WEBVTT 00:00:00.000 --> 00:00:02.339 Losing money on digital ads usually isn't because 00:00:02.339 --> 00:00:05.120 the ads themselves don't work. It happens because 00:00:05.120 --> 00:00:07.639 we hit that boost button way too fast. We're 00:00:07.639 --> 00:00:10.000 operating completely without a compass. Oh, totally. 00:00:10.140 --> 00:00:12.199 We treat it like a casino. Right. Throw money 00:00:12.199 --> 00:00:16.420 down and hope the algorithm figures it out for 00:00:16.420 --> 00:00:19.039 us. Welcome to the deep dive. If you've ever 00:00:19.039 --> 00:00:21.460 felt like your advertising budget is just disappearing 00:00:21.460 --> 00:00:24.600 into a black hole, this is absolutely for you. 00:00:24.739 --> 00:00:27.960 Yeah. It's a painful feeling. It really is. Today, 00:00:28.120 --> 00:00:31.660 we are exploring a complete AI advertising strategy 00:00:31.660 --> 00:00:34.719 workflow. We're using Claude Code. We'll look 00:00:34.719 --> 00:00:37.079 at how to stop the guessing game completely. 00:00:37.140 --> 00:00:39.560 We're mapping out customer pain points. Right. 00:00:39.700 --> 00:00:41.840 And we're going to build an agency in a box, 00:00:42.020 --> 00:00:45.119 one that creates actual campaign structure before 00:00:45.119 --> 00:00:47.700 a single dollar is spent. Because that invisible 00:00:47.700 --> 00:00:50.799 structure behind the scenes is, well, it's everything. 00:00:51.310 --> 00:00:53.929 Yeah. The visual ad you see on your phone. Yeah. 00:00:54.009 --> 00:00:56.310 That's really just the tiny tip of a massive 00:00:56.310 --> 00:01:00.049 strategic iceberg. Exactly. Let's start by unpacking 00:01:00.049 --> 00:01:02.670 the real problem businesses face today. The guessing 00:01:02.670 --> 00:01:05.829 game. Right. We tend to treat ad platforms like 00:01:05.829 --> 00:01:08.430 a magical vending machine. Yep. You've got a 00:01:08.430 --> 00:01:11.530 product you poured your soul into. So you jump 00:01:11.530 --> 00:01:14.269 onto Meta. You throw 50 bucks at a campaign. 00:01:14.769 --> 00:01:17.019 And you just wait for the sales to roll in. But 00:01:17.019 --> 00:01:18.900 they don't. No, they usually don't. And when 00:01:18.900 --> 00:01:21.640 those sales inevitably don't happen, we immediately 00:01:21.640 --> 00:01:25.260 look for a scapegoat. People always want to blame 00:01:25.260 --> 00:01:27.719 the platform. Yeah. It's much easier than blaming 00:01:27.719 --> 00:01:30.400 their own lack of preparation. You hear it all 00:01:30.400 --> 00:01:32.920 the time, right? People declare that Meta is 00:01:32.920 --> 00:01:35.280 completely dead. Oh, constantly. Or they say 00:01:35.280 --> 00:01:38.299 TikTok traffic is too low quality. They complain 00:01:38.299 --> 00:01:40.719 LinkedIn is too expensive unless you're some 00:01:40.719 --> 00:01:43.879 massive enterprise company. Right. But I mean, 00:01:43.879 --> 00:01:47.620 the platform is almost never the actual bottleneck. 00:01:47.640 --> 00:01:49.780 No. The algorithm is just doing exactly what 00:01:49.780 --> 00:01:52.340 you told it to do. Which is usually nothing specific. 00:01:52.599 --> 00:01:55.480 Exactly. Yeah. The fundamental problem is a total 00:01:55.480 --> 00:01:57.900 lack of strategic direction. It's kind of like 00:01:57.900 --> 00:02:00.340 walking into a pitch black room. You throw a 00:02:00.340 --> 00:02:02.659 handful of darts and just hope you hit a customer's 00:02:02.659 --> 00:02:05.439 wallet. What we actually need is to turn on the 00:02:05.439 --> 00:02:08.340 stadium lights. We need to use a sniper rifle. 00:02:08.580 --> 00:02:10.639 That's a much better way to look at it. Or, you 00:02:10.639 --> 00:02:12.259 know, think of it like building a house. Okay. 00:02:12.419 --> 00:02:14.080 You would just buy a bunch of bricks, throw them 00:02:14.080 --> 00:02:16.159 in a yard, and expect a mansion. Right. You need 00:02:16.159 --> 00:02:19.060 a blueprint. Exactly. To hit your target, you 00:02:19.060 --> 00:02:21.439 need a meticulous blueprint. You need to conduct 00:02:21.439 --> 00:02:24.520 deep audience research. You have to build out 00:02:24.520 --> 00:02:27.680 distinct customer personas and plan your platforms. 