WEBVTT 00:00:00.000 --> 00:00:02.620 Imagine you have a massive, complicated project 00:00:02.620 --> 00:00:05.740 ahead of you. Instead of opening 50 different 00:00:05.740 --> 00:00:08.779 research tabs yourself, manually cross -referencing 00:00:08.779 --> 00:00:11.839 everything, you just press a single button. And 00:00:11.839 --> 00:00:16.100 instantly, 50 digital workers just appear. They 00:00:16.100 --> 00:00:18.379 analyze all the necessary data at the same time 00:00:18.379 --> 00:00:20.600 connect all the dots and then they hand you a 00:00:20.600 --> 00:00:23.019 finished organized report in five minutes flat 00:00:23.019 --> 00:00:25.100 Yeah, that's the difference we're talking about 00:00:25.100 --> 00:00:27.339 today We're really moving from just chatting 00:00:27.339 --> 00:00:29.940 with an AI to actually managing an autonomous 00:00:29.940 --> 00:00:32.780 digital workforce That's the core shift and it's 00:00:32.780 --> 00:00:35.079 all about business leverage a massive amount 00:00:35.079 --> 00:00:38.450 of it. Welcome to the deep dive Our mission today 00:00:38.450 --> 00:00:41.329 is to give you a clear roadmap for getting that 00:00:41.329 --> 00:00:44.009 leverage. We're going to unpack the technical 00:00:44.009 --> 00:00:46.549 side of the wide research feature that makes 00:00:46.549 --> 00:00:48.909 this possible. Look at some real world examples, 00:00:48.969 --> 00:00:50.969 everything from market research to finding your 00:00:50.969 --> 00:00:53.909 next client, and define this digital general 00:00:53.909 --> 00:00:56.090 contractor model that makes it all profitable. 00:00:56.170 --> 00:00:58.229 And I think the initial context is crucial here. 00:00:58.310 --> 00:01:01.009 Your standard AI, like a typical chat bot, it's 00:01:01.009 --> 00:01:03.009 linear. It's just a step by step conversation. 00:01:03.270 --> 00:01:05.489 One thing at a time. Exactly. These autonomous 00:01:05.489 --> 00:01:08.620 agents we're focusing on, They're parallel. They're 00:01:08.620 --> 00:01:11.040 built for action. So we're really focused on 00:01:11.040 --> 00:01:14.459 how this technology scales your time, not just 00:01:14.459 --> 00:01:16.780 how fast you can write an email. Precisely. OK. 00:01:16.819 --> 00:01:19.000 Let's unpack this a little. When we talk about 00:01:19.000 --> 00:01:21.620 an autonomous agent versus, say, a standard tool 00:01:21.620 --> 00:01:25.280 like CHAT GPT or Claude, what is the fundamental 00:01:25.280 --> 00:01:27.659 distinction there? They're both running on large 00:01:27.659 --> 00:01:30.480 language models, right? They are. The defining 00:01:30.480 --> 00:01:32.180 difference isn't really the intelligence. It's 00:01:32.180 --> 00:01:35.000 the way they operate. The mechanism. Standard 00:01:35.000 --> 00:01:38.260 LLMs work linearly. You ask it to summarize one 00:01:38.260 --> 00:01:40.760 website, you wait, then you ask it to summarize 00:01:40.760 --> 00:01:42.379 a second one. You're bidding for each step to 00:01:42.379 --> 00:01:44.920 complete. Exactly. If you need 100 data points, 00:01:45.200 --> 00:01:48.099 you're basically waiting 100 times. But an autonomous 00:01:48.099 --> 00:01:50.120 agent, especially the ones using these new tool 00:01:50.120 --> 00:01:52.640 use updates, like in the Manus AI system we looked 00:01:52.640 --> 00:01:56.620 at, it's parallel. It doesn't just talk, it acts. 00:01:57.260 --> 00:01:59.700 It's an LLM that's been given permission to use 00:01:59.700 --> 00:02:02.920 the browser, to click links, to save files. So 00:02:02.920 --> 00:02:05.340 the LM itself now has a toolkit, like a virtual 00:02:05.340 --> 