it does go into the 1:12:01 other ones, but it goes into why yeah, I 1:12:03 probably wouldn't choose these. They're 1:12:04 not the best. Now, why is this a 1:12:06 giveaway that this is actually a brand 1:12:08 pushing? Well, because if we look at the 1:12:09 rankings here, you can't click on any of 1:12:11 these, right? But you can click on this 1:12:13 one. And if you click on this one, goes 1:12:14 to a custom landing page that then takes 1:12:16 you through a quiz funnel. You go 1:12:18 through a quiz funnel and then you get 1:12:19 pushed through to the conversion. So, 1:12:21 this is a really good funnel that 1:12:23 they're running here. I really like it. 1:12:24 And almost to make sure they're a little 1:12:25 bit compliant, I imagine. As you scroll 1:12:28 down, this takes you to the same funnel, 1:12:29 but as we continue going down, look at 1:12:30 all the green ticks here. So many green 1:12:32 ticks, one red. Only available online. 1:12:34 And then him and hers, oh, widely 1:12:37 available online. Oh, that's the best 1:12:38 thing about it. All of these negatives, 1:12:40 but you can still click here and it 1:12:41 takes you to an Amazon affiliate link. 1:12:44 And so they're like, "If people are not 1:12:45 clicking on us, then we may as well 1:12:46 still redirect them somewhere." And then 1:12:48 obviously you can see they've put the 1:12:50 the into reviewing all of these and all 1:12:51 the reasons as to why they wouldn't buy 1:12:53 them. This is a super effective funnel. 1:12:54 I imagine this is doing quite well for 1:12:56 them. If your target demographic is 50 1:12:57 plus, you can serve on Facebook feeds, 1:12:59 this ad type crushes. It's static image 1:13:02 ads with long-form copy. And the reason 1:13:03 why they do so well is they go through a 1:13:05 massive story that speaks to an unaware 1:13:08 audience and slowly pulls them into 1:13:10 being problem aware, pulls them into 1:13:12 being solution aware, then pulls them 1:13:13 into being product aware, and then 1:13:15 finally, they're all the way at the 1:13:17 stage in which they can convert. These 1:13:19 ads will crush on cold audiences, but it 1:13:22 takes an enormous amount of skill to 1:13:23 actually be able to write this copy 1:13:25 properly. Like, yes, you can be AI 1:13:26 assisted, but if you don't know 1:13:28 fundamentals of how to emotionally 1:13:29 appeal to a specific target demographic, 1:13:31 then it's not going to work. And this is 1:13:33 ultimately where getting good at 1:13:34 copywriting really matters. If you can 1:13:36 get good at copywriting and you can get 1:13:37 good at pulling people through the 1:13:39 stages of customer awareness, then that 1:13:41 is ultimately how you can create these 1:13:43 static image ads with long-form copy 1:13:45 that are going to absolutely crush. Now, 1:13:47 in addition to that, to keep it 1:13:48 contextual to 2026 in the time in which 1:13:50 this video is being made, a type of 1:13:52 video asset that does incredibly well 1:13:54 right now on older target demographics 1:13:56 is a VSL, a video sales letter. The 1:13:58 reason why these types of ads do so well 1:14:00 is because they pull people through all 1:14:02 of the stages of awareness and they're 1:14:03 highly educational. You can also really 1:14:05 leverage AI within them. We make a lot 1:14:08 of just AI VSLs end-to-end, and they do 1:14:11 quite well. And then some of them do 1:14:12 unbelievably well and hold hundreds of 1:14:14 thousands of dollars in ad spend. So, 1:14:15 what we're going to do now is I'm going 1:14:17 to pop up a few examples of VSLs and me 1:14:19 breaking them down so you can see an 1:14:21 example of what this ad type looks like. 1:14:23 >> IBS, leaky gut, or bloating, you're 1:14:26 probably missing this one thing. 1:14:29 >> That's a really good VSL. First off, it 1:14:31 starts with a persona callout. If you 1:14:33 have these issues, keep watching. 1:14:34 >> Doctor probably told you to just manage 1:14:36 it with a course of pills. That's 1:14:37 because 1:14:38 >> Then it's going into objection handling, 1:14:40 the common way to fix it. 1:14:41 >> They're treating the symptom, not the 1:14:43 cause. But here 1:14:44 >> Then it problem educates and opens a 1:14:46 curiosity gap. Okay, I thought it was 1:14:48 actually going to cause us to fix the 1:14:50 problem, but this ad is saying it's not. 1:14:52 All right, I'm going to keep watching. 1:14:53 >> Here's what they're not telling you. 1:14:55 Your gut isn't broken, it's starving. 1:14:57 Your gut is an ego 1:14:58 >> I don't love that because it's an AI 1:15:00 script. So, that's called contrast 1:15:02 negation, which is where you say it's 1:15:03 not X, it's Y. You see a lot of this in 1:15:05 a lot of scripting these days. I 1:15:06 probably would take that out. 1:15:08 >> an ecosystem of trillions of bacteria. 1:15:10 When the good bacteria die and bad 1:15:12 bacteria take over, everything 1:15:13 collapses. When you get inflammation, a 1:15:16 leaky gut, and your body stops absorbing 1:15:18 nutrients. This 1:15:19 >> Really good educational piece. 1:15:21 >> is dysbiosis, and it's not rare. 40% of 1:15:24 people have it. Most of them don't even 1:15:25 know. You might be among them, but 1:15:27 here's the problem with 1:15:29 >> Really good way to enforce that no, you 1:15:31 do have this problem because when you 1:15:33 prescribe a problem to someone, 1:15:35 particularly when it's like a disease or 1:15:37 it's a physical medical issue, the 1:15:39 immediate reaction is, well, I probably 1:15:41 don't have this. I imagine this is rare. 1:15:43 So, instantly increasing the TAM by 1:15:44 knowing that 40% of people actually have 1:15:46 this problem is a really good way to 1:15:47 keep people watching to go, okay, maybe 1:15:49 I do have the problem. Tell me a little 1:15:50 bit more about it. 1:15:51 >> with every gut supplement you've tried. 1:15:53 They just throw probiotics at it. That's 1:15:55 like sending a single soldier into a war 1:15:57 zone. One probiotic can't survive in a 1:15:59 toxic gut. It dies. It doesn't colonize. 1:16:02 It doesn't help. I AM 8 daily ultimate 1:16:04 essentials doesn't just add probiotics. 1:16:07 First, digestive enzymes clear 1:16:09 >> I don't love the bridge. If the ad 1:16:11 team's watching this, I would extend for 1:16:13 another 15 seconds and educate around 1:16:15 how the I AM 8 ingredients actually fix 1:16:17 the issue and then introduce I AM 8 as 1:16:20 the solution that has those ingredients. 1:16:22 >> and undigested food. Second, prebiotics 1:16:24 create a safe environment for good 1:16:26 bacteria. Third, 10 billion CFU of 1:16:28 probiotics flood in and colonize. 1:16:31 Fourth, postbiotics strengthen your gut 1:16:33 lining and stop the leaks. 1:16:34 >> So, the 10 to 15 seconds would be 1:16:36 saying, uh the way that you get 1:16:38 probiotics to work is that you have 1:16:39 prebiotics and postbiotics around it, 1:16:41 and that's actually what creates the 1:16:43 system to work. Fortunately, Biome Aid 1:16:45 has put all of those three solutions in 1:16:46 one sachet, so you can get them all at 1:16:48 once, and it can fix the issues that we 1:16:50 agitated at the start. 1:16:51 >> It's a complete military operation, not 1:16:53 a single soldier. 1:16:54 >> That's an AI script once again, contrast 1:16:56 negation. I'll try to avoid that. 1:16:58 >> And it works. In a 12-week clinical 1:17:00 trial, 85% of 1:17:01 >> And then it goes into social proofing, 1:17:03 it keeps going on. Overall, really good 1:17:04 ad. Just some small points of 1:17:05 improvement. 1:17:06 >> This ad has 25 million views. It's been 1:17:08 running for 3 to 4 years now. It's 1:17:10 absolutely crushing, and I'll show you 1:17:12 what it does so well. It takes a really 1:17:14 common format of VSLs, and then it adds 1:17:16 objection handling UGC to the front. So, 1:17:19 let's click play and start to watch the 1:17:20 ad. 1:17:21 >> I want to talk to you about something 1:17:22 that's been making waves on social media 1:17:23 lately. You've probably seen this guy 1:17:26 all over your feed claiming that cutting 1:17:28 carbs is 1:17:29 >> So, they're using the split screen for 1:17:30 engagement and to add more context to 1:17:32 the video. 1:17:32 >> necessary for weight loss. No need for 1:17:34 long hours on the treadmill, either. 1:17:36 Now, I know it sounds too good to be 1:17:37 true, but let me tell you, there's more 1:17:39 to this than meets the eye. 1:17:40 >> I decided When you're watching this, you 1:17:41 think that this guy is obviously going 1:17:43 to disagree with the cutting carbs 1:17:45 because he's coming from a news 1:17:46 authority perspective, and it kind of 1:17:48 feels like a collab video. Like, this 1:17:50 guy is talking about how you don't need 1:17:51 to eat carbs, it's crazy. Like, what is 1:17:53 he talking about? 1:17:54 >> I decided to delve into this topic and 1:17:55 do some research. Turns out, this guy is 1:17:57 a celebrity trainer who's gained a 1:17:58 massive following for challenging 1:18:00 mainstream advice. He claims that what 1:18:02 we've been told about weight loss is 1:18:03 wrong, completely wrong. And you know 1:18:05 what? Makes a lot of sense when you 1:18:06 listen to what he has to say. So, 1:18:09 >> So, that's the first time in which he 1:18:10 validates the person he's talking about, 1:18:12 but everything through that first 28 1:18:14 seconds was contrarian hooks, contrarian 1:18:16 takes over and over again. 1:18:18 >> Celebrities don't cut carbs. 1:18:20 >> That ain't how you lose weight. I don't 1:18:22 do that. None of my clients do that. 1:18:25 Nobody that has ever asked me how to 1:18:26 lose weight have I ever told them to do 1:18:28 that because it simply just doesn't 1:18:30 work. For anyone who doesn't know their 1:18:31 metabolic type, listen up. The fastest 1:18:34 way to burn fat and get shredded is not 1:18:36 keto, it's not paleo, it's not 1:18:39 carnivore, it's not vegan. 1:18:41 >> Note that this creative is just going 1:18:44 very, very hard on contrarian takes. 1:18:47 It's pretty much the entire strategy of 1:18:48 this whole ad is like, how do we put as 1:18:50 many contrarian takes in here as 1:18:51 possible to continue to string people 1:18:53 along as we kind of educate and we kind 1:18:55 of hint towards that we have our own 1:18:57 system. And then eventually we obviously 1:18:59 bridge into that system. So they're 1:19:00 creating a curiosity gap through this 1:19:02 whole thing. Okay, if no one needs to 1:19:04 cut carbs, then what is the actual 1:19:06 solution? 1:19:07 >> And it's definitely not super intense 1:19:09 exercises like this. Cuz listen, if 1:19:11 you're trying to get in shape and I 1:19:13 don't care what it's for, a wedding, 1:19:14 vacation, summer, don't care. If you're 1:19:16 trying to get in shape fast, you need to 1:19:18 follow a plan for your body type. And 1:19:20 there is a very simple breakdown of body 1:19:23 types. There's three body types. Uh you 1:19:25 have the skinny person who's trying to 1:19:27 put on muscle. Then you have someone 1:19:28 who's kind of in the middle, maybe has a 1:19:30 little bit too much fat 1:19:31 >> turns going to go into the educational 1:19:33 piece. It's effectively going to sell 1:19:34 his system in the way that he views 1:19:36 weight loss and body types and 1:19:37 programming, etc. This is such a good 1:19:40 ad. In the last 5 days it's got 2,000 1:19:42 likes. Let's break down exactly why. 1:19:45 >> So you hop on chest is 1:19:46 >> opens with a contrarian hook and 1:19:48 obviously the visuals pull you in. 1:19:50 >> What I would say if I was a caveman that 1:19:52 wasn't aware of all the way 1:19:54 >> builds contrast and then goes into an 1:19:55 educational piece. 1:19:57 >> is that you can boost it naturally. 1:19:58 Being low T can be caused by many fac 1:20:00 >> switches over to a different format, 1:20:02 which is the faceless format, which is 1:20:04 better in it to education, particularly 1:20:06 on organic and we're trying to make this 1:20:08 ad feel organic. 1:20:09 >> High stress, vitamin deficiencies, not 1:20:11 enough fat in your diet, not enough good 1:20:13 sleep, microplastics, high estrogen, 1:20:15 high inflammation. It's become a serious 1:20:17 problem for young men who often feel 1:20:19 like there's only one solution. 