Founder’s Guide to AI: Building a Marketing Engine with AI

Hey everyone, I hope you're having an awesome day. My name is Daniel. I'm a part of the Founder Institute marketing team. And today we are back with part 2 of our AI for startup founders workshop series. And today we'll be talking about creating a content marketing engine using AI. Last time around, we had over 1,000 people who signed up, a ton of really great interactions. Trevor went through the nitty-gritty of everything that he's been working on, and he'll be doing the same today. Before I let Trevor take over the stage, a couple of things to remember for you all. This is part 2 of our workshop series, so there is definitely a part 3. I encourage everyone to join the next one. And in case you were— you missed the first one, I'll share a link where you can actually watch the replay and also register for the next one. And Folks, if you have any question, please drop them in the chat and I will get to them in the end with Trevor. And I think we should be good to start. Just the last thing, if you are enjoying everything that's being said, there's an emoji button at the bottom of your screen. Please do hit it. Let us know what you think, how you're feeling. Trevor is a great friend of FI. He is a startup mentor. He's somebody who's scaled businesses. He's helped launch so many, so many startups, and right now he is fully focused on white coding and building with AI. So he is the perfect person for this. So super, super excited for this one. Trevor, the stage is yours.

Phenomenal. Thanks everybody. Let me throw this to the screen here. Boom. All right, let's talk about Building trust at scale. Now, if you were part of the previous workshop, Workshop 1, you would get that there is a big idea you have to put into how you build these days. This big idea is that the cost of effort has dropped practically to zero. Despite some of the technical challenges of the demo that I was trying to run during our workshop last time, I'm able to launch practically anything in a day, right? And frequently you can launch it in a few minutes. Indeed, I've got just recently launched, uh, and also netted a cool domain for it, greenchemistry.ai, right? Uh, something that analyzes chemical protocols, uh, and looks for how to make them more environmentally friendly. Uh, launched a little while ago. Scan.ai, which analyzes your cloud skills before you install and run them to make sure that you don't install malware by mistake, because the number one top-loaded cloud code skill last week was indeed malware that grabbed your crypto wallet and transferred money out of it. So like these sorts of things, you can build and launch almost immediately. I bought— I know I'm running this on localhost right now, but I bought greenchem.ai this morning, I think. And turn this on for a demo that my wife, who owns a nonprofit, is doing. I mean, she's demoing this to somebody today, so we need a domain for it. Let's go. Right? Like, the ideas are not the moat. That I can come up with anything and launch it happens almost instantaneously. And you don't have to be a very good coder at all. Goodness knows. I'm a terrible coder. So the way you build a moat, actually, I just alluded to it, right? Greenchemistry.ai is not something that I care about. I'm building it, but my wife is CEO of a nonprofit around the green chemistry space. She's putting out content for almost 2 years already about green chemistry and about environmental chemistry. She's part of a nonprofit whose mission is to reduce the carbon impact of all chemistry, right? That person putting a piece of product out around their area of expertise, that has trust. That people will see you care about in a way that Trevor, the guy who vibe codes and builds stuff— I built this for her to have and will continue to build and maintain and run it. Since you have to ask, yes, it is a real background behind me. There we go. See, these are not fake books. Just saw the hongobai in chat. So any of that, right? The trust you build, the wall you set up around yourself isn't the features, because anybody could launch green-chemistry.ai today and compete head-to-head with my brand new little startup, right? But unless they've got years of time and experience building in this space, people won't want to buy from them because they'll go, ah, you vibe code something good for you. Anybody could vibe code it. I don't believe you care about the problem. And the, the way you talk about what you do and who you talk to about what you do is what's going to make or break your business. And we have a problem here because I can say anything at scale on the internet these days, right? I can blow a million tweets out there that all say a slightly different thing, all written with AI. All of them are relatively garbage, right? AI produces the average of the entire internet unless you give it better instructions. That means in general, you're going to tend to get the same kind of output that everybody else gets. And we've all seen that AI slop on the internet. You know, when you see a quote that says, oh, it's not this, but that. In response to something, or, hey, the real fundamental shift is colon something, or you're absolutely right, explanation point, any of this stuff these AI tells we see. So people don't want to see that. They want to see you, and they want to see you talking about the specific problem that they have. So There are many ways to grow and scale a business, right? Content leads the way in a way that I don't think people really get for a few reasons, right? Paid ads are— I'm not going to knock them because CrowdTamer, the agency, for 15 years has been and continues to be a paid media agency. But there's a catch here, which is it used to be you could, with some level of expectation of success, say, I made a thing, buy it. And that's your ad. And some people would come. But now that is not the way to drive people into a buying conversation because they go, uh, another ad for a thing I don't want to buy. I have enough of those in my life already. And they keep scrolling. They go away, right? And paid ads are a pay-to-play, by definition, platform. If I stop giving money, I won't have any more growth. Now that's okay as long as the growth is profitable. But the way you used to be able to make your ads has fundamentally changed. Now, SEO can be a great long-term play, but most of us in this call today don't want to wait 9 months for traffic to start to show up and convert. Anyway, if I build an SEO engine around a business that I don't yet have validated, I might get hundreds or thousands of visitors and then discover they're the wrong one. And what you do when you combine ideas in SEO or PR, ideas in ads, ideas in content marketing, and combine them all together, we begin to create something where we have a positive feedback loop in place. But this used to be something I would tell people takes 6 to 9 months to roll out. It was slow. Now, it worked every single time. It was very dependable, but that slowness doesn't really fly in the year 2026. What you do is you take your content, and I'll show you how. I'm going to buy a set of ads live here today to show you how this works, and we'll even, we'll put real budget behind it and hopefully not show my credit card number while we do so, so that you can see How to buy an ad right away. When you do these ads, when you create content leading all of this, what you get is better ad performance, better outbound performance if you're doing outbound, and you get a sales cycle that's shortened because you build trust upfront. So let's talk about this. The first bit is you have to find a niche, and I know everyone says, I know my niche. But then inevitably, when I talk to a founder at Founder Institute, they say, hey, yeah, I service small business owners. And I'm like, cool, what type of business? Oh, you know, knowledge workers. Okay, well, there's 200 million of those. Like, narrow it down, keep it smaller. And so I don't want something that's that broad. I want to see— whoops— I want to see something like, hey, this Tool I have built could work for a number of different audiences. I'm going to try a few very small niches. I want to try people who run lawn care businesses in the Pacific Northwest that have between 100 to 300 clients. That's tiny, but that means you can say exactly what their problem is and understand, hey, I need to make content just for you, because if I make content just for you, then you're going to trust me. Better than somebody who's running very generic software ads. And you're more likely to care about my solution. And you're more likely to try my answer than you are any other one, because anybody can write the solutions now. It becomes, do I believe that you care about me? Now, the other part of this