AI has many use cases for the marketing world, and this week Paul Roetzer and Mike Kaput are talking about AI for advertising.
From allocating ad budgets to optimization, ad versioning, targeting, copywriting, and more, AI can help advertisers and agencies more than many realize.
This episode is the first in a series of AI use cases by marketing discipline, based on chapters of their new book, Marketing Artificial Intelligence: AI, Marketing, and the Future of Business. Paul and Mike talk about case studies including CAT HOLT, Vanguard, and more. Big thanks to our sponsor, AiAdvertising!
[00:05:33] Why is advertising one of the top disciplines for AI in marketing?
[00:09:32] AI is already infused in your paid digital and paid social
[00:15:09] Buying paid digital in 2018…and how AI is changing that process
[00:19:22] HOLT CAT/AiAdvertising example
[00:21:32] Vanguard/Persado example
[00:28:37] Mike and Paul discuss other AI-powered tools for advertisers
Disclaimer: This transcription was written by AI, thanks to Descript, and has not been edited for content.
[00:00:00] Paul Roetzer: Welcome to The Marketing AI Show. The podcast that helps your business grow smarter by making artificial intelligence approachable and actionable. You'll hear from top authors, entrepreneurs, researchers, and executives, as they share case studies, strategies, and technologies that have the power to transform your business and your career.
[00:00:20] My name is Paul Roetzer. I'm the founder of Marketing AI Institute. And I'm your host.
[00:00:29] Thanks for joining us for episode 17 of The Marketing AI Show. Before we get started, I want to take a moment as always to tell you about our show sponsor. AiAdvertising. AiAdvertising is the first campaign performance platform that closes the loop, eliminates guesswork and connects advertising investments to sales revenue.
[00:00:51] They leverage zero first and third-party data to identify customer segments and predict winning creative. There are automated design tools, easily scaled, multiple versions for all ad formats and channels. While the sophisticated measurement capabilities provide user level analytics that continually drive creative optimization and performance.
[00:01:12] AiAdvertising is shifting the paradigm, leveraging AI to fuse marketer and machine, something we talk a lot about. Learn more at aiadvertising.com. Now on to episode 17. And I am joined today again by Mike Kaput , our chief content officer at Marketing AI Institute and my co-author for the upcoming book, Marketing Artificial Intelligence: AI, Marketing, and the Future of Business.
[00:01:42] Welcome back to the show. That's great to be back. Thanks. We're getting good, man. What is it for four or five in a row? I think as far as a month, we did it. We made it, we made a month, like we're locked process, like getting to the gym, you get there like four times in one week and you're, you're like back in the groove.
[00:01:57] So we are, we are back at it. Uh, so today's going to be, uh, usually what we do is we just talk, Mike and I once a week talk about big ideas, news trends, something going on, but for the next couple of months, Yeah. So we think about this. We're going to start drilling into, to content from our book. So the book comes out June 28th.
[00:02:15] It's available for pre-order. Now you can check it out. marketingaibook.com, but the way the book is structured is we start off with the big picture telling the story of Microsoft and Amazon and Google. And their work in AI over the last couple of decades and sort of how that's parlayed into the innovation we're seeing today in language, vision technology, um, gets into language, vision, and prediction.
[00:02:37] It gets into like how to get started with AI, but the middle part of the book is the piloting AI section. And what we did there is. 10 key categories within marketing, advertising, communications, e-commerce social media, data and analytics on and on. And we dedicated a chapter of the book to each of those.
[00:02:57] So if you're an advertising, there is an AI and advertising or advertising an AI chapter of the book, same with e-commerce and content marketing, all these others. And so we thought would be interesting is to just take an episode each episode for the next 10 episodes and drill into each of those areas.
[00:03:11] So today, AI and advertising. And so Mike has written a ton of the content for us on these topics. Over the years, we started the blog in 2016. So we've published, I don't know about last guy, like eight, 900 articles on artificial intelligence, many of them category specific or related to specific areas of marketing.
[00:03:32] So that's. If someone was reading and they do advertising, there there's things that resonate with them. So we try to kind of cover this AI and strategies is what we call it. Um, so the AI for advertising pillar page. So the way we think about our content is, you know, we have pillar pages and then off of that, like that's the main page and it tends to be what.
[00:03:53] 1500 to 3000 words. It's like robust covering of a topic. And then it splits off into like tools and technologies use cases, case studies, expert profiles. So we kind of follow a topic clustering strategy with our blog. I think it's safe to say maybe one of these days. A podcast about our content strategy.
[00:04:12] Um, but the AI for advertising page has over 35,000 views in the last two years, which in, in our site, we get about 24, 20 5,000 visitors a month to the site. So with 35,000 views, Probably in the top 10, I would imagine pages on the site. So we know that there's a lot of interest in this area. So it was kind of natural, not only to build it into the book, but to do an episode dedicated to it.
[00:04:38] So again, if you're interested in content marketing, e-commerce social media communications, uh, SEO, there's going to be upcoming episodes on each of those. Keep coming back each week, if you're interested in those topics. Um, but we want to do today is really just big picture, explore what is going on in advertising.
