WEBVTT 00:00:00.000 --> 00:00:02.799 You know that feeling. It's 11 p .m. The house 00:00:02.799 --> 00:00:04.799 is dead quiet. You're staring at your monitor 00:00:04.799 --> 00:00:07.740 and the blue light is just burning a hole into 00:00:07.740 --> 00:00:10.279 your retinas. You've got Facebook Ad Manager 00:00:10.279 --> 00:00:12.800 open in one tab, a half -written blog post in 00:00:12.800 --> 00:00:15.140 another, and you're trying to figure out why 00:00:15.140 --> 00:00:17.879 your conversion rate has totally flatlined. You 00:00:17.879 --> 00:00:20.579 are trying to be the writer, the data scientist, 00:00:20.660 --> 00:00:22.660 the designer, and the strategist all at once. 00:00:22.760 --> 00:00:26.280 And it is just exhausting. And the classic solopreneur 00:00:26.280 --> 00:00:27.960 shuffle. You're doing everything, but it feels 00:00:27.960 --> 00:00:30.809 like you're accomplishing nothing. Exactly. It's 00:00:30.809 --> 00:00:32.770 the paralysis of the modern marketer. I mean, 00:00:32.850 --> 00:00:35.049 we have more tools than ever, but somehow we 00:00:35.049 --> 00:00:37.869 have less time. The cognitive load of switching 00:00:37.869 --> 00:00:41.789 hats going from like creative artist to cold 00:00:41.789 --> 00:00:44.570 analyst in five minutes is massive. It burns 00:00:44.570 --> 00:00:46.549 you out faster than the actual work does. It 00:00:46.549 --> 00:00:48.250 does. And that context switching is the killer. 00:00:48.369 --> 00:00:51.409 It's not the work. It's the mental tax of changing 00:00:51.409 --> 00:00:54.039 gears. That is exactly what we are unpacking 00:00:54.039 --> 00:00:56.460 today. We're doing a deep dive into a really 00:00:56.460 --> 00:00:58.859 fascinating concept from marketing expert Corey 00:00:58.859 --> 00:01:01.539 Haynes. It's about stopping that late night shuffle. 00:01:01.840 --> 00:01:04.560 We're looking at how to build an AI marketing 00:01:04.560 --> 00:01:07.159 dream team. And I want to be clear, right up 00:01:07.159 --> 00:01:09.459 front, we are not talking about just asking chat 00:01:09.459 --> 00:01:12.000 GPT to write a blog post and, you know, hoping 00:01:12.000 --> 00:01:14.079 for the best. Right, because that's the mistake 00:01:14.079 --> 00:01:17.060 everyone makes. They treat AI like a magic eight 00:01:17.060 --> 00:01:19.579 ball. You just shake it and hope for a good answer. 00:01:19.799 --> 00:01:22.920 What we're looking at today is treating AI like 00:01:22.920 --> 00:01:28.030 a staff. We're talking about installing 10 specific 00:01:28.030 --> 00:01:31.329 specialized personas or skills that act like 00:01:31.329 --> 00:01:34.010 individual employees with very specific job descriptions. 00:01:34.430 --> 00:01:36.650 A 10 -person marketing department for free. I 00:01:36.650 --> 00:01:39.010 love the sound of that. The roadmap for this 00:01:39.010 --> 00:01:40.269 deep dive is pretty cool. We're going to build 00:01:40.269 --> 00:01:42.329 this team layer by layer. We'll start with the 00:01:42.329 --> 00:01:44.829 creatives, move to the outreach engine, head 00:01:44.829 --> 00:01:46.569 into the science lab, which I'm a little scared 00:01:46.569 --> 00:01:48.829 of, to be honest, and finally, meet the boss. 00:01:48.950 --> 00:01:50.829 The science lab is where the money is made, so 00:01:50.829 --> 00:01:53.069 don't be too scared. But let's start where every 00:01:53.069 --> 00:01:55.959 business starts. the creative team. Let's do 00:01:55.959 --> 00:02:00.000 it. So, expert number one, the page conversion 00:02:00.000 --> 00:02:03.280 optimizer. This one hits close to home because 00:02:03.280 --> 00:02:05.879 I've definitely spent hours picking fonts for 00:02:05.879 --> 00:02:08.599 a landing page, thinking that was the work. We 00:02:08.599 --> 00:02:10.500 all have. We obsess over the aesthetics, the 00:02:10.500 --> 00:02:12.800 colors, the vibe, the logo placement. But this 00:02:12.800 --> 00:02:14.879 agent, the page