is you can't only bet on people who run lawn care businesses in the Pacific Northwest, because if that niche doesn't pan out, or if you've exhausted that niche, you need to have more than one arrow in your quiver. So the real idea behind this content engine is you find several niches and you use these niches, one, to test, and then two, to scale. And because you're not depending on organic reach, one of the big problems with YouTube's algorithm in general is when I make a piece of content, YouTube wants to serve people who've watched that content more of the exact same content. If I have different types of content, my YouTube reach organically is trash. And you can actually see that if you check out CrowdTamer's LinkedIn, uh, YouTube channel. We have, I want to say, 190 subscribers now, right? This is not a channel that organically gets recommended. I made $200K on YouTube last year with advertising, not with organic reach, because I pick a niche, I put content out in that niche. And then I make a different niche and put content out there. And if you've watched niche one and you see content for niche two, you're going to scroll. It's not for you. But so my organic reach is throttled, and that's okay because I'm making content that makes me money. I'm not trying to be the next viral hit. So then we have to have content for a number of niches, but then we have a volume problem. How do I make enough good content for 3 or 5 or 7 niches? Well, You start with shorts. One-hour recording, something actually kind of like this video I'm doing right now. If you can block out an hour of your time, you can make, depending on your skill and fluency in front of the camera, somewhere between 5 to 8 good 45-30-second clips. I've been doing this for a very long time, and my background originally, way back in the days, is theater. So I tend to make more like 12 to 15 in an hour. Because I come into it knowing what I want to say, and I usually get it on the first or maybe the second take. But that's me. You do you, right? If you can get 5 good clips, that means 1 hour gets you about half of what you need to really make enough content to keep a niche satisfied. So look for 10 to 15 shorts in one niche, and that's enough. Now, when we begin to put paid distribution into this, we're going to have enough where we don't just burn through all of our creative before we drive people into a buying journey. So then once I've got a couple of shorts pulled together, I want to drive you to a VSL. That's video sales letter, but there's a catch. In the old school ClickFunnels style marketing, marketing as they taught it 5 years ago and before, Your VSL is a landing page. They used to call it a squeeze page, which is a fairly unfriendly term. And the idea is the VSL was there to get you to say, hey, I want this $8 thing. And the squeeze page would slowly lead you to not pay $8, but then maybe $80 or maybe $300. And that's how they actually made their money, by convincing you, yeah, $8 would be cool, but $300 bucks would be way better. We're not doing that. Our video sales letter still sits on YouTube because YouTube rewards you for not driving traffic off of YouTube. And so we only want to have that last video, the one that is the go sign up, go book a demo, go do whatever. That's the only one that drives you off the platform. The other ones are all just watch content on YouTube because your cost per view is way cheaper when you have ads set for that. So VSL drives interest. You drive them to a page. That page's job is to convert. If you've done this right, shorts with paid distribution got your audience interested. The video sales letter with paid distribution has the audience now willing to consider. They go to your landing page, they do a free thing, they book a demo, they get a trial, they do something. And that's your job to sell them from there. So this loop can work in any niche because I'm not depending on you seeing me organically and finding trust over time. Instead, I'm depending on you're in this niche, I have targeted my ads however I wish to target my ads, and then kind of flooded the end zone with 10 to 15 pieces of content. So over the course of a month, you're going to keep fricking seeing me. Keep scrolling on the Shorts and going, ah, it's that mustache again. Dang that guy. Next one. Right? But that prevalence means when you start to watch my content more, I drive you to my VSL. You go, fine, fine, fine. I'll continue. Let's go. So that loop is very modular, very rebuildable. Let's talk about how you do this. And I'm going to grab a video real quick here in Descript. Whoop. And we're going to pick, uh, where is one that I can upload here? Uh, so Descript, there's a bunch of tools that are similar to this. What I want to show you here is how easy it is to edit a video, because if you are accustomed to back in the day, you had to use Adobe Premiere or something. It was a very complex tool to make video edits. Well, here's a much simpler one. If you're able to edit a Google Doc, you're able to edit a video these days. So I've got my tooling here, right? I've got a video uploaded. How long is this thing? The full length of it is 31 minutes. Now, this is way too long for a video sales letter, even if you're a lunatic like me who makes 10-minute sales letters sometimes on YouTube. You don't want to make one that's 30 minutes long. That's too long. So I can just read this and say buyer trust is at an all-time low. Sure. Uh, I want to go to right about there. We go there. I'm going to make this. Its own project, uh, new composition, and we're going to call this— this is the wrong one— going to call this my video sales letter. Then there are a few features. We're going to— video sales letter. You can see how good my typing is these days. Okay. There's a few features here built into Descript that also make things very handy. The first one is in the AI tools. I can do a, I can throw on eye contact so that I'm looking at the camera the whole time, even if I was reading notes. For those of you who really feel awkward and have to, oh yeah, what tool am I using here? Descript. Does it show up anywhere on the, on the title here? Uh, there you go. You see the upper corner, Descript. I would note this tool is like $40 a month, I think, $50 a month. It is quite cost effective given how easy it is to use. So I can also, with AI tools, ask it to do things like edit for clarity, right? I've got a 4-minute video and I ramble a bit. You can see a hmm here. I'm probably looping and reiterating at least once or twice in here. I can just say, clean this up. I don't want to edit crazy, but I'm also going to allow you to— what is this here? Hang on. There we go. I haven't assigned a speaker yet. I've actually put my voice in here. Oh, but not on this device. Okay. So normally, When you have Speaker installed, which I have it on the web but not on my local client here, you can actually have the AI reword you and try to paper in some of the gaps with you, your voice being put over whatever the AI thinks you need to say. It's actually kind of alarming. I can't enable it because I'm running this locally here on my Mac, but you leave Overlord to think, it's going to have a go at tightening up the script without any human intervention required. Then you can also do things like, and I have successfully in the past done, right, take the, um, the background away. This background's pretty busy. You can see I'm in the midst of redecorating. I finally got my encyclopedias down from behind me, but I haven't yet got all of my new bookshelves up in place yet. When that's all done, it'll hopefully look a little better back behind me as opposed to that kind of disordered mess, which I guess more or less reflects my brain. Uh, regardless, as this runs through, it is done, right? You can see the grayed-out lines here. So the problem is that the scope, hmm, right? It cut here in the last 6 months of 2025. Anything else that cut? It's cut a word here or there and shaved this from 4 minutes down to 3 and knocked a minute off of it. Okay, then the other AI tool I can use that might be useful here is I could say remove the filler word. Now I expect it already removed most of them, but let me see if it finds any more still. It's got the kind of, so there we go. I can remove that filler word. Oop, and it skips past it just to keep the velocity up for the speech. And then with all that done, I can say, awesome, that's what I wanted to see. Video sales letter. This is February 2026, and I wish to download this to a local export. Let's go. So that is all you need to do. To turn talking into the camera into a first version of a video sales letter. There's a lot more I could do, right? I could cut myself out from the background and throw slides behind me. There's auto green screen feature built