WEBVTT 00:00:00.000 --> 00:00:02.419 So I think we have to address the elephant in 00:00:02.419 --> 00:00:04.120 the room. I mean, it feels like everybody is 00:00:04.120 --> 00:00:07.440 whispering about it right now. Is AI the next 00:00:07.440 --> 00:00:10.199 dot com crash? Are we just watching this massive 00:00:10.199 --> 00:00:12.419 market bubble inflate knowing it's going to pop? 00:00:12.939 --> 00:00:16.280 Or is the foundation actually? You know solid 00:00:16.280 --> 00:00:18.260 this time. That is the essential question, isn't 00:00:18.260 --> 00:00:20.000 it? The headlines are just all over the place 00:00:20.000 --> 00:00:22.579 one day It's total fear the next it's a trillion 00:00:22.579 --> 00:00:25.260 dollar boom. So our mission today is really to 00:00:25.260 --> 00:00:26.879 cut through that noise we've been digging into 00:00:26.879 --> 00:00:30.239 the actual spending data and some Some really 00:00:30.239 --> 00:00:32.479 fascinating reports from places like MIT and 00:00:32.479 --> 00:00:34.100 Wharton to get to the truth, right? We're looking 00:00:34.100 --> 00:00:36.659 for the real story here. Yeah, because the next 00:00:36.659 --> 00:00:40.259 12 to 24 months That period is going to be the 00:00:40.259 --> 00:00:42.679 critical test for this whole market and for you 00:00:42.679 --> 00:00:45.560 listening The core insight is figuring out where 00:00:45.560 --> 00:00:48.340 the real sustainable value is especially if you're 00:00:48.340 --> 00:00:50.579 running a small business or an agency Exactly. 00:00:50.979 --> 00:00:53.000 So we've structured this deep dive as a kind 00:00:53.000 --> 00:00:55.600 of survival playbook We're gonna start by defining 00:00:55.600 --> 00:00:57.520 this idea of circular spending which sounds a 00:00:57.520 --> 00:01:01.219 bit scary then we'll compare today to that dark 00:01:01.219 --> 00:01:04.640 fiber crash back in 2000, and then deliver a 00:01:04.640 --> 00:01:07.159 really practical five -point plan for how you 00:01:07.159 --> 00:01:10.159 can secure your success, no matter what the stock 00:01:10.159 --> 00:01:12.060 market does. OK, so let's start with the scale 00:01:12.060 --> 00:01:14.219 of this investment, because the number is just 00:01:14.219 --> 00:01:16.400 huge. We're talking about big tech, Microsoft, 00:01:16.739 --> 00:01:20.500 Google, Metta spending around $400 billion every 00:01:20.500 --> 00:01:23.939 single year on AI infrastructure, so data centers, 00:01:24.159 --> 00:01:26.340 chips, all of it. And that leads us right to 00:01:26.340 --> 00:01:28.620 the paradox, which is where all this bubble anxiety 00:01:28.620 --> 00:01:30.579 comes from. These giants are spending way more 00:01:30.579 --> 00:01:32.099 money building all this stuff than they're making 00:01:32.099 --> 00:01:34.560 back right now and, you know, clear customer 00:01:34.560 --> 00:01:37.640 revenue. The money flow looks immense, but how 00:01:37.640 --> 00:01:39.939 much is actually new money? Which brings us to 00:01:39.939 --> 00:01:41.859 this term that's making people really nervous. 00:01:42.519 --> 00:01:44.799 Circular spending. It's kind of an accounting 00:01:44.799 --> 00:01:47.620 trick, but it shows a fragile foundation. The 00:01:47.620 --> 00:01:49.560 old lemonade stand analogy doesn't really cut 00:01:49.560 --> 00:01:52.140 it. No, it doesn't. What's really happening is 00:01:52.140 --> 00:01:54.640 that the big cloud providers are investing in 00:01:54.640 --> 00:01:56.459 startups. They're giving them huge amounts of 00:01:56.459 --> 00:01:59.340 capital. But, and this is the key part, those 00:01:59.340 --> 00:02:01.659 startups