1:20:21 >> Af 1:20:21 >> So notice they'll problem agitate and 1:20:23 then they'll educate. Problem agitate, 1:20:25 then educate. So there's all the 1:20:26 problems, here's the solution. This 1:20:28 isn't the solution for you. Goes back 1:20:30 into problems. 1:20:31 >> After all, global testosterone levels 1:20:33 have dropped over 30% in the last 50 1:20:35 years. Young, fit, healthy 1:20:37 >> So, there's the education, and then 1:20:39 here's the problem agitation. 1:20:40 >> guys will have the testosterone of a 1:20:41 70-year-old that has never seen the 1:20:43 inside of a gym in their life. It's a 1:20:45 grim side effect of our modern society 1:20:47 with low test taking away the feeling of 1:20:49 being a real man. A consequence of this 1:20:51 is a wave of young men, even teenagers, 1:20:54 hopping on TRT out of desperation and 1:20:57 because they don't want to fall behind. 1:20:58 But, injecting test is a minefield of 1:21:00 side effects. Rapid hair loss, high 1:21:02 estrogen causing gyno, mood swings, 1:21:04 >> objection handling around the core 1:21:06 premise of the hook, which was injecting 1:21:08 test, and so now he's explaining why 1:21:09 that's not a good idea. 1:21:10 >> There's infertility and acne, not to 1:21:12 mention a big-ass needle injected 1:21:14 multiple times a week. In amongst this 1:21:16 chaos, how do I, a lifetime natural, 1:21:19 have more testosterone than most guys 1:21:21 running cycles? Whilst I'm dialed with 1:21:23 my sleep, training, recovery, and diet, 1:21:25 so many guys who are low T. However, my 1:21:27 secret weapon is a Nattie Plus all of 1:21:30 me. 1:21:30 >> And there you go, there's the product. 1:21:31 So, the product's been introduced after 1:21:32 about a minute and a half into the 1:21:34 creative, which is why it's done so 1:21:35 well. It's why it's scaling so well. 1:21:37 It's on an influencer profile. People 1:21:38 know this influencer. It's likely 1:21:40 prioritizing getting served to his own 1:21:41 audience. And so, when people are 1:21:43 watching this, they're not knowing it's 1:21:44 an ad. They're thinking it's 1:21:45 educational. And honestly, it's a great 1:21:47 educational piece. It goes into 1:21:48 everything you need to know about 1:21:49 testosterone and why you would want to 1:21:51 get behind and buy this product. And it 1:21:52 does it in a way that's very 1:21:54 entertaining. It's fast cuts. It's 1:21:55 constantly changing formats. It's built 1:21:58 native for the platform, which is why 1:22:00 it's scaling so well. 1:22:01 >> One important PSA here is DPAs, okay? 1:22:04 DPAs, dynamic product ads, those are the 1:22:07 carousel ads that will dynamically pull 1:22:09 in the product that someone visited on 1:22:10 the website, as long as your pixel is 1:22:12 set up correctly, pushing back in data 1:22:14 on which pages were viewed. DPAs cause a 1:22:17 death spiral in a lot of brands. I see 1:22:19 this in fashion particularly, uh which 1:22:21 is that DPAs have really good attributed 1:22:23 ROAS. The ROAS number is very high. Now, 1:22:26 the ROAS actually isn't very high once 1:22:29 you delineate out existing customers. 1:22:31 So, if you remove existing customers, it 1:22:33 drops precipitously. And then, you also 1:22:36 want to delineate the attribution model 1:22:37 down to incremental attribution. And so, 1:22:41 you can do this by clicking on the 1:22:44 breakdown button in the top Actually, 1:22:46 sorry. You click on the columns button, 1:22:48 compare attribution settings, 1:22:51 incremental attribution, apply. And 1:22:53 then, you'll get a split out of what the 1:22:55 incremental ROAS is on the ad. A quick 1:22:57 lesson on how incremental attribution 1:22:59 works is that every single day, Meta is 1:23:01 targeting, let's say, 10,000 people in 1:23:03 this square. It will hold out 10% of 1:23:05 users, and it will look at these people 1:23:07 that didn't see your ads, how many of 1:23:08 them converted. Let's say, 1% of them 1:23:10 converted. Over here, the people that do 1:23:12 see your ads, 2% of them convert. Okay, 1:23:15 great. What we can do, what's called in 1:23:17 data science, is a 1:23:18 difference-in-difference calculation, 1:23:19 which is simply minusing the control 1:23:21 versus the treatment. This equals 1%. 1:23:24 So, your ads are actually having a 1% 1:23:26 impact, not a 2%. And so, when we see a 1:23:28 five ROAS on that DPA, let's say as an 1:23:31 example, we actually need to adjust that 1:23:33 down by 50% to 2.5. So, that's what 1:23:35 incremental attribution is doing. That's 1:23:37 what it's going to show you on the ROAS 1:23:38 here. And the ROAS in your DPAs is going 1:23:40 to be way lower than you think it is, 1:23:42 particularly on new audiences. I am 1:23:43 still yet to see an ad account where 1:23:45 this is a great scalable ad type on cold 1:23:47 audiences. Now, is it an ad type that 1:23:49 can work on cold audiences? For sure. 1:23:52 Can you spend a thousand dollars a day 1:23:53 on it? 1:23:54 For sure. We have some ad accounts that 1:23:56 do. 1:23:57 Can you spend fifty thousand dollars a 1:23:58 day on it? Absolutely not. Like, you 1:24:01 just can't scale DPAs that well on cold 1:24:03 audiences. Most people are way 1:24:04 overspending cuz the ROAS number looks 1:24:06 good. So, just be super careful. Cuz 1:24:07 what ends up happening there is because 1:24:09 DPAs sit at the bottom of the funnel 1:24:11 down here, is that people start 1:24:12 reallocating their budgets from these 1:24:14 top-of-funnel efforts that have bad 1:24:17 return on ad spend, and they just start 1:24:19 moving budget down here. And when you 1:24:21 start moving budget down here, you start 1:24:23 losing the quantity of users that are 1:24:24 coming through from the top of funnel, 1:24:26 and you start scaling something that 1:24:27 isn't scalable, and the whole funnel 1:24:29 topples over, and return on ad spend 1:24:30 declines, and everything goes poorly. 1:24:32 So, in fact, when you're thinking about 1:24:33 funnel budget allocation, you want to be 1:24:35 thinking about 70% to 90% of budget 1:24:38 going towards top of funnel. You want 1:24:40 20% ish of budget going to middle of 1:24:43 funnel, and then you want really like 5% 1:24:45 to 10% of budget going to bottom of 1:24:46 funnel. There's actually a better way to 1:24:48 calculate bottom of funnel budgeting, 1:24:49 which I'm going to show you later in the 1:24:50 video. But, this is rough split on how 1:24:53 it should look. Now, what's interesting 1:24:54 these days is that you used to do this 1:24:56 through campaign structure. So, you 1:24:57 would have your top of funnel ads, and 1:24:59 you would just set your budget here, 1:25:00 you'd have your middle of funnel, you'd 1:25:01 set your budget here, and then you'd 1:25:02 have your bottom bottom of funnel. Uh 1:25:03 the issue right now is there actually is 1:25:05 a lack of campaign structure. You 1:25:06 shouldn't be structuring like this. You 1:25:07 should just consolidate these down, and 1:25:09 you might be able to have a bottom of 1:25:10 funnel that sits separate. And so, 1:25:11 because of that, the actual budget 1:25:13 allocation is coming from the creative 1:25:15 production volume. And so, you want 70% 1:25:17 to 90% of the volume of assets you make 1:25:20 to be at the high stages of awareness, 1:25:22 and you want 20% of them to be at the 1:25:24 middle to bottom of funnel stages of 1:25:26 awareness. So, this is more so a 1:25:28 percentage allocation of creative 1:25:30 production than it is a percentage 1:25:32 allocation of budgets across a campaign 1:25:33 structure. I would say one of the most 1:25:35 underutilized 1:25:37 formats in 2026 is partnership ads. Now, 1:25:39 I think people are catching on to this 1:25:40 actually, and the arbitrage is probably 1:25:42 coming towards its end. So, by the time 1:25:43 you watch this video, if you're way into 1:25:45 the future, this might not be the case 1:25:46 anymore. You might have missed it. But, 1:25:48 partnership ads, what they are is they 1:25:49 run through the creator's profile. So, 1:25:51 you'll have your profile here, so your 1:25:53 company name on the top of the post, and 1:25:55 then you'll have and and then the 1:25:57 influencer's profile right here. Now, 1:25:59 these days there's actually a setting 1:26:00 that allows them to change the 1:26:02 combinations, and so sometimes you'll 1:26:03 just see the company, sometimes you'll 1:26:04 just see the influencer, sometimes 1:26:06 you'll see both, and it will split test 1:26:07 and decide on whatever the best mix is. 1:26:09 Uh but, the reason why these ads do so 1:26:11 well is that the targeting uses the 1:26:14 audience on both handles. And so, if 1:26:16 this is an influencer that's actually 1:26:18 somewhat contextually relevant to your 1:26:19 target demographic, which I hope it is, 1:26:21 this ad will go and reach their 1:26:22 audience, and it will use their audience 1:26:24 data to find new people. And so, you're 1:26:26 effectively leveraging the influencer's 1:26:28 audience to help in your targeting on 1:26:31 cold, not even just targeting on warm. 1:26:33 And that's why they end up doing so 1:26:34 well. They also feel super authentic. 1:26:36 When you get an ad from a uh influencer 1:26:38 profile like this, it doesn't feel as 1:26:40 much so like an ad, so it doesn't feel 1:26:41 like you're getting sold to. So, when 1:26:42 you're getting an organic piece of 1:26:43 content, it resonates better. And some 1:26:45 accounts that we're working on, we have 1:26:46 ad spend up to 40% of total account 1:26:48 spend running through partnership or 1:26:50 collaborator ads. And so, this is a 1:26:52 really important thing that you should 1:26:53 have in your account. If you are not 1:26:55 running them, you were just at a 1:26:56 disadvantage to competitors. All right, 1:26:57 so going into the testing framework and 1:26:59 how this then translates into the ad 1:27:01 account, this is what the testing cycle 1:27:03 looks like, which is that you build a 1:27:05 hypothesis, which is you define what 1:27:07 you're actually testing. So, when we're 1:27:09 introducing new creative, generally the 1:27:11 test is the concept, and particularly 1:27:13 one of those four variables that we're 1:27:14 changing in the creative. Then we want 1:27:16 to build a bunch of ads under that 1:27:17 concept. One big mistake that a lot of 1:27:20 people make is that they come up with a 1:27:21 concept, they build it all out, it looks 1:27:23 amazing, and then they make one ad, or 1:27:24 they make two ads, or they make three 1:27:25 ads. Like, really, if you want to test a 1:27:27 concept properly, you want at least 10 1:27:29 plus ads, considering that most people's 1:27:31 hit rates in their account are around 1:27:32 about 5%. And so, technically speaking, 1:27:34 if you have a new concept coming in, you 1:27:36 would need to do, if you had a 5% hit 1:27:38 rate, you'd need to do at least 20 ads 1:27:40 to expect a winner. And so, thinking 1:27:41 about your hit rate in the account as a 1:27:43 function of how much volume you need to 1:27:45 do to be able to validate the hypothesis 1:27:47 is critical. Then you want to go and 1:27:48 launch, and I'm going to show you with 1:27:50 the exact structure that you should 1:27:51 launch into. You then want to go and 1:27:52 analyze, particularly against the 1:27:54 hypothesis. So, what did we expect? What 1:27:56 How much spend did we want to go 1:27:57 through? How many conversions did we 1:27:58 expect? What was the ROI? And then how 1:28:00 did that compare? If it worked, we want 1:28:02 to be looking at how we can employ 1:28:03 strategies to scale, and then we want to 1:28:05 go down and iterate. And this iteration 1:28:07 process gets faster and faster the 1:28:09 bigger your ad spend. So, now when we 1:28:11 think about account structure and 1:28:12 introducing tests into the account, this 1:28:14 is where there becomes a lot of 1:28:15 flexibility. And this is where we prefer 1:28:17 not to actually make broad sweeping 1:28:20 account structure recommendations 1:28:21 because the fundamentals of account 1:28:23 structure is that the complexity of the 1:28:25 ad account should be a product of the 1:28:27 complexity of the business. And so, if 1:28:29 you run a startup that's doing $10,000 a 1:28:32 month in revenue and you have one 1:28:33 product and you have one persona that 1:28:35 you're trying to crack, the account 1:28:36 structure should be unbelievably simple. 