in here. I could add intro effects, exit effects. Like, there's a lot you can do, but for a first attempt, that's really all you need to do. Do you have a hook that's strong enough to get people's interest? And if your hook is any good, then can you build this into something more compelling? I would note I also have, uh, recently run, uh, a couple other video sales letters. The hook on this one is definitely very mid, but I'm game to try this and see how this goes. It's okay, this is downloading, running locally. It'll be a few minutes. Let's not watch while that slowly loads. Let's talk about what it is you need to do. So first up, right, try that. A lot of people, especially their first 2 or 3 times recording this kind of thing, they stumble, they feel awkward, they don't like it. That tool makes it really easy to, uh, blend all this together. I didn't show you studio sound because I'm not going to try and do a sound broadcast through Airmeet while we're all watching. That's not very compelling video, but I have a fancy pants Yeti mic that I use for all of my audio. That doesn't mean that you have to have something that fancy. And if you just have the webcam mic or a lavalier mic you clipped on or whatever, and the sound quality is mid, there's an easy way to make it sound good because paradoxically, the most important part of a good video is the audio. People will watch a video where the lighting is only kind of mediocre quality, like this one right here, but if my sound was terrible, you'd all alt+F4 out and find something better to do with your life. Being able to make changes right to the text is the first time I had that happen. What was it, man? 2 years ago, I think, when they rolled out. That felt like sorcery compared to how it used to be. And what you need to think about, there's two types of content we want, right? There's Shorts and there's VSLs. That was a VSL. You can also make a Short, and what you want to do then is find a hook, an opening sentence that's really strong. Now, spoiler alert here, at the end of this presentation, because I always have downloadable goodies for you guys when you come to see one of my workshops, I have a Claude skill. That you can run, or I mean, any AI agent can run, that will look at a transcript that you have created from a video, and it will say, hey, these are the hooks you should have used. Now, if you're working on video, when you record the transcript and you throw it into cloud, if you haven't gone anywhere yet, you can just real quick re-record those hooks so that you can say, oh man, I had all these great ideas for better hooks than the ones that I did. Well, I'm still in front of my camera. Let me just record them again real quick. And hey, by the way, even if I read them off my screen down here, that AI eye gaze trick can make it look like I'm still talking directly to the camera. That first 3 seconds of any video, but particularly a YouTube Short, is important. Some of the best videos I have right now, I'll show you in a moment, but one of them is targeted to followers of Tony Robbins. I think that one might have stopped running finally. Uh, it says, hey, Tony Robbins is, uh, the world's best motivational speaker, but he buys ads like an idiot. Here's how you don't make the mistakes that Tony does. And you're like, oh, okay, if I'm following Tony Robbins's YouTube channel, I'm going to pay attention now because I'm mad at you. I like Tony. You just called him an idiot. Let's fight. Uh, another one is, I built $273 million marketing engine in the last 7 years. And they all start the same 3 things. You're like, okay, what are the, what are the 3 things, Trevor? I'll watch. Come on, right? A hook is something that'll make you pay attention for long enough. I can hopefully sink my teeth into you. Okay. Oh, my image tag broke here. I was just mentioning to Daniel that I found a bug in here and fixed it, but not here, right? What we get with that video which should now be downloaded. Let me check my downloads folder. Is indeed. You can see here my video sales letter, which is done at awkward typo. Comes from doing this live. The first thing you need when you want to run a YouTube ad campaign, which is what we're going to do next, is we're going to do this to studio.youtube. S-T-U-D. There we go. Is we're going to upload our sales letter to YouTube. So here we go. Create, upload a video. I've got this one right here. And the name doesn't really matter because people won't see this, but we're going to call this Buyer trust is your number one problem for 2026. And then learn how to build buyer trust and reduce your cost per lead through the CrowdTamer's Content Engine Marketing System. There we are. I could make a YouTube thumbnail. I won't here. You don't want to watch me do it. I'm curious to see what the AI-generated thumbnails look like these days. I usually make mine AI generated. I just make them in Gemini and I use my face. What you will find is because this is an ad, it's not something that someone sees the thumbnail on before they actually click on it. The thumbnails themselves don't matter, right? This is going to run as a pre-roll ad one way or the other. So, okay, it's not made for kids. Next, uh, no subtitles, no end screen, no ad card. These ads don't have any of those things, and there's no music, so I won't be in trouble. We'll do this unlisted because I don't necessarily want this just running organically. Okay, this will be a few minutes while it runs. Now what we're going to do next is we're going to prepare, and I've got an ad account where I just recently started to monkey around with some new video stuff in here. And we're going to build a new campaign. So what you're going to want here is you're going to want to build a two-stage campaign. The first stage you're going to build is for reach. That's the one where you want to say, hey, I'm going to get as many people as I can seeing my best content. And for that, what you're going to want to do is not necessarily the content with the most views. You're rather going to want to say, hey, of all of the content on this channel. I want to see my average percentage viewed— I beg your pardon— over my time frame here should be, uh, we'll do the last 90 days here. You can see when I've been running some ads to promote my content. And what I want to do is sort by this, right, so that we're seeing what gets the most views here, because the stuff that's sticky is the stuff I want people to watch, not the stuff that organically had the most views. You can see some of this stuff already had ads running, so the view count is immaterial for me here. If my view count is, you know, 2 or whatever, that's not very useful for me. But I can see this video, I'm not too surprised, organically did fairly well. And my percentage viewed at 50% is acceptable. And I bring this up because I need the link for this. And then this also will do okay. And then that isn't from my brand, so it's not very useful. That's also not for my brand, not very useful. Okay. Let me also change my timeframe. Give you running ads a while back. Hey, there we go. What else was getting numbers in reach? Let me do by views. There we are. Okay, that one reach dropped off. There we go. Sure. 1, 2, 3, and I want to see one more with at least— nah. There we go. There's 1, 2, 3, there's 4. Okay, so what we're going to do, we're going to build ourselves a new campaign. And this is an awareness and consideration campaign. We're looking for video views. We're doing this on YouTube. And we're going to start new. I'm not going to use the old one here. We're going to call this, uh, Founders Institute Test Campaign February 2026. I did the wrong keyboard shortcut. That's for my Windows PC. Founders Institute Test Campaign February 2026. There we go. All right, uh, what I want to run these as is— these are going to be— sure, want all of these target CPV We're going to run a daily spend. So I recommend about $100 a week is enough to give you a ton of info. And in 1 week, 2 weeks— yeah, this is me, I promise you— in 1 or 2 weeks, you will— on my iPad, I've of course— there we go. I could— on my phone as well, my iPad not being here. There we go. Come on. There we are. So $100 or so a week will give you enough data that in a week or two, you will understand exactly whether or not this, this campaign is finding the right people for you or not. I want to run this only in the US. I am going to turn off the Google Display Network. The Google Display Network is the closest thing I can think of to internet cancer for your eyeballs. I would never recommend anybody run something on there unless they are quite sophisticated at building blacklists and blocking advertisers from seeing their stuff, because Google Display Network is where all of your spend goes to die. So, all