conversion optimizer, it doesn't 00:02:14.879 --> 00:02:17.620 care about your font choice. It puts on a critic 00:02:17.620 --> 00:02:20.460 hat. It's looking for friction. Define friction 00:02:20.460 --> 00:02:22.939 in this context for me. Friction is anything 00:02:22.939 --> 00:02:25.139 that causes even a millisecond of hesitation 00:02:25.139 --> 00:02:27.840 in the user's brain. And the problem is, as the 00:02:27.840 --> 00:02:30.020 business owner, you're blind to it. You built 00:02:30.020 --> 00:02:31.500 the house. You don't see the cracks. You know 00:02:31.500 --> 00:02:34.020 what you meant to say. This AI agent acts like 00:02:34.020 --> 00:02:36.240 a psychologist. It looks at your headline and 00:02:36.240 --> 00:02:39.500 asks, does this actually address a fear or is 00:02:39.500 --> 00:02:41.840 this just a clever pun that nobody understands? 00:02:42.000 --> 00:02:44.240 The source material had a great example of this 00:02:44.240 --> 00:02:46.719 regarding a productivity app. The generic headline 00:02:46.719 --> 00:02:49.879 was, organize your life better, which sure sounds 00:02:49.879 --> 00:02:52.870 fine. But the optimizer just tore it apart. It 00:02:52.870 --> 00:02:54.969 did. It suggested changing it to stop pulling 00:02:54.969 --> 00:02:58.389 all -nighters. Whoa. That is a visceral difference. 00:02:58.729 --> 00:03:01.550 It is. Organize your life is a vague promise. 00:03:01.889 --> 00:03:04.009 It's vitamin marketing. Stop pulling all -nighters 00:03:04.009 --> 00:03:06.949 is a painkiller. It implies the AI understands 00:03:06.949 --> 00:03:09.050 the consequence of the problem. That's what this 00:03:09.050 --> 00:03:11.930 expert does. It shifts you from talking about 00:03:11.930 --> 00:03:15.090 features to talking about benefits. It forces 00:03:15.090 --> 00:03:17.530 you to answer this so what question. It raises 00:03:17.530 --> 00:03:20.030 a probing question for me, though. Why is it 00:03:20.030 --> 00:03:22.340 so? incredibly hard to critique our own work. 00:03:22.879 --> 00:03:24.879 Why do we need a robot to tell us our headline 00:03:24.879 --> 00:03:28.699 is boring? It's context blindness. We are so 00:03:28.699 --> 00:03:30.740 deep in our own context. We know the product. 00:03:30.819 --> 00:03:33.300 We know the history. We know the jargon that 00:03:33.300 --> 00:03:36.319 we assume the visitor knows it, too. The AI provides 00:03:36.319 --> 00:03:38.639 objective distance. It doesn't care about your 00:03:38.639 --> 00:03:40.719 feelings. It cares about clarity. It simulates 00:03:40.719 --> 00:03:42.860 the cold and patient eyes of a stranger who just 00:03:42.860 --> 00:03:45.280 landed on your site. That is valuable. OK, so 00:03:45.280 --> 00:03:47.620 the optimizer fixes the structure, but we still 00:03:47.620 --> 00:03:50.039 have BlinkPabe syndrome. We need to fill the 00:03:50.039 --> 00:03:52.740 empty boxes with words. Enter expert number two, 00:03:53.080 --> 00:03:56.139 the professional copywriter. Now I have to play 00:03:56.139 --> 00:03:58.479 devil's advocate here. I've seen generic AI writing. 00:03:58.539 --> 00:04:01.599 It loves words like unlock, unleash, and tapestry. 00:04:02.099 --> 00:04:04.060 It sounds like a bad high school valedictorian 00:04:04.060 --> 00:04:06.819 speech. It does. In today's fast paced digital 00:04:06.819 --> 00:04:10.180 landscape, it's just fluff. The reason AI does 00:04:10.180 --> 00:04:11.900 that is because it's trained to be agreeable 00:04:11.900 --> 00:04:14.900 and safe. But the professional copywriter persona 00:04:14.900 --> 00:04:17.620 is different because it's forced to use frameworks. 00:04:18.060 --> 00:04:20.800 It doesn't just write text. It constructs an 00:04:20.800 --> 00:04:22.959 argument. You mentioned frameworks. The deep 00:04:22.959 --> 00:04:25.360 dive highlights things like ADA and PAS. Can 00:04:25.360 --> 00:04:28.720 we break those down? Sure. ADA is classic. Attention, 00:04:29.180 --> 00:04:33.139 interest, desire, action. But PAS is where the 00:04:33.139 --> 00:04:35.600 magic happens, especially for pain point marketing. 