1:28:39 You don't need 100 different campaigns. 1:28:41 On the other hand, if you're an $800 1:28:43 million fashion retailer with 50 new 1:28:45 SKUs coming out per week and you have 1:28:48 all of these different business units 1:28:49 and regions within the business, well, 1:28:51 the ad account's probably going to be 1:28:52 pretty complex because it needs to match 1:28:54 that complexity in some degree so that 1:28:55 we can align the commercial goals of the 1:28:57 business with the structure that is 1:28:59 being run in the ad account. And so, the 1:29:00 structure I'm about to give you is a 1:29:02 very broad sweeping, this will work if 1:29:05 you're a small to mid-size business, but 1:29:07 it's probably not exactly what you 1:29:08 should run right now and moving into the 1:29:10 future as it will likely evolve as the 1:29:13 business dynamics evolve. But to keep it 1:29:15 simple, you generally want a testing 1:29:17 campaign. Now, whether this is an or 1:29:20 a CBO is honestly going to come down to 1:29:22 personal preference and I'll tell you 1:29:23 how to think about this problem in a 1:29:25 moment, but then underlying this, you 1:29:27 have your concepts at the ad set level. 1:29:29 So, this is concept one, this is concept 1:29:32 two, 1:29:33 this is concept three, 1:29:35 etc., etc. Now, you're either setting 1:29:37 budgets across these manually using an 1:29:39 or the CBO's just going to 1:29:41 distribute budget. Under here, you're 1:29:42 going to have all your ads. A common 1:29:44 question I get is, "Well, okay, we go 1:29:46 and put four ads or five ads under here. 1:29:48 How many ads should we launch and when 1:29:49 we come up with new ads for this 1:29:50 concept, do we launch it in here or do 1:29:52 we launch a new ad set?" Which is both 1:29:54 really good questions. In terms of ad 1:29:56 volume, this is really open based on 1:29:58 budgets, but you generally want a 1:29:59 minimum of three ads and then a maximum 1:30:03 of probably 50 to 100, but honestly, the 1:30:05 limit is kind of the sky based on your 1:30:07 ad spend. And so, I wouldn't really be 1:30:09 thinking about upper funnel upper 1:30:10 limits. I would just be thinking about 1:30:12 lower limits, which is a minimum of 1:30:13 three. 1:30:14 When you then want to go and introduce 1:30:15 more ads, this is based on whether 1:30:17 you're hitting KPI. So, what is going to 1:30:20 happen when you introduce more ads under 1:30:22 this ad set is you were going to reset 1:30:24 the learning phase to some degree. 1:30:26 Because there is sequencing that is 1:30:27 occurring across all of these different 1:30:29 creative. And then when you go and 1:30:30 launch more ads in here, you disrupt the 1:30:32 sequencing and it relearns how to 1:30:34 distribute your spend. Now, that is not 1:30:36 good if the ad set's performing well. 1:30:39 If this ad set is crushing it and you're 1:30:40 really happy with it and you're scaling 1:30:42 it up, don't go and mess with the 1:30:44 learning phase. Don't go and inject a 1:30:46 bunch of new ads. You might mess the 1:30:48 performance up. And I've actually 1:30:49 personally done this years ago and so 1:30:51 that's why I'm very big on not touching 1:30:53 stuff that's working. If it's not 1:30:54 hitting KPI, and this is way 1:30:56 underperforming, 1:30:58 go for your life. Do whatever you want. 1:30:59 You can turn ads off. It doesn't matter 1:31:01 cuz yes, you're going to break some 1:31:02 sequencing, but the sequencing isn't 1:31:03 working. So, who cares? Go and launch as 1:31:05 many ads as you want in here, right? Do 1:31:07 whatever you want cuz it's not a KPI. 1:31:09 So, you need to change something. But if 1:31:11 things are working, definitely don't go 1:31:12 and launch more ads in here. If you go 1:31:14 and make more ads for this concept, 1:31:16 let's say the concept's crushing it, so 1:31:17 you go and make 20 more ads, go and 1:31:19 launch another ad set, right? And call 1:31:21 it concept one and then like shoot two. 1:31:24 Or obviously, whatever naming convention 1:31:26 you want to use to be able to clarify 1:31:28 the difference between these, but then 1:31:29 also be able to filter that they are the 1:31:31 same concept. Now, the question becomes, 1:31:32 well, how do you scale? Let's say 1:31:33 concept one is performing well. What do 1:31:36 we do? Well, number one, increase 1:31:37 budgets. Never turn this off. A big 1:31:39 mistake that people make is they take 1:31:41 something that's winning and they turn 1:31:42 it off and they go and launch it in like 1:31:43 a scaling campaign. It's such a rookie 1:31:45 mistake, just fundamentally speaking, 1:31:47 because if something works, don't touch 1:31:49 it. Leave it. 1:31:50 Increase the budget. Now, in addition to 1:31:52 that, sure, you [snorts] can take these 1:31:54 post IDs and you can wrap them around 1:31:56 into a scaling campaign. That's fine. 1:31:58 And that will work. It works in a a of 1:32:00 accounts. We run scaling campaigns in a 1:32:02 lot of accounts, a lot of accounts we 1:32:03 don't run scaling campaigns. Once again, 1:32:05 it's dependent on the actual ad account 1:32:06 and the business. You can run a scaling 1:32:07 campaign, you can take these, you can 1:32:09 post them over here. If it doesn't work, 1:32:10 don't do it. Just keep it in the 1:32:11 testing. And then next to these two, the 1:32:13 only third campaign that I would usually 1:32:14 recommend is to have a retargeting 1:32:17 campaign that is generally on existing 1:32:20 customers. So, it gives you the ability 1:32:22 to exclude existing customers from here 1:32:25 and exclude existing customers from 1:32:26 here. So, these can surely be cold and 1:32:28 that allows you to control budgets. 1:32:29 Depending on how long the customer 1:32:30 journey is as well, you might want an ad 1:32:32 set here on 90-day website visitors as a 1:32:34 mid-funnel. But, this is only if you 1:32:36 think that you need to increase 1:32:37 frequency on engaged audiences. Now, 1:32:39 what are the metrics that matter for 1:32:41 being able to measure the performance of 1:32:42 creative? Let's list them out in order 1:32:45 of reliability. Number one is amount 1:32:47 spent. Now, this is counterintuitive. A 1:32:49 lot of people look at this and go, "Why 1:32:50 would budget distribution to an ad be 1:32:52 the most important proxy for performance 1:32:54 over extended time periods?" And the 1:32:56 reason being is due to the sequencing 1:32:58 that we talked about before. Right? So, 1:33:00 if you have a bunch of different ads, 1:33:01 attribution's only going to go to the 1:33:03 last ad. So, if we use ROAS as the 1:33:04 primary KPI, we're going to miss the 1:33:07 fact that Meta's actually putting all 1:33:08 your budget here. Why? When this only 1:33:10 has a 1.7 and this has a 2.2, why would 1:33:12 all the budget go here? Probably because 1:33:14 this ad is causing purchases elsewhere 1:33:17 in the account. And so, this is actually 1:33:19 a great top-of-funnel asset that likely 1:33:21 has a low frequency that's cutting 1:33:22 through on new audiences. And so, if 1:33:23 you're going to look at these three ads, 1:33:25 I would say that this is the highest 1:33:27 performer if it has double the spend of 1:33:28 the other ones. Even though this is a 1:33:30 higher ROAS, sure. But, if this could 1:33:31 scale, it would have more spend in it 1:33:32 and it doesn't. And so, amount spent 1:33:34 becomes the best proxy for performance 1:33:36 when looking at creative. Number two, 1:33:39 sure, is ROAS or CPA. You just want to 1:33:42 be making sure you're looking at 7-day 1:33:43 click or incremental attribution. Number 1:33:45 three is going to be CPC and CTRs. This 1:33:49 is a leading indicator for the 1:33:52 effectiveness of the ad. Obviously, you 1:33:54 can have good CTRs and good CPCs, but 1:33:57 the conversion rate of the ad can suck 1:33:58 and that's not a good ad. And so this 1:34:00 obviously isn't a primary indicator of 1:34:02 the quality of a creative. That is where 1:34:04 you have to use multiple different 1:34:06 metrics in conjunction to be able to 1:34:08 assess quality. 1:34:09 Um but it is a good leading indicator, 1:34:11 particularly on the fringes. So like if 1:34:13 we look at CPCs as an example, and let's 1:34:15 say this is what your CPCs on average 1:34:17 look like across all of your ads. If 1:34:19 above this point is let's say a $3 CPC, 1:34:22 and none of these ads have ever been 1:34:24 profitable, we can use this as a leading 1:34:26 indicator to know that these ads at very 1:34:28 early spend will actually never work cuz 1:34:29 it's too expensive to drive traffic. The 1:34:31 next two are hook rates and hold rates. 1:34:33 These only apply to video assets, 1:34:34 obviously, because the hook rate is how 1:34:36 many people make it past 3 seconds in 1:34:38 the video, and the hold rate, there's 1:34:40 different definitions of hold rate. We 1:34:41 actually look at a lot of different 1:34:42 definitions in our reporting, but as a 1:34:45 generalization, this is how many people 1:34:46 make it past 15 seconds. Obviously, hook 1:34:48 rates and hold rates are not correlated 1:34:52 to conversion performance, but they are 1:34:54 a diagnostic tool that allows us to 1:34:56 iterate on creative based on what we 1:34:58 think is letting us down. 1:35:00 And so we may have a creative that we 1:35:02 thought was going to do really well, but 1:35:04 the hook rate is terrible against our 1:35:05 average. And so we try to switch the 1:35:08 hook out to try to fix that to see if 1:35:09 the body actually performs. Same thing 1:35:11 applies to hold rate. People might not 1:35:12 be holding through the bridge. Turns out 1:35:14 the bridge was actually done poorly, so 1:35:16 we redo that part of the video and then 1:35:18 we can iterate accordingly. So before I 1:35:19 said I would explain versus CBO and 1:35:22 the risk profile here. So the reason why 1:35:24 you would choose one over the other is 1:35:26 actually based on risk tolerance. So 1:35:27 what happens when you use CBO is that 1:35:29 it's going to play in to Pareto's 1:35:32 principle. This applies literally 1:35:34 everywhere I've ever seen it, which is 1:35:36 that 80% of something will drive will be 1:35:39 driven from 20%. So this applies into 1:35:42 companies where 20% of people will drive 1:35:44 80% of revenue generation. This applies 1:35:46 into the ad account, which is that 20% 1:35:48 of ads will drive 80% of revenue. 20% of 1:35:51 ads will hold 80% of revenue. Crazy 1:35:53 thing about Pareto's principle is that 1:35:55 it compounds in on itself and you see 1:35:57 this within every single data set I have 1:35:59 done on ad accounts breaking down their 1:36:01 ad profile. Which is that of this 80%, 1:36:05 80% of this which comes out to 64% of 1:36:08 total is driven by 20% of this, which 1:36:10 this comes out to 4%. So, 4% of ads will 1:36:14 hold about 64% of total spend and total 1:36:17 revenue generation within the business. 1:36:19 Now, CBOs play into Pareto's principle. 1:36:23 And what that means is that if you just 1:36:24 allow the campaign to distribute budgets 1:36:27 however it wants, it will naturally end 1:36:29 up distributing budgets like this. Where 1:36:31 majority of all of your budget in the 1:36:32 account will go to 4% of your ads. Now, 1:36:35 that's fine cuz that's going to yield 1:36:36 the best efficiency and performance, but 1:36:39 it is super risky. Because if these ads 1:36:41 fatigue, you are in a very bad position. 