right, uh, no political ads, nothing rated here. This ad group, sure, whatever, we're going to call this, uh, first layer ad campaign. Now, this is where YouTube gets to be interesting because I can target by a lot of different ways here. That's going to be very relevant for your business. Everything before now is the same for everyone. This is where it's going to be different for you. I know in my case, I might want to search for things such as change my demographics here, right? One thing I do always do is switch that to be whatever age to 64 and not 65 plus, because YouTube and most of the ad platforms treat a 67-year-old as exactly the same as a 95-year-old. And of course, that 95-year-old is almost certainly not in the market for a business solution the same way how a 67-year-old very possibly could be. So, okay, in here We get the ability to do a few things. One is to just to pick and say, hey, I've got a— whoop— why don't you just pick and say, hey, like you're interested in this segment. Cool. I want to advertise by that. That's a pretty big segment and you won't run out of views. But the other thing you can do that is more interesting to me is to say, I don't want that segment. What I want to do is look by keyword. And I want to say something like As an example for me, one that works well frequently, HubSpot pricing. I don't sell a CRM, but if you're in market for HubSpot pricing, odds are pretty good you're thinking about a B2B marketing solution. You're trying to figure out your marketing action as a B2B marketer. So if I advertise on someone who searched for that, and this is what's relevant, right? You search Google for this. And then what you see is I'm not running when you search that term. You don't see me next to that search term. What happens is after you search for that for 30 days, you're going to see my content on YouTube when you go on YouTube for whatever reason. You're, you know, taking a break in the evening and you're scrolling, and then you see someone saying, hey, you've got this business problem and here's how you solve it. You go, oh wow, that is relevant to me. Even though nothing about what I'm talking in my ads is HubSpot pricing or CRM or any of those things. Now, for many of you, I mentioned you wanna find a smaller niche than people who are doing business services in general, right? This starts to target down where I can say, uh, I know you're early in figuring out a new B2B marketing motion, so This is way smaller than just everybody who is B2B, right? These sorts of terms let me narrow down my marketing. Or I can say, uh, we're going to do, uh, business content marketing agency. These are fairly broad. I'm not going to go deep into my research, although another one of the skills that you'll get at the end of this workshop will help you figure out how to dial down into a narrow niche. Then you can choose those words, put them in as some of how you're searching and doing findability here. So topics here are pre-selected clusters by Google. They are fairly broad. For the purposes of what we want to be doing, I would tend to avoid those. The last thing that can be really powerful is you can do by YouTube channel placement. Now, you need to have at least 10 channels in an audience group to make it where you can do a YouTube channel placement. But I'll give you a hint: if one of your channels is Tony Robbins and 9 of your channels is somebody no one's ever heard of with 200 subs, guess who's going to see most of your videos? Tony Robbins's people, not those random folks. So if you know of a particular YouTuber whose audience speaks directly to you, Put them in and then put in 9 other randos. There's your reach right now. You have almost all of your views are targeted by that one, not by the other randos. So any of these ways to target lets you say, okay, I've got my content. Now I need to grab— this is a video I made about don't be concerned about looking stupid because look at me. I'm lying in a snowbank recording a video. So I can then— oops, I have to grab the link. I switch between Mac and PC several times a day, which is a great way to forget how all of your shortkeys work. But there we go. I'm going to grab this, put this in here. I'm going to grab this. Grab this here, and we're gonna grab this last one and put this here. Okay, so now we have only 4, and I recommend, if you recall, 5 to 10. But you don't need to see me search and paste and grab more videos from other channels I've been in and all the other stuff. This is enough to give you the idea, right? You put these all together. Then you say, where do I want them to go? Well, not to toot my own horn here, but that's my WordPress login. Don't go there. Go here. So I can say, I want you to go to this page, but then this is something a lot of folks get wrong. They say, great, go there. And then I don't get any data. Oops. I pasted this in the wrong place because I'm busy talking to y'all. Hang on. There we go. Paste this here. Now you can say, great, I'm sending people there. That's my homepage. That's where I want them to go. Well, two things. One is no, you should be sending these folks to a specific targeted page. Now, for the purposes of the demo we're doing here, I haven't made a page. I could make any one of a number of them. But the idea is if you are, uh, working in lawn care services, you make a lawn care service particular, uh, specific page. And any of us can vibe code landing pages in a minute now, right? I tend to use, uh, I think I have Lovable hooked up to landing.crowdtamers.com. And then any page I make there just auto-deploys and I can route to those specific pages. Uh, the technology to build landing pages is an old and very solved one. Build it how you want. But the other thing people get wrong is they go, great, there's my link. And what you didn't do, I'll give you this, right? We have UTM Link Builder. Is Google put this tool together? About 200 years ago. You can tell from the look of it, right? I can say this is the URL we're going to. My campaign is Founder Institute 2026. My source is YouTube. My medium is display. My campaign name is Business Exec Founders. And then this generates a link with all of this other data. Now, when I am running a campaign all by itself, I probably don't need all that data. But here's the thing. Once you start to run 5 or 6 campaigns at once, you 100% want to make sure that you are differentiating your different campaigns with this data so that you can then go look in your analytics. And say which one sent traffic to my website. Now, there's another catch here. Hang on, we're going to do that for now. And this is, yes, ad number 1. Cool. So my target CPV is going to do $0.50. This is quite high. I could honestly do $0.20. I'll still be fine. There we go. So there's 2 things here. One is I just put a URL in here and you're like, great. So when I run these ads, I'm going to see clicks, even though this is a reach campaign, I'm going to see clicks on these ads go to that destination. And that's a lie. And Google will cheerfully completely lie to you because you're going to see, let me go back to, is this campaign live enough to have any clicks yet? Hang on, let me clear this up. Um, we're running, uh, this one, I think, is the other one we have. No, all of them. Campaign status, all my enabled ones. Yeah, but where's the other one? There we go. This one. Oh, this just turned on. All right, so when you're running these You will find clicks, true view. Where can I show you a campaign that won't get someone mad at me? Sure, I know what I can do. Do I have the right— these guys? We ran a small campaign from a little while back, which was running through here. I want to show you the stats here, and because they would lie to you, right? They say 83 clicks. And these 83 clicks, you're like, oh, okay. Uh, I sent 83 clicks to my website. No, you did not. Those 83 clicks on YouTube represent a click anywhere in the ad, on the ad, expand the text, doing anything. The only way you're going to know what clicks went to your website is to go look at the actual analytics on your website side of things. That's why it's important that you actually tag these things so you can properly monitor and measure them. Because you might think, wow, my cost per click here, right? $150, 83 clicks. I mean, we're like $1.90 a click. Do we have that? Where are we? $1.80 per click. That wasn't bad for mental math. But it's not clicks to your website, so don't get too excited. Okay, so I built and run this campaign— not this campaign. I built this campaign live right now. I have built the first part of it. This is the reach part, the part that says, hey, I exist. And then if you watch my content, I'm trying to build trust over time. I should have 5 to 10 videos in this to start with, right? We're using 4 for now. It'll