are then contractually obligated to 00:02:01.659 --> 00:02:04.780 spend that exact same money buying cloud services 00:02:04.780 --> 00:02:07.519 back from the company that just invested in them. 00:02:08.300 --> 00:02:11.759 Ah, so company A gets to record massive revenue 00:02:11.759 --> 00:02:14.520 from the cloud sales to company B and their stock 00:02:14.520 --> 00:02:16.759 price goes up. Right. But it's just internal 00:02:16.759 --> 00:02:19.180 financial movement. It's not really new money 00:02:19.180 --> 00:02:21.710 coming in from Main Street. Exactly. The growth 00:02:21.710 --> 00:02:25.330 is fueled by these self -generated deals. And 00:02:25.330 --> 00:02:27.889 that concentration is incredibly dangerous. If 00:02:27.889 --> 00:02:30.750 you look at the S &P 500, something like 75 % 00:02:30.750 --> 00:02:33.569 of the recent gains are driven by just the magnificent 00:02:33.569 --> 00:02:36.490 seven tech stocks. So if that financial circle 00:02:36.490 --> 00:02:38.689 breaks, like if a huge startup suddenly can't 00:02:38.689 --> 00:02:41.330 pay its giant cloud bill, the drop would be fast 00:02:41.330 --> 00:02:43.699 and it would be widespread. But if this circular 00:02:43.699 --> 00:02:46.300 spending is so risky, does that mean the entire 00:02:46.300 --> 00:02:48.560 utility of AI is just some kind of illusion? 00:02:49.020 --> 00:02:51.360 No, not at all. The utility is very real. The 00:02:51.360 --> 00:02:53.500 financial model is just highly, highly concentrated 00:02:53.500 --> 00:02:55.800 right now. OK, so the financial engine is risky. 00:02:56.139 --> 00:02:58.319 But what about the technology itself? Have we 00:02:58.319 --> 00:03:01.500 seen this kind of massive infrastructure overspending 00:03:01.500 --> 00:03:03.500 before? Oh, absolutely. We have to go back to 00:03:03.500 --> 00:03:05.939 the last big tech bubble. The dot -com crash 00:03:05.939 --> 00:03:08.539 in the early 2000s. That whole era was defined 00:03:08.539 --> 00:03:11.379 by dark fiber. Right. These massive networks 00:03:11.379 --> 00:03:13.599 of fiber optic cables that were laid under the 00:03:13.599 --> 00:03:16.659 ground and across oceans. The idea was that everyone 00:03:16.659 --> 00:03:19.259 would need this incredible bandwidth instantly. 00:03:19.419 --> 00:03:21.539 But they were too early. The users weren't there. 00:03:21.599 --> 00:03:24.080 The apps didn't exist. And those expensive cables 00:03:24.080 --> 00:03:27.270 just sat there. Dark. And that premature bet 00:03:27.270 --> 00:03:30.270 caused a spectacular crash. What's so dramatically 00:03:30.270 --> 00:03:32.750 different today is the speed of adoption. The 00:03:32.750 --> 00:03:34.389 infrastructure being built now is absolutely 00:03:34.389 --> 00:03:38.069 not dark. I mean, whoa. Just imagine scaling 00:03:38.069 --> 00:03:41.090 to hundreds of millions of people using ChatGPT 00:03:41.090 --> 00:03:43.370 every single week. Every week. Or millions of 00:03:43.370 --> 00:03:45.530 developers using tools like Claude every day 00:03:45.530 --> 00:03:48.469 to write and test software faster. The cables 00:03:48.469 --> 00:03:51.349 are lit up. The demand is not speculative. It 00:03:51.349 --> 00:03:54.139 is real. Yeah, the consumer demand is undeniable. 00:03:54.340 --> 00:03:57.560 That's bucket one. But bucket two business applications, 00:03:58.060 --> 00:04:00.780 B2B, that's where it gets much more complicated. 