1:36:43 You don't have backups. If we go back to 1:36:46 the start of the video, which is 1:36:47 portfolio management, CBOs are terrible 1:36:50 for portfolio management. Okay? Because 1:36:52 you're going to over leverage into a few 1:36:55 ads. This also becomes an issue when you 1:36:57 want to control sell through rates 1:36:58 across multiple different products. And 1:37:00 so, you might have a large product 1:37:02 portfolio that all sit under the one CBO 1:37:04 and then all of the spend just goes to a 1:37:06 few to particular products. Because 1:37:08 Pareto's principle applies into a 1:37:10 product level as well. 20% of products 1:37:12 will drive 80% of revenue. And so, your 1:37:14 entire business will end up skewing and 1:37:16 increasing the risk profile, which you 1:37:17 might be fine with, but that's the 1:37:19 trade-off that you have to make. So, 1:37:21 when you think about these two, 1:37:22 CBO is increased risk, 1:37:25 but likely increased ROI. This is 1:37:28 decreased risk, but likely slightly 1:37:31 lower ROI. And it depends on which 1:37:33 position you want to be in. I 1:37:34 personally, if I'm running the business, 1:37:36 I actually choose this every day of the 1:37:39 week, dependent on like what season and 1:37:42 where you are in the business. And so, 1:37:43 there are periods in time in which I 1:37:45 would actually switch over to CBO if I 1:37:47 need a quick spike in efficiency, 1:37:49 efficiency is isn't where we need it to 1:37:50 be. I don't care about testing and 1:37:52 de-risking right now. We just need ROI 1:37:54 to come through. But then, ideally, you 1:37:56 want to be in a steady state over here 1:37:58 where the business is de-risked across 1:38:00 more ads, across more products, and a 1:38:02 lot more stuff is working so that if 1:38:04 something stops working, you aren't in a 1:38:05 terrible position. Uh this next point 1:38:07 I'm particularly calling out cuz we work 1:38:09 with a lot of nine-figure large 1:38:10 retailers here in Australia. And if any 1:38:12 of them are watching, then they'll know 1:38:14 that this problem exists, which is that 1:38:16 in a lot of large retailers who a lot of 1:38:18 my audience is spent on, there is this 1:38:20 90/10 problem with desire versus pain 1:38:23 point led content. And this 1:38:24 fundamentally comes from old-style 1:38:26 marketing, which is campaign shoots. And 1:38:28 so, a lot of these large uh retailers in 1:38:31 fashion or whatever industry it is, they 1:38:34 were built around campaign shoots. And 1:38:36 campaign shoots do one thing, and that 1:38:38 is that they build desire. And so, they 1:38:40 have models, they have shots that make 1:38:42 you want to desire having the product, 1:38:44 but play to pain points at all. And so, 1:38:47 you end up with the ad account being 1:38:48 significantly overweighted towards 1:38:50 desire rather than towards pain. And at 1:38:52 the end of the day, the best performance 1:38:54 marketing assets are those that sit on 1:38:57 pain. And so, I would strongly recommend 1:38:59 looking at your current creative mix 1:39:01 through this lens and thinking about how 1:39:03 do we actually start to shift this ratio 1:39:06 over to 60/40. Now, I personally think 1:39:08 that the best brands are probably closer 1:39:11 to 70/30 here. You can almost think of 1:39:13 this in a way of like this is 1:39:15 performance marketing driven assets. 1:39:17 This is branding driven assets. Now, one 1:39:19 last thing in the testing framework is 1:39:20 assets for retargeting. So, how should 1:39:23 you be thinking about retargeting ads on 1:39:25 bottom of funnel? This is something I've 1:39:26 been saying for a very long time. This 1:39:28 is probably one of my first YouTube 1:39:30 videos that we put out, which is that on 1:39:32 your bottom of funnel retargeting 1:39:33 campaigns, you want to think about 1:39:36 objection handling. This is the primary 1:39:38 purpose of a product aware ad. And you 1:39:40 can think about this mathematically, 1:39:42 which is that let's say you have a 5% 1:39:44 conversion rate on the website. That 1:39:46 means that 95% of people did not buy. 1:39:49 Now, the question is, why didn't these 1:39:52 people buy? What were their objections? 1:39:54 Maybe 20% of them had a price objection. 1:39:58 It was just price too high. They never 1:40:00 will buy at that particular price. Now, 1:40:02 they might buy in the future when you're 1:40:03 on sale, but right now they won't buy. 1:40:05 20% of people 1:40:06 don't trust the brand. There wasn't 1:40:08 enough trust credibility markers on the 1:40:10 website or the ads that they 1:40:11 particularly saw to reinforce them to 1:40:14 actually purchase the product. 20% don't 1:40:17 like the quality. They think the quality 1:40:19 is probably going to be bad. Now, this 1:40:20 is arguably a trust issue, but these are 1:40:22 really two different things and they 1:40:23 should be come at in two different ways. 1:40:25 And then maybe another 20% of people are 1:40:27 just not ready to buy yet. Okay, so it's 1:40:29 too soon. It was the first ad they've 1:40:32 seen. It was the first ad they've 1:40:33 clicked on. They just need a little bit 1:40:35 more time to purchase. So, we want to 1:40:37 understand this mix. We want to 1:40:38 understand why people aren't buying, and 1:40:40 then we want to handle all of these 1:40:42 objections in our bottom of funnel 1:40:43 creative. And so, on the price 1:40:45 objection, we want to be showing 1:40:46 discounts. We want to be value 1:40:48 anchoring. We want to even going for a 1:40:50 creative that's like a cost per use 1:40:52 angle. These actually crush on bottom of 1:40:54 funnel. So, if you're selling like a 1:40:56 supplement, you compare it to a coffee. 1:40:58 It's a cost the price of a coffee with 1:40:59 the same caffeine, etc., etc. If they 1:41:01 don't trust the brand, we want to come 1:41:03 in with social proofing. How many 1:41:05 customers do we have? How many five-star 1:41:06 reviews do we have? Do we have any 1:41:08 user-generated content testimonials that 1:41:09 we can use here? If they're doubting the 1:41:11 quality of the product, we want to show 1:41:13 UGC of the product actually in use. We 1:41:15 want before and afters. We want 1:41:17 demonstrations. And if it's simply too 1:41:19 soon for them to buy yet, we want to be 1:41:21 staying present and top of mind with 1:41:23 educational or lifestyle content. And 1:41:24 then this is what the mix should be in 1:41:26 all of our bottom of funnel assets. And 1:41:28 this is how you ultimately yield strong 1:41:30 conversion rates once people have been 1:41:31 moved through the funnel. The question 1:41:33 is, does diversity or volume matter more 1:41:35 when it comes to creative? So, you want 1:41:36 to be thinking through volume through 1:41:38 this hierarchy, which is that volume 1:41:40 matters the most, then concept quality, 1:41:43 then hook strength, then editing 1:41:44 quality, then UGC talent, then 1:41:46 storytelling all the way down the 1:41:47 bottom. Now, it's not to say that any of 1:41:49 these components aren't important, but 1:41:51 it's the fact that volume ends up 1:41:53 creating quality, and hit rates are 1:41:55 limited within an ad account. And so, a 1:41:57 way to conceptualize through this idea 1:41:59 is that if you take the best creative 1:42:01 strategists, you can pull 10 of them out 1:42:03 of 10 different businesses, and you can 1:42:05 show them five ads that are all 1:42:07 objectively good creative, and you tell 1:42:09 them which of these five was the winner. 1:42:12 Which one produced the most spend or the 1:42:14 most return? None of them will be able 1:42:16 to guess it. It will just be a complete 1:42:18 guessing game. They'll a- 1:42:20 assign reasons as to why they believe 1:42:21 this creative is better than the other 1:42:23 ones. Might have been due to the concept 1:42:24 quality, might have been due to the 1:42:25 hook, but the reality is is that people 1:42:28 are not very good at guessing the top 1:42:30 end level of creative performance. And 1:42:33 so, instead, you want to understand 1:42:35 where the bar of quality is, and just 1:42:37 ensure that ads are reaching that 1:42:39 standard. And so, if all of these five 1:42:41 ads sit above this bar of quality, and 1:42:44 we could objectively say that this is a 1:42:45 seven out of 10. So, if an ad is better 1:42:48 than a seven out of 10, 1:42:50 it meets that bar, and it is 1:42:52 indistinguishable from any other 1:42:54 creative in being able to try to guess 1:42:56 its performance. Where you end up dying 1:42:58 in creative strategy is trying to win 1:43:01 the game of understanding which ads will 1:43:04 perform well at a high level. No one is 1:43:06 able to objectively do this well. I've 1:43:08 never seen anyone be able to pick 1:43:09 winners. And so, you instead want to 1:43:10 become very good at understanding what 1:43:12 is the minimum bar of quality, what does 1:43:14 a good ad look like, and then how do we 1:43:16 reach that level of quality, and then 1:43:18 maximize volume above it. Where people 1:43:20 get volume completely wrong, and the 1:43:21 example that I gave way back at the 1:43:23 start, which was people were 1:43:24 misunderstanding activity for creative 1:43:26 strategy, is that they were putting tons 1:43:28 of ads into the account, but they didn't 1:43:30 have an understanding internally of what 1:43:32 the bar of seven out of 10 actually 1:43:34 looks like. And so when you go and do an 1:43:36 ad breakdown of all the creatives that 1:43:37 they were putting into the account, 1:43:38 which was thousands a month, the ads 1:43:40 just weren't very good. They didn't have 1:43:43 any of the elements of a good concept. 1:43:45 They didn't have any of the elements of 1:43:46 a good hook. They didn't have any of the 1:43:48 elements of formats that were relevant 1:43:50 to the stage of awareness of the 1:43:51 creative that they were trying to 1:43:52 produce. And so because of that, yes, 1:43:55 they put tons of volume in the account. 1:43:56 Yes, they followed this hierarchy, the 1:43:58 most important thing on it. But it was 1:44:00 just junk volume. It was just a bunch of 1:44:02 ads that aren't even good. 1:44:04 And so instead, you need to focus your 1:44:05 efforts on how do we understand where 1:44:06 the bar is, then how do we maximize 1:44:08 volume, and then how do we start 1:44:10 step-stoning down through all of the 1:44:12 different priorities to create a good 1:44:14 ad. The question off the back of this 1:44:15 then becomes, okay, well, how much 1:44:16 volume do we actually need then? And 1:44:19 there's a really easy, simple answer, 1:44:21 which I'll give to you, and then there's 1:44:22 a more complicated answer that you'd 1:44:23 have to go and build a financial model 1:44:25 out for. And if you're an e-commerce 1:44:26 brand, uh we'll give it to you 1:44:27 completely for free. We'll put the link 1:44:29 in the YouTube description below. 1:44:31 So the simple answer is, for every 1:44:33 $1,000 in monthly spend, you want one 1:44:36 new ad. 1:44:38 And so if you're spending 30K a month, 1:44:39 30 new ads. 50K a month, 50 new ads. Now 1:44:42 this is in AUD. Also translate this into 1:44:44 USD. Roughly that ad volume is fine. Now 1:44:47 the better way to do this is to back 1:44:49 propagate based on hit rates and 1:44:51 expected spend per ad. So what you can 1:44:53 do is you can calculate the average 1:44:54 spend of an ad. So if you know that 1:44:56 across the last year of ads, you've 1:44:58 launched, let's say, 500 ads, 1:45:01 and you've spent $500,000, 1:45:03 well then each ad here, on average, 1:45:06 spent $1,000. 1:45:08 Now if you know the average return of an 1:45:10 ad, which you can look at either your 1:45:12 acquisition MER, which is new customer 1:45:14 revenue divided by ad spend in the 1:45:15 business, or you could just look at 1:45:17 ROAS, but if you're going to look at 1:45:18 this, look at it on a 7-day click basis. 