be fine for the purposes of what I'm doing today. Now we have to build the next campaign, which is I want to send traffic to my site or If you're more advanced and you've got a little more spend, you can do leads or conversions, either of these, right? But when you're just starting out, you probably haven't got all of your tracking working. Website traffic is much easier to track and optimize for than sales and conversions actually are. So if you start with website traffic here, your spend will be lower. And then what you want to see is, uh, why are you not liking this? What are my two actions? Oh, they're all inactive. Interesting. It's possible because I haven't run campaigns from this account for a while, it thinks they're inactive. I know they are functioning and they are reporting. We're going to run video ads. Don't, don't do PMAX. If you start to run Google Ads, Google will tell you, hey, you should run PMAX campaigns. They make a lot of money. They do make a lot of money if you're spending several thousand dollars a month. We're talking about spending $100 a week. Don't run PMax, run a video ad campaign. Don't run Google Display Network, right? We're going to do a new campaign. This campaign is going to be built on— you want to send clicks over. Yeah, we're going to send my daily budget here will be again, we'll do $15. And it— you can ignore this budget recommendation for now. You will, you will tune things where you want them eventually. And, uh, not political ads. Then we're going to run these in, uh, you want to make sure you do presence and not presence or interest because Google is happy to take your US-only targeted ads and run them in, I don't know, Vietnam, people who are interested in the US if you don't throttle and control that. And then there are definitely, when you are more advanced, you can say things like, hey, there's nobody who is, uh, going to be doing business buying decisions, uh, between the hours of, let me find my time here. We'll go, uh, 6 AM to, we'll do maybe 9 PM, right? Anybody who's running outside those hours in their time zone, they're not having a good day already. If they're on the internet this late, don't, don't show them ads. Cool. Then what we're going to do is our ad group here running in this area, right? And we're going to choose YouTube, not here. Okay. Then with this, we've got a video ad, and we're going to pull from our most recent uploads. Hang on, let me go back. We just put our video sales letter up, right? That is content. That's this guy. There we go. So here, paste. Nope, paste. There we go. That's that video. And we're not going to add any more in this campaign because all we want is our, um, This is a new account. Well, I can upload. Then we have my brand lion is there. You vex me. I need to send the brand line without the text. Give me one second. Not that one, this one. It is there. Okay, cool. Now we have both of these. Nope. There's my logo. Then text. We have, uh, get all of your content marketing solved for you. Oh, this is shorter than that. Yep. Build trust in 6 weeks. Build trust in 6 weeks. That fits. Get all your content marketing done for you in 1 hour a month. That may be too long. There we go. Then description. You need to build trust with buyers. We're the experts to help. Then this is Crowd Tamers. I can add a few sitelinks if I have them built. Ah, this is too new account. Okay. Then this, the last thing I want to show you is Aha. We have in our targeting here. I was busy scanning past things and didn't see it. Let me go back to here real quick because we need to set in the ad group, the targeting needs to be for folks who've seen your video. There we go. Add an audience. We can add a new audience. And this data can be a couple different segments you can pull from, right? You can, one, create custom audiences, as you've seen I already have, but I can also do things like you have, you know, YouTube users, Shorts ad viewers, right? Now what this is, is this video only shows to people who saw that ad already, my YouTube Shorts, right? This way I'm not spending to show my video sales letter to people who don't know who I am. They've already seen a handful, 2, 4, whatever number of these videos, and now they see this longer form video. When you're running this at scale, what I tend to find is I will get, even on this 4-minute long video, 60, 70, 80% of people watch all the way to the end because they've already seen my content and liked it. There are a few pieces that are important though. Right, the first important thing here is— oh, audience name, uh, Shorts Watchers. There we go. Now we've got this ready. This, and then publish. Right, what's the problem with this? Oh, there's no video put in yet. Awkward. Okay, we will grab that. And then paste it in. There we go. That all looks good. Review. Publish. Okay. So what we've done now is in 20 minutes, we've gone from, I don't have anything ready to, I've got shorts that to be fair, I already recorded on my channel, but you could record your own and then publish them. And then those shorts, what we're going to do is every week we're going to look at them and see how they perform and go, oh, these videos have a really bad view-through rate. Let me stop them. No one likes them. Or these videos have a better than 40% view-through rate. Let me keep running them because they're what's seeding my audience for the actual video sales letter. Right? So $15 a day on my, uh, awareness campaign, and then $15 a day on my conversion campaign. Now, right now, the conversion campaign will serve to zero people because no one's seen my Shorts ads yet. But over the next couple of days, as more and more folks see the Shorts ads, these video sales letters will start to serve as well. So that is how you build your two-stage audience and then begin to figure out who's responding and how to monetize. So let's get from here back to— I can close you out now, I'm done with you. Let me get back to ImpressFlow, which is— well, we'll just do it here. Oh, there it is, first tab, and present. All right, so I kind of went through all this, so we'll blow past this. But the key thing here is your spend doesn't have to be intense, and in a week you already know enough, right? You'll be $70, maybe $150 if you're running both conversion and awareness, and you'll have learned something. So what matters? I saw a question come by. What's the stats I care about? There's a couple of key things here, right? Your view-through rate is crucial for the performance of these videos for two reasons. The first reason is if people aren't watching your video, stop showing it to them. You're wasting your money. The other thing is, and I think this is unique on YouTube, if people tend to watch more than 50% of your video, then YouTube will organically recommend your content to other people like the people who watched your video. Which means if my content is good, then I get free views out of it. That means YouTube kind of shoots itself in its own foot, but really all YouTube wants at the end of the day is content people like. And if your paid content is something people like, they don't care, man. They will show it organically for you. That's what you want to see. And organically, more than 80% view-through rate tells you this is something that is interesting to folks at some scale, right? I've got videos with 5,000, 10,000 views organically because the view-through rate was good and people just watched it. But that's not how I make money reliably, right? So keep that number high. Better than 30, 40% view-through rate is crucial for this to begin to work. I already mentioned click-through rate is a bald-faced lie. Do not believe it. But instead, look at your website analytics platform to see what your traffic is doing. Now, your spend is going to eventually drift into a couple areas, right? And you're going to have 2 or 3 types of content that you want to build out. Awareness content. I showed you that, right? That's the YouTube Shorts. You may discover, as you build over time, that your view-through rate and your engagement on your conversion content, on your video sales letter, remains bad. The usual cause of that is that people don't yet have in their mind, they haven't made the switch from this is interesting to I want to know more about this product. And so you can add a layer in. So if we have our shorts ads audience, we can create another audience of folks who've watched the shorts ads to show them more shorts ads, but those shorts will be thinkier. There'll be a little bit more like, not just, here's a fun thing, or here's Trevor lying in a snowbank, but instead it will be, this is a problem and this is how you solve it, and I've told you how in 45 seconds. And then that builds more trust in your expertise, and then you serve conversion