00:04:36.279 --> 00:04:39.139 Problem agitation solution. PAS is interesting 00:04:39.139 --> 00:04:42.040 because it seems almost Negative. You start with 00:04:42.040 --> 00:04:44.040 the problem. You have to. You have to validate 00:04:44.040 --> 00:04:46.439 the user's struggle. The example from the deep 00:04:46.439 --> 00:04:49.220 dive was for a meal prep service. A generic prompt 00:04:49.220 --> 00:04:51.980 would just say, buy our healthy food. The copywriter 00:04:51.980 --> 00:04:54.279 agent, using P .A .S., starts with the problem. 00:04:54.720 --> 00:04:56.600 It's 6 p .m., the kids are hungry, and you're 00:04:56.600 --> 00:04:59.480 exhausted. Okay, I feel seen. Right. Thin the 00:04:59.480 --> 00:05:01.839 agitation, twisting the knife a little. Do you 00:05:01.839 --> 00:05:03.660 really want to feel the guilt of the drive -through 00:05:03.660 --> 00:05:05.920 again? Do you want that sluggish feeling tomorrow 00:05:05.920 --> 00:05:09.170 morning? Oof. That guilt trip is effective. And 00:05:09.170 --> 00:05:11.990 then, only then, does it offer the solution. 00:05:12.550 --> 00:05:15.149 Dinner on the table in five minutes. No guilt, 00:05:15.509 --> 00:05:19.490 just health. See, it's not just words. It's an 00:05:19.490 --> 00:05:21.870 emotional journey. I can see how some people 00:05:21.870 --> 00:05:24.129 might think using a framework makes the writing 00:05:24.129 --> 00:05:26.629 feel robotic, though, like painting by numbers. 00:05:27.370 --> 00:05:30.350 Doesn't it lose the soul? That's a common worry. 00:05:30.629 --> 00:05:33.629 But think of the framework as the skeleton. It 00:05:33.629 --> 00:05:36.709 holds the creature up. You, the human, provide 00:05:36.709 --> 00:05:38.889 the flesh and blood, the empathy, the specific 00:05:38.889 --> 00:05:41.550 details. The AI just ensures the bones are in 00:05:41.550 --> 00:05:43.329 the right place so the argument doesn't collapse. 00:05:43.790 --> 00:05:45.490 Fair enough. So we've got the structure, we've 00:05:45.490 --> 00:05:47.550 got the persuasive copy, we're ready to hit publish 00:05:47.550 --> 00:05:49.230 route. Absolutely not. I knew you were going 00:05:49.230 --> 00:05:51.810 to say that. You need expert number three, the 00:05:51.810 --> 00:05:54.209 copy editor. And this is crucial. The writer 00:05:54.209 --> 00:05:56.449 and the editor are two different brains, even 00:05:56.449 --> 00:06:00.009 in a human. When you're writing, you're in creative 00:06:00.009 --> 00:06:03.870 mode generative, messy, flowy. If you try to 00:06:03.870 --> 00:06:06.110 edit while you write, you block the flow. But 00:06:06.110 --> 00:06:08.550 if you publish without editing, you look unprofessional. 00:06:08.689 --> 00:06:10.629 This is the strict teacher with the red pen. 00:06:10.930 --> 00:06:13.769 It is. It's looking for passive voice. It's looking 00:06:13.769 --> 00:06:16.470 for sentences that are over 20 words long. It's 00:06:16.470 --> 00:06:18.709 ensuring the tone is consistent. There's a prompt 00:06:18.709 --> 00:06:20.970 in the source material that literally asks for 00:06:20.970 --> 00:06:24.290 a before and after of a paragraph. And it's shocking 00:06:24.290 --> 00:06:26.029 how much better the after is just by cutting 00:06:26.029 --> 00:06:28.379 the fluff. So why not just let the copywriter 00:06:28.379 --> 00:06:30.860 agent edit itself? Why the separation? Why can't 00:06:30.860 --> 00:06:33.120 I just say, write this and make it good? Because 00:06:33.120 --> 00:06:35.680 of what you could call hallucinated brilliance. 00:06:35.980 --> 00:06:38.920 The AI thinks what it just wrote is perfect. 00:06:39.500 --> 00:06:41.779 By engaging a separate agent, a separate chat 00:06:41.779 --> 00:06:44.839 window or context, you force a modality shift. 