1:45:21 And let's say your average ROAS is a 1:45:23 three, 1:45:24 and you've got $1,000 in mean spend, 1:45:28 well then we know that the average 1:45:29 return, or you could call this the 1:45:31 expected value, 1:45:33 if we're using probabilistic math, 1:45:35 is then three thousand dollars in 1:45:38 revenue. So, every time we launch an ad, 1:45:41 we should expect on average, on a larger 1:45:43 data set, that we generate three 1:45:44 thousand dollars in revenue. So, then we 1:45:45 can start to back propagate this into 1:45:48 our actual revenue targets. Now, more 1:45:50 specifically, I would actually delineate 1:45:52 this down into new customer revenue to 1:45:54 be specific to acquisition. So, if you 1:45:56 have a 300K 1:45:59 monthly new customer revenue target, you 1:46:02 take 300K, you divide by your 3,000, 100 1:46:05 creatives per month 1:46:08 are required. Now, there's a few nice 1:46:10 things about this. Number one, you can 1:46:11 flex the actual volume of creative in 1:46:14 conjunction with spend, which is what 1:46:16 you should be doing. As you go through 1:46:17 the year, and your forecast changes 1:46:21 in terms of expected new customer 1:46:22 revenue, your creative volume should 1:46:24 flex accordingly. So, this should be 1:46:26 exactly what your revenue looks like, 1:46:28 but also what your creative volume looks 1:46:30 like. Now, there are two disadvantages 1:46:31 to this simplified model, which is that 1:46:33 number one, it doesn't take into 1:46:34 consideration the average time of an ad 1:46:37 spending, and so these ads might 1:46:39 generate three thousand dollars in 1:46:41 revenue, but it might take four months. 1:46:42 And so, that needs to be factored in cuz 1:46:44 obviously volume needs to be a lot 1:46:45 higher then. And it also doesn't factor 1:46:47 in the difference between a winner and a 1:46:48 loser. If you got 100 losing ads this 1:46:52 month, you will generate barely any 1:46:54 revenue. If you got 100 winners, you'll 1:46:56 generate like multiple millions in new 1:46:58 customer revenue. 1:47:00 And so, you need a forecast losing ads 1:47:01 versus winning ads separately, and then 1:47:03 you need to add the time component. 1:47:05 We've built a model that does all of 1:47:07 this. Once again, if you're an e-com 1:47:08 brand and you want it, you can click the 1:47:10 link in the description, uh fill out the 1:47:12 form, and we'll send it out to you. Now, 1:47:13 when we're thinking through creative 1:47:14 budget, how much budget should go 1:47:16 towards production versus actual 1:47:18 distribution of the assets? If you're at 1:47:20 four to 30K a month in spend level, 1:47:23 there is no reason as to why your 1:47:25 content shouldn't be super founder led. 1:47:27 And the actual budget here on production 1:47:29 is your time rather than money. Because 1:47:33 you really don't need to go and pay for 1:47:34 production on this kind of spend level. 1:47:36 This can all be done literally through 1:47:38 statics in most businesses, but even 1:47:40 just founder led videos can get you to 1:47:42 30K a month in spend easily. And 1:47:44 honestly, I would challenge that the 1:47:45 founder of a business at this size 1:47:47 should be creating content because 1:47:49 content and creative is the largest 1:47:50 lever that exists within the business. 1:47:52 And so, understanding what good looks 1:47:54 like, what good creative looks like, 1:47:55 getting an understanding for how to 1:47:56 piece it together, what the editing 1:47:58 process looks like, what the ideation 1:48:00 process looks like, gives you the 1:48:01 ability to then KPI and train people 1:48:03 moving forward on one of the most 1:48:05 critical roles in the business. As you 1:48:06 then move into 30 to 100K per month in 1:48:09 ad spend, you want to be thinking of 1:48:10 allocating 25% of media budget towards 1:48:13 production. So, at 100K a month, that's 1:48:14 25K a month into creative production. 1:48:17 This is way more than like 99% of brands 1:48:20 are actually spending. And that's also 1:48:22 one of the reasons why brands don't 1:48:24 actually scale and do that well. It's 1:48:25 because they significantly under weight 1:48:27 the allocation that needs to go into 1:48:28 their media budget. Now, this is an 1:48:29 uncomfortable range because the cost of 1:48:32 production is so high excessive to the 1:48:35 scale of the business right now that it 1:48:36 feels off. It feels like this is too 1:48:39 high, but it does democratize as you 1:48:41 achieve more scale. Once you crack 1:48:43 through 100K plus, this percentage 1:48:45 really should asymptote down to around 1:48:47 about 10% of media budget. And so, as a 1:48:50 function of revenue in the business, 1:48:52 let's say that you're at a 25% MER, 10% 1:48:55 of this would therefore be 2.5% of total 1:48:58 revenue in the business. So, 2.5% of 1:49:00 total revenue should be going towards 1:49:01 media budget. This kind of spend 1:49:03 allocation allows diversity in the 1:49:05 creative that you're putting in. This is 1:49:06 enough budget to have UGC going into the 1:49:09 account, to have partnership ads going 1:49:11 into the account, to have a full-time 1:49:13 either international designer or someone 1:49:15 part-time uh in-house on statics, and it 1:49:18 can likely support a low-cost uh agency 1:49:21 in some capacity to supplement with 1:49:23 another creative type in here as well. 1:49:24 Maybe it's VSLs or something. I'll give 1:49:26 you a real example of a client that 1:49:28 unfortunately we had to part ways with 1:49:30 about uh 6 to 8 months ago because we 1:49:33 couldn't get this concept across to 1:49:35 them. And this is why I stress this so 1:49:36 much in a lot of content that we put out 1:49:38 now, and it's really important that 1:49:39 people understand this particularly 1:49:40 through the initial discussions when 1:49:42 they start talking with us because we 1:49:44 really don't want to onboard brands that 1:49:46 aren't across this concept. So, this 1:49:48 brand was spending 350k 1:49:52 per month on Meta when they started 1:49:53 working with us. Their CAC was declining 1:49:56 pretty steadily over the course of the 1:49:57 last 2 years. 1:49:59 Now, they only had about 15 ads live 1:50:03 within the account on a 350k a month 1:50:05 spend, which is crazy. And so, right out 1:50:07 the gates, we made a bunch of media 1:50:09 buying changes. The account was super 1:50:11 over-segmented. The previous media buyer 1:50:13 was doing so much over the top stuff 1:50:15 because he didn't have anything to work 1:50:17 with from an ad and creative 1:50:18 perspective. And so, when you don't have 1:50:20 anything that you can do on creative, 1:50:22 what do you do? Well, you go to the next 1:50:23 thing, and that's let's create 25 1:50:25 different campaigns. Let's use big caps, 1:50:27 cost caps. Let's try everything to try 1:50:29 to squeeze these 15 ads, which 1:50:31 ultimately isn't the lever that's going 1:50:32 to double this business. And so, when we 1:50:34 did the audit, we identified this as the 1:50:35 bottleneck. And what we did immediately 1:50:37 is as we came in, we went and took all 1:50:40 old performers and started relaunching 1:50:41 and reworking them a little bit. Even 1:50:43 though at the time we weren't doing 1:50:45 creative production in house, we just 1:50:46 threw this on top cuz we knew that this 1:50:49 was the limiter and that we had to get 1:50:50 this done. So, we took the ad account up 1:50:52 to about 60 ads. Nowhere near where it 1:50:54 needs to be, but this is all we could do 1:50:56 in terms of high-quality assets that 1:50:57 have been turned off in the past. Now, 1:50:59 what the results ended up being for this 1:51:01 brand is we actually turned them around, 1:51:04 which was one of the most insane case 1:51:05 studies that we've done cuz we did it 1:51:07 with no new creative. And we actually 1:51:09 took them from 350k a month in spend. We 1:51:11 increased their spend by about another 1:51:13 40k and we decrease their CAC by about 1:51:15 20 to 30%. And so new customer revenue 1:51:18 expansion off the top of my head, 1:51:20 there's a case study out there on this 1:51:21 somewhere, I think it was about like 1:51:23 28%. And so we were able to turn this 1:51:25 brand around substantially because of 1:51:27 this. Now the issue was that this 1:51:30 founder was very much so of the opinion 1:51:33 that this spend on Meta is a revenue 1:51:35 driver and he spent towards creative 1:51:37 production is a cost center. And so 1:51:39 because of that, they refused to do any 1:51:42 more creative production. We intro'd 1:51:44 them to multiple agencies, they went and 1:51:45 chatted to them. Uh they actually signed 1:51:47 on and then pulled out a lot of a lot of 1:51:49 those deals. And so we weren't able to 1:51:51 get any more ads, 1:51:52 any more creative. And so because of 1:51:54 that, there's only so far that you can 1:51:56 push media buying over time and so 1:51:58 performance started to come down. And 1:52:00 then as performance started to trickle 1:52:01 down due to these 60 ads fatiguing, ads 1:52:04 started to need to be turned off, spend 1:52:05 started getting distributed away from 1:52:07 them and the account slowly fell back to 1:52:09 what it was previously. Now the 1:52:10 proposition that we put forward to this 1:52:12 brand well before any of that happened 1:52:14 was that if you took just $10,000 of 1:52:18 this budget, just 10k out of it, and you 1:52:21 decrease from Meta and you moved it over 1:52:23 to creative production, this could get 1:52:25 us with an agency that we were talking 1:52:26 to about 40 assets. Not all unique, some 1:52:29 were just hook swaps, some were statics, 1:52:30 but it would give us 40 assets. 1:52:32 This would allow us to almost double the 1:52:33 amount of active ads in the account, 1:52:35 which would likely unlock at least a 5% 1:52:38 improvement in efficiency. Now a 5% 1:52:41 improvement in efficiency on 350k is the 1:52:44 equivalent of times 2.5%, a 17.5k 1:52:49 saving. So we could either drop 17.5k 1:52:52 out of this budget, which we already 1:52:54 did, we dropped 10, so we've effectively 1:52:55 made 7 and a half grand, or we could 1:52:58 keep the budget where it is and this 1:52:59 would provide us with an incremental 3x 1:53:02 was their return at the time on this, so 1:53:04 we would make an extra 40 to 50k a 1:53:06 month. And so you would actually get 1:53:07 revenue expansion in this business by 1:53:09 keeping spend the same but 1:53:10 redistributing it over to production. 1:53:13 It's a similar concept to just changing 1:53:14 between channels, but instead of it 1:53:15 being a channel, it's creative with an 1:53:18 output. But unfortunately, this founder 1:53:19 was very set on the fact that they 1:53:21 didn't need new creative, that we could 1:53:23 just repurpose these existing assets, 1:53:25 and that would continue the business 1:53:26 going. And this all definitely could 1:53:28 have been fixed well ahead of time 1:53:30 simply through a slight allocation of 1:53:32 this budget over into production. So, 1:53:34 definitely avoid being that brand that 1:53:37 sees creative as a cost center rather 1:53:39 than a revenue driver. Now, one last 1:53:41 tactical piece of advice here that 1:53:43 applies right now as of April 2026, this 1:53:47 might change as the algorithms change, 1:53:49 this might not apply into the future, 1:53:51 but one of the best ways to quickly test 1:53:53 high volumes of assets right now, we 1:53:55 actually do this on all of our own ads, 1:53:57 is that you post them as trial reels and 1:53:59 you cut the CTA off. So, if there's a 1:54:00 call to action at the end, cut the call 1:54:02 to action out, throw it up as a trial 1:54:04 reel on the business page that's going 1:54:05 to run through or the personal page 1:54:06 that's going to run through, and then 1:54:08 what you will get back is two things. 1:54:09 Number one, you'll see how viral does 1:54:11 the post actually go, does it get any 1:54:13 views, does it not, so you're going to 1:54:14 have comparative data. But number two, 1:54:15 you're also going to get retention 1:54:16 graphs, which is going to show you how 1:54:18 you've retained viewers throughout the 1:54:21 ad. Now, the retention graph is going to 1:54:23 show you was the hook bad, was the 1:54:24 bridge bad, was the body bad, and this 1:54:27 lets you iterate quickly on the next 1:54:29 shoot without needing to actually push 1:54:30 the ad straight into the ad manager. And 1:54:32 you can do this before it's perfected. 