00:02:27.860 --> 00:02:30.219 You literally have to map out a multi -stage 00:02:30.219 --> 00:02:33.120 funnel before writing a single word. Let's ground 00:02:33.120 --> 00:02:35.340 this. Let's use the razor company example from 00:02:35.340 --> 00:02:37.240 our sources. Oh, that's a great example. If you 00:02:37.240 --> 00:02:40.319 skip the strategy, a bad ad usually ends up saying 00:02:40.319 --> 00:02:43.020 something incredibly generic. Right. It just 00:02:43.020 --> 00:02:46.159 says, high quality razors delivered right to 00:02:46.159 --> 00:02:48.840 your door. Which is grammatically fine. Sure. 00:02:49.020 --> 00:02:52.159 But it's deeply boring. Right. I mean, it doesn't 00:02:52.159 --> 00:02:54.419 actually give me a compelling reason to stop 00:02:54.419 --> 00:02:56.319 scrolling. Yeah, it doesn't make me pull out 00:02:56.319 --> 00:02:58.860 my credit card. Not at all. It completely fails 00:02:58.860 --> 00:03:01.419 to agitate a real pain point. A good ad goes 00:03:01.419 --> 00:03:04.539 so much deeper into the human experience. Exactly. 00:03:04.800 --> 00:03:07.699 It talks about the frustration of massively overpaying 00:03:07.699 --> 00:03:10.520 for name brands. Or the physical discomfort of 00:03:10.520 --> 00:03:13.620 using dull blades. Yeah. Or my absolute favorite 00:03:13.620 --> 00:03:16.219 angle we found in the source material, the sheer 00:03:16.219 --> 00:03:19.539 indignity of waiting for a pharmacy employee 00:03:19.539 --> 00:03:23.139 to unlock a plastic razor case. Oh, man. That 00:03:23.139 --> 00:03:25.719 is the worst experience. It really is. Standing 00:03:25.719 --> 00:03:27.539 there for 10 minutes, you're just trying to buy 00:03:27.539 --> 00:03:30.819 a basic grooming product. It's universally infuriating. 00:03:30.860 --> 00:03:33.479 So infuriating. And when you tap into that specific 00:03:33.479 --> 00:03:36.960 frustration, your ad instantly resonates. But 00:03:36.960 --> 00:03:39.479 finding those nuanced psychological angles, that 00:03:39.479 --> 00:03:41.580 takes a massive amount of investigative work. 00:03:41.719 --> 00:03:44.439 It takes serious brainpower. Yeah. And hours 00:03:44.439 --> 00:03:46.580 of parsing through customer feedback. Right. 00:03:47.080 --> 00:03:49.180 And that is exactly why most business owners 00:03:49.180 --> 00:03:52.199 skip it. The research phase feels entirely overwhelming. 00:03:52.300 --> 00:03:54.400 You just want to launch something today. Yeah. 00:03:54.520 --> 00:03:56.919 Which naturally makes me wonder about the human 00:03:56.919 --> 00:04:00.439 element here. Why do human marketers so consistently 00:04:00.439 --> 00:04:03.080 skip this foundational step when we know it works? 00:04:03.319 --> 00:04:06.479 Because proper strategy is incredibly tedious. 00:04:07.020 --> 00:04:09.939 I mean, people are exhausted. They just want 00:04:09.939 --> 00:04:12.120 the quick dopamine hit of seeing a live campaign 00:04:12.120 --> 00:04:14.699 running on their dashboard. It feels like progress, 00:04:14.900 --> 00:04:17.199 even if it's wasteful. We blame the platform 00:04:17.199 --> 00:04:20.100 because building a real blueprint is hard. Yeah, 00:04:20.100 --> 00:04:22.500 precisely. But this friction is exactly where 00:04:22.500 --> 00:04:25.160 Claude Code fundamentally changes the paradigm. 00:04:25.399 --> 00:04:27.620 It really does. We're moving away from treating 00:04:27.620 --> 00:04:30.500 AI as a simple chatbot. We're turning it into 00:04:30.500 --> 00:04:33.779 a structured, highly disciplined strategist. 