00:06:45.139 --> 00:06:48.300 You are explicitly telling the system, stop creating, 00:06:48.500 --> 00:06:50.959 start critiquing. It acts as a quality control 00:06:50.959 --> 00:06:53.790 gate. I like that. A quality control gate. OK, 00:06:53.790 --> 00:06:55.389 so that's our creative team. We've built the 00:06:55.389 --> 00:06:57.089 asset. Now we need to get it in front of people. 00:06:57.470 --> 00:06:59.589 Let's move to segment B, the outreach engine. 00:06:59.610 --> 00:07:01.449 And we have to start with the granddaddy of them 00:07:01.449 --> 00:07:04.769 all, email. Expert number four. And before anyone 00:07:04.769 --> 00:07:07.589 rolls their eyes, email is not dead. I feel like 00:07:07.589 --> 00:07:09.649 we hear email is dead every single year, usually 00:07:09.649 --> 00:07:11.470 from people trying to sell us a course on TikTok 00:07:11.470 --> 00:07:14.509 ads. Exactly. And every year, the data proves 00:07:14.509 --> 00:07:17.949 them wrong. The source mentions a return of $36 00:07:17.949 --> 00:07:21.689 for every $1 spent on email marketing. That is 00:07:21.689 --> 00:07:24.519 astronomical. but only if you do it right. Most 00:07:24.519 --> 00:07:27.600 people just blast B -U -Y now emails until everyone 00:07:27.600 --> 00:07:29.740 unsubscribes. Guilty as charged. So what does 00:07:29.740 --> 00:07:31.819 the email specialist agent do differently? It 00:07:31.819 --> 00:07:34.160 builds sequences. It understands the narrative 00:07:34.160 --> 00:07:36.420 arc. It takes a stranger and moves them from 00:07:36.420 --> 00:07:39.459 who are you to I trust you to take my money. 00:07:40.180 --> 00:07:43.279 It maps out, for example, a four -day flow. Day 00:07:43.279 --> 00:07:46.699 one is pure value, no pitch. Day two might be 00:07:46.699 --> 00:07:49.120 a horror story. A horror story? A vampire's a 00:07:49.120 --> 00:07:51.300 business horror story. What happens if you don't 00:07:51.300 --> 00:07:53.680 solve the problem? For a real estate agent, maybe 00:07:53.680 --> 00:07:55.139 the story's about a couple who tried to sell 00:07:55.139 --> 00:07:57.980 their home alone and lost $20 ,000. It raises 00:07:57.980 --> 00:08:00.220 the stakes. And then day three? Social proof, 00:08:00.600 --> 00:08:02.120 showing that others have survived the horror 00:08:02.120 --> 00:08:04.459 story by working with you. And the probing question 00:08:04.459 --> 00:08:06.879 here, what is the single most critical component 00:08:06.879 --> 00:08:09.100 the AI focuses on? Because I can write a great 00:08:09.100 --> 00:08:11.360 email, but if nobody opens it? The subject line, 00:08:11.579 --> 00:08:14.589 always. It's the gatekeeper. If the subject line 00:08:14.589 --> 00:08:17.509 is boring, the brilliant email inside never exists. 00:08:18.250 --> 00:08:21.810 The AI balances curiosity with clarity. It avoids 00:08:21.810 --> 00:08:24.430 clickbait, but it demands attention. It creates 00:08:24.430 --> 00:08:26.209 that open loop in your brain that you just have 00:08:26.209 --> 00:08:29.569 to close. OK, so email nurtures the people we 00:08:29.569 --> 00:08:31.569 know. But what about the billions of people who 00:08:31.569 --> 00:08:34.649 don't know us? We need Google. Enter expert number 00:08:34.649 --> 00:08:37.889 five, the SEO auditor. Now, SEO usually feels 00:08:37.889 --> 00:08:40.769 like math homework. It's scary. It feels technical. 00:08:41.039 --> 00:08:43.320 People think they need to be coders, but the 00:08:43.320 --> 00:08:47.000 SEO auditor skill acts like a librarian. Imagine 00:08:47.000 --> 00:08:49.580 your website is a book. If the cover is torn, 00:08:49.720 --> 00:08:51.980 the title is missing, and the pages are glued 00:08:51.980 --> 00:08:54.320 together, the librarian Google is going to hide 00:08:54.320 --> 00:08:55.460 it in the basement. They aren't going to put 00:08:55.460 --> 00:08:57.659 it in the front window. This agent checks the 00:08:57.659 --> 00:08:59.299 condition of your book. It's checking