00:04:00.879 --> 00:04:02.759 And that's where the nervousness really lies. 00:04:03.000 --> 00:04:06.159 So if the demand is real, why are some investors 00:04:06.159 --> 00:04:08.460 still so nervous about the business application 00:04:08.460 --> 00:04:11.900 side? Because enterprise adoption is just. It's 00:04:11.900 --> 00:04:14.580 complicated. And it's leading to these wildly 00:04:14.580 --> 00:04:17.029 conflicting success reports. Which brings us 00:04:17.029 --> 00:04:20.170 to this tale of two studies. We have these massive 00:04:20.170 --> 00:04:23.110 contradictory reports from two of the most respected 00:04:23.110 --> 00:04:26.149 institutions out there, MIT and Wharton. OK, 00:04:26.189 --> 00:04:28.250 let's start with the sobering one, the MIT report. 00:04:28.509 --> 00:04:31.589 It published the staggering claim that 95 % of 00:04:31.589 --> 00:04:34.009 AI projects inside big companies enterprises, 00:04:34.050 --> 00:04:36.839 they fail. They either get canceled or just don't 00:04:36.839 --> 00:04:39.699 deliver any value. 95%. I mean, that is just 00:04:39.699 --> 00:04:41.639 staggering. How can that number possibly exist 00:04:41.639 --> 00:04:43.720 at the same time companies are reporting massive 00:04:43.720 --> 00:04:46.980 ROI? Well, when you dig into the why, it's usually 00:04:46.980 --> 00:04:50.060 structural. Big companies are like these giant, 00:04:50.240 --> 00:04:52.579 slow cruise ships. They're running on old systems. 00:04:52.680 --> 00:04:54.740 They have layers and layers of management. And 00:04:54.740 --> 00:04:57.139 when they try to build some custom AI solution, 00:04:57.579 --> 00:04:59.990 they just get stuck. They move too slow. OK, 00:05:00.029 --> 00:05:02.410 but then you've got the Wharton Report. And they're 00:05:02.410 --> 00:05:06.990 saying that 75 % of companies using AI are seeing 00:05:06.990 --> 00:05:10.769 a clear positive return on investment, a positive 00:05:10.769 --> 00:05:13.730 ROI. It's so hard to square those two numbers. 00:05:13.870 --> 00:05:16.500 Here's the secret. It's in the methodology. MIT 00:05:16.500 --> 00:05:18.899 was looking at companies attempting custom development, 00:05:19.180 --> 00:05:21.639 trying to build their own AI -bron from scratch. 00:05:21.759 --> 00:05:24.899 That's the 95 % failure zone. Ah, OK. And Wharton? 00:05:25.000 --> 00:05:27.360 Wharton was looking at companies using off -the 00:05:27.360 --> 00:05:29.279 -shelf tools, just teaching their employees how 00:05:29.279 --> 00:05:31.740 to use Chat GPT better, or integrating simple 00:05:31.740 --> 00:05:34.000 existing AI services. So it's a huge difference 00:05:34.000 --> 00:05:36.319 between trying to build an AI versus just trying 00:05:36.319 --> 00:05:39.480 to use an AI. Exactly. And the most vital takeaway 00:05:39.480 --> 00:05:41.660 here for anyone selling AI services is what they 00:05:41.660 --> 00:05:44.560 call the speedboat advantage. Wharton found that 00:05:44.560 --> 00:05:47.019 smaller companies, the ones between $50 million 00:05:47.019 --> 00:05:51.180 and $250 million in revenue, they are winning. 00:05:51.600 --> 00:05:55.240 Big time. They have a 79 % success rate. They're 00:05:55.240 --> 00:05:57.550 speedboats. They can turn fast. They don't have 00:05:57.550 --> 00:05:59.629 all those management layers and old systems holding 00:05:59.629 --> 00:06:02.269 them back like the giant battleships. So smaller 00:06:02.269 --> 00:06:04.290 businesses are actually three times more likely 00:06:04.290 --> 00:06:06.829 to succeed. So how does this disparity in success 00:06:06.829 --> 00:06:09.949 rates immediately change how an independent AI 00:06:09.949 --> 00:06:12.610 service provider should operate? It's simple. 