content. But in general, your conversion content's only 5 or 10% of your overall reach. When you're really optimizing things, you don't have a $15 a day budget on awareness and a $15 a day budget on conversion. You have a $35 or $40 a day budget running, and then the ads are split across it all. The way you build it and the way you optimize it takes place over time. So why have I been talking about this for an hour? Why should you care about this? Because the economics on this are solid. And it is really hard in my experience to find an engine that's this dependable on other channels. So I run a B2B business, right? This is something that I have over the last 18 months built out to where, not in this new channel where I just launched these ads, but in, uh, my actual main ad channel here, my cost per lead fluctuates between $20 to $80 per lead, which on the B2B side is pretty good, but these are relatively high intent. That means my cost per customer is about $415, somewhere up to $470, depending on the kindness of the ads gods these days. My average client pays me $1,500 a month on their first month. That means I get my money back on month one. I even get my money back after labor. On month 1. And I've got tons of clients who see very similar results. Some of them are charging $20, $40, $80K for a big contract, and they're still able to hit— they're more like the $80 cost per lead, right? But if your client is worth $20K, $30K, and you can get, you know, one a month signed on a very modest budget, that's phenomenal. This works B2C too. I've used this B2C too. It is harder. And it is harder to the point where don't pay somebody to do it at small scale. When I get somebody who shows up in my, uh, sales funnel, they say, hey, I run an e-commerce storefront, we sell shoes. I say, cool, don't hire me because you need to be spending like $40 grand a month for this engine to work for that type of persona. And that's a relatively rare amount of spend because you have to hire the agency to manage everything. That's You can manage it yourself. Your margins get way better, right? You can get a cost per lead down to about $2 if you make good content. That means your cost per paying customer, $12 thereabouts. If you've got especially a multiple purchase action on the B2C side, this can work. It is a narrower margin. I like this way better B2B than B2C. On B2B, it is stupidly valuable. On B2C, it's yet another tool in your arsenal. Okay. This also can go— go away prompt. This also can go crazy places. You can make fully AI-generated video these days if you, the founder, are not comfortable on video. There are a few things you can do that are dangerous though, right? UGC-style AI videos can be great. The top video on the channel that I have right now, this one, where was my analytics, was a test I did for somebody else's business. Let me do advanced mode and we will do this. All right. This video here is UGC-looking content for a different brand. And the performance on it is quite nice. It wasn't my, hang on, what's my view-through rate? This one. Uh, average percent of view. There we go. The content performance now is probably the best of anything I've got running right now. Yeah, very, very high, right? But the catch here is you can easily say things that will get you in trouble with Johnny Law. Because if you say with AI-generated video, hey, I tried this tool out and I found it phenomenal. Well, they didn't. It's AI. It's fake. And the SEC and the FCC are going to start pursuing people who use AI-generated stuff that says false claims because that's not truth in advertising. So be careful when you make AI-generated stuff. You could say, you should try it out. This can achieve these results. That's perfectly fine language, but don't have it say me. If not a real human. You can also use— and I actually have, uh, one of the videos that I had in Descript which I, I didn't showcase. I won't because the founder hasn't approved it yet to go live, uh, is I— the founder tried to record a video sales letter and they just couldn't make themselves do it. And so I went, don't worry, buddy, let me take the first 30 seconds of the video sales letter that you made that was rough but usable. And I went to AI generate the entire next 3 and a half minutes of that video, where we have natural looking gestures, where we have that person's voice cloned, as well as their face cloned. So it still sounds like them talking. And then we made the full video. That way, if you are an awkward founder, you stumble, you don't feel comfortable at this, upload a short video of you talking, grab out the audio. You can use AI to generate this stuff at scale as well, and you're still the founder, so it's still building trust and faith in you. You can also go fully nuts, right? VO3.1, SeedDance 2— there are tools that let you make Hollywood-style insane cinematic videos for the price of a pizza, maybe a somewhat decent dinner out. Things that used to cost 60 grand to film and take months to prep and cast and light and location scout and all that, you can do in an afternoon in your boxer shorts sitting on your couch. Your limitation, like everything else I've been talking about, the cost of effort is dropping to zero, right? So then how you want to talk about your product and what you want to showcase your product is limited only by what you can think of. And crazy Hollywood-style cinematic videos might do it for you, except that anybody can make those videos, right? So don't build your whole campaign around a wild computer-generated avatar if you can build it around you and your interest instead. Unless your skills are that you can speak to your audience so well because you know their pains so well that nobody else could generate equally good, crazy, crazy computer animated videos. So I've got, uh, all you Founder Institute people here. I'll say the most frightening words in the English language: special assignment. I've got homework for you, right? And the homework for you is build your way to out-trust your competitors. You can start today. Right. And what you do here is take a couple hundred bucks or just $100. That's all you've got right now that you can put towards it. And start by grabbing these files. Take out your phone unless you're on your phone, in which case, good luck. Although we will send these after the fact. This is the link to download the, uh, skills and the files and all the setup documentation and everything so you can start to do your own video effort, make your own recordings, and produce your own high-quality enough content to begin to build and make money with this. Then your special assignment is— whoops, that went the wrong way— special assignment is get started, right? Find a niche, name your niche. One of these skills is your niche finding skill. It helps you go out and figure out what niche you might want to go after. Then record your shorts, sit down. You can just record it on, I know Mac has Photo Booth, Windows has some similar application. If that won't work, you can sign up for Descript and record right into Descript. There's any way you want to capture the video. right? Record an hour of content. Now, if you want to prep for it, there's a skill included in your download that is your interview skill. It is how to actually draw out from you interesting things to say. And we're going to make your VSL, right? Take your campaign, spend on it. Then there's specifically— Crowdtamers works with— what do we got now? 22 founders we're working on. Every month we work on VSLs with them, and I have taken the ideas of how I tell people to make a VSL, the stages, the steps, all the elements, that's in that download too. So you can go through and ask your AI to help you write a VSL for the niche that AI just helped you find. You make your content, you put your spend behind it, you get in front of people. The founders who start today, who take this seriously, who tomorrow block out an hour of their time to record this video, and on Thursday block out an hour of their time to follow through the steps of what I just did and buy their ads, will by this time next week, so February the 20th, the 31st, which doesn't exist, so that's March the 2nd and March the 3rd, they will learn, they will move faster, they will be able to make money while the money-making is still good. Now, if you want to get this presentation for whatever reason, there's the QR code for this, or you can just type in the link. This has the QR code for this and for the Gumroad link in case you've missed it previously. But that's my, that's my bit. That's my discussion, right? That's the thing I want to empower you to do and tell you to go solve right away. Now, I would hope, while I've been talking, that some questions have popped into the Q&A tab. Let me ask you, Daniel, what's going on?