for things 00:08:59.299 --> 00:09:01.980 like keywords and H1 tags, right? It's deeper 00:09:01.980 --> 00:09:05.360 than just keywords. It scans for semantic gaps. 00:09:05.899 --> 00:09:08.000 For example, you might write a great post about 00:09:08.000 --> 00:09:10.950 hiking boots. But if you never use the phrase 00:09:10.950 --> 00:09:14.629 ankle support or waterproof or traction, Google's 00:09:14.629 --> 00:09:17.389 algorithm gets confused. It thinks, is this really 00:09:17.389 --> 00:09:19.549 about hiking boots? How can you talk about boots 00:09:19.549 --> 00:09:23.230 without mentioning traction? The AI spots those 00:09:23.230 --> 00:09:26.230 gaps instantly. So it's not that SEO is just 00:09:26.230 --> 00:09:28.490 for coders. It's that the technical checks are 00:09:28.490 --> 00:09:31.690 just tedious. Correct. The agent handles the 00:09:31.690 --> 00:09:33.970 tedious technical check, leaving you to focus 00:09:33.970 --> 00:09:36.169 on the content quality. Now here's where it gets 00:09:36.169 --> 00:09:39.159 really interesting. Expert number six. The programmatic 00:09:39.159 --> 00:09:41.840 SEO strategist. When I read this part of the 00:09:41.840 --> 00:09:44.980 deep dive, my jaw hit the floor a little bit. 00:09:45.340 --> 00:09:47.340 We're talking about scale here that humans can't 00:09:47.340 --> 00:09:50.059 really do manually. This is the big leagues tactic. 00:09:50.159 --> 00:09:52.740 I mean, think about TripAdvisor or Yelp. When 00:09:52.740 --> 00:09:54.820 you search best hotel in Austin, you get a page. 00:09:55.179 --> 00:09:57.399 Best hotel in Boston, you get a page. Do you 00:09:57.399 --> 00:09:59.620 think they have a human writer typing out those 00:09:59.620 --> 00:10:01.679 thousands of pages one by one? That would take 00:10:01.679 --> 00:10:04.500 a century. Exactly. They use programmatic SEO. 00:10:04.860 --> 00:10:07.399 They create one master template, a head term, 00:10:07.610 --> 00:10:10.929 plus a variable, best hotel and city name. Then 00:10:10.929 --> 00:10:14.129 they plug in a database. This AI agent helps 00:10:14.129 --> 00:10:16.190 you design that architecture. It's like building 00:10:16.190 --> 00:10:19.049 a factory instead of handcrafting a chair. Perfectly 00:10:19.049 --> 00:10:21.610 put, it helps you identify the patterns in your 00:10:21.610 --> 00:10:24.370 niche. If you run a job board, you don't write 00:10:24.370 --> 00:10:26.970 500 articles. You build a structure for remote 00:10:26.970 --> 00:10:29.850 job title jobs and let the data fill the pages. 00:10:30.029 --> 00:10:31.570 But I have to stop you there. Doesn't Google 00:10:31.570 --> 00:10:34.129 hate duplicate content? Isn't that a penalty 00:10:34.129 --> 00:10:36.350 waiting to happen? If I just swap Austin for 00:10:36.350 --> 00:10:39.370 Boston, isn't that just spam? That's the catch. 00:10:39.929 --> 00:10:42.110 And that's what the strategist ensures you avoid. 00:10:42.570 --> 00:10:44.509 You cannot just copy -paste the same text and 00:10:44.509 --> 00:10:47.110 swap the city name. You need unique data points 00:10:47.110 --> 00:10:50.210 for every page. Average salary in Austin versus 00:10:50.210 --> 00:10:53.190 Boston. Cost of living. Top industries. The AI 00:10:53.190 --> 00:10:55.809 helps you map out what unique data you need so 00:10:55.809 --> 00:10:58.600 each page is valuable on its own. That is wild. 00:10:58.759 --> 00:11:01.139 It really shifts your thinking from writer to 00:11:01.139 --> 00:11:03.039 architect. You're building systems. And that's 00:11:03.039 --> 00:11:04.919 the theme here. You're becoming a manager of 00:11:04.919 --> 00:11:07.759 systems, not just a doer of tasks. Speaking of 00:11:07.759 --> 00:11:10.679 systems, let's enter the science lab, segment 00:11:10.679 --> 00:11:13.419 C. This is where we stop guessing and start measuring. 