Cody Schneider | On the Homefront with Jeff Dudan #72

Welcome everybody to the home front. I am Jeff Duden and I'm here with Cody Schneider today and we are super excited to learn about AI digital marketing and some of this fast dynamic changes that are going on right now across all sectors of business. And Cody's got an incredible background he's going to share with us. Welcome Cody.

Jeff, thanks for having me man. Super excited to be here.

So yeah, awesome. Thanks for being on Cody. We always like to start, can you tell us a little bit about your background, kind of where you grew up and who Cody Schneider is?

Yeah, yeah. So I'm from the northwest, originally from the small town up in North Idaho called Coeur d'. Alene. It's about two hours from the Canadian border. Yeah, spent a lot of time there and then a little bit of time in Denver, Colorado during high school. But that's kind of the origin. Grew up in an entrepreneurial household. There was no real way that I was going to escape that. Both parents were just kind of profiteers. I don't. That's the only way. I didn't really know how to describe it. My dad did kind of like just logistics for forever so brokering a lot of like international like products and then my mom was in the medical space. So yeah, I've had real jobs and have also probably had way more not real jobs throughout my career. But the big thing that I'm doing now and kind of building is in this AI like media space. But before that I worked in the E comm world, did a lot of stuff there and like programmatic SEO. So basically what that is is building out like hundreds of thousands of landing pages for these like large brands and then also worked in a, or at a B2B digital marketing agency where we worked with like manufacturing goods companies. So think about like people that manufacture like doors, windows, wood products, anything for like home building. That was kind of our target customer and what I did there was help them transition from like traditional media and traditional marketing into digital marketing and really did all their digital strategy stuff and then that I don't know how it led into working for like early stage startups. So I ended up in out in San Francisco working for a company called Rupa Health. We ended up ended up, I was employ six there and ended up helping take them from a $20 million valuation to 110 million dollar valuation in about six months. One of the things we did there was basically just like spin up this massive media arm so like podcasts, live classes, chopping that up into clips, distributing across all Social channels, writing blog posts, et cetera, you, you name it, we probably did it. And then building out the team and kind of infrastructure and the workflows and processes around that. So that's kind of what I'm obsessed with is I see where the market's going for like most companies is they have to create some type of digital media component of their business if they want to stay relevant. And so we're really looking at and exploring how are companies doing that, what are the tools they're using and really what are the workflows that they're building and the processes they're building based off of those tools. So yeah, happy to dive deeper on any of that, but that's kind of the high level.

When you were growing up, what were your platforms of choice?

Yeah, yeah, so I learned graphic design pretty young. I had this art teacher who taught me how to screen print T shirts. And then from that I took a Photoshop Illustrator class. And then I was just basically hawking bootleg band T shirts and anything else that people wanted to buy like throughout the school and then kind of built up notoriety within that. And you know, for me it was like this like aha moment. I remember I sold this like it was like a roller Hockey Club, like 25 shirts. And I bought them at Michael's for I think it was like, you know, at the time like 25 bucks. And then yeah, like ended up selling it to him for like a couple grand or something. And I was like this is a no brainer, like why wouldn't I do this all day long? So yeah, that, that actually is how I ended up in E commerce. I got into the print on demand world really early. So I was doing graphic design work, figured out how to do data analytics on these platforms like Etsy.

Right.

What I mean by that is basically sign like finding bestselling products that were on those platforms and then creating like similar products based off of what was performing well and then using these outsourced companies to, to print and ship and do all kind of the production of those goods. But yeah, that was kind of the. My origin into all of this was really that for a lot of people I feel like it's the same, they like sell, you know, sweatshirts or T shirts and then somehow they up in B2B software, which is the most ridiculous thing.

But yeah, well look, it goes a couple of ways. So when I was growing up, all of our side hustles were, you know, we pressure washing or service type businesses, washing cars, things like that. And Then my kids as they grew up on Etsy, I got a Kid, you know, 3D printing aftermarket RC car parts. I love it, selling them on Etsy and you know, seeing what cars are coming out. And then before the manufacturer could make all the accessories they would design some on some CAD software and then print it and then people would come and buy it. And you know it's a, it's a, you know a lot of parents are saying, looking at what their kids are doing online but man it, it's these side hustles turn into real material advantages especially when it comes to demand, demand generation, understanding how to, how to, you know, just make money on the Internet and do these types of things. So I'd like to talk a little bit about AI and you know, I've been traveling a lot lately and I've been going to different masterminds and different groups and I've just come to believe that marketing is changing faster than anybody really understands in terms of how customers are acquired, how the customer journey is automated, how content's produced, where it's placed, how awareness for things is being generated in consumers minds. Even even though they might have not thought they were a buyer and suggested. It's, it's really psychological, it's really powerful. Where is your head with that? I mean is, you know, how fast? If I'm a, if I'm a business and I've been marketing the same way for the last 10 years but I've been making incremental changes in my business. I'm going to do a blog maybe I kind of had a podcast. I still have a diversity of lead sources. I'm advertising on these different platforms. I think a lot of that's going to change really, really fast for huge swaths of the economy. What's your view on that?