00:06:13.050 --> 00:06:15.610 Stop chasing the giants. The mid -market is where 00:06:15.610 --> 00:06:17.670 the money and the success are found. The data 00:06:17.670 --> 00:06:19.610 is pretty clear on this. The best strategy for 00:06:19.610 --> 00:06:21.769 an agency is not to go after the Coca -Colas 00:06:21.769 --> 00:06:23.910 and the Nikes of the world. They're the ones 00:06:23.910 --> 00:06:26.279 failing. Right. And the good news from the MIT 00:06:26.279 --> 00:06:28.560 report is that when those big companies do work 00:06:28.560 --> 00:06:31.680 with an outside vendor, an agency, their success 00:06:31.680 --> 00:06:34.959 rate doubles. So the agency model is the necessary 00:06:34.959 --> 00:06:37.779 bridge. Companies just can't do this alone. OK, 00:06:37.779 --> 00:06:39.620 let's jump into the playbook for survival, then, 00:06:39.620 --> 00:06:42.120 based on this data. Point one, avoid the enterprise 00:06:42.120 --> 00:06:45.079 trap. You cannot stress this enough. An enterprise 00:06:45.079 --> 00:06:47.399 sales cycle can take six to 12 months. It's just 00:06:47.399 --> 00:06:50.019 endless meetings. And because of that 95 % internal 00:06:50.019 --> 00:06:52.300 failure rate, the project often gets canceled 00:06:52.300 --> 00:06:54.839 before you even get your second check. The huge 00:06:54.839 --> 00:06:57.480 waste of time and resources. So the smarter move 00:06:57.480 --> 00:06:59.839 is to focus on the mid -market. These are companies 00:06:59.839 --> 00:07:03.540 with, say, $1 million to $50 million in revenue, 00:07:03.639 --> 00:07:06.939 maybe 10 to 200 employees. Exactly. They're big 00:07:06.939 --> 00:07:09.339 enough to pay well for real value. We're talking 00:07:09.339 --> 00:07:13.339 $5 ,000 to $20 ,000 projects to start. And crucially, 00:07:13.459 --> 00:07:15.939 they're small enough to say yes quickly. You 00:07:15.939 --> 00:07:18.319 get to bypass all that bureaucracy. It's about 00:07:18.319 --> 00:07:20.759 the velocity of your deals. OK. Playbook point 00:07:20.759 --> 00:07:23.850 two. Get obsessed with ROI. The hype days are 00:07:23.850 --> 00:07:25.870 over. You know, where you could sell AI just 00:07:25.870 --> 00:07:27.689 because it was cool and futuristic. That window 00:07:27.689 --> 00:07:29.889 has closed. This has to be maybe 18 months tops. 00:07:30.189 --> 00:07:32.230 Now clients are asking the hard question. Show 00:07:32.230 --> 00:07:35.389 me the money. If I pay you $10 ,000, how much 00:07:35.389 --> 00:07:37.250 will I make back or save in the next quarter? 00:07:37.629 --> 00:07:40.529 You have to sell money saved or money made, not 00:07:40.529 --> 00:07:42.810 just technology. And we have to teach clients 00:07:42.810 --> 00:07:45.709 how to even think about this. Like, take a simple 00:07:45.709 --> 00:07:48.329 dental clinic. You can literally calculate the 00:07:48.329 --> 00:07:51.300 annual cost of their manual work. Let's say two 00:07:51.300 --> 00:07:53.680 receptionists spend four hours a day each just 00:07:53.680 --> 00:07:56.759 answering simple appointment questions. At $20 00:07:56.759 --> 00:08:00.199 an hour, that's over $20 ,000 a year. And by 00:08:00.199 --> 00:08:02.259 showing them that specific number, that specific 00:08:02.259 --> 00:08:04.139 cost, and then guaranteeing you can automate 00:08:04.139 --> 00:08:06.740 80 % of those simple tasks, well, you're showing 00:08:06.740 --> 00:08:09.660 them $16 ,000 a year in savings. Suddenly, your 00:08:09.660 --> 00:08:12.500 $5 ,000 fee isn't a cost. It's a clear investment 00:08:12.500 --> 00:08:15.199 with a fast return. That is the language they 00:08:15.199 --> 00:08:17.199 understand. Okay, point three in the playbook. 