00:02:07.420 mouse and keyboard. That's a perfect way to put 00:02:07.420 --> 00:02:10.280 it. And this wide research feature, it lets the 00:02:10.280 --> 00:02:13.620 AI delegate. So instead of running 50 tasks one 00:02:13.620 --> 00:02:16.310 after another, it just... executes them all at 00:02:16.310 --> 00:02:18.889 the same time. Like opening up 50 lanes of traffic 00:02:18.889 --> 00:02:21.669 on a highway instead of just one. Exactly. That 00:02:21.669 --> 00:02:24.289 analogy of moving from a conversational assistant 00:02:24.289 --> 00:02:27.430 to a managerial role really is the key. We're 00:02:27.430 --> 00:02:30.210 moving beyond just generating text and into managing 00:02:30.210 --> 00:02:32.669 action. So if you had to boil it down, what's 00:02:32.669 --> 00:02:35.110 the number one reason to choose an agent over 00:02:35.110 --> 00:02:38.389 a chatbot? To synthesize huge amounts of complex 00:02:38.389 --> 00:02:41.389 data instantly. OK, so it's about making sense 00:02:41.389 --> 00:02:43.430 of complexity, not just fetching facts. I get 00:02:43.430 --> 00:02:45.449 it. Yeah. And this is where it gets really interesting 00:02:45.449 --> 00:02:48.270 for me. How does that parallelism and that speed 00:02:48.270 --> 00:02:50.810 translate into reports that are actually useful 00:02:50.810 --> 00:02:53.349 and save you time? Well, the agent goes beyond 00:02:53.349 --> 00:02:56.129 just simple retrieval. It synthesizes the information 00:02:56.129 --> 00:02:58.610 into what are basically decision ready reports. 00:02:59.310 --> 00:03:01.229 It's the difference between being handed 20 articles 00:03:01.229 --> 00:03:04.009 to read and being handed a single Excel sheet 00:03:04.009 --> 00:03:06.729 that compares all 20 side by side. based on the 00:03:06.729 --> 00:03:08.770 criteria you set. It saves you the mental energy 00:03:08.770 --> 00:03:11.090 of comparison. It saves you all of that mental 00:03:11.090 --> 00:03:13.590 load. Yeah. Let's walk through that laptop case 00:03:13.590 --> 00:03:15.509 study from the sources. I think it illustrates 00:03:15.509 --> 00:03:19.389 the complexity so well, because trying to analyze 00:03:19.389 --> 00:03:23.909 20 different laptops by hand is a nightmare of 00:03:23.909 --> 00:03:26.830 switching between tabs. It is. And the prompt 00:03:26.830 --> 00:03:29.229 didn't just ask for the basic stuff like price, 00:03:29.469 --> 00:03:31.789 RAM, or processor speed. And this was the critical 00:03:31.789 --> 00:03:34.479 part. The prompt specifically demanded the agent 00:03:34.479 --> 00:03:37.620 find one negative review about overheating on 00:03:37.620 --> 00:03:40.460 a form like Reddit for each of those 20 laptops. 00:03:40.500 --> 00:03:43.580 Whoa. So that's taking unstructured data from 00:03:43.580 --> 00:03:46.439 multiple different websites 20 times over and 00:03:46.439 --> 00:03:48.680 doing it all simultaneously. And it's that filtering, 00:03:48.740 --> 00:03:51.280 that action -oriented part that makes it so valuable. 00:03:51.340 --> 00:03:53.439 It ends up creating this full comparison spreadsheet 00:03:53.439 --> 00:03:55.500 in about three minutes. You don't just get specs. 00:03:55.659 --> 00:03:58.379 You get filtered, actionable intelligence that 00:03:58.379 --> 00:04:00.340 a person would take, what, three, four hours 00:04:00.340 --> 00:04:03.199 to compile. At least. And the second case study 00:04:03.199 --> 00:04:06.039 on marketing shows the same power. Instead of 00:04:06.039 --> 00:04:08.219 guessing about social media trends, the agent 00:04:08.219 --> 00:04:11.259 was told to research the top 50 coffee influencers 00:04:11.259 --> 00:04:14.159 on Instagram. And it looked at their most liked 00:04:14.159 --> 00:04:16.660 photos from the last month. Yeah, and it categorized 00:04:16.660 --> 00:04:19.060 the style of each photo. You know, was it latte 00:04:19.060 --> 00:04:22.399 art, cozy reading, coffee shop vibe? And then 00:04:22.399 --> 00:04:25.290 based on that real... proven engagement data, 00:04:25.610 --> 00:04:29.009 it generated 20 specific image ideas for a brand. 