Yeah, I mean I think it's redefining like any base technology changes the constraints, right. And this is a base technology. It's just like what mobile did to how people discover things previously traditionally before we had the iPhone and the phone in everybody's pocket, it was like you would do a search on desktop and what you needed to do was way different than you know, when mobile happened. The same thing is happening here. I think the biggest, like what I'm, I'm a data junkie, right. So like I'm just looking at like the trends that are naturally happening and adjusting based on that. So like some of the things that we're seeing, especially when it comes to like local or small businesses or service businesses that are in specific geographies is a majority of people are starting to use social is the way that they find those businesses because they don't trust Google Maps. They don't trust like Google. If you look at like the data on Gen Z in particular, majority of their searches start on like TikTok or Instagram to find like, you know, companies that are in their geography. Like, especially when it's like anything that's food, you know, like service companies related, it's like that's the origin of their search and then maybe they go to another place to like find contact information, et cetera. So just like thinking about that and that as like a larger shift in the mindset of the consumer and it kind of makes sense, right? Like there's so much like spam that exists in these traditional channels. Like looking at Google, like again, I started in the SEO, the SEO world, like majority of searches that are related to like H Vac companies is some guy in the middle of, you know, nowhere, Arkansas that has 50 different websites that he ranks for like all these tier two cities across the US Right? So he's just like, you know, capturing leads and then selling those to like local H Vac companies. In contrast, I think people are starting to understand that they're dissatisfied with that experience. And what they're looking for is more of a human connection that has a face. So if, you know, here's Sarah and she is like, you should go to this restaurant. I had an unbelievable experience, you know, you know, it's in Denver, Colorado, blah, blah, blah. She shows you like the food and the ambiance and that like you're. Here's this person that is like, you can directly connect to that. If you have a bad experience, you're like, oh, well, I don't trust Sarah anymore. Not some, you know, person behind a website that I have no idea who they are. So I think that's just like consumer preferences are changing. It really depends on the industry that you're in. Like if you're in B2B in particular, like we're just seeing podcasting be this unbelievable, like distribution mechanism for these companies. Because you can have the founder of the company or the CEO of the company one hour a week, they sit down and they talk with an industry expert about a topic that they know their customer is going to find interesting and relevant. And that's all the mental energy they have to expend. And then we take that long form piece of content. You chop it up into clips for social. You make blog posts based off of the insights, the key takeaways, you write social posts for LinkedIn and Twitter and wherever else your target customer is and suddenly like your entire, your marketing team has all of the marketing content that they need for an entire week, right? And so I think that like those, these things are how we're seeing kind of the market change, but it really just depends on like the industry that you're in. Like, are you local, like, are you national, are you selling? B2B is a long sales cycle? Is it short? Is it, you know, purchasing decision? And we could talk about like kind of all those different ways that we're seeing these AI tools use. The biggest thing that I would say just to kind of frame this conversation is like for a lot of companies that we talk to, they're like, oh, I'm going to change my whole internal workflow and redesign. It's like this almost massive adoption and we're just like, whoa, whoa, whoa, stop. First off, what are you doing right now? What can we automate away? And a lot of the times 50% of a company's workflow can be automated away with these AI tools. So okay, sweet. That can be spun up in three days, you know, 72 hours. You just automated away, you know, 50% of this like time consuming task. And so that can be a massively, just like instant, you know, impact full thing that these companies can do. And so anyways, I, I always try to kind of preface. This is like, that's like a, a great jumping point for these conversations around this stuff.

So Anyway, you mentioned TikTok and Instagram. What's your experience on Snap?

Yeah, Snap. My problem with Snap is like, it's not like personally, it's not used as a search engine is the biggest thing we've seen really great results on the ad side in particular, like, I mean, just incredibly cheap advertising. There's actually really cheap advertising on Twitter right now. It's kind of a cesspool. But like you can get just like, I mean we've seen some ads that we've ran on Twitter get like 0.0001 cent clicks to the site, right? And it's like the traffic is like not super targeted but like I can't get, you know, first I can get a thousand people to go to a website for like $0.01 basically is the idea. Like that's like mind blowing, right? So with Snap, I think about that like a lot of the times it's like the, the challenge is that the demographics of the, of the site are not as Robust as these other places. Like I can get a really broad spectrum of people on Instagram, on TikTok. They've got, you know, from a psychographic and a demographic standpoint, it's way broader. But I honestly love to learn like if you guys are experiencing like seeing some success on Snap, I think the biggest thing I've seen it be really good at is like lead gen for like Internet money, like how to make like, you know, start your first business, that type of thing. I have some friends that they ran a show on Snap and it was like literally they just interviewed like people about their companies and then it was like into a training program for them. But yeah, just kind of sharing what I know.

Yeah, I was, I was an event last week and big YouTuber was there, a couple big YouTubers were there and they were talking about Snap and that it's one of the easiest places right now where they're making a lot of money totally on their content.