00:11:14.080 --> 00:11:16.419 Expert number seven is the analytics tracking 00:11:16.419 --> 00:11:19.980 specialist. This is where dreams go to die or 00:11:19.980 --> 00:11:23.110 to thrive. Most people live on vanity metrics. 00:11:23.389 --> 00:11:26.309 Likes. Likes, views, impressions. They feel good. 00:11:26.330 --> 00:11:28.490 They give you a little dopamine hit, but they 00:11:28.490 --> 00:11:31.049 don't pay the rent. The analytics specialist 00:11:31.049 --> 00:11:34.509 forces you to look at KPIs, key performance indicators, 00:11:34.850 --> 00:11:38.899 leads, cost per acquisition. conversion rate. 00:11:39.139 --> 00:11:40.820 The source mentioned something really specific 00:11:40.820 --> 00:11:42.980 here, filtering out internal traffic. I laughed 00:11:42.980 --> 00:11:45.080 because I am definitely the person who visits 00:11:45.080 --> 00:11:47.539 my own website 50 times a day just to see how 00:11:47.539 --> 00:11:49.379 it looks. And you are ruining your own data. 00:11:49.620 --> 00:11:52.000 If you have a hundred visitors and 50 of them 00:11:52.000 --> 00:11:54.139 are you, your conversion rate looks terrible. 00:11:54.279 --> 00:11:55.980 You think your business is failing but really 00:11:55.980 --> 00:11:58.940 you're just obsessing. The AI acts as a data 00:11:58.940 --> 00:12:01.200 translator. It helps you set up filters and tools 00:12:01.200 --> 00:12:03.799 like Google Analytics 4 which is notoriously 00:12:03.799 --> 00:12:05.440 confusing so you're looking at real customer 00:12:05.440 --> 00:12:08.600 behavior, not your own anxiety. So once we have 00:12:08.600 --> 00:12:11.480 clean data, we start asking questions. What if 00:12:11.480 --> 00:12:14.039 I changed this headline? What if the button was 00:12:14.039 --> 00:12:18.379 red? This brings us to expert number eight, the 00:12:18.379 --> 00:12:21.480 A -B testing designer. This is about the scientific 00:12:21.480 --> 00:12:23.559 method. Most people just change something on 00:12:23.559 --> 00:12:25.759 their site and see what happens, but that's dangerous. 00:12:26.019 --> 00:12:27.919 Maybe sales went up because it's payday, not 00:12:27.919 --> 00:12:30.039 because you changed the button. Right. Correlation 00:12:30.039 --> 00:12:33.580 is not causation. Exactly. The A -B testing designer 00:12:33.580 --> 00:12:36.470 calculates Statistical significance. Explain 00:12:36.470 --> 00:12:39.590 that like I'm five. Okay. Imagine you flip a 00:12:39.590 --> 00:12:42.009 coin three times and it comes up heads three 00:12:42.009 --> 00:12:45.110 times. You might conclude, wow, this coin only 00:12:45.110 --> 00:12:48.090 lands on heads. That's a bad business decision. 00:12:48.450 --> 00:12:50.470 You need to flip it a hundred times to know the 00:12:50.470 --> 00:12:53.500 truth. The AI tells you. Based on your current 00:12:53.500 --> 00:12:56.200 traffic, you need to run this test for 14 days 00:12:56.200 --> 00:12:58.340 to be sure it's not just luck. What's the danger 00:12:58.340 --> 00:13:00.840 of testing without this agent? You call a winner 00:13:00.840 --> 00:13:03.240 too early. You make a permanent change that actually 00:13:03.240 --> 00:13:04.960 hurts your business, all because you had one 00:13:04.960 --> 00:13:08.220 lucky afternoon. The AI prevents that. OK. Expert 00:13:08.220 --> 00:13:11.220 number nine. We're in the home stretch. The user 00:13:11.220 --> 00:13:13.919 is ready to sign up. They're on the form. And 00:13:13.919 --> 00:13:18.419 then they leave. This is the sign up flow optimization 00:13:18.419 --> 00:13:21.500 expert. The last mile problem. You've done all 00:13:21.500 --> 00:13:23.659 the work. and you lose them at the finish line 00:13:23.659 --> 00:13:25.419 because you asked for their phone number. Why 00:13:25.419 --> 00:13:27.399 do we do that? Why do we ask for so much info? 00:13:27.860 --> 00:13:29.879 Greed. Pure data greed. We want to know everything 00:13:29.879 --> 00:13:32.820 about them, just in case. But every extra field 00:13:32.820 --> 00:13:36.080 is cognitive load. It makes the user think, why 00:13:36.080 --> 00:13:37.259 do they need my phone number? Are they going 00:13:37.259 --> 00:13:39.740 to call me in? And once they start thinking, 00:13:39.799 --> 00:13:42.700 they stop clicking. We want zero thinking. We 00:13:42.700 --> 00:13:44.980 want a slippery slide. So this expert looks at 00:13:44.980 --> 00:13:47.500 your form and just starts cutting. Ruthlessly. 