00:04:29.230 --> 00:04:31.769 It's completely bypassing the human research 00:04:31.769 --> 00:04:33.990 bottleneck. Right. It takes the work of a whole 00:04:33.990 --> 00:04:36.230 social media research team and just collapses 00:04:36.230 --> 00:04:38.449 it into a few minutes. That's real leverage. 00:04:38.949 --> 00:04:41.870 Wow. I mean, imagine scaling that. Not just to 00:04:41.870 --> 00:04:45.040 50 influencers, but to a Billion queries across 00:04:45.040 --> 00:04:47.579 really complex data sets all looking for these 00:04:47.579 --> 00:04:50.779 subtle market signals. The speed is it's mind 00:04:50.779 --> 00:04:53.079 -boggling It really is, you know I have to admit 00:04:53.079 --> 00:04:55.220 I still wrestle with prompt drift myself when 00:04:55.220 --> 00:04:57.699 I ask a standard LLM for huge amounts of data 00:04:57.699 --> 00:04:59.339 the instructions kind of get muddy by the end 00:04:59.339 --> 00:05:01.660 but it sounds like these agents handle that complexity 00:05:01.660 --> 00:05:04.360 by just Delegating the subtasks out from the 00:05:04.360 --> 00:05:06.579 start. They do they handle the organization before 00:05:06.579 --> 00:05:08.639 the results even come back to you So beyond just 00:05:08.639 --> 00:05:12.430 fetching information What unique value do these 00:05:12.430 --> 00:05:14.790 agents really provide? They synthesize disparate 00:05:14.790 --> 00:05:17.509 data into decision -ready reports. So they save 00:05:17.509 --> 00:05:19.449 you the time of actually having to think and 00:05:19.449 --> 00:05:22.589 compare. OK, that makes sense. So we understand 00:05:22.589 --> 00:05:24.970 how the agent can create these valuable reports. 00:05:26.310 --> 00:05:28.889 Let's pivot to the money -making side. How do 00:05:28.889 --> 00:05:31.870 we actually use this speed to find clients and 00:05:31.870 --> 00:05:35.569 generate revenue? OK, so the key here is to shift 00:05:35.569 --> 00:05:38.670 your focus. You want to find people who are already 00:05:38.670 --> 00:05:41.199 asking for help. We call this the low hanging 00:05:41.199 --> 00:05:44.339 fruit strategy. It avoids all the pain of cold 00:05:44.339 --> 00:05:47.120 outreach. And we're targeting business owners 00:05:47.120 --> 00:05:49.839 in those high value local service industries, 00:05:50.079 --> 00:05:52.939 right? Like home renovation, plumbing, roofing, 00:05:53.500 --> 00:05:55.720 people who are active in public Facebook groups. 00:05:56.120 --> 00:05:58.000 Exactly. They're literally raising their hands 00:05:58.000 --> 00:06:00.620 and telling you their problems. The specificity 00:06:00.620 --> 00:06:02.560 of the prompt is what makes this work so well. 00:06:02.670 --> 00:06:05.470 We tell the agent to target public posts in these 00:06:05.470 --> 00:06:08.509 groups from the last 14 days. And that 14 -day 00:06:08.509 --> 00:06:11.310 filter is key. It's critical. It ensures the 00:06:11.310 --> 00:06:13.709 pain is immediate and fresh. And we're looking 00:06:13.709 --> 00:06:15.810 for specific stresses, like getting more leads, 00:06:15.970 --> 00:06:18.870 fixing my website, or help with Google Ads. So 00:06:18.870 --> 00:06:21.509 the output