I know their RPM is super high, like the payout that they're getting to the creators because they're trying to secure them away from. Everybody's playing this game, right? It's like why is like elon Musk courting Mr. Beast to come on Twitter and post his shows there? Like he's giving him these exclusive advertising rights and like the reason for that is because they're realizing like if we don't have these mega creators on platform and aren't paying them these high RPMs, like we just don't have the staying power with our user base.

So yeah, yeah, for sure. Another thing that was interesting to me, who would you think the third largest search engine is, man?

I would say Pinterest just off the top of my head, but Amazon, Yeah, totally. I 100% see that. I mean it's where people go like you start in two places to buy products, right? Especially if it's like a transactional activity, like it's Google or Amazon, right? We always looked at Pinterest. I did some work in another space and people, that's where they like plan future purchases. So it's kind of this like you know, up the funnel from like that. I'm not trying to make this purchase decision immediately, but the Amazon totally makes sense. I think that's super interesting actually. I know that a majority of ad spend dollars have gone away from Google shopping ads to their. My friend works for like one of the major Amazon like marketing consulting agencies and she's like, we're. Every year we just get like 20% more of the budget that they have allotted for like paid ad spend is being dedicated to Amazon in particular.

So yeah, yeah, I think so. What was said was you go to Amazon and put in a search for something that you want to create a product around and then don't go to the five star reviews, go to the two and three star reviews because those are the review reviews that are going to tell you what people didn't get from the product that they bought. So it'll identify the gaps in the market just by saying, hey, you know, I wish it would have had more of this or I wish it would have had more of that. And then you can target a product to those users right there. So just interesting stuff, man. It's all, it's all out there. If you, you know, so much more. I don't know if you ever read the book How Google Works. No, I haven't, but it was, it was how basically, you know, how they built Google and they identified very early that they needed creatives. And so like just massive numbers of high level creatives there. And really with AI coming around now. And so now, yeah, it used to be like, okay, well spend 10 years learning how to code, spend all this time learning how to do this. Well now, you know, or even graphic design, right? Kid Rock was on Theo Vaughn and he's like, he's like, let me show you this picture I made. I asked AI to make me with this, with that and the other thing. And it was beautiful picture, right? And it took, it took a minute. And so now it's coming back to the creatives who are, you know, who know how to ask the right questions or who can, who can think, you know, more broadly about how to use these tools. And so because you know what, what I think Google needs is they need more money. They need more money. They don't have enough money. So it's playing, I don't know what.

Is going on over there. It's like the biggest fumbling of the best bag I've ever seen. It's actually crazy like talking about, just like with this whole AI piece, like Google has been years ahead of all of these companies and right now they can't release a product that's stable that we can even use. And we build on, on top of all of these platforms, right? Like anything that we can get our hands on, we're incorporating into the tools that we're building. And like, it's the most insane thing to me. Like this is like Google is one of the most profitable money printing companies that's ever existed. And they just have so much atrophy at the company that they can't actually get products out the door that like consumers like really want to use. And it's just crazy to see this happen. And these start. I mean, it's classic though, right? Where it's like you have a smaller, nimbler company that can navigate and change course like way faster. And so because of that they end up winning. But it just, it's insane to me because, like they've had DeepMind for years. They've had all of these sub organizations within the larger ecosystem of Alphabet and Google and they've almost just hamstringed all of them until now. When they're seeing this existential crisis, I think it's going to be written about. If they don't in the next two years solidify themselves as being one of these big players, their market share is going to be absolutely eaten up and eroded. So that's just personal opinion based off of what we're seeing on the front lines.

But yeah, well, so fascina. Fascinating that you say that. So I've started using a tool called Perplexity.

So good.

Are you?

No, yeah, yeah, it's so good. We're all about it. Yeah, yeah.

I mean, I thought you said no good. And I was like, well, oh no.

Yeah, I can't use that anymore. No, it's great.

But like, it's great, you know, and basically what it does is it, it takes the searches, but then it makes sense of the searches. So if you're like, if I was researching you on Perplexity, which somebody did, and I got this great write out on you and it was, and it wasn't just a bunch of rando facts. Right? It was a. It's, it's really. How do you use it?

Yeah, we use it for a lot of like, similar stuff, like research or just like aggregating information. Like, for example, I'm doing some consulting for a company right now, they're in the fine arts space and they wanted to basically create a directory of like, I think it's close to like 20,000 galleries that are in the United States that sell fine art and then like make a landing page for each of them. So we're using Perplexity's API to basically like collect all the information that we can find related to the gallery's name, pulling that back into a database and then writing a blog post based off of the data that we collected from Perplexity. And what we found is that like, I Mean, that used to be, if I wanted to do this, like, with a team of people that's like 50 people all offshore that I'm managing, like, it would probably take six to 12 months and we're going to do this in like 30 days for them. I mean, it's the most. Like, even saying it feels unreal. Like, it feels like cheat codes. So. Yeah.

