Balancing AI and Human Insight | John Horn

David McBee: Hello and welcome to Simpli.fi TV, the web series and podcast for agencies, marketers, media buyers, and business owners. I am David McBee. Our guest today is John Horn, CEO of StubGroup, a digital advertising agency. John has been in the marketing field for close to 15 years. He has created digital advertising courses studied by over 100,000 students, and his YouTube videos have received over two and a half million views. In his spare time, John enjoys spending time with his wife and two little boys reading and traveling. John, welcome to Simpli.fi TV. John Horn: David, thank you so much for having me. David McBee: Yeah, I'm so happy you're here. We're going to talk AI, which everyone loves to talk about. So 2024 kind of was like the year of how the heck do I even leverage AI? And that's been especially true for marketers. So my first question to you is, can you tell us how AI is changing the way advertisers approach targeting and personalization in the digital landscape? John Horn: Yeah, there's so many ways I could go with that question. Let me try and give you one story as an example. So we had a business that came to us recently. They're in the financial lending space and they're saying, "Hey, we've been doing Google Ads for a long time. All of our competitors do Google Ads and it's really hard. We're always breaking even at best, and then trying to find other ways to upsell and cross sell and try to make some profit. But then there's this other business that's come and they're doing amazing. They've got like 40% margins and how are they doing that?" And so we start doing some reverse engineering digging into things with them and pretty confident that what they're doing is they are using AI better. And specifically what they believe they're doing is they are importing data into Google, not just about what leads are coming in the door, but which leads are turning into orders, into paying customers. And Google's AI is really good at then taking that information saying, "Let me get more people like that," or, "Let me not bid on the people who maybe are filling out the initial form but not turning into a customer." And so that's just one microcosm example, but there's so many places where people are leveraging really good down the funnel data and then using AI to find the pattern and do the pattern recognition that can make those campaigns more successful. David McBee: So instead of the old tactics of optimizing to a click or optimizing to a form completion, they're going the next step and optimizing to actual customer using AI. John Horn: 100%. David McBee: Yeah, that's kind of brilliant. So how would you say AI is enhancing the customer targeting to make the ads feel more relevant and less intrusive? John Horn: I think what we're seeing there, maybe I'll use Meta as an example, is that the AI is figuring out what types of content people are engaging with and therefore giving them more of that and doing that in a really compelling fashion, so you've got on the Meta ad side of things, let's say you've got these Advantage Plus campaigns and that is rolled out where the AI takes over most of the targeting. You're no longer going in there and saying, "Okay, I want to target people who own cars in this geography and match these psychographics." It's more of, "Hey, here's our creative Meta and go forth and conquer." And Meta is able to, not always, but in many cases is able to see, "Okay, we're actually targeting the right people because these are the people who are engaging with the ads. And then we're learning from that and showing more and more." And so that kind of personalization or customization of the ad copy is delivered by the AI in a way that the humans just weren't as good at going in and making manual lists. And I think we see that type of thing happening both on Meta as well as other platforms. David McBee: It sounds almost like they're using the same algorithms for the organic things that we see on Meta to show us the ads. Is that kind of what you're saying? John Horn: Yes, I believe so. 100%. David McBee: So I think that's good, unless you're in a microcosm. Everything I do on Instagram has to do with Jeeps and going off-road, but at the same time, I also still clearly want to buy sneakers or something that's not related to that. So I wonder how that plays out with their algorithms. John Horn: I think there it's one of those things where as humans, if we're identifying audiences, we may be missing out on that saying, "Okay, hey, we want to go after people who are interested in sneakers, and you might not be in that audience yet," but as Meta goes through and starts serving these ads to different people, they realize, "Oh wait, hey, David's responding to this ad. Interesting. Let's see what's unique about David and his behavior, his search history. Let's find other people who are similar." And so again, at times, not always perfect, but at times the AI is doing a much better job at finding the right person based upon that than the preconceived notions we have as marketers going into things and trying to tell a system, just go after this because we believe it's going to work best. David McBee: Yeah, I always tell advertisers we start with your instincts. We start with what you think, and then we let the data tell us the real solutions. All right, so AI is obviously very powerful. Where do you see its limitations in advertising and are there areas where human intuition is still playing a crucial role? John Horn: It still has a lot of limitations, a lot of places where humans need to play that role. A lot of examples, maybe one I'll give is on the data input side of things. So like we talked about earlier, AI can be really good at taking the data and extrapolating, but I've seen time and time again where something goes wrong with the data you're feeding into, let's say Google for example, we had a client