Awesome, awesome. Thank you, Trevor, so much for that presentation. Folks, if you liked what you heard, please show us with your emojis, tell us in the chat. Trevor, I think you can see a bunch of hearts on your screen and thumbs up hearts, a ton of them. Awesome, awesome. Okay, so quick things here. I've shared the link to the skills and resources in the chat and I've pinned it in case you want it right away, you can do that. But we will send out a follow-up email with recording to this, the resources as well as the presentation. So you don't need to worry about saving all these links.

You should be good.

It'll be sent out in the next 48 hours. So Trevor, we don't have a ton of time, but I'll run through some quick questions. I want to start with a lot of people with AI, especially when it comes to content, there's this concern about making slop, right? So as a founder who's creating, who wants to create good content, like how do you determine whether or not what you're making is slop?

So there's two ways. One is you exercise your human like judgment upfront. And you say, hey, that's a stupid question. But the other way, and I think the most indisputable way, is look at your view-through rates. If people are watching your content, it does its job. If people are watching more than 30% of your content, or 80% if it's organic, right? Then it's doing its job. Now, that doesn't mean that some content that people watch isn't kind of garbage. We all have that show that we watch kind of guiltily, right? Whatever it may be. I don't know, Buffy Vampire Slayer, which I guess is showing how old I am, but whatever it is, right? Like, fine. But as long as people are consuming your content and then moving their way through the engine all the way to your landing page, your content's good enough. Don't worry about it.