Yes. Yeah. I mean, but like, that goes right back to the point, is that. And here we are just cracking the door into broadly distributed AI 100 and people, creatives, and then people that are in the right position are taking advantage of it in these ways. And if you're a. If you're a $10 million H vac business and you've been doing it the same way, you don't have a bunch of creatives, you know, you're using standard tools, you run a tight business. It's. It's a nice, profitable business. I think things are just going to start getting away from people and they're not going to figure out. They're not going to be able to figure out how. How they're not winning anymore on some of these things.

I think that's a part of it. I mean, with perplexity, like, what we've seen, because we've been thinking about, like, okay, how do you rank on perplexity, right? Like, how do I show up in the result for, like, somebody does a perplexity search. And so the early data that we found, and it's still like early days with all this, but what we found is that, like, the larger, like, digital mass you have on the Internet, the more likely you are to show up in these searches for the AI, right? So when we think about how the AI is trained, it's basically just like scraping all of the available Internet or all of the available information on the Internet and then has this massive corpus of data that it's training on top of. And so when you get like a search result back, say when you're using perplexity, what it's doing is it's basically pulling its general knowledge of information from everything that it scraped and then reporting, you know, hey, we think these five companies are the best results for, you know, XYZ location, right? So how do you show up more often, like, in that search, when somebody searches, you know, H Vac Denver as an example, it's just like having more content out there than everybody else. Like about, you know, doing H Vac or anything in related that's in that, like, you know, to that. To that to that, to that niche. And so that's what we were just seeing in the data. And it makes sense kind of logically like based off of how these AI models are trained. Like of course, like if you imagine it like you have 50, you have 100% of a database and 50% of that is based off of XYZ company. And then you know, you make a query related to that and there's these other 10, you know, or we'll say five companies that take up that other 50%. So they're each like 10% of that source data. Of course it's going to reference the one that's 50% that it has the most training on top of. Right. Like that's where it's going to go. Because it looks at me, it looks at the information that's available in public for that company as being the measurement of trust that that company is real. And like it should report that as being the best answer to the search results. So anyways, still super early, but something I wanted to throw out there.

Yeah. So what's the path to profitability for a company like Perplexity? I just started hearing about it a couple of candidly, a couple of really high end conferences where executives were talking about how they're saving time and you know, everyone ever. Is there a conference that exists now without your top 10 AI tools that you're using?

I mean it's every industry because I, yeah, it feels like a, you know, like a meteor hitting earth event. Right. Like it's like every, I think in for a lot of people, like they're super afraid, like is this going to replace my job? Is this going to make my company, you know, go under like these type of. I don't think it's happening that fast like just based off of what we're seeing. Especially in like our enterprise, like sales, they like want to use this. They're still trying to figure out how to use it, how to get their teams to use it. And the biggest thing is the lawyers. Right. We've been in the sales deal or a cycle right now for nine months with this company that you would know the name of. And for them, we've been talking to their lawyers just like, okay, who owns the data? How is the data trained on where does the data go? Show me your data pipelines, all of these things. Things. Right. And so I think that there's going to be a way slower period of adoption for these companies just based off of like the interactions that we're having. I could be totally Wrong. But that, that's kind of the what we're seeing on, on, on from our, you know, sales flows.

So yeah, lawyers will slow everything down 100%. But so, so, but back to you. So the connection I'm trying to get to is Perplexity. AI is free.

Yep.

And I don't know if they have a paid. But you just go right to it and you, you, you put in the information, you ask the right question, you get a great answer. So then is it now they're gonna, so people are saying, I'm using this instead of Google. So now you're gonna get a mass of people using this. And then, and then it's, Google's not paid, it's still free. But then the ads are going to come and then people are going to be paying to get themselves to show up there and it's just going to be Google, you know, 2.0.

I think that's probably the path forward. What we've been seeing, like ads are really not good from what we've seen within Chatbot uis.

Right.

There's a lot of, we have some friends building companies like super small scale scale and compared in comparison to Perplexity. But like what they found is it's like it's hard to serve ads in the traditional way where it's like a text ad based off of like whatever the question keyword that people asked. So what they found to be more effective is having affiliate relationships with like companies so that when they send that traffic to the site, they get like some type of commission based off of the traffic that they send. So I could see something like for Perplexity where that's how this relationship is. It's like, okay, yeah, it's pulling this general knowledge from the Internet and then it has all of these affiliate like relationships with every company or like every tool or service or software that it like shows up. And then any of that traffic that it sends off of its platform to you, you pay them back some type of like commission. Right.

Okay, so just like a toll booth. Just like, just like kind of like it used to be like a pay. Not a pay per click, but a.

Pay per per action or whatever.

Yeah, kind of. Yeah, kind of pay per visit.

Totally. And I think that.

And then you can bid it up. Yeah, then you bid it up and then they'll serve you. They'll serve you up more.