1:54:34 So, if there's some issues with the 1:54:35 captions or there's some stuff that 1:54:36 isn't great and needs to go in for 1:54:37 another round, you can just take the V1, 1:54:40 throw it as a trial reel, cut the call 1:54:41 to action off in app, and get feedback 1:54:44 while that round two version is actually 1:54:45 getting worked on. And then you could 1:54:47 pass even more feedback on during that 1:54:49 round two within 24 hours before you've 1:54:52 even got the final asset. And so, we 1:54:53 found this is a really effective way to 1:54:55 take a bulk amount of creative and 1:54:57 quickly test, particularly on hook 1:54:58 variations as well. So, we can take like 1:55:00 20 hook variations, throw it up as trial 1:55:02 reels, what hooks actually perform the 1:55:03 best, and then only put five ads into 1:55:05 the account rather than wasting 20 ads 1:55:07 and spending money on trying to figure 1:55:08 it out. So, now we move into creative 1:55:10 fatigue. So, there's two concepts to 1:55:12 understand with creative fatigue. Number 1:55:13 one is the two reasons why creatives 1:55:15 fatigue, and the other concept is to 1:55:17 understand the maximum spend threshold 1:55:20 of an ad over time. So, when you think 1:55:22 about launching an ad, and on the Y axis 1:55:24 we'll put spend, and on the X axis we 1:55:27 have time. 1:55:28 You launch the ad, it does well. And so, 1:55:31 one of two things happen. Either you 1:55:32 start manually spending more on it, so 1:55:34 you start increasing budgets, or Meta 1:55:36 will go and automatically distribute 1:55:38 spend to it. And so, spend goes up and 1:55:39 up and up and up. Eventually, there will 1:55:41 be a limit in how much daily spend that 1:55:44 ad can hold. 1:55:46 You'll reach that limit, you'll probably 1:55:47 go past it, you'll course correct, and 1:55:50 then you'll find the maximum daily spend 1:55:52 limit of the ad. Let's say it's at like 1:55:53 $500 a day or $1,000 a day. 1:55:56 From there, this ad will continue 1:55:58 spending moving forward in time. It will 1:55:59 eventually start fatiguing. It will 1:56:02 start falling off, you'll either start 1:56:03 pulling budgets or the CBO will pull 1:56:05 budgets out of it, and then eventually 1:56:06 you will just turn it off entirely or 1:56:08 Meta will stop distributing spend. 1:56:10 Now, this area under the curve here is 1:56:13 the total spend capacity of the ad. So, 1:56:15 you can go and do derivative math and 1:56:18 just get an understanding of how much 1:56:19 area there is there, and you will have 1:56:21 total spend capacity. Now, what's 1:56:24 important to understand is number one, 1:56:26 every ad has a total spend capacity, 1:56:28 every ad will fatigue, every ad will get 1:56:30 turned off. Number two is you can change 1:56:32 what this graph looks like over time. 1:56:34 And so, if you want, you can squeeze ads 1:56:37 harder in the short term, but they will 1:56:40 fatigue faster. And so, you could take 1:56:42 this ad all the way up to here in daily 1:56:44 spend, but it just means it's probably 1:56:45 going to fatigue and be turned off here. 1:56:47 Or you could go even more aggressive and 1:56:49 you could actually launch it at a high 1:56:50 budget, take it way up, but then it's 1:56:51 going to fall off. Or you could 1:56:53 obviously do the opposite. And you 1:56:55 could, under portfolio management 1:56:57 theory, which is what we did at the 1:56:58 start, is that you actually just want to 1:57:00 introduce this ad, but keep it at $100 a 1:57:02 day, and you'd rather have it live for 6 1:57:05 months than die off within 6 weeks. Now, 1:57:08 this is ultimately more so going to be a 1:57:10 product of the inventory constraint and 1:57:12 demand within the business. So, if you 1:57:14 have a lot of inventory that you need to 1:57:15 sell through for the current 1:57:17 profitability position in that month, 1:57:19 then obviously you should go with option 1:57:21 one here and just squeeze the winning ad 1:57:23 to sell through as much product as 1:57:24 possible, even if it's at a degraded 1:57:26 MER. On the other hand, if you don't 1:57:28 have much inventory, um or if you're 1:57:30 waiting for more inventory to happen, or 1:57:32 if you're growing too quickly that your 1:57:33 cash conversion cycle can't actually uh 1:57:35 keep up with the growth rate, well, then 1:57:37 why would you do this? This would be a 1:57:39 ridiculous decision. You would instead 1:57:41 want to operate at higher efficiency and 1:57:42 let the ad breathe for 6 months. 1:57:44 However, if you're working with an 1:57:46 agency, they don't have any of that 1:57:48 wider context within the business unless 1:57:50 they're asking for it unless you're 1:57:51 actively giving it to them, and so they 1:57:54 will see an ad that's winning, and they 1:57:55 will go, "Let's squeeze the hell out of 1:57:57 it." Because of two things. Number one, 1:57:59 we just started working with this brand, 1:58:00 and this will give us a great 1:58:01 30-to-90-day case study. And number two, 1:58:04 is well, this is giving us the best 1:58:05 ROAS, and we report on ROAS or 1:58:07 efficiency and profit, and so let's do 1:58:09 this. But the issue is it completely 1:58:11 screws up the long-term longevity of the 1:58:12 ad account and the business because 1:58:14 you've over-prioritized short-term gain 1:58:16 for the long-term stability in the ad 1:58:18 account and the inventory purchasing. 1:58:21 So, you're just going to go out of stock 1:58:22 here, and you're going to ruin a lot of 1:58:23 what's going on. Then the two drivers of 1:58:25 creative fatigue. Number one, is the ad 1:58:27 itself just becomes stale. 1:58:29 So, this is a function of it's just 1:58:31 reached its maximum spend capacity, and 1:58:33 it will ultimately just achieve at the 1:58:34 end of its life cycle. 1:58:36 Um this could also be just a function of 1:58:38 the ad has the same visuals, it has the 1:58:40 same concepts, it's targeting the same 1:58:42 audience as the other creatives in the 1:58:44 account, and so you'll get much more 1:58:46 faster fatigue because you're targeting 1:58:47 the same audience with the same value 1:58:49 propositions. The second reason ads 1:58:51 fatigue is that the audience is actually 1:58:53 just too small. When we went through the 1:58:55 persona component of uh this video, we 1:58:58 talked about how specificity in the 1:59:00 persona is critical to being able to 1:59:02 script a better ad that resonates with a 1:59:04 very specific target demographic. This 1:59:06 is also going to generate way better 1:59:08 results cuz you're talking to a few 1:59:10 people rather than a lot. Making very 1:59:12 good top of funnel unaware ads is 1:59:14 incredibly difficult. The skill ceiling 1:59:16 is very high. Most people shouldn't try 1:59:17 to do it unless you're already a $50 1:59:20 million plus brand. Now, 1:59:22 with that, if you speak to a very small 1:59:24 persona, there is obviously going to be 1:59:27 a maximum daily spend capacity that you 1:59:29 can spend talking to that persona. And 1:59:31 so, if we go back to the example before 1:59:34 of men trying to run sub-4-hour 1:59:37 marathons that have gone to the gym, 1:59:39 that are trying to cut 5 minutes off 1:59:40 their time using creatine, very specific 1:59:43 target demographic. Now, there's still a 1:59:44 lot of people there. There are a lot of 1:59:45 people are trying to run marathons and 1:59:46 trying to run sub-4, but you're not 1:59:48 going to be able to spend, particularly 1:59:49 in the Australian market, $10,000 a day 1:59:51 on that creative. It's just not going to 1:59:53 happen. And so, you need to find out 1:59:55 what the spend capacity is, keep it 1:59:57 there, rotating creative so that once 2:00:00 they use fatigue, we can continue to hit 2:00:01 that audience over and over again, 2:00:03 because that is an audience that churns, 2:00:05 which is really important. So, people 2:00:06 will enter into that audience, then run 2:00:08 a marathon and enter out of it. So, 2:00:09 there's constantly new people coming in 2:00:11 that we can continue to saturate. But, 2:00:14 we don't want to oversaturate, 2:00:16 over-prioritize, overspend, drive up 2:00:18 frequency, and then just ruin the 2:00:20 profitability of us targeting this 2:00:21 audience altogether. Now, an obvious key 2:00:23 learning here is that broader audiences 2:00:28 will scale much further. 2:00:30 And so, if you have a top of funnel 2:00:32 unaware ad that speaks to a much larger 2:00:34 TAM, total addressable market, you will 2:00:36 obviously be able to spend a lot more 2:00:38 money on that. In a B2B context, an 2:00:40 example of this is that if I have a 2:00:42 persona call out at the front of one of 2:00:44 our ads that says e-commerce CFOs in 2:00:47 Australia earning over a hundred million 2:00:50 dollars a year in your business. That is 2:00:52 such a tiny pool. There's like how many 2:00:54 people fit that particular persona in 2:00:56 Australia? It's probably a thousand, 2:00:59 maybe two thousand. And so because of 2:01:01 that, when I go and put that ad in 2:01:02 market, I can't spend a lot on that ad. 2:01:04 There would be no point of me going and 2:01:05 putting ten thousand dollars a day 2:01:07 behind that creative because I would 2:01:09 then just be hitting the thousand people 2:01:10 like six hundred times a day. Makes no 2:01:12 sense. But then if I expand the persona 2:01:15 callout to e-commerce founders earning 2:01:17 over a hundred K a year, well then 2:01:19 that's a much larger target demographic. 2:01:21 So the daily spend capacity improves. 2:01:24 Now does that mean I should just do 2:01:25 broad callouts? Absolutely not because 2:01:28 broad callouts means you have to have 2:01:29 better creative that resonates with more 2:01:30 people. And so you're actually better 2:01:32 off having the more specific persona 2:01:35 callouts that force you to create better 2:01:36 assets and then move into broader 2:01:38 audiences once you hit a particular 2:01:40 scale. In terms of tactical things that 2:01:42 you can do when creative fatigue, there 2:01:45 are changes that you can make. Number 2:01:46 one is you can make iterations. And so 2:01:48 you can take an asset that's already 2:01:50 done well, which I mentioned before, a 2:01:51 hero ad that started to fatigue, and you 2:01:53 can just rotate new hooks on and you'll 2:01:55 be able to get more spend through that 2:01:56 asset. Number two is you can move the 2:01:58 ads over into a cost cap campaign and 2:02:01 you can try to get more spend through 2:02:02 here. Now this will also work. It will 2:02:04 fatigue the creative faster and it's a 2:02:06 short-term tactical media buying 2:02:08 strategy. It's not something that's 2:02:10 really going to materially change the 2:02:11 business. Number three, you could go and 2:02:13 duplicate ad sets. It's not something 2:02:15 that I would recommend. I'm not even 2:02:17 going to go and put it down here. I know 2:02:18 it's something that people still argue 2:02:20 works. I think it's too tacky as a media 2:02:22 buying strategy. It used to work back in 2:02:24 the day and I could explain the concept 2:02:25 as to why it works and why it doesn't 2:02:27 now. But if it's something you want to 2:02:28 do, you can do it, but I wouldn't 2:02:29 recommend it. You can take the post ID 2:02:31 of the creative so it retains all of the 2:02:33 engagement and you can go and move that 2:02:36 into another campaign like a scaling 2:02:37 campaign to try to get more spend 2:02:39 through it. This sometimes works, but 2:02:41 once again, like this isn't going to 2:02:42 save the account. And then you could 2:02:44 rotate, which is kind of a variation of 2:02:45 number one. You could rotate formats. 