00:08:17.420 --> 00:08:20.670 Shift from builder to optimizer. This is a common 00:08:20.670 --> 00:08:23.149 mistake. Agencies build a system, they hand it 00:08:23.149 --> 00:08:25.009 over, and they walk away. That is the crucial 00:08:25.009 --> 00:08:27.730 mistake that leads to churn. AI is not static. 00:08:27.790 --> 00:08:29.910 It needs cost and supervision because it makes 00:08:29.910 --> 00:08:32.610 mistakes. It hallucinates. And that just means 00:08:32.610 --> 00:08:35.309 the AI makes stuff up. It gives confident but 00:08:35.309 --> 00:08:38.169 wrong answers. The real long -term value is in 00:08:38.169 --> 00:08:40.429 monitoring it, fixing those bad answers, and 00:08:40.429 --> 00:08:42.570 making the system smarter over time. So you tell 00:08:42.570 --> 00:08:44.809 the client upfront, building this is just step 00:08:44.809 --> 00:08:47.549 one. My contract includes monthly optimization 00:08:47.549 --> 00:08:50.559 to make sure it keeps getting better. And honestly, 00:08:50.779 --> 00:08:53.879 I still wrestle with prompt drift myself. Keeping 00:08:53.879 --> 00:08:57.360 an AI consistent over months is hard work. So 00:08:57.360 --> 00:08:59.399 to optimize, you can run a simple analysis on 00:08:59.399 --> 00:09:02.080 the chat logs and find whether the AI was vague 00:09:02.080 --> 00:09:05.639 or wrong. Then you can suggest three specific 00:09:05.639 --> 00:09:08.159 rule changes to its knowledge base to make it 00:09:08.159 --> 00:09:11.139 more accurate next month. That's real, tangible 00:09:11.139 --> 00:09:13.539 value. So what's the easiest way to secure those 00:09:13.539 --> 00:09:15.679 long -term client relationships based on this 00:09:15.679 --> 00:09:17.879 optimization focus? You start with low stakes 00:09:17.879 --> 00:09:19.799 entry points like training, then you immediately 00:09:19.799 --> 00:09:22.019 shift the conversation to a monthly retainer 00:09:22.019 --> 00:09:24.519 for that optimization work. Which brings us to 00:09:24.519 --> 00:09:27.379 playbook point four. Become an AI transformation 00:09:27.379 --> 00:09:29.879 partner and start with training. Because sometimes 00:09:29.879 --> 00:09:32.419 a business just isn't ready for a complex build. 00:09:32.759 --> 00:09:34.779 Their data is a mess, their team is scared of 00:09:34.779 --> 00:09:38.169 the tech. And the Wharton study confirmed this. 00:09:38.590 --> 00:09:40.669 Training provides the fastest, most immediate 00:09:40.669 --> 00:09:43.909 ROI. If you teach 10 people how to use a tool 00:09:43.909 --> 00:09:46.210 correctly, how to write a good prompt, how to 00:09:46.210 --> 00:09:49.070 check the answers, you can get an immediate 20 00:09:49.070 --> 00:09:52.350 % productivity gain. That's massive value for 00:09:52.350 --> 00:09:54.789 a very low cost. So the strategy is to offer 00:09:54.789 --> 00:09:58.409 an AI workshop or an AI audit first. It's way 00:09:58.409 --> 00:10:00.330 cheaper and easier to sell than a big development 00:10:00.330 --> 00:10:02.509 project. You get your foot in the door, you build 00:10:02.509 --> 00:10:04.750 trust, you prove your value, and then you can 00:10:04.750 --> 00:10:07.409 sell the bigger services later. Exactly. You 00:10:07.409 --> 00:10:09.610 could pitch a two -hour workshop for a real estate 00:10:09.610 --> 00:10:12.269 agency on writing listing descriptions faster. 00:10:12.649 --> 00:10:14.769 You show them three wow moments live, saving 00:10:14.769 --> 00:10:17.090 them 30 minutes right there, and give them five 00:10:17.090 --> 00:10:19.429 pre -tested prompts to take home. You've just 00:10:19.429 --> 00:10:22.570 become the indispensable authority. And finally, 00:10:22.769 --> 00:10:25.490 playbook point five. Make retainers your default. 