Totally, totally. But I, you know, I hope that's where it goes personally because I just think that like I don't, I don't Know if I love the ads, like I think there's a better way to do these types of ads. Like, like if I could have perplexity convincing somebody that like my product is the best product for the query that they just made. So it's like almost doing sales for me and it's like doing that in real time, like in a chatbot format. Like I would pay all day long for that. Right, right. And it has the capability to be able to do that. So what's stopping you know, a huge like, you know, like anybody, you know, anything from like Procter and Gamble, like what's the best? Blah, blah, blah, Toothpaste. Right. And, or like you know, the best all natural toothpaste. And then it's upselling a brand that Procter and Gamble owns. Right. Like that this is the best, you know, like that, that is something that I could see definitely happening like with the tool and it's like current form, but I don't really know. I think right now like what, you know, the rumblings we're hearing with all these AI companies is there's so much funding and liquidity to put money into them that they're just like get as big as you can as fast as you can. So like perplexity is growing at like a stupid rate right now.

Right.

And they're just, I mean they raised like another round of funding. Like I think it was like four months after their previous round, which is like unheard of in venture capital. And it's like largely because they're like, yo, we just like we're growing so fast. Like we have like we can't even get the lights on without more money. Right, right. And so for them like it, it's just going to be like how much of Google's market share can they eat up? And then once they kind of start to see that plateau, like that hockey stick growth is starting to like, you know, reach its ceiling, then at that point it's figure out profitability. And Google did the same thing, right? Like they, they bought Double Click. It was like, I want to say like probably four or five years into their initial, like their initial launch. Like they weren't even profitable for the first like five years until they introduced, you know, bidding on self service, bidding on keywords and then you could actually show those ads up. But that was an acquisition of a separate company that they basically rolled into them. So I think that we're going to see, you know, history repeats itself. Time is just a flat circle and we're going to see that kind of same thing occur here. It might be a little bit different, but there's only so many ways to make money on the Internet and like ads is one of the big ones. Right. So. Yeah.

You know what's interesting though, if you look at each, I mean, ultimately there's not a lot of winners.

100%.

I mean, I mean it's, you know.

Pepsi and Coke, right?

Huh? Yeah.

Market share. Yeah.

You got YouTube, you got Google, you got Facebook, you got Instagram, you got, you know, then what? Right, There's. So we'll see if, you know, AI gives some people a toehold who can break out first and maybe, you know, get a big piece, grab a, you know, bite a big hunk of Google and. And then they'll buy it and kill it 100%.

That's probably what will happen. I know Google is talking about releasing, it's called sge. I think it's like search Engine engagement. I can't remember what it stands for, but it's basically their version of perplexity where it's like.

Sure.

Pulling in the results and then citing sources and that whole thing. Yeah. Again, it's not live. It's been eight months of them talking about this like, where are we? So, and in that time frame, Perplexity is, you know, they've triple, tripled in the last, I think six months.

So well, either be, either be the best or get in the best way. And they'll, they'll open the closet, they'll just push vine over to the side, put it right in there and close the door for another decade. But it's ridiculous. But so, hey, well Cody, tell us what you're doing now.

Yeah, yeah. So right now we're building this company called Swell AI. It helps you turn long form content like videos, audio into blog posts, newsletters, clips for social. We're also just starting to launch social scheduling. This will be out in about a week or two. Basically like how we're seeing companies use us is they build their entire kind of like content marketing engine on top of us. So like our goal is basically instead of you needing 10 different content people to run all these different sources, you can have one 10x marketer and they're just sitting on top of the tool like this, basically creating all of the content that you could ever want for your brand. So we work with, we actually started out with podcasting. That was like the initial kind of wedge that we did into the market. We're starting to work now more and more with like traditional media broadcasting that type of thing. And then also with like B2B companies that are, all of them are spinning up some type of, like, show that's talking about the industry, and we're helping them kind of do that. But our big thing, what we're focused on, is like helping people build workflows, where the analogy I keep using is like, hey, provide raw materials to the AI, which is in the form of, you know, a 45 minute video. And then from that you have templates that are created that just automatically get generated based off of that source file. So the analogy is take crude oil and we turn it into plastics and gasoline and diesel and aggregate and all these other forms that you could want out of that, out of that initial piece of content.

All right, so I have a specific. If you're not a podcaster, you can put your head down on your desk because this is a specific question. We'll come back in five minutes. But all right, so generating AI content out of a long form piece of content and getting the hook, right, are there challenges with that or what's the solution for that?

Yeah, so the biggest challenge is that the AI is trained off of the general information of these social media platforms, right? So when we think about, like the average post that happens on social, like, it's pretty bad. So, like, think about it as like a bell curve, right? And it's like you're pretty far down the, like, is this good content scale? So the challenge with it is basically like helping it identify, like, what good content is, like, within the episode. So the trick, there's kind of two ways to do this. And this is like, what we're doing and what we're trying to do in the future. So the first way is basically asking, hey, like, what's a key insight from this episode? It's pretty good at identifying that. And then we say, okay, now that you found that key insight, go find me a hook and an insight that's related to that key insight from this transcript and then return back the timestamps of that information that you just found. So almost breaking it up into steps. Because if I just ask it, like, go find me a hook, like within the transcript, it will bring back the most random, irrelevant information.