00:04:33.839 --> 00:04:36.240 It represents a massive shift in how we interact 00:04:36.240 --> 00:04:38.600 with large language models. We're talking about 00:04:38.600 --> 00:04:42.019 setting up a comprehensive agency in a box. right 00:04:42.019 --> 00:04:44.339 on your local machine. Let's define the actual 00:04:44.339 --> 00:04:46.819 technology driving this. This system operates 00:04:46.819 --> 00:04:50.300 entirely inside Visual Studio Code. That is simply 00:04:50.300 --> 00:04:53.040 a free coding workspace where AI reads files. 00:04:53.339 --> 00:04:55.019 Right. It's just a digital environment. Yeah. 00:04:55.100 --> 00:04:57.360 Gives the AI a place to organize its thoughts. 00:04:57.680 --> 00:05:00.220 And it can review your documents without uploading 00:05:00.220 --> 00:05:02.199 them to a public server. The fascinating part 00:05:02.199 --> 00:05:05.839 is how it operates. It utilizes 15 distinct analytical 00:05:05.839 --> 00:05:10.930 skills. And it relies on five parallel AI agents. 00:05:11.009 --> 00:05:13.970 And parallel AI agents are just multiple AI programs 00:05:13.970 --> 00:05:17.069 working on different tasks at once. Right. And 00:05:17.069 --> 00:05:19.129 that parallel architecture is the secret sauce 00:05:19.129 --> 00:05:23.910 here. It completely transforms a linear workflow 00:05:23.910 --> 00:05:27.209 into a synchronized engine. I have to offer a 00:05:27.209 --> 00:05:29.509 vulnerable admission here, though. Even after 00:05:29.509 --> 00:05:31.850 studying all this prompt engineering, I still 00:05:31.850 --> 00:05:34.209 wrestle with blank chat boxes myself. Really? 00:05:34.470 --> 00:05:37.149 Yeah. I open a new browser tab, I just stare 00:05:37.149 --> 00:05:39.329 at them and my mind goes blank on what to prompt. 00:05:39.629 --> 00:05:42.610 Oh, you are definitely not alone. Blank canvas 00:05:42.610 --> 00:05:45.209 syndrome is a huge hurdle. It is. When you have 00:05:45.209 --> 00:05:47.490 infinite possibilities, you usually end up doing 00:05:47.490 --> 00:05:50.009 nothing at all. Or you do something overly simplistic. 00:05:50.389 --> 00:05:52.569 Right. If you just ask a normal web -based AI, 00:05:52.850 --> 00:05:55.129 write me a Facebook ad for my razor company, 00:05:55.329 --> 00:05:59.250 it fails miserably. Horribly. You get these generic, 00:05:59.410 --> 00:06:02.529 overly enthusiastic slogans. Yeah. It usually 00:06:02.529 --> 00:06:04.569 sounds like a very confused robot trying to sell 00:06:04.569 --> 00:06:07.610 you soap. Because it lacks any strategic constraints. 00:06:07.829 --> 00:06:10.829 Exactly. A standard chatbot is just guessing 00:06:10.829 --> 00:06:13.490 at what sounds good. But with Claude Code... 00:06:13.709 --> 00:06:16.949 Inside this workspace, you give it a highly strict 00:06:16.949 --> 00:06:20.149 role. You force it to analyze your budgets, your 00:06:20.149 --> 00:06:22.629 sales funnels, and your landing pages simultaneously. 00:06:23.449 --> 00:06:26.389 By locking it into those parameters, the output 00:06:26.389 --> 00:06:29.670 is incredible. Oh, yeah. It outputs fully realized 00:06:29.670 --> 00:06:32.990 campaign structures. It writes tailored ad copy 00:06:32.990 --> 00:06:36.230 for every specific social platform. Right. It 00:06:36.230 --> 00:06:38.709 handles the mathematical budget allocation. It 00:06:38.709 --> 00:06:42.110 even generates preliminary ROI projections. And 00:06:42.110 --> 00:06:45.189 those projections often include your expected 00:06:45.189 --> 00:06:49.410 ROAS. Let's clarify that term. ROAS is just the 00:06:49.410 --> 00:06:53.060 revenue made per ad. dollars spent. Right. Which 00:06:53.060 --> 00:06:55.019 helps you understand if the campaign is actually 00:06:55.019 --> 00:06:57.360 sustainable. So looking at the architecture of 00:06:57.360 --> 00:07:00.120 this engine, why must these five specific agents 00:07:00.120 --> 00:07:03.459 run in parallel instead of just working sequentially 00:07:03.459 --> 00:07:05.540 down a list? Because the strategic pieces are 00:07:05.540 --> 00:07:08.000 heavily interdependent. And if they run sequentially, 00:07:08.160 --> 00:07:10.540 the budget agent might allocate funds in a way 00:07:10.540 --> 00:07:12.339 that doesn't align with the creative agent's 00:07:12.339 --> 00:07:14.620 video concepts. Running them together ensures 00:07:14.620 --> 00:07:17.019 your funnel mechanics and your ad copy match 00:07:17.019 --> 00:07:19.040 up perfectly in real time. They build the puzzle 00:07:19.040 --> 00:07:21.629 pieces at the exact same time. Exactly. Knowing 00:07:21.629 --> 00:07:24.230 how the architecture functions in theory is great, 00:07:24.389 --> 00:07:27.769 but the application is what matters. So what 00:07:27.769 --> 00:07:30.129 actually happens when you press go on a real 00:07:30.129 --> 00:07:32.629 world business? This is where the abstract concepts 00:07:32.629 --> 00:07:35.949 finally become incredibly tangible. Let's dig 00:07:35.949 --> 00:07:38.089 into the Dollar Shave Club demo from the source 00:07:38.089 --> 00:07:41.589 material. Perfect. As a user, you open the workspace 00:07:41.589 --> 00:07:45.269 and run one simple command. You just type slash 00:07:45.269 --> 00:07:49.610 ads, ad strategy, dollarshaveclub .com. It is 00:07:49.610 --> 00:07:52.689 remarkable that an entire agency workflow kicks 00:07:52.689 --> 00:07:55.490 off from just that single line of text. It really 00:07:55.490 --> 00:07:58.930 is. The moment you hit enter, the AI instantly 00:07:58.930 --> 00:08:02.209 enters a deep discovery phase. It autonomously 00:08:02.209 --> 00:08:05.069 browses the target website. It studies the underlying 00:08:05.069 --> 00:08:07.470 business model. And then it generates a massive 00:08:07.470 --> 00:08:09.509 strategic output report. Right. It starts by 00:08:09.509 --> 00:08:11.189 grading your current setup, right? It gives you 00:08:11.189 --> 00:08:13.689 a baseline to work from. It issues specific readiness 00:08:13.689 --> 00:08:17.589 scores. And audience clarity scores. It tells 00:08:17.589 --> 00:08:19.550 you exactly where your current messaging is weak. 00:08:19.709 --> 00:08:22.529 But the most powerful part of the report is the 00:08:22.529 --> 00:08:24.949 persona generation. Oh, absolutely. In this demo, 00:08:25.110 --> 00:08:29.350 it created a highly specific buyer persona. It 00:08:29.350 --> 00:08:31.870 called him the professional groomer. Let's unpack 00:08:31.870 --> 00:08:34.710 him for a second. Who exactly is this guy? And 00:08:34.710 --> 00:08:37.679 why does he matter? According to the AI's analysis, 00:08:38.019 --> 00:08:41.379 he is male, generally 28 to 40 years old. Okay. 00:08:41.480 --> 00:08:43.879 He lives in the suburbs, usually with the family. 00:08:43.940 --> 00:08:46.379 He wants to look sharp for his career. But his 00:08:46.379 --> 00:08:50.000 absolute primary driver is saving time. And having 00:08:50.000 --> 00:08:52.659 that psychological profile. changes your copy 00:08:52.659 --> 00:08:54.879 completely. Completely. You aren't just selling 00:08:54.879 --> 00:08:56.960 a sharp piece of metal anymore. You are selling 00:08:56.960 --> 00:08:59.659 time. You are selling a frictionless morning 00:08:59.659 --> 00:09:02.059 routine for a guy who just wants his weekends 00:09:02.059 --> 00:09:05.000 back. Once it understands that guy, it maps out 00:09:05.000 --> 00:09:07.759 the entire architectural funnel. Right. It breaks 00:09:07.759 --> 00:09:09.899 the customer journey down into four distinct 00:09:09.899 --> 00:09:12.580 logical stages. Walk us through how it structures 00:09:12.580 --> 00:09:14.480 those four stages, because beginners usually 00:09:14.480 --> 00:09:17.320 get this entirely wrong. Yeah, they do. First 00:09:17.320 --> 00:09:20.470 is the awareness stage. This copy strictly focuses 00:09:20.470 --> 00:09:23.190 on introducing