you get is this clean list of fresh 00:06:21.509 --> 00:06:23.850 leads. It includes their name, their business, 00:06:23.910 --> 00:06:26.689 and their exact problem. Something like John 00:06:26.689 --> 00:06:29.209 Smith. plumber asking how to get his website 00:06:29.209 --> 00:06:32.329 to show up on Google. You know their pain point 00:06:32.329 --> 00:06:34.610 before you even say hello. You know their exact 00:06:34.610 --> 00:06:36.449 pain point. You're not guessing. Why is that 00:06:36.449 --> 00:06:39.310 14 day filter so important for the lead quality? 00:06:39.490 --> 00:06:41.990 It ensures the leads are fresh and their pain 00:06:41.990 --> 00:06:45.129 is immediate. Got it. Fresh problems that need 00:06:45.129 --> 00:06:47.829 solving right now. OK, so we found the lead. 00:06:47.930 --> 00:06:50.550 We know their problem. Now how do we convert 00:06:50.550 --> 00:06:52.889 them without sounding like every other salesperson? 00:06:53.149 --> 00:06:55.709 This is where we shift from selling to helping. 00:06:56.209 --> 00:06:58.569 We use what's called the competitor spy method. 00:06:58.990 --> 00:07:01.350 The whole goal is to use the agent to provide 00:07:01.350 --> 00:07:04.610 instant, undeniable value for free. So let's 00:07:04.610 --> 00:07:06.870 stick with our example, Sarah the Roofer. The 00:07:06.870 --> 00:07:09.069 agent would research her biggest local competitor, 00:07:09.310 --> 00:07:12.129 say, top -tier roofing. Right. And the research 00:07:12.129 --> 00:07:14.649 prompt has to be super specific. We'd ask the 00:07:14.649 --> 00:07:17.069 agent to analyze the competitor's website, figure 00:07:17.069 --> 00:07:19.370 out what keywords they rank for on Google, look 00:07:19.370 --> 00:07:21.949 at their Facebook ads, and then compare all of 00:07:21.949 --> 00:07:24.170 that to Sarah's business to find three specific 00:07:24.170 --> 00:07:26.529 things she's missing. And the results that come 00:07:26.529 --> 00:07:29.410 back are things she can actually use, like the 00:07:29.410 --> 00:07:32.189 competitor is ranking for emergency roof repair 00:07:32.189 --> 00:07:35.430 and Sarah is not. Or the competitor is using 00:07:35.430 --> 00:07:38.629 video testimonials and Sarah has no video content 00:07:38.629 --> 00:07:41.509 at all. Or maybe they have a big get a quote 00:07:41.509 --> 00:07:44.620 button on every page and Sarah's is buried. These 00:07:44.620 --> 00:07:47.019 are tangible gaps a business owner can understand 00:07:47.019 --> 00:07:49.779 immediately. Instantly. And the final step is 00:07:49.779 --> 00:07:51.860 simple. You send a friendly message with a quick 00:07:51.860 --> 00:07:55.360 free PDF report detailing those three specific 00:07:55.360 --> 00:07:58.180 actionable changes. This one move makes you look 00:07:58.180 --> 00:08:00.339 like an absolute genius and a genuine expert 00:08:00.339 --> 00:08:02.360 who's there to help. So what's the ultimate goal 00:08:02.360 --> 00:08:04.819 of the competitor's spy report? To provide so 00:08:04.819 --> 00:08:07.060 much value that your outreach looks like help, 00:08:07.279 --> 00:08:09.500 not a sales pitch. OK, it's about leading with 00:08:09.500 --> 00:08:12.300 value to build trust. I like that. That's it. 00:08:12.560 --> 00:08:14.860 OK, so what does this all mean for actually doing 00:08:14.860 --> 00:08:16.839 the work? Because I can hear a listener thinking, 00:08:17.000 --> 00:08:19.600 that's great, but I don't know how to fix a website 00:08:19.600 --> 00:08:22.420 or run Google Ads. And that is the pivot point. 00:08:22.519 --> 00:08:25.240 That's the key insight that makes this whole 00:08:25.240 --> 00:08:27.980 thing scalable. You don't have to be the worker. 