Yeah.

And so by saying, like, hey, like, what's the key takeaways list? Five key takeaways. Now go find me quotes from this transcript based off of those. For each of these five key takeaways that support these five key takeaways and return the transcript, or to return the timestamps of the like words from that, from that quote, we can then take those timestamps, you then go find the clips, it automatically goes and finds those clips and then basically, you know, that's what gets serve. That's kind of the hacky way to do this right now where all of this is headed and where it's going to go is in the future like somebody's going to train a model based off of only the top performing clips that are on these social media platforms. So the data set that goes into it is not just the general knowledge of that entire platform, it's just like the top 10%.

Yeah, it's curated winners.

Exactly, exactly. And so this is the same thing that we saw with like AI SEO. So like we did, we ended up spending this company out of swell called Draft Horse AI. It's like a programmatic SEO writer. Basically you give it a list of target keywords and it writes like hundreds of blog posts based off that. So how like what it's doing is writing from the general knowledge of the like AI, which is again the general knowledge. It's not like the like you know, the thought leadership or the leading top percent. So what we found is that if you actually will like go and give it constraints on what it can write from. So like hey, here's like this keyword. So say we're trying to rank for the keyword phrase like, like Hotel Marketing Strategies 2024. So you go find what's ranking on the top the front page for like the, you know, the top five search results for that keyword. You extract all the text from those and then you say okay, AI now write based off of this source data that we already know is ranking on Google. And we've seen that we can get some of these blog posts and landing pages to go to page one in 72 hours. But you could scale that up where it's like hey, instead of training off of the general knowledge of the Internet, I'm just going to train it off of the front page of Google because that's only the winners. So now you have the structure of, here's how a blog post, like a winning blog post is supposed to be structured. Everything that we create, everything that you write from now on, that's a blog post is basically fine tuned or trained off of that. So that's how we see all this content stuff heading. Everybody's using these general models because it's extremely expensive to build custom AIs. Our friends own this company, it's called SF Compute. If you go to sf compute.com, it's like the dumbest website ever. It's just like text. But they're basically what they do is they sell training time on GPUs. GPUs are basically like the compute that's necessary to train these models. And the training time is becoming more and more pricey because everybody's trying to do this, right? But if I wanted to go train that model that I'm talking about, that's like the. Find all of the top clips. That's like a million five. Like, that's close to the quote we got for being able to do that, right? So, and that's just like a single very small use case, right? It's just like sharing viral clips and then thinking about, okay, now I have to do this for every, every one of these subsets. So what's probably going to happen is like, companies will come out and they'll train these models and then they'll give access to other companies like us that like, we basically like an API that we can call to like find those like viral clips or you know, do social media, data analytics or you know, whatever that is. And then they'll just incorporate the, they're called, we're calling them like agents. Like, they'll incorporate these like digital like robot thinkers into their applications through these like other companies that, that have trained on that, that specific data set that like provides like the, the output outcome that people are looking for. So there's a lot of information. Sorry for the monologue as well. I always.

No, no, man, that's fascinating because what it, it comes down to. So our experience is we've been doing the podcast almost a year and by the way, just cracked 100 on Apple Business and top 25 for entrepreneurship.

So congrats, man. That's huge. Those categories are so competitive, it's crazy.

Yeah, man. And something's going on in Great Britain over in the uk. I don't know what's going on over there, but like we, we, you know, we. So our. In Q2 here, we're going to be looking at international markets and seeing what we need to do to just go over there. But because we, we just have some random markets that have just picked up the show. So I love it.

I love it.

Yeah, man. But like, so we were doing this, you know, small, small, small and mighty team and we're, we're early and we're going through the episodes or they're going through the episodes and they're picking the clips that are of interest. To them. And it's like, that's a good clip. That's a hot take. And, and then, you know, as we, we had to. We really needed to double capacity to continue to do the things we're doing. So we go to an outsource and we've got. Which is, you know, and, and who's more sophisticated than us. And we're doing the, you know, the clips and, and using AI. And we were using AI, but were. Sometimes the clips don't make sense.

Totally.

You know, they're just because it's like, it's a good clip, it's tight, it's. It's a conversation, it's a contained conversation. But, you know, unless we. No human labored over it, over it to see if it was really interesting or not. And there's. So your AI will serve those types of clips up, but yet, you know, the band, the personal bandwidth to actually have people going through these things and saying that it's unrealistic. So, you know, when will the AI really be able to cross into that, you know, I'll misuse a word into that singularity to where now it's actually smart enough to consider these things whether they're interesting or not.

Totally. And I think it's going to come down to, like, there's a lot of, like, everybody's trying to solve this problem. So we just talked to this researcher out of, I want to say, Oxford, you know, way above my pay grade. There's way smarter than us on like this type of stuff.

Sure.