where they're uploading that offline information into Google and something went wrong, Google glitched, and was no longer seeing the value that we were uploading and the campaigns went haywire because the AI doesn't know that there's a problem. They're like, "Oh, well that's not converting. Let's stop bidding on that. Let's go over here," and it required us as humans to say, "Wait a second, things are going haywire. Why? Oh, it's because of this glitch. Here's how we need to fix it. Here's how we need to massage the data and try and explain to the system what went wrong and how they need to re-optimize." And so you've got to have those human guardrails of looking at things. As well of course, there's so many areas where humans need to be directing the AI to begin with, of which of our services do we want to advertise? Who is our target audience? Where do we fit in the marketplace? And so I'm not scared that AI will replace marketers, but I do love that many of the things we used to do as marketers have been done by AI or being done by AI now, so we can focus on other aspects of our roles. David McBee: Yeah. It kind of reminds me of the old programming term GIGO, like garbage in, garbage out. If you give the AI wrong information, it's just going to, like you said, go haywire. So do you have any favorite AI marketing stories or examples of advertisers who are really leveraging it well and seeing success? John Horn: Yeah, I would say that client, actually, I gave it as an example where things went haywire. When things aren't going haywire, when the data is not garbage that we're putting in, we've seen tremendous improvement in results from back when we used to optimize towards phone calls that lasted a certain amount of time to now we're optimizing towards are those phone calls turning into sales, how much value is aside to those sales and putting that data into Google and Google and using their AI and their machine learning is doing a great job, generally speaking of getting much more efficient calls, giving us the calls that are higher quality and not getting us traffic from those calls that maybe people were on the line for a while, but they weren't turning into sales. David McBee: What kind of customers does StubGroup work with generally? John Horn: A wide variety. Everything from e-commerce to lead generation, local, national, big, small. We don't niche down at this point. We have a pretty wide customer base. David McBee: All right. Good to know. Do you have any favorite AI tools and maybe split that into your favorite free AI tools and your favorite paid AI tools and what they do? John Horn: Yeah, I mean, I'll give the boring answer, which is ChatGPT, both free and paid. I use the paid version, but I mean, it's so powerful, especially with search just being released very recently. They're going to be giving Google a run for their money from the search perspective. And I just find myself more and more often to be like, "Oh, I need this. Cool. Let me plug that into ChatGPT and oh, I've got it." So that's probably the AI tool that I use as a tool the most often. And then in terms of other tools, Perplexity also, I've recently made my first purchase after using Perplexity to research things like, "Okay, yep, cool, that makes sense." And my favorite thing on there too is after I did that, I asked Perplexity, "Hey, give me coupon codes for this site." And I saved 20% off with one search. It gave me all the coupon codes that worked. So kudos to Perplexity. David McBee: What about for creatives or images? John Horn: Midjourney is something we've used a lot here at StubGroup. Some things we like about it, some things we don't. We've used Claude as well for some image generation. I think Midjourney is the one that we've used the most. And also we're starting to use Google's Product Studio, which is a tool they've kind of built into their advertising platform, which can be really good at doing things like changing up backgrounds for product images and things like that and doing that relatively at scale for no additional cost. That's a pretty cool tool. David McBee: Awesome. I hope everyone's writing these down. I'm definitely going to go check out Perplexity. Definitely want to save some money. All right. I guess before we go, what is your best advice for marketers, advertisers, even your fellow agencies when it comes to AI? John Horn: I would say be careful of the buzzword. I see so many vendors, businesses right now who are just adding AI on to what they do and then clients are like, "Oh, well it's magical. It'll work. It's AI," and it's not. And we have to have these conversations. So just be transparent and honest with your advertisers, your clients about what AI is and isn't. And then don't get too caught up in the hype of the vendors who are promising to make your life amazing because of their new AI tool. It may make life better, but it's not magic. So trust but verify. David McBee: I like that. That's really good advice. All right. Well, what is the best way for viewers to learn about StubGroup and maybe reach out to you? John Horn: Best place would be our website, stubgroup.com. David McBee: Are you on LinkedIn? Can users follow you there? John Horn: Yeah, John Horn on LinkedIn. Also our YouTube channel, search StubGroup. We put a lot of free video content out through there. David McBee: Perfect. Well, thank you so much for being my guest today, John. John Horn: Absolutely. Thank you so much for having me. David McBee: And thank you guys for watching Simpli.fi TV. Please help us out with a like, comment, or a review and be sure to follow or subscribe to be informed about new episodes. Simpli.fi TV is sponsored by Simpli.fi, helping you to maximize relevance and multiply results with our industry leading media buying and workflow solutions. For more information, visit simply.fi. Thanks for joining us today. I'm David McBee, be awesome and we'll see you next time.