00:08:28.199 --> 00:08:31.079 you become the general contractor. A general 00:08:31.079 --> 00:08:33.460 contractor manages the project. They hire the 00:08:33.460 --> 00:08:35.740 plumber, the electrician, the painter. They don't 00:08:35.740 --> 00:08:38.620 do the work themselves. Your job is to manage 00:08:38.620 --> 00:08:41.720 the communication and connect the client to the 00:08:41.720 --> 00:08:44.299 right solution. And there are two main ways to 00:08:44.299 --> 00:08:46.799 do that once you land the client, right? Option 00:08:46.799 --> 00:08:49.309 A is the easy way, software. using platforms 00:08:49.309 --> 00:08:51.389 like Go High Level, which have all these pre 00:08:51.389 --> 00:08:53.570 -made templates for industries like plumbers 00:08:53.570 --> 00:08:56.149 and roofers. Exactly. You can charge the client, 00:08:56.350 --> 00:08:59.450 say, $500 for a professional website setup that 00:08:59.450 --> 00:09:01.950 takes you maybe an hour using one of those templates. 00:09:02.470 --> 00:09:04.909 You're providing the solution, but the software 00:09:04.909 --> 00:09:07.830 is doing the heavy lifting. And option B is Arbitrage, 00:09:08.330 --> 00:09:11.389 the outsourcing model. Yeah. So for something 00:09:11.389 --> 00:09:14.330 more complex, like a Google Ads campaign, you 00:09:14.330 --> 00:09:17.490 might charge the client $1 ,000. Then you go 00:09:17.490 --> 00:09:20.590 on a site like Upwork and hire an expert freelancer 00:09:20.590 --> 00:09:24.409 for $300 to do the work. You keep the $700 profit. 00:09:24.549 --> 00:09:26.990 The agent found the lead, the freelancer does 00:09:26.990 --> 00:09:29.450 the work, and you just manage the process. You 00:09:29.450 --> 00:09:33.409 manage the flow and you capture the margin. So 00:09:33.409 --> 00:09:36.110 stripping it all down, what's the core value 00:09:36.110 --> 00:09:38.649 proposition of being the general contractor? 00:09:39.129 --> 00:09:42.279 Connecting the client to the solution. not executing 00:09:42.279 --> 00:09:44.620 the technical labor yourself. It's about project 00:09:44.620 --> 00:09:46.860 management and quality control, not being the 00:09:46.860 --> 00:09:50.080 technician. Oh, okay. I see. Now, getting paid 00:09:50.080 --> 00:09:52.440 once is great, but to build a real business, 00:09:52.720 --> 00:09:54.990 you need recurring revenue. We have to talk about 00:09:54.990 --> 00:09:57.389 the service after the sale model. Absolutely. 00:09:57.610 --> 00:10:00.750 After that initial setup, that $500 website fix, 00:10:00.889 --> 00:10:03.610 you offer monthly retainers. And these are services 00:10:03.610 --> 00:10:05.590 that mostly run on low maintenance automation. 00:10:05.750 --> 00:10:08.850 Things like website hosting for, say, $50 a month, 00:10:09.509 --> 00:10:11.669 or reputation management to get them more five 00:10:11.669 --> 00:10:14.529 -star reviews for $100 a month, or maybe content 00:10:14.529 --> 00:10:17.549 posting for their social media, $200 a month. 00:10:17.990 --> 00:10:20.850 The math really shows the power here. Just... 00:10:20.720 --> 00:10:24.039 10 clients paying you an average of $300 a month. 00:10:24.639 --> 00:10:27.100 That's $3 ,000 in monthly recurring revenue. 00:10:27.210 --> 00:10:30.149 And because the AI tools can handle so much of 00:10:30.149 --> 00:10:32.549 the automation, the actual time you spend on 00:10:32.549 --> 00:10:34.750 it is just a few hours a week once you're set 00:10:34.750 --> 00:10:37.190 up. The time commitment is incredibly low for 00:10:37.190 --> 00:10:39.690 the return. We do have to address the cost, though, 00:10:39.769 --> 00:10:41.769 for someone just starting out. If an advanced 00:10:41.769 --> 00:10:44.610 tool that costs around $30 a month is too much 00:10:44.610 --> 00:10:46.870 at the beginning, you can use free alternatives. 