What they're doing is basically combining like linguistic AI models with just the general AI models that we're seeing be the most popular, like ChatGPT, Claude, et cetera. So what that ends up looking like is like they can use a linguistic model to find what they call, like, entities. So an entity is like an idea within a corpus of text. Right. So that can be like, oh, here's a cohesive thought. And then they go, okay, now AI, like, is there a hook within this cohesive thought? And they're like, yes, no, there's a hook. Okay, is there an insight within this cohesive thought? Yes. No, there is. And then that basically, like, say they started with 100, that filters out 15 of them. And so it's really. What it's looking like is going to happen is like the layering of all of these things together. The universal model. I think we're just like, we're probably just like two years out from this. I say all this, right. And like GPT5 could come out tomorrow and, like, my business basically evaporates overnight. Like, we have no idea what's going on. Right. Like, that's kind of like the world that we're actually living in with all this stuff. Yeah. So with that said, though, I think what we're seeing a lot of companies do is they are basically like layering all of these different models together that do some, like, small task of that entire process and then, you know, linking those together and it makes sense. Right? Like, how does a human work? Right. It's like, cool. Here's my checklist of 10 things. Like, this is how I go through. Like, how do I systematize and make a, you know, like a decision tree. Yeah, exactly.

You go down the decision tree. Like, okay, we got. We went from here to here, and this makes sense. Now we're going to refine it this way. Then we're going to lay it against this, and we're going to lay. So you use. You. You layer multiple technologies on top of it. And, you know, like, we're getting. With our, you know, like our. With our expanded team now, which is great, we're getting just as many good clips as we did before, but we're getting three times or four times as many. And, you know, not all of them are winners, but like I just said, we've just. We've cracked up into the top 25 and the top 100. So, like, it's. It's working. And we wouldn't have been able to do it doing it the way we did it. So now the question is, is, all right, well, if this is good, what does great look like?

100.

And you know, how does. How can everything be. Be a winner? Does. Do we have to go back and in. In, you know, interject that 10x marketing human in there and say, oh, if we just change this hook to say this, then it's going to be a bet. Because it's all. I said this. You know, I had these YouTubers at this conference, man, and it was just like. And they were giving examples of like, what, you know, one guy, any. Any channel he puts up, goes to 500, 000 subscribers. And he's like, here's. And it was. In 10 minutes, he's like, here's how I do it. And it was just. And it was just right on, man. It was just like, right. I mean, it's like. And it's. It's the hook and then this and then it's that. And if people watch it more than once and then you get this and that and the other thing and they, you know, it's down to a science. But like, I think Mr. Beast has just to produce what he's got. He's got over 100 employees.

100%.

Yeah.

So, and like that's, that's a level.

And they're maximizing tools probably 100%.

100%. Yeah, yeah. I think with like where again, just what, where we're seeing this go is like there's always going to be a human in the loop that is just like has some type of intuition. Like the best people that we know that are doing content, like, yeah, they may use AI tools, but they're just using it for like, cool. I have an hour and a half of video. Like, give me five minutes that I can look through and pull out hooks and insights. Right. Or like pull out those, like those, those highlighted sections. There's a likelihood like that it could get. Again, we don't. With all this, I have no idea. Right. Like again, GPT5 could come out tomorrow. I could give it a transcript with timestamps and it just pulls out the best like most banger clips that you could ever find. So. Right, that, that, that could happen, you know, like very soon. It's just super costly and it, for a lot of, like, for a lot of companies, it doesn't make sense to train a model off of that, like just that specific thing. So again, we could go like kind of in the weeds now with this. So but I think the thing that I'm realizing more and more with all this is click volume is this huge thing. So we're seeing some of our top performing clients. What they're doing is they're creating multiple accounts across multiple of the channels. So imagine 30 Instagram accounts, 30 TikTok accounts, 30 YouTube shorts accounts, and then they're doing just combinations of clips for all of those. And then they see, oh, this one, you know, clip 187 went viral on TikTok on this channel. Let's take that. We're going to remix it a little bit and let's distribute that across all of the TikTok channels or all of the, you know, Instagram reels. And so there's like this compound learning that's happening but now does, well, does swell.

Give them those capabilities.