the hidden problem the user didn't 00:09:23.190 --> 00:09:26.559 realize they had. Second is consideration. This 00:09:26.559 --> 00:09:29.220 stage uses logic to explain why your specific 00:09:29.220 --> 00:09:32.519 product is actually the superior solution. That 00:09:32.519 --> 00:09:34.600 logical bridge is crucial. What comes after they 00:09:34.600 --> 00:09:36.799 understand the solution? Third is the conversion 00:09:36.799 --> 00:09:39.659 stage. This is where you finally push the hard, 00:09:39.799 --> 00:09:42.740 specific offer to get the sale. And finally, 00:09:42.879 --> 00:09:46.759 retargeting. This stage uses urgency or discounts 00:09:46.759 --> 00:09:48.960 to bring back the people who clicked your site 00:09:48.960 --> 00:09:51.440 but abandoned their cart. It separates the psychological 00:09:51.440 --> 00:09:53.980 messaging perfectly. It really does. Inexperienced 00:09:53.980 --> 00:09:56.700 marketers usually try to smash all four of those 00:09:56.700 --> 00:10:00.139 complex stages into one single overwhelming Facebook 00:10:00.139 --> 00:10:02.519 post. Yeah. Ask for marriage on the first date, 00:10:02.600 --> 00:10:05.559 you know. Well, imagine scaling to five parallel 00:10:05.559 --> 00:10:08.139 agents instantly mapping out a four stage funnel 00:10:08.139 --> 00:10:11.159 for a brand new product to sex silence. It's 00:10:11.159 --> 00:10:13.019 just staggering to think about the time saved. 00:10:13.389 --> 00:10:15.049 But before we dig into how you actually install 00:10:15.049 --> 00:10:17.149 this system, I have a quick question. Are these 00:10:17.149 --> 00:10:19.769 AI ROI projections actually guaranteed returns? 00:10:20.149 --> 00:10:23.029 Oh, absolutely not. You should never treat AI 00:10:23.029 --> 00:10:25.769 projections as guaranteed money. They are strictly 00:10:25.769 --> 00:10:28.309 starting baselines to help guide your initial 00:10:28.309 --> 00:10:31.129 budget planning. You still need to gather real 00:10:31.129 --> 00:10:34.070 campaign data once the ads are live. It gives 00:10:34.070 --> 00:10:36.429 us a clear compass, not a magic crystal ball. 00:10:36.529 --> 00:10:38.490 That is the perfect way to frame it. Let's take 00:10:38.490 --> 00:10:40.190 a quick pause to hear from our sponsor, and then 00:10:40.190 --> 00:10:42.509 we'll get into the actual installation. Sponsor. 00:10:43.300 --> 00:10:47.840 Okay, we are back. We just mapped out this massive 00:10:47.840 --> 00:10:50.620 multi -stage strategy. We did. So let's talk 00:10:50.620 --> 00:10:52.539 about the actual engine running under the hood. 00:10:53.200 --> 00:10:55.899 I want to demystify the technology so it doesn't 00:10:55.899 --> 00:10:58.600 feel like magic. Sure. How do these various skills 00:10:58.600 --> 00:11:01.639 actually constrain the AI? The entire engine 00:11:01.639 --> 00:11:04.679 runs on what we call skill .md files. Let's define 00:11:04.679 --> 00:11:06.779 that clearly. Those are plain text files that 00:11:06.779 --> 00:11:09.559 give the AI strict rules. Right, because large 00:11:09.559 --> 00:11:11.379 language models are naturally unpredictable. 00:11:12.539 --> 00:11:15.059 These plain text files act as heavy guardrails. 00:11:15.159 --> 00:11:19.019 You were giving Claude a strict, unbendable playbook 00:11:19.019 --> 00:11:21.360 rather than just a loose conversational prompt. 