00:13:48.059 --> 00:13:50.159 Do you really need their job title right now? 00:13:50.340 --> 00:13:54.909 No. Delete it. It often suggests breaking forms 00:13:54.909 --> 00:13:57.629 into steps. Get the email first. Then once they're 00:13:57.629 --> 00:14:00.509 committed, ask for the name. It reduces the perceived 00:14:00.509 --> 00:14:02.549 effort. I love the idea of reducing cognitive 00:14:02.549 --> 00:14:05.190 load. Just let them in. Exactly. Get the relationship 00:14:05.190 --> 00:14:06.690 started. You can ask for the phone number later. 00:14:06.889 --> 00:14:09.590 So we have our creatives, our outreach, and our 00:14:09.590 --> 00:14:12.289 scientists. But a team of specialists needs a 00:14:12.289 --> 00:14:14.990 leader. If the SEO guy wants to write robot text 00:14:14.990 --> 00:14:17.350 and the poppy writer wants to write poetry, who 00:14:17.350 --> 00:14:19.769 wins? That's expert number 10. The marketing 00:14:19.769 --> 00:14:22.129 strategist. Your boss. The boss. The strategist 00:14:22.129 --> 00:14:25.850 sees the forest, not the trees. It aligns everything 00:14:25.850 --> 00:14:28.269 to a business goal. The example in the deep dive 00:14:28.269 --> 00:14:30.549 was a flower shop. The strategy isn't worrying 00:14:30.549 --> 00:14:32.990 about the H1 tag. It's looking at the calendar. 00:14:33.269 --> 00:14:36.149 It's saying, OK, January is retention, February 00:14:36.149 --> 00:14:39.169 is Valentine's prep, March is partnerships. I 00:14:39.169 --> 00:14:41.830 was surprised by that. Can AI really understand 00:14:41.830 --> 00:14:44.750 something as human as seasonality? Surprisingly, 00:14:45.009 --> 00:14:47.470 yes, because. Seasonality is just a pattern of 00:14:47.470 --> 00:14:50.230 human behavior. The AI has access to vast amounts 00:14:50.230 --> 00:14:52.629 of historical data. It knows when people buy 00:14:52.629 --> 00:14:54.870 flowers and knows when people join gyms. And 00:14:54.870 --> 00:14:58.429 it knows when people buy grills. It maps tactics 00:14:58.429 --> 00:15:00.529 to those timelines. So we have this incredible 00:15:00.529 --> 00:15:02.990 team of 10. But here's the practical question, 00:15:03.110 --> 00:15:04.990 the technical implementation. How do I actually 00:15:04.990 --> 00:15:07.049 hire these robots? Do I need to be a coder? Do 00:15:07.049 --> 00:15:09.289 I need to install Python? This is the best part. 00:15:09.490 --> 00:15:13.289 No, you don't. These experts are just text files. 00:15:13.639 --> 00:15:16.259 They're open source prompts written in Markdown. 00:15:16.600 --> 00:15:18.779 Markdown. It's just a fancy way of formatting, 00:15:19.059 --> 00:15:21.919 test, using bolding, bullet points, headers. 00:15:22.700 --> 00:15:25.720 It helps the AI understand the structure of the 00:15:25.720 --> 00:15:28.039 instructions. So it's literally copy and paste. 00:15:28.100 --> 00:15:30.580 It is copy and paste. You go to the repository, 00:15:30.740 --> 00:15:33.409 Corey Haynes has these on GitHub. You open the 00:15:33.409 --> 00:15:36.370 file for Copywriter, you copy the text, and you 00:15:36.370 --> 00:15:38.750 paste it into your chat with Claude or ChatGPT 00:15:38.750 --> 00:15:40.409 before you start working. It's like handing a 00:15:40.409 --> 00:15:42.529 job description to a temp worker right as they 00:15:42.529 --> 00:15:44.769 walk in the door. Precisely. You're saying, for 00:15:44.769 --> 00:15:47.250 the next 20 minutes, you are not a generic AI. 00:15:47.629 --> 00:15:50.330 You are a professional copywriter using the PAS 00:15:50.330 --> 00:15:52.789 framework. Go. I have to admit something here. 00:15:53.129 --> 00:15:55.350 I often struggle with what they call prompt drift. 