Awesome. Okay, so we have a question here from June. And June asked the, uh, you know, the process that you talked about, is that more for ads or can you also apply the same things to organic content?

Well, so this is all organic content. Like ideally you're publishing this all to your channel and then you may also be putting ad spend behind it, right? If you go look at the CrowdTamer's channel at youtube.com/@launchtoday, All the shorts that I put into the ad campaign we just launched are released organically. They're just also boosted through paid because organic already tells me if people are watching this organically and they're in my niche, this is where it gets a little bit tricky. Then I know it's good content. You can do the same thing over on LinkedIn side. And if I take my content and publish it on LinkedIn, I get no reactions and no likes. That's probably not very good content, right? If I get at least a few, and LinkedIn is very, very tight with its reach these days. The same content that used to get me 8,000 views a year ago now gets me 300. Because I actually mentioned this in, uh, well, one, in one of the Shorts on my YouTube channel, but also I think in the last workshop, right? LinkedIn's content has exploded. They've had 64x more content But the same size user base, everyone's using ChatGPT to publish random text spam, right? You making video are able to cut through the noise more than just another text post that's the same quality crappy AI slop as everybody else. So publish it organically if you can, if it's not somehow opposed to the current audience that you've got, but also promote the stuff that you want to make money from.

Awesome. Okay, so we have more of a marketing question, less so AI. So we have, let me just see, Kajal. Kajal asks if they had to pick one channel, there's, you know, doing this for the first time, what channel should they double down or do they maybe experiment with a bunch of different channels?

I mean, obviously YouTube I'm recommending in this particular video. I only use I use two channels for my business, right? I use YouTube, I use LinkedIn. And do I publish on Meta? Yeah, I think I still do, right? Because it's one extra checkbox on the social media scheduling tool. Do I publish on X? Yeah, I do. I get one client or so a year off of X, but I mostly, I don't post on X mostly for the money. I do it because it's just a place I've been in for forever. The same with Reddit. I've been on Reddit for, I think, 8 years. Have I ever landed a client from Reddit? I don't think so. That's not my, that's not my acquisition channel, right? My acquisition channels: LinkedIn, YouTube. And by only picking 2, I can focus my efforts and deliver well there, rather than try to be on 20 channels and do a mediocre job of most of them.

Okay, cool. Uh, so John has a question here, which I think is super useful for anyone who's doing this. So should you make a landing page on your own domain, or should you— or do you use like one of those throwaway services, maybe something like Card or even Replit? I know you can connect your domain, but like, how would you think about it?

So for an advertising point of view, it doesn't matter too much, except if I'm at, you know, landing or try.crowdtamers.com. Right. Which is where I think my domain, my landing pages used to go. Try.crowdtamer.com/something. At least I know this is the business, right? Card.com/random jumble of numbers. Little bit lower trust. The setup time and the setup complexity is so low these days. Like I set up greenchemistry.ai today. Right. And I have, again, I'm not a very good dev. My AI writes the code. It pushes it to GitHub. I have my GitHub repo synchronized to Vercel. Vercel automatically deploys it for me. I don't do anything. And then I went into Vercel and said, hey, I want a custom domain. It said, hey, you, you put this number in on the domain name server in Namecheap and to copy and paste and enter. That was it, right? The technical hurdles to have your own domain or subdomain hosting these days are so low. Just do it. It takes 10 minutes and then it looks a little bit more professional than all the jokers who didn't.

Awesome. So we have one more question and then we will get into networking, folks. If you do have the time to stick around, I do recommend staying for networking. You'll get to meet founders from across the world, talk shop, talk AI, that's all up to you. Okay, so the question here is, sorry if I missed it, can you share which specific AI-based campaigns you ran that had like the highest ROI versus maybe not so great ROI?

Sure. So while I talked about this is AI-led content, right? There's two things here that I see in the chat are going by. One of them is you didn't use, you didn't use a lot of AI, Trevor. And I'm going to argue with you because everything around Descript was AI selection of the best part of that video, AI editing of that video. And then when I'm working on how do I generate my business ideas, how do I use a Claude skill? I showed that on the last workshop. I could have showed it again here, but we already ran pretty tight on time, right? Using AI to help you formulate your ideas, using your human brain to say, of the 6 things you propose, I like these 3, let's go. Recording the video, having AI do the edit work for you. So you just click 1 or 2 buttons and say, great, export and publish. And then I would note the entirety of Google's ad buying system, like how they place ads now is all AI. So these campaigns use AI not in the sense of I'm talking to ChatGPT, But in the sense of AI is functionally crucial to the creation and delivery of every single asset you see throughout the entire sequence. So that said, uh, campaigns I am running right now that are doing well, uh, I'll give you one. So one of my clients right now, they do, um, workflow automation. They are an AI integration agency. Super crowded space. There's at least 10 major competitors that I'm aware of and probably 300 tiny ones. So what you don't want to do, what they have been doing, is saying, hey, uh, you're in the software as a service business, let me advertise my automations to you, because everyone is fighting over that. The AI wins that I'm getting for these guys are, hey, Hey, you're a dentist. And right now it takes you 120 days to get the money for your dental surgery from either your patient or in the rare cases they have dental coverage, the dental insurer. But you can automate the entire process away, be fully HIPAA compliant, and then no longer have a 3-person team chasing $1 million in backlogged payments that you're just kind of floating. Right? That content to somebody who owns and runs a dental group of a couple million dollars. There's a few thousand of those in the US. Those folks go, wow, I didn't know I could shorten my, my pay period from 120 days to 60 days. Is that worth 15, 20, 40, 100 grand to me? Heck yeah. Right? So AI campaign that's doing really well is finding an underserved niche. And going after them. It is the same for me as a marketer, right? The people who I'm landing are not SaaS agencies so much these days. I'm finding deep technical expertise, professional services, uh, people who sell to large businesses. These people have tended to get kind of missed by a lot of the marketing agencies. And being able to say, hey, this engine at this price brings you this many people who pay attention to you every single week. They go, tell me more. Right? And so my average life cycle for a sale went from 5 weeks down to, I mean, honestly, if I don't close a deal within 72 hours, I just forget to follow up with people these days. I probably have 20 contacts in my CRM that I should have followed up with, but just They land easily enough. Why fight?

Awesome, awesome. Thank you, Trevor, so much for taking out the time to do this, and thank you to all the folks who joined us. Uh, don't forget, we do have the third part of the workshop series, so in case you did, uh, miss the link, I'll share it in the chat again. We'll have all the links that you need in the follow-up. So before we conclude, A last reminder, we do have networking open for as long as you're around. So do stick around and join. Trevor, before we close out, any final words for the good folk here?

Yeah, I always try and remind people, all of this stuff only works if you do it. And I don't want you to watch this and go, hey, that sounds really cool. Let me close the tab and go have a coffee. I want you to be energized. I want you to go solve this problem. If you can't solve the problem, hire someone to solve it. Not even necessarily CrowdTamer, although of course I want you to hire CrowdTamer, but go hire someone if you can't solve it. Because this is the difference right now between success and failure in 2027. And I can't like impress on you how important it is to find a niche that you can own that no one else is fighting over.

Awesome, awesome. There you have it, folks. Uh, do the work. I hope you all have an awesome day ahead. We'll see you in the next one.

Take care, y'all.