2:02:48 And so if this is a video that's doing 2:02:50 really well, you could try to translate 2:02:52 it into an image. If it's a VSL that's 2:02:54 doing really well, you could try to 2:02:56 translate it into UGC. If it's an image 2:02:59 that's doing really well, you could try 2:03:01 it translated into an unaware long-form 2:03:03 copy Stan image, which is actually 2:03:05 completely different strategy. So 2:03:07 they're really the four options as a 2:03:08 creative fatigue. It is just worth 2:03:10 noting that once a creative fatigue 2:03:11 fatigues, there's not like a whole lot 2:03:13 you can do about it. You're That's why 2:03:14 you need the portfolio management 2:03:16 approach, and you need other ads that 2:03:17 are doing well so that you can fall back 2:03:19 on them. There are things that you can 2:03:20 do though just to squeeze a little bit 2:03:22 more out of the back end of a creative, 2:03:24 which is what these tactics are. Tying 2:03:26 all of that back into portfolio 2:03:28 management, on account spending greater 2:03:30 than 50k per month, really the job of 2:03:32 the media buyer is portfolio management 2:03:35 across the ads and diversifying risk 2:03:38 here. Portfolio management against the 2:03:40 products, and assuring commercial 2:03:43 objectives within the business is 2:03:45 aligned with what is actually occurring 2:03:47 within the ad account. And so the 80/20 2:03:49 Pareto principle's rule applies into 2:03:52 ads, and so the question becomes, do we 2:03:54 want this naturally occurring within the 2:03:56 account, or do we want to force back 2:03:57 against it through introducing some 2:03:59 structure so that we're not 2:04:01 over-leveraged on ads? 2:04:02 Same thing applies into products. And 2:04:04 then on commercials, this becomes, are 2:04:06 we making a balance sheet or a P&L play? 2:04:09 And what I mean by that is that agencies 2:04:11 will always optimize to maximize profit. 2:04:14 Okay, the idea is how do we have the 2:04:15 highest efficiency, how do we have the 2:04:17 highest profit contribution? That is 2:04:19 what they are KPI'd on. But often in a 2:04:21 lot of businesses, if you do that, if 2:04:23 you over-optimize towards profit, you 2:04:24 end up end up under-optimizing towards 2:04:27 the balance sheet, and you end up in a 2:04:29 position where you have a lot of unsold 2:04:30 inventory sitting on the balance sheet 2:04:32 that isn't selling through. And yes, you 2:04:34 are making profit, and yes, you are 2:04:35 spitting out cash, but the cash gets 2:04:37 tied up in the balance sheet quite 2:04:39 quickly. And so, you could have a P&L 2:04:41 that says $2 million in profit, but then 2:04:44 you have a balance sheet that has $3 2:04:46 million in unsold inventory, and this 2:04:48 business is actually negative $1 million 2:04:49 in cash because all of this cash that 2:04:52 they produced just moved into unsold 2:04:53 inventory. And this is fixable through 2:04:55 paid media, through the account 2:04:57 structure. You just need to restructure 2:04:59 the account to sell through this stock. 2:05:01 It's not going to give you the best 2:05:02 return. Your efficiency isn't going to 2:05:03 be that good. You're going to have to 2:05:04 make creatives around this stock as 2:05:06 well, which is going to be a hard pill 2:05:07 to swallow, but it's going to allow you 2:05:09 to get out of this position and 2:05:10 rebalance the business. And this is 2:05:12 ultimately what the media buy needs to 2:05:13 be across. If the media buy isn't across 2:05:15 inventory, that is not good. Creative 2:05:18 strategy is worthless without the 2:05:20 production. And so, moving into creative 2:05:22 production, teams briefing, and the 2:05:24 process, I want to start with the three 2:05:26 content types. Number one is winning 2:05:28 content replications. So, this is taking 2:05:31 your winners and replicating them in 2:05:33 some capacity, okay? Whether that's 2:05:34 iteration on the hook, whether this is 2:05:36 translating them into different formats, 2:05:38 you're taking concepts, you're taking 2:05:40 ideas, you're taking creatives that have 2:05:41 already worked in the past, and you're 2:05:43 replicating on them. This is by far the 2:05:45 highest impact content type that you can 2:05:48 do across paid as well as organic. If 2:05:51 you have organic content that's 2:05:52 performing well, what most people don't 2:05:54 do enough is just repost the same stuff, 2:05:56 redo the same stuff over and over and 2:05:58 over again because every single time you 2:05:59 do it, it performs just the same, but it 2:06:01 reaches a new audience. So, this is 2:06:03 where a lot of your effort should go. 2:06:05 Then you move into iterative content. 2:06:06 This is trying to fix underperformers 2:06:08 through changes in hooks, changes in 2:06:10 formats. You're working on tangential 2:06:13 creative ideas. So, if you have a 2:06:15 concept that's working, you're making a 2:06:16 slight change to a new persona, or 2:06:18 you're introducing a new offer, and 2:06:19 you're trying to get it to work. This is 2:06:21 the second highest priority prioritized 2:06:24 by leverage. And then down the bottom 2:06:26 here is net new concepts. This is 2:06:28 testing completely new new completely 2:06:31 new angles, completely new concepts. You 2:06:34 should be doing this and you should be 2:06:35 allocating time and energy to it. How 2:06:37 much you allocate is based on the 2:06:39 performance up here as well as your risk 2:06:41 profile. What I would not recommend is 2:06:43 some people do 0% here. This is a 2:06:45 terrible idea because all of this will 2:06:47 eventually fatigue and fall off and 2:06:49 you'll be in a very terrible position. 2:06:50 So, you need to constantly be finding 2:06:51 new stuff to then flow up to the top of 2:06:53 the funnel here. But, I would say that 2:06:55 most people actually overweight towards 2:06:58 net new concepts. I would say that I 2:07:00 actually do this myself in my own 2:07:02 organic content, which is that probably 2:07:04 80 to 90% of the content I put out is 2:07:07 net new. It's new concepts, it's new 2:07:09 ideas, it's new stuff I've never spoken 2:07:11 about before and 10 to 20% is just the 2:07:14 exact same stuff over and over again. 2:07:15 And what ends up happening is all of the 2:07:17 winning content replication always does 2:07:19 unbelievably well, holds 80% of the 2:07:21 views and this stuff doesn't really work 2:07:22 that well. It does allow for continuous 2:07:25 winners to be pushed up here, but this 2:07:27 is probably an over-allocation. And so, 2:07:29 my recommendation, generally speaking, 2:07:31 this is going to change contextual to 2:07:33 the business, is that you want about 20 2:07:36 to 30% of effort down here, another 20 2:07:39 to 30% of effort down here, and about 50 2:07:43 to 60% effort up here. Now, when it 2:07:45 comes to briefing in creators, the key 2:07:48 here is that the quality of the output 2:07:50 of the creator is going to be a product 2:07:52 of the quality of the brief. Okay, so 2:07:54 you want to do two things when you're 2:07:55 sourcing creators. Number one is you 2:07:58 want to check what they've made in the 2:07:59 past. You want examples and if you're 2:08:01 not doing this, that's kind of crazy. 2:08:03 That this is kind of self-evident, but 2:08:05 you want examples of what they've 2:08:05 produced in the past. But, not only 2:08:07 that, this is kind of where the 2:08:08 additional source is is that you go 2:08:10 under that and you get the brief 2:08:13 of the example that they created. Cuz 2:08:15 that's going to give you a direct 2:08:17 one-for-one comparison of how does this 2:08:18 creator produce against a brief? Was the 2:08:21 brief super in-depth and therefore we 2:08:23 need to match that level? Did this 2:08:24 creator make something amazing with a 2:08:26 terrible brief? Okay, then we know that 2:08:28 we can give them relatively broad 2:08:29 strokes and they're going to pull it 2:08:30 into something that's going to be good. 2:08:32 Now, generally speaking, I want the 2:08:34 highest quality, most in-depth in-depth 2:08:36 brief possible. If they don't want to 2:08:37 hold to the script, that's fine. Shoot 2:08:39 one that's held to the script, shoot one 2:08:41 that feels more natural. We can choose 2:08:43 the best, we can run them both. Um but 2:08:45 you really want examples and the briefs 2:08:47 associated with those examples to be 2:08:49 able to set yourself up for success when 2:08:51 you're going and building out creative 2:08:52 briefs. When it comes to building out 2:08:54 teams around production, I've seen a lot 2:08:56 of different teams that all produce 2:08:58 enormous amounts of volume, but there's 2:08:59 a few different key hires that you 2:09:01 should think through when you're 2:09:02 thinking about maximizing volume within 2:09:04 a team. Uh number one is a video editor, 2:09:06 specifically for video ads. 2:09:10 Now, if this person can be AI native or 2:09:12 someone that uh has built a lot of AI 2:09:14 creatives in the past, this is even 2:09:16 better because this just unlocks the 2:09:17 ability to make way more assets and work 2:09:20 with way more footage that doesn't 2:09:22 actually exist yet. 2:09:23 Another one is a designer for statics. 2:09:26 Okay, now both of these can be based 2:09:28 anywhere globally. Both of them can be 2:09:31 part-time contractors based on actual 2:09:33 outcome or they can be full-time within 2:09:34 the business. Then as a third, you want 2:09:37 a creative strategist. Now, this is a 2:09:39 new role that didn't really exist two to 2:09:42 three years ago and therefore there 2:09:43 aren't actually many creative strategist 2:09:45 out there. So, what do you actually look 2:09:47 for when hiring for this role? Well, you 2:09:50 you look for number one, copywriters 2:09:52 because creative strategy is primarily 2:09:55 copywriting. They're writing the 2:09:57 scripts, they're writing the brief for 2:09:59 the static image, they're ideating on 2:10:00 new personas. This is exactly what 2:10:02 copywriters have been doing except 2:10:04 copywriters are now redundant because 2:10:06 LLMs will just produce most of it. And 2:10:08 so, these people are coming up with the 2:10:10 ideas and then they're moving this into 2:10:12 a Claude scale, which needs to be built 2:10:14 by you and brand purposed to then output 2:10:16 an actual script that then goes into 2:10:18 production. The next option is you just 2:10:20 poach a creative strategist from 2:10:22 somewhere else. There's enough floating 2:10:23 around now that you can probably find 2:10:25 one. Or the third is you actually 2:10:28 backfill this role from a designer or a 2:10:30 video editor. And so, you train a 2:10:31 designer or a video editor editor up 2:10:34 into becoming a creative strategist. And 2:10:36 this actually allows for faster feedback 2:10:38 loops cuz they can ideate on all the 2:10:40 ideas and then they can design and then 2:10:42 they can obviously brief in for any 2:10:43 video edits. These are really the two 2:10:45 pathways into creative strategists as it 2:10:47 currently stands. Now, in terms of a 2:10:48 side note here, what a lot of people 2:10:50 will do with their creative strategists, 2:10:52 which depending on the scale and volume 2:10:54 that you're at, depends on whether you 2:10:55 want to do this or not, is they will 2:10:57 actually pay out a percentage of ad 2:10:59 spend 2:11:00 based on the creatives that they 2:11:02 obviously come up with and then brief 2:11:04 in. And so, the creative strategist is 2:11:06 directly incentivized to maximize hit 2:11:08 rates because if their ads don't get 2:11:09 spend and don't perform, they don't get 2:11:11 paid. Typically, this sits at about 3% 2:11:14 of ad spend. So, you'll pay out the 2:11:15 creative strategist on top of a base 2:11:17 salary with a percentage variable model 2:11:20 that is obviously directly proportional 2:11:22 to the results within the account. Now, 2:11:24 I've seen this work really well on 2:11:25 mid-size brands. When you get really 2:11:27 big, this gets outrageous cuz these 2:11:28 creative strategists end up getting paid 2:11:30 like $80,000 a month and so then the 2:11:32 model kind of breaks and you end up 2:11:33 moving away from it. A really important 2:11:35 part of this whole component of all 2:11:38 production is where does AI fit into the 2:11:41 mix? When you break down the creative 2:11:43 process end to end, you've got ideation 2:11:46 as step number one. You then have 2:11:48 briefing. You then move into the actual 2:11:50 production and then you move into the 2:11:52 publishing into the account and 2:11:54 ultimately the analysis process that 2:11:56 feeds back into here on future ideation. 