00:15:55.870 --> 00:15:58.610 I start a chat, and five minutes later the AI 00:15:58.610 --> 00:16:01.129 is giving me generic advice again. It forgets 00:16:01.129 --> 00:16:03.559 who it's supposed to be. We all do. I still wrestle 00:16:03.559 --> 00:16:06.340 with it myself. It's a known issue with large 00:16:06.340 --> 00:16:09.200 language models. Their context window gets muddy. 00:16:09.679 --> 00:16:12.379 That's why this method is so powerful. By pasting 00:16:12.379 --> 00:16:14.539 that persona definition at the start, often called 00:16:14.539 --> 00:16:17.860 a system message, you anchor the AI. You create 00:16:17.860 --> 00:16:20.659 a boundary. It keeps the employee in character. 00:16:20.899 --> 00:16:23.000 That anchoring concept is key. So if I switch 00:16:23.000 --> 00:16:25.759 tasks, I should probably start a new chat. 100%. 00:16:25.759 --> 00:16:27.779 Don't ask the copywriter chat to check your analytics. 00:16:28.159 --> 00:16:30.860 Open a new window. Hire the new expert. Keep 00:16:30.860 --> 00:16:33.299 the context clean. Okay, let's unpack the big 00:16:33.299 --> 00:16:36.159 picture here. We've gone from being a frantic 00:16:36.159 --> 00:16:38.759 solopreneur staring at a blank screen at 11 p 00:16:38.759 --> 00:16:42.059 .m. to being a manager of experts. That is the 00:16:42.059 --> 00:16:44.350 shift. The future of work isn't about working 00:16:44.350 --> 00:16:46.669 harder or learning every single skill yourself. 00:16:47.149 --> 00:16:50.289 You cannot possibly master SEO and copywriting 00:16:50.289 --> 00:16:52.769 and data science and strategy. It's impossible. 00:16:53.070 --> 00:16:55.669 You'll die trying. Right. The ship is knowing 00:16:55.669 --> 00:16:57.850 which expert to call off the bench. It's about 00:16:57.850 --> 00:17:00.049 being the conductor of the orchestra, not trying 00:17:00.049 --> 00:17:02.149 to play every instrument at once. You don't need 00:17:02.149 --> 00:17:04.750 to hire 10 people. You just need to install 10 00:17:04.750 --> 00:17:07.789 skills. That's a powerful reframing instead of 00:17:07.960 --> 00:17:10.859 I have to do this, it's I need to direct this. 00:17:11.319 --> 00:17:13.279 Exactly. So for the listener who is feeling a 00:17:13.279 --> 00:17:15.279 mix of excitement and maybe a little overwhelmed 00:17:15.279 --> 00:17:17.680 because we just threw 10 personas and a lot of 00:17:17.680 --> 00:17:20.400 technical terms at them, what is the move? What 00:17:20.400 --> 00:17:23.700 do they do tonight? The move is to do less. Do 00:17:23.700 --> 00:17:26.539 not try to overhaul your entire business tonight. 00:17:27.039 --> 00:17:30.279 Do not try to set up programmatic SEO and GA4 00:17:30.279 --> 00:17:33.319 tracking in one sitting. You will crash. Just 00:17:33.319 --> 00:17:36.480 pick one. Pick one. What is the one thing you've 00:17:36.480 --> 00:17:39.019 been avoiding? Is it that landing page that isn't 00:17:39.019 --> 00:17:41.019 converting? Great. Just use the page conversion 00:17:41.019 --> 00:17:43.559 optimizer. Give it 20 minutes. That's it. I love 00:17:43.559 --> 00:17:45.920 that. Start with the friction. Fix one thing. 00:17:46.059 --> 00:17:48.599 And once you see the result, when you see that 00:17:48.599 --> 00:17:51.339 headline transform or that email sequence right 00:17:51.339 --> 00:17:53.720 itself, you'll realize you aren't alone at that 00:17:53.720 --> 00:17:56.119 desk anymore. You have a team. The future belongs 00:17:56.119 --> 00:17:58.339 to those who build the best team, whether those 00:17:58.339 --> 00:18:00.900 team members are biological or silicon. Couldn't 00:18:00.900 --> 00:18:02.740 have said it better myself. That's it for this 00:18:02.740 --> 00:18:05.019 deep dive. Go meet your new employees. We'll 00:18:05.019 --> 00:18:05.599 see you next time.