00:11:21.700 --> 00:11:23.820 Our sources break these skill files down into 00:11:23.820 --> 00:11:26.720 four main categories. You have strategy, creative 00:11:26.720 --> 00:11:29.840 copy, funnels, and landing page audits. Each 00:11:29.840 --> 00:11:32.139 of those file categories is designed to solve 00:11:32.139 --> 00:11:34.559 a totally different mechanical part of the advertising 00:11:34.559 --> 00:11:36.679 campaign. Let's talk about the installation process 00:11:36.679 --> 00:11:40.159 outlined here. You download Visual Studio Code. 00:11:40.730 --> 00:11:43.129 You add the cloud code extension directly from 00:11:43.129 --> 00:11:46.110 Anthropic. You drop these plain text skill files 00:11:46.110 --> 00:11:49.389 into a local folder. You open a session and you 00:11:49.389 --> 00:11:52.110 type the slash command. It sounds like a lot 00:11:52.110 --> 00:11:55.049 of technical steps, but it's actually a very 00:11:55.049 --> 00:11:58.149 linear, straightforward process. I have to push 00:11:58.149 --> 00:12:00.269 back slightly on the accessibility here, though. 00:12:00.490 --> 00:12:03.610 The research claims this is easy, but. Using 00:12:03.610 --> 00:12:06.629 phrases like visual studio code and local extensions 00:12:06.629 --> 00:12:09.929 sounds absolutely terrifying to a non -coder. 00:12:10.070 --> 00:12:12.789 Ah, fair. Is a local plumber or a neighborhood 00:12:12.789 --> 00:12:14.809 dentist really going to be able to set this up? 00:12:14.950 --> 00:12:17.870 I completely understand why it sounds intimidating. 00:12:18.190 --> 00:12:21.419 But they absolutely can use this. You truly do 00:12:21.419 --> 00:12:23.899 not need to be a software developer. Visual Studio 00:12:23.899 --> 00:12:26.940 Code is literally just a digital desk. It is 00:12:26.940 --> 00:12:29.279 merely the piece of furniture where the AI sits, 00:12:29.519 --> 00:12:31.440 opens its notebook, and reads your instructions. 00:12:31.700 --> 00:12:34.700 So a freelance graphic designer or a local gym 00:12:34.700 --> 00:12:37.259 owner can set this up on their laptop without 00:12:37.259 --> 00:12:40.419 breaking anything. 100%. The scale of the budget 00:12:40.419 --> 00:12:43.259 changes between a local gym and a massive software 00:12:43.259 --> 00:12:46.320 company. but the technological foundation stays 00:12:46.320 --> 00:12:49.299 the exact same. That makes the barrier to entry 00:12:49.299 --> 00:12:52.159 feel much more approachable. Yeah. I do have 00:12:52.159 --> 00:12:54.259 one question about those specific skill categories, 00:12:54.480 --> 00:12:57.320 though. Why do we need a dedicated landing page 00:12:57.320 --> 00:13:00.960 audit skill if the AI already wrote a high -performing 00:13:00.960 --> 00:13:03.919 ad? Because the ad creative only does one job. 00:13:04.039 --> 00:13:06.500 It gets the click. If a user clicks your brilliant 00:13:06.500 --> 00:13:08.899 ad, but your website takes 10 seconds to load, 00:13:09.039 --> 00:13:11.759 or it has a confusing checkout process, they 00:13:11.759 --> 00:13:14.240 bounce immediately. The landing page is what 00:13:14.240 --> 00:13:16.740 actually secures the revenue. A great ad is useless 00:13:16.740 --> 00:13:19.100 if a broken website scares them away. Exactly. 00:13:19.179 --> 00:13:22.139 If you don't fix the page, you are just setting 00:13:22.139 --> 00:13:24.320 your budget on fire. So we have built the local 00:13:24.320 --> 00:13:27.259 engine. We have run the deep analytical strategy. 00:13:27.620 --> 00:13:31.059 Now we need the actual words. We need the visceral 00:13:31.059 --> 00:13:33.340 language that will stop a distracted person from 00:13:33.340 --> 00:13:35.500 scrolling right past your brand. This is where 00:13:35.500 --> 00:13:38.169 we finally enter the creative output phase. The 00:13:38.169 --> 00:13:40.649 documentation outlines another very specific 00:13:40.649 --> 00:13:44.429 command for this. Right. You type slash ads hooks 00:13:44.429 --> 00:13:48.669 dollarshaveclub .com. But there is a crucial 00:13:48.669 --> 00:13:51.830 non -negotiable rule attached to this step. You 00:13:51.830 --> 00:13:54.269 must never ask the AI for creative hooks before 00:13:54.269 --> 00:13:56.429 the deep strategy is completely finished. Never. 