2:12:00 Now, a lot of people are going and 2:12:02 applying AI and they're building clawed 2:12:03 skills for this right here, the ideation 2:12:06 process. This right now, at least at the 2:12:08 time of recording, is not the place 2:12:10 where you want to automate because the 2:12:12 reality is is that all of the LLMs just 2:12:14 output the mean. And so, no matter how 2:12:16 much context you provide, no matter how 2:12:18 much documentation you have, no matter 2:12:20 how much you build out your skills to go 2:12:22 and scrape Reddit and all of these 2:12:23 different forms to come up with unique 2:12:24 ideas, ultimately the ideation process 2:12:27 here ends up not being strong enough to 2:12:29 differentiate from competitors, and you 2:12:31 still need a creative strategist to 2:12:32 really do a lot of heavy lifting here to 2:12:34 validate the concepting. 2:12:36 When you then move into briefing and 2:12:38 script writing, this is when this can be 2:12:40 heavily leveraged with AI. That's when 2:12:42 you want to build a Claude skill to be 2:12:44 able to walk you through like a 2:12:46 20-question questionnaire about who's 2:12:48 your persona, who's the concept, who's 2:12:49 the offer, do you want this to be 2:12:51 long-form, what stage of awareness is it 2:12:53 at? And so, you're answering all of 2:12:54 these questions, and then ultimately at 2:12:56 the end of it, it pumps you out a script 2:12:58 or a brief that's going to be 99% of the 2:13:00 way there, and you just have a 1% tweak. 2:13:03 On production, there's a lot of leverage 2:13:04 right now in AI as well. It's obviously 2:13:06 going to speed up workflows, but in VSLs 2:13:08 and static image ads as well, you can 2:13:11 pretty much get away with AI for 90% of 2:13:13 it. Then in the analysis process, this 2:13:15 is also where AI becomes a little bit 2:13:18 dangerous because it won't over weight 2:13:20 to the hierarchy of metrics that you're 2:13:21 looking at, and won't sit towards the 2:13:23 importance of the account structure and 2:13:25 aligning the context of the creative to 2:13:28 the commercial objectives of the 2:13:29 business. So, there is leverage here. 2:13:31 You can use it in analysis. We use it in 2:13:34 analysis. But there is so much context 2:13:36 required that it still misses a lot of 2:13:38 the time, and if you're using AI 2:13:40 analysis to make core decisions within 2:13:42 the business, 2:13:43 a lot of the time it actually misleads 2:13:44 you in the wrong direction. And so, with 2:13:46 AI application, I would try to skip the 2:13:48 outsides and really maximize throughput 2:13:51 on the middle. Want to make a quick note 2:13:52 here on content-first brands, which is 2:13:56 brands that are starting with organic 2:13:57 content first before they even touch 2:13:59 paid. I think this is the approach for 2:14:02 '26 and '27 because it's not due to the 2:14:05 fact that organic content outperforms 2:14:07 paid, but it's due to the fact that 2:14:09 organic content forces you to become 2:14:11 good at creative, which then compounds 2:14:13 into paid when you know how to make 2:14:15 content that people actually want to 2:14:16 watch. And so my recommendation to 2:14:18 smaller founders that are watching this 2:14:20 video is that I would be posting 2:14:22 organically every single day. I would be 2:14:24 getting good at figuring out what good 2:14:25 content looks like, see what actually 2:14:27 gains views, look at the retention 2:14:28 graph, look at the share and save rates 2:14:31 because this is ultimately going to 2:14:32 improve your ability to make ads that 2:14:34 are actually at a good quality. Another 2:14:36 side note here is that when it comes to 2:14:38 creators, you want to make sure that 2:14:41 they're regional to where you're 2:14:42 actually targeting. So as an example of 2:14:44 this, in Australia, you want Australian 2:14:46 accents. When you go to the US and you 2:14:48 try to translate an Australian accent 2:14:50 into the market, it just does not 2:14:51 perform as well. So you need native US 2:14:53 accents. Uh UK to AU actually translates 2:14:57 decently well, but still native accents 2:14:59 always outperform. So you want to make 2:15:00 sure the creative is a native to the 2:15:02 actual region that you're targeting. It 2:15:03 can be the exact same script, it just 2:15:06 has to be a local creator. And then one 2:15:07 final recommendation when it comes to 2:15:09 production is on copywriting. 2:15:11 Copywriting is still one of the highest 2:15:13 leverage skills when it comes to 2:15:15 building out video assets scripts and 2:15:18 also building out static images. 2:15:21 Particularly long-form copy static 2:15:22 images. There are four books on 2:15:24 copywriting that I have read and I 2:15:26 recommend you read all of them if you're 2:15:28 wanting to become good at creative 2:15:29 strategy and copywriting. Number one is 2:15:31 Breakthrough Advertising by Eugene 2:15:33 Schwartz. Number two is Scientific 2:15:34 Advertising by Claude Hopkins. Number 2:15:36 three is Ogilvy on Advertising by David 2:15:39 Ogilvy. And then number four is 2:15:41 Influence, the Psychology of Persuasion. 2:15:44 If you read those four books, you will 2:15:45 become exponentially better at being 2:15:47 able to actually script out copy that 2:15:50 will convert better through ads. So if 2:15:51 we now go and put it all together, the 2:15:53 foundation is that you want to 2:15:54 understand how the ad platforms work. 2:15:56 Understand that the creative is the 2:15:58 targeting. Concept separation is what's 2:16:00 going to allow you to reach new 2:16:01 audiences. Then when it comes to 2:16:02 concepts, you want to understand how to 2:16:04 split up a concept, start with personas, 2:16:06 move into angles, then move into the 2:16:08 offer, and as the final component, you 2:16:10 want to bring in the format. Hooks drive 2:16:12 80% of the view through within an ad, so 2:16:14 it's really important to spend time in 2:16:16 optimizing that portion of the ad, and 2:16:18 usually hook testing across multiple 2:16:20 different videos. The three components 2:16:21 of a hook is the visual, the copy, and 2:16:23 the audio. You can change any three of 2:16:26 these variables and improve performance. 2:16:28 Then, you move into the format. You want 2:16:30 to start with the concepting first, then 2:16:31 always move into formats. Statics are 2:16:33 great for quick testing on new concepts, 2:16:35 and then they can be validated and moved 2:16:37 into videos. Statics also are incredibly 2:16:39 underrated, and most people don't think 2:16:41 they perform well because it's a skill 2:16:42 issue. I have seen static image ads 2:16:44 spend $500,000 in lifetime spend. And 2:16:47 you want to always be careful on DPAs. 2:16:48 DPAs are good, but they generally 2:16:49 overspend and the ROAS number over 2:16:51 inflates them. When it comes to testing, 2:16:53 you always want to be testing one 2:16:54 variable at a time. You want amount 2:16:56 spent to be the best proxy for 2:16:58 performance, then ROAS and the tertiary 2:17:00 metrics. And the actual structure of the 2:17:02 ad account should be a product of the 2:17:03 complexity of the business. When it 2:17:05 comes to thinking through volume versus 2:17:07 quality, you want as much volume as 2:17:08 possible, as long as it meets the bar of 2:17:10 quality that we would classify it as a 2:17:13 good ad. If it's a good creative, no one 2:17:15 will be able to guess which one will 2:17:16 win, so put them all into the account. 2:17:17 When ads fatigue, it's because you're 2:17:19 either trying to push too much spend 2:17:21 through a niche audience, and it's 2:17:22 driving up frequency, or the ad has 2:17:25 simply reached its lifetime spend, it's 2:17:27 falling off, and you need to diversify 2:17:29 by thinking through portfolio management 2:17:31 risk. When it comes to production, we 2:17:33 ran through all of the relevant and tame 2:17:35 structure that you should think through, 2:17:36 as well as the fact that you want to be 2:17:38 putting majority of your production cost 2:17:40 towards iteration, rather than net new 2:17:42 concept design, because that's going to 2:17:43 drive most of the performance. And then 2:17:45 if we do one final breakdown of spend 2:17:48 per month, and the recommendations on 2:17:50 how you should approach creative, at 4 2:17:52 to 30k a month, the founder's creating 2:17:54 the content, it can be done on an 2:17:55 iPhone, you want it to come across 2:17:57 natural, all the leverage is going to be 2:17:58 in how strong the scripts are. You You 2:18:00 to be working on three core concepts, 2:18:02 you really don't need more than that at 2:18:03 this scale. You just want to do these 2:18:05 well. 2:18:06 You want a high testing allocation. 2:18:08 Realistically, 60 to 80% of your 2:18:09 budget's going to go towards testing 2:18:10 because you just don't have enough spend 2:18:12 yet to even scale winners. And then you 2:18:14 can use trial reels as well as a hack 2:18:16 for validating ideas. Once you move into 2:18:18 30 to 100k per month, you want two to 2:18:21 three UGC creators. 25% of this spend 2:18:24 needs to be going towards production. 2:18:26 You're going to move into five to six 2:18:27 concepts. You want to start branching 2:18:29 out and de-risking. You're going to have 2:18:31 an and then likely a scaling 2:18:33 campaign. You're sitting at 30 to 100 2:18:35 new creatives per month here, or you're 2:18:37 using our creative calculator and it'll 2:18:38 give you an exact amount. And then as a 2:18:40 little hack at the bottom, you want to 2:18:42 be introducing partnership ads. And then 2:18:44 if you're at over 100k per month in ad 2:18:46 spend, you really want a full creative 2:18:48 team. Someone who's doing scripting, you 2:18:50 want an editor, you want four to six 2:18:52 creators, and you want a designer. Now, 2:18:54 these could be part-time, but you want 2:18:56 these resources in place so that you 2:18:58 have the production machine in place to 2:19:01 churn out creative. That's really what 2:19:02 you want to be looking for at this ad 2:19:03 spend level, is you need a system that 2:19:05 consistently produces creative on an 2:19:07 ongoing basis. If that isn't in place, 2:19:09 you're just constantly going to be 2:19:10 scrambling and creative will be the 2:19:12 bottleneck. You want at least 10% as a 2:19:14 minimum budget going towards creative at 2:19:16 this spend level. You want eight to 10 2:19:18 active concepts within the account. You 2:19:20 want at least 100 new assets per month 2:19:22 based on the basic 1k per ad formula. 2:19:25 And you want a financial model in place 2:19:26 that's tracking your cost per asset, 2:19:28 ideally across the different sources. 2:19:30 So, how much is these creative pieces 2:19:33 costing you? And then, what is the 2:19:35 average profit contribution per ad on 2:19:37 these creatives? So, we can start to see 2:19:39 if the cost per asset of these different 2:19:41 content types is actually worth it. And 2:19:43 you want to be tracking ROI at an 2:19:44 individual ad level based on ad spend 2:19:47 against the cost of production. For most 2:19:49 people right now, creative is the 2:19:50 constraint on meta. You don't know what 2:19:53 good looks like until you start doing 2:19:54 volume, building out concepts, 2:19:56 scripting, getting ads in the account, 2:19:58 and beginning that feedback loop. So, as 2:20:00 a wrap-up, if you're a performance 2:20:02 marketer and you have at least 2 years 2:20:03 experience, please reach out to us in 2:20:04 the description below. We're always 2:20:06 hiring. And if you're an e-commerce 2:20:07 brand or a retail business doing at 2:20:09 least 5 to 10 million dollars a year at 2:20:10 a minimum, you can also click below or 2:20:12 reach out to us and we'll do a 2:20:13 completely free audit where everything 2:20:15 that you've learned in this video, we 2:20:17 will actually translate directly into 2:20:19 the ads that you're running, directly 2:20:20 into your ad account. We'll pick it 2:20:21 apart so that you can know how to unlock 2:20:23 the current bottleneck that's stopping 2:20:25 you from growing. And also, please 2:20:26 subscribe.