00:10:47.110 --> 00:10:50.289 Oh, absolutely. Tools like Google Gemini or the 00:10:50.289 --> 00:10:53.409 free version of Chat GPT are plenty powerful 00:10:53.409 --> 00:10:55.330 to get started. The main trade -off is just going 00:10:55.330 --> 00:10:57.970 to be efficiency. It's slower. It's much slower. 00:10:58.250 --> 00:11:00.269 You're looking at maybe an hour or two for a 00:11:00.269 --> 00:11:03.090 report that an agent does in five minutes. And 00:11:03.090 --> 00:11:05.769 you have to do a lot of manual copy pasting instead 00:11:05.769 --> 00:11:08.409 of getting a clean Excel file. But you can still 00:11:08.409 --> 00:11:10.529 get started. You could use a really targeted 00:11:10.529 --> 00:11:13.809 prompt in Gemini like, list five roofing businesses 00:11:13.809 --> 00:11:16.309 in my city that don't have a website link on 00:11:16.309 --> 00:11:19.029 their Google Maps profile. That gives you a starting 00:11:19.029 --> 00:11:21.029 list right there. It's a totally viable path. 00:11:21.110 --> 00:11:23.669 So what's the recommended starter path for the 00:11:23.669 --> 00:11:26.429 broke entrepreneur? Use the free tools to land 00:11:26.429 --> 00:11:28.710 that first client, then use that money to buy 00:11:28.710 --> 00:11:31.570 the advanced paid tools. Reinvest the first win 00:11:31.570 --> 00:11:35.029 to unlock speed and leverage. Simple. Got it. 00:11:35.370 --> 00:11:40.440 So. What does this all really mean? I think the 00:11:40.440 --> 00:11:43.039 biggest insight today is just the incredible 00:11:43.039 --> 00:11:45.139 leverage this technology gives you. For sure. 00:11:45.379 --> 00:11:47.980 Two years ago, this exact business model would 00:11:47.980 --> 00:11:50.159 have required a whole team. You'd need a researcher, 00:11:50.360 --> 00:11:52.539 a writer, a designer, a secretary. And today, 00:11:52.720 --> 00:11:54.440 the Autonomous Agile lets you become that one 00:11:54.440 --> 00:11:58.080 person army. And there is a massive window of 00:11:58.080 --> 00:12:00.139 opportunity right now because most people are 00:12:00.139 --> 00:12:03.960 still just using linear chatbots. for basic tasks. 00:12:04.159 --> 00:12:06.120 They're still just chatting. They are. So using 00:12:06.120 --> 00:12:08.799 agents for this kind of deep parallel research 00:12:08.799 --> 00:12:11.419 and client finding, it gives you a huge advantage 00:12:11.419 --> 00:12:13.779 in any local market. So we've given you the plan. 00:12:13.960 --> 00:12:16.980 Find the low -hanging fruit, spy on their competitors, 00:12:17.679 --> 00:12:20.259 send that free report to build trust, and then 00:12:20.259 --> 00:12:22.279 be the general contractor who connects them to 00:12:22.279 --> 00:12:25.539 the solution. That's the playbook. And as a final 00:12:25.539 --> 00:12:27.500 thought to leave you with, consider this. If 00:12:27.500 --> 00:12:29.919 we can manage a workforce of 50 digital interns 00:12:29.919 --> 00:12:33.490 with just one prompt, How quickly will the very 00:12:33.490 --> 00:12:36.190 definition of work itself be redefined for both 00:12:36.190 --> 00:12:39.190 the client and for us the contractor? What human 00:12:39.190 --> 00:12:41.570 skills will truly matter in a world where so 00:12:41.570 --> 00:12:44.049 much of the execution is just outsourced entirely 00:12:44.049 --> 00:12:45.850 to these parallel processes?