00:13:56.590 --> 00:14:00.090 Why is that specific sequence so critical to 00:14:00.090 --> 00:14:02.070 the workflow? Because if you ask for the catchy 00:14:02.070 --> 00:14:05.009 hooks first, the AI has no foundation. You just 00:14:05.009 --> 00:14:07.870 get generic clickbait garbage. Truly, scroll 00:14:07.870 --> 00:14:10.190 -stopping hooks only come from deep customer 00:14:10.190 --> 00:14:12.929 insight. And the strategy document provides that 00:14:12.929 --> 00:14:15.149 exact insight. That makes complete sense. The 00:14:15.149 --> 00:14:17.370 AI needs to read the strategy first to understand 00:14:17.370 --> 00:14:20.029 the psychology. Once it does, it generates 20 00:14:20.029 --> 00:14:23.210 to 25 distinct hooks, and it intentionally bases 00:14:23.210 --> 00:14:26.070 them on core psychological angles. It leans heavily 00:14:26.070 --> 00:14:28.549 into powerful human drivers, like frustration, 00:14:28.909 --> 00:14:31.870 burning curiosity, or undeniable social proof. 00:14:32.240 --> 00:14:34.259 Let's look at the specific creative examples 00:14:34.259 --> 00:14:36.480 generated in the source material. Yeah, let's 00:14:36.480 --> 00:14:39.240 do it. The first one is categorized as the locked 00:14:39.240 --> 00:14:42.820 razor case hook. I genuinely love this angle. 00:14:42.899 --> 00:14:46.299 It is so relatable. The AI wrote this hook. You 00:14:46.299 --> 00:14:48.139 shouldn't need to wait 10 minutes for someone 00:14:48.139 --> 00:14:50.679 to open a razor case. It is incredibly visceral. 00:14:50.860 --> 00:14:53.259 I mean, almost everyone has experienced that 00:14:53.259 --> 00:14:56.960 exact annoyance at a pharmacy. It taps into pure, 00:14:57.120 --> 00:15:00.059 unfiltered frustration. The second example takes 00:15:00.059 --> 00:15:02.320 a slightly different angle. It's the dull blade 00:15:02.320 --> 00:15:05.919 hook. The copy reads, the moment you drag a three 00:15:05.919 --> 00:15:07.840 -week -old blade across your face and accept 00:15:07.840 --> 00:15:10.799 the burn. Again, it paints a very vivid, painful 00:15:10.799 --> 00:15:13.720 picture. It targets a specific physical reality 00:15:13.720 --> 00:15:15.960 that men accept just because they're too lazy 00:15:15.960 --> 00:15:18.480 to go buy new blades. But the context of where 00:15:18.480 --> 00:15:21.059 you place these words matters deeply. Oh, absolutely. 00:15:21.320 --> 00:15:23.240 You can't just slap that exact sentence onto 00:15:23.240 --> 00:15:25.240 every single social network and expect it to 00:15:25.240 --> 00:15:27.820 work. Right. And the AI agents are specifically 00:15:27.820 --> 00:15:29.860 programmed. to know this. If you are running 00:15:29.860 --> 00:15:32.840 ads on Meta, the AI knows you need split screen 00:15:32.840 --> 00:15:36.259 visual scripts. If you are on TikTok, it knows 00:15:36.259 --> 00:15:39.240 you need a rapid three second problem and solution 00:15:39.240 --> 00:15:42.740 setup. Exactly. If you are on YouTube, it scripts 00:15:42.740 --> 00:15:45.940 a sharp arresting first sentence before the user 00:15:45.940 --> 00:15:48.539 can hit the skip button. The engine formats the 00:15:48.539 --> 00:15:51.539 core psychological hook to match the native language 00:15:51.539 --> 00:15:54.419 of each specific environment. But we need to 00:15:54.419 --> 00:15:57.659 highlight a massive warning explicitly stated 00:15:57.659 --> 00:16:01.299 in our research. Yes. Clawed code is an incredibly 00:16:01.299 --> 00:16:04.440 powerful first draft machine. It is absolutely 00:16:04.440 --> 00:16:07.360 not the final step of the process. It reliably 00:16:07.360 --> 00:16:09.879 gets you to the 80 % mark, but you still have 00:16:09.879 --> 00:16:12.340 to step in and do the rest. human beings must 00:16:12.340 --> 00:16:15.059 rigorously edit this final output right you have 00:16:15.059 --> 00:16:18.200 to infuse the ai structure with real gritty customer 00:16:18.200 --> 00:16:20.759 language right you find the specific words by 00:16:20.759 --> 00:16:23.600 reading through niche reddit threads or analyzing 00:16:23.600 --> 00:16:25.980 your one star reviews or digging through your 00:16:25.980 --> 00:16:28.259 customer support tickets the ai is incredibly 00:16:28.259 --> 00:16:30.620 smart but it doesn't actually know what razor 00:16:30.620 --> 00:16:33.200 burn feels like on a tuesday morning which brings 00:16:33.200 --> 00:16:35.480 up a fairly critical question about automation 00:16:36.409 --> 00:16:38.950 If this AI system is so strategically brilliant, 00:16:39.169 --> 00:16:42.269 why not just connect it via API and let it post 00:16:42.269 --> 00:16:44.850 the finished ads directly to the platforms? Because 00:16:44.850 --> 00:16:47.789 the AI doesn't truly feel human frustration, 00:16:48.190 --> 00:16:52.850 joy, or relief. It just mathematically mimics 00:16:52.850 --> 00:16:55.350 those concepts. Human beings have to provide 00:16:55.350 --> 00:16:58.370 the final layer of empathy to make sure the emotional 00:16:58.370 --> 00:17:01.460 tone actually rings true to another human. AI 00:17:01.460 --> 00:17:04.059 gives us speed, but humans provide the necessary 00:17:04.059 --> 00:17:06.660 empathy. That is the essential partnership. The 00:17:06.660 --> 00:17:08.720 machine builds the house, but you have to make 00:17:08.720 --> 00:17:11.099 it a home. Let's step back and synthesize everything 00:17:11.099 --> 00:17:14.000 we've covered today. We have traversed a massive 00:17:14.000 --> 00:17:16.180 amount of ground in this deep dive. We really 00:17:16.180 --> 00:17:18.440 have. We went from staring blankly at a blinking 00:17:18.440 --> 00:17:21.880 cursor all the way to generating platform -specific 00:17:21.880 --> 00:17:24.400 psychological hooks. Digital advertising feels 00:17:24.400 --> 00:17:26.420 like a total gamble when you don't have a reliable 00:17:26.420 --> 00:17:28.680 system. You end up throwing money at a wall and 00:17:28.680 --> 00:17:30.460 just praying the algorithm takes pity on you. 00:17:30.619 --> 00:17:32.940 Right. But implementing a local engine like Cloud 00:17:32.940 --> 00:17:35.819 Code completely changes that paradigm. It provides 00:17:35.819 --> 00:17:37.799 the architectural rigor that most businesses 00:17:37.799 --> 00:17:40.759 lack. It takes the chaotic guesswork. out of 00:17:40.759 --> 00:17:43.359 the equation entirely by building that invisible 00:17:43.359 --> 00:17:45.240 structure we talked about at the top of the show. 00:17:45.420 --> 00:17:47.940 It starts with the deep, sometimes uncomfortable 00:17:47.940 --> 00:17:51.759 psychology of your actual buyer. It moves mathematically 00:17:51.759 --> 00:17:55.039 down to the exact budget allocation across different 00:17:55.039 --> 00:17:58.059 platforms. It ensures that every single ad you 00:17:58.059 --> 00:18:01.140 run has a distinct, measurable purpose within 00:18:01.140 --> 00:18:03.660 a logical funnel. It stops you from hitting that 00:18:03.660 --> 00:18:05.720 dangerous boost button without a plan. I want 00:18:05.720 --> 00:18:07.720 you to pull up the last digital ad you ran for 00:18:07.720 --> 00:18:09.700 your business. Look at the copy in the target. 00:18:09.769 --> 00:18:12.410 getting closely? Yeah. Did you know exactly who 00:18:12.410 --> 00:18:15.329 that specific message was for? Did you know precisely 00:18:15.329 --> 00:18:18.109 what stage of the funnel it belonged to? If the 00:18:18.109 --> 00:18:19.829 honest answer is no, you were probably just throwing 00:18:19.829 --> 00:18:22.769 darts in the dark. If an AI agent can now build 00:18:22.769 --> 00:18:25.309 our entire strategic framework in a matter of 00:18:25.309 --> 00:18:27.450 seconds, does that mean the future of marketing 00:18:27.450 --> 00:18:30.130 relies entirely on our human capacity to feel 00:18:30.130 --> 00:18:33.069 and our ability to beautifully edit that final 00:18:33.069 --> 00:18:37.190 10 % beat? Something to think about. Out to your 00:18:37.190 --> 00:18:37.430 music.