00:10:25.789 --> 00:10:28.090 This is your safety net, especially when the 00:10:28.090 --> 00:10:30.509 economy gets a little rocky. It's the fundamental 00:10:30.509 --> 00:10:33.029 difference between long -term growth and that 00:10:33.029 --> 00:10:35.429 feast or famine cycle we all hate. When things 00:10:35.429 --> 00:10:37.990 get tight, one -time projects are the first thing 00:10:37.990 --> 00:10:40.230 to get cut from a budget. But services that are 00:10:40.230 --> 00:10:42.169 keeping the business running and cutting costs, 00:10:42.470 --> 00:10:44.490 those almost never get canceled. So you don't 00:10:44.490 --> 00:10:47.100 call it maintenance. That sounds... Passive, 00:10:47.419 --> 00:10:50.620 boring. You charge for optimization and evolution. 00:10:51.539 --> 00:10:54.240 Yes. Focus the value on active, forward -looking 00:10:54.240 --> 00:10:56.980 work. You say, we do a weekly review for accuracy, 00:10:57.159 --> 00:10:59.559 we update the AI with your new pricing, and we 00:10:59.559 --> 00:11:02.519 send a monthly report on lead quality. If you 00:11:02.519 --> 00:11:05.620 charge, say, $1 ,500 a month, you frame it as 00:11:05.620 --> 00:11:08.080 a non -negotiable investment that prevents losses 00:11:08.080 --> 00:11:10.360 and keeps them ahead of the competition. OK. 00:11:10.360 --> 00:11:12.440 Let's pull all of this together. The headlines 00:11:12.440 --> 00:11:14.860 are screaming doom, but the data is telling a 00:11:14.860 --> 00:11:16.740 very different, very specific story. We have 00:11:16.740 --> 00:11:20.220 two key facts. First, the utility of AI is absolutely 00:11:20.220 --> 00:11:22.580 here to stay. The demand is real and it's growing. 00:11:23.059 --> 00:11:25.440 And second, it's the small and medium businesses, 00:11:25.679 --> 00:11:27.940 the nimble speedboats that are the big success 00:11:27.940 --> 00:11:31.059 stories. A 79 % success rate compared to that 00:11:31.059 --> 00:11:34.620 95 % failure rate for enterprises. And your role 00:11:34.620 --> 00:11:37.019 as the agency, the service provider is confirmed. 00:11:37.139 --> 00:11:39.539 You're the necessary bridge. Companies can't 00:11:39.539 --> 00:11:42.480 do this alone. The stock market bottle for the 00:11:42.480 --> 00:11:45.779 Magnificent 7 might pop because of circular spending, 00:11:46.220 --> 00:11:49.740 but the actual utility of AI in the real economy 00:11:49.740 --> 00:11:52.700 is going to remain robust. Absolutely. The future 00:11:52.700 --> 00:11:55.179 belongs to those who prepare. So we encourage 00:11:55.179 --> 00:11:57.460 you, just pick one action point from this playbook 00:11:57.460 --> 00:12:00.299 this week. Maybe practice calculating that dollar 00:12:00.299 --> 00:12:03.039 ROI for a potential client, or draft up a training 00:12:03.039 --> 00:12:05.480 workshop offer. Just take one step to become 00:12:05.480 --> 00:12:08.330 an optimizer, not just a builder. And here's 00:12:08.330 --> 00:12:10.330 a final provocative thought for you to consider. 00:12:10.789 --> 00:12:13.610 If the failure rate for custom AI projects inside 00:12:13.610 --> 00:12:18.029 massive global enterprises is 95%, what foundational 00:12:18.029 --> 00:12:20.370 business assumption does that reveal about traditional 00:12:20.370 --> 00:12:22.730 corporate structures in the face of truly rapid 00:12:22.730 --> 00:12:24.850 technological change? A question worth thinking 00:12:24.850 --> 00:12:26.730 about. Thank you for diving deep into the sources 00:12:26.730 --> 00:12:28.830 with us today. Stay curious. We'll see you next 00:12:28.830 --> 00:12:28.990 time.