Yeah, they can go and make clips like so basically like what we built into it is like a text editor that I can select a section and then it, I hit create clip and the clip is made and then it's like, okay, I want to auto crop that so that it goes into a portrait. Sorry, it goes from landscape into portrait. Cool. That's done. So they can come back. Like how we're viewing this is like you're going to have this library of content that you sit on top of. We're building the social scheduling feature so that we can pull in the data of best performing content because then we can create like a learning viral loop. So we saw like, oh, like here's the top performing clips from all of March, the top 10%. Like, they're all about these ideas, like, what's the insight? Et cetera. All right, let's go back to our back catalog. Let's find from our 500 episodes other clips that are similar to the most popular ones. And let's go make those right. And then we can just surface those up. We're imagining it as almost like this Pinterest dashboard that just based off of the data that you're. Because with all this, I mean, that's what the human is doing, right? Like the best social media managers I know, and this is what I used to do when we were at Rupa Ha Health is like, we would publish a ton of content, we would see what was working, we would go back to our, our back catalog and we would find more content like that. And then that creates this growth loop. We're just trying to do that same thing, but with these AI tools that now suddenly, like, I don't need 25 people to run this whole, you know, content production marketing department. Now I can do that with maybe one or two people or a team of, you know, three or whatever. But yeah, so we're so early on this stuff. Like, we just added clips like three months ago. Like, we, for a while we were like, yo, we're just going to only focus on writing content. But our customers kept asking for this feature and now they're asking for social. And then the other thing that they're looking for is like SMS and email newsletter marketing. We're seeing SMS for podcasts. The open rates be like 90% and like the clear rates be like 40%. So sites like Community, they're like hacking together versions of this. So Community is one basically like any of these, like bulk SMS newsletter tools. We're just like, what people have asked for is like, hey, the ability to build that into. That's. That's kind of on our product roadmap as well. But anyways, a lot of inside baseball there.

But yeah, yeah, man. No, that's fantastic. Well, I'm tracking who's a great customer for. For swell.

Yeah.

Customer.

Yeah. One of our larger ones is Wondery. We just started working with their team. We're pretty excited about that. What are they?

What kind of business are they?

Yeah, so they. They're an Amazon company. They work with a lot of, like, Audible and like, their production. They also, like, have some of the larger podcasts in the US okay, so.

Like online retail promotion.

Yeah, so they make like, a lot of like, custom shows. Like, they'll do like paired shows, for example, with like, Disney or something like that, and then they, you know, sell ad space on top of it. Yeah, so that's one of our larger clients that we're starting to work with. And then we work with like a lot of just like SMBs, a lot of kind of mid range podcasts as well. It's like a broad spectrum between, you know, solo podcasters up to like podcast networks. It's like on the podcasting side, so.

All right, well, you want to take a look at the home front and see what you can do for us?

Yeah, I would love to help you guys however I can. So we'll get some locked in with your team. That'd be fun to do. So.

All right, well, so I can. For anybody else who wants to come on the podcast, Cody got one lead from this podcast already, so here it is.

Awesome. Jeff, thank you for having me, man. I really appreciate it.

Yeah, yeah, 100%. So a couple, couple final questions here. If you had. You know, I'm always, we're always interested in informing people's entrepreneurial journey. So if there's somebody right now that's out there thinking about a career in this or building a business or something like that, if you had one sentence to speak into their life, what would that be?

Whoa, tough one. The one sentence. I don't know. But my framework for all of this now is like, find some skill that you can teach yourself online, that you can sell online. Do some type of consulting with that. So, like, learn how to make podcast clips as an example. Then you build an agency where it's like, hey, unlimited podcast clips for $1,000 a month. Go get 30 clients and you're making 30 grand a month. And then from that, like, parlay it into, okay, how do I build some type of business that has, like, whether it's a course or a software, where it's like a build once, sell, you know, a thousand times. Looking back on like my own personal, I guess, like business journey, like, you know, in hindsight, very obvious. That's what happened, right? Like, I look at what you're, you know, you're talking about what your kid's doing. Like, I think that that is going to lead to a massive company. Like, he understands, like supply and demand. How do I do product research, customer service, all of these things, right? And it's like the small scale thing, but all those learnings are going to be parlayed into the next venture, which is going to be bigger, which is going to be parlayed into the next venture, which is going to be bigger. And so I think that's the biggest thing for like, a lot of people. Like, they try to go for this like moonshot day one, and it just like kills the entrepreneurial spirit, right? Like, they go and they're like, I'm going to build a rocket company or whatever, right? Like just that it's a ridiculous one, but that's like kind of the level that they go to. But in contrast, it's like you're talking about like a window washing company or a power washing company. Like, you can start something like that where it's just you, you have no employees and easily get that to 10 grand a month, which is like hustling for two years. And then at that point, okay, you have cash flow, you can go build that next thing where you have some people that are making you money. And then from there, okay, how do I then go build a product that I can build once, sell a thousand times. Right. And do that over and over again. And that to me is the evolution of this. So that's what I'd love to leave with people is like, kind of like if you're trying to start, that's the way that I would do it. Looking back on just my kind of, you know, experience and journey.

So, yeah, get into the game. Get in the game. Well, Cody, this has been awesome. I've learned a lot and I really appreciate your perspective on the rapidly changing marketing environment with AI how can people get in touch with you?

Yeah, so I'm active on Twitter and LinkedIn. If you Google Cody Schneider, I should come up. If I'm not, I'm doing my job really terribly, terribly. And then our company is swellai.com you can go and sign up for free and take it for a test drive. So those are kind of the core places that you can get in touch with me.

So, Cody Schneider, thank you so much for being on.

Jeff, thanks for having me. Really appreciate it.

Absolutely. Everybody, thank you for listening. This has been Jeff Duden and we have been on the home front.