[00:04:54] What are the opportunities for AI to make it more efficient on lot creativity. And so to do that, I want to kind of turn it over to Mike to get us started. Because again, he's written a lot of the content on this and you say, look, what is going on in advertising today? What are the opportunities that we're seeing for AI and you know, what are some of them we'll kind of drill into like the use case in technology.
[00:05:14] So I'm like give us the high level what's going.
[00:05:16] Mike Kaput: Yeah. Sure. Um, so I think AI is really fundamental to advertising today. You know, like you said, we covered. All sorts of different, um, categories and pillars in marketing from content marketing to sales, SEO. What have you, uh, advertising is really in my mind, one of the top areas that AI is not only useful, but essential and.
[00:05:41] The reason for that is because today's advertising already runs on AI. So we all know that advertising digital advertising has changed from how we used to advertise in the past, but what's changed even since digital started becoming the predominant form of advertising. Is that. The volume of advertising content and the speed at which advertisers must reach consumers in moments that can be, you know, only seconds on a mobile app, on a website.
[00:06:12] What have you, this requires AI to power it. I mean, every programmatic ad exchange, every social media site that allows you to advertise, they all rely on. Almost exclusively on machine learning to do everything from placing an ad in front of a user to regulate in bids in real time to even some of them have started offering suggestions on ad copy on audience segments and on how to structure your.
[00:06:42] To reach the most people in the best way possible. So, you know, Mehta, Facebook, Amazon, Google, any of these people, all use machine learning to serve ads, programmatically, any of the third party exchanges also use AI already. So on the kind of, um, infrastructure side, AI is everywhere in advertising, but I think what a lot of marketers might not always understand immediately is that.
[00:07:09] In order to deal with that volume of content and the volume of data generated by 24 7 real-time advertising at scale, which is what digital advertising is today. You also should be thinking about an honestly, probably need some form of artificial intelligence to really. Uh, keep up so to speak. I mean, it's hard enough, uh, creating a few different ad variations and, um, you know, setting budgets, creating campaign plans.
[00:07:40] I mean, there's plenty of human specialists out there that do advertising reasonably well, but it's not at scale and it's not at the speed or the velocity that today's advertising demands. I mean, once you get beyond a handful of different. Ad creatives audiences across a few different platforms. This starts growing into exponential amounts of variations in your advertising.
[00:08:07] And frankly, it's just impossible for a human to manage. So humans have a very serious limit when it comes to doing advertising well, and I think as more and more people start using smarter technology in their advertising, it's like, how can you even compete against a larger company or a smaller company using AI?
[00:08:27] With just a human team. So we see, you know, today tons of advertising use cases for AI. So it does everything from, you know, creating ads is a huge one. There's AI out there we've talked about that can write ads for you that can create. Ad creative for you. Um, there's systems out there that will actually predict which ads will work before you launched them, which is pretty incredible using the past performance of your campaigns or other campaigns that it has access to.
[00:09:00] And there's even a certain AI systems will actually automatically. Kind of optimize budgets, audiences, and performance so that you kind of, you don't necessarily put your advertising on autopilot, but it's essentially a machine understanding what the machines on the other end of the transaction, uh, want and need in order to succeed at advertising.
[00:09:23] So there's really just a ton of innovation in the space. And there's a ton of use cases people are doing with AI in advertising.
[00:09:32] Paul Roetzer: A couple of things jumped out to me. You know, talking though the one is just that idea that it's already infused. Like if you're buying ads on LinkedIn, on Twitter, on Tik TOK, on Facebook, on Google you're, you're being affected by AI, by the machine learning algorithms that are deciding.
[00:09:51] You know what shows up and when, and who is targeted to, and things like that. Uh, even to the example of our Facebook page managers out there have seen it like, Hey, run this ad, but they create the ad for you. There's not hundreds of Facebook employees sitting there designing ads. They can just use their AI to whip these things up, and then they can build predictive models of what's going to work and how likely you are to buy the ad and run the ad and things like that.
[00:10:13] Like AI is just everywhere within the ad space. So even if you don't go find. AI tools, your a basic understanding of how this stuff works still is almost required to do what you do. The other thing that comes to mind, and this is real for us. And again, we're at the Institute is a five person company. Uh, my agency that I sold last year, we were about 15 employees or so, and we had clients in a lot of different industries that didn't have $50,000 a month budgets.
[00:10:43] They may have 5,000 a month. 500 a month. Like they, they weren't, um, you know, spending enough that we could either justify building a paid practice internally. It wasn't a core competency of ours or the outside partners we would try and bring in to help drive leads. You know, we create all this content and we need paid to drive leads to it.
[00:11:06] Um, they wouldn't touch anything under 10 or 20,000 a month in ad. And so, you know, all these small and medium sized businesses across the country or world that we're, where do you go for help? And so we looked at like, we used to use WordStream was one and they've since actually I think taken away the self service version of it.
[00:11:23] And now you have to work with their consultants, unfortunately, but we used to use WordStream because it was using some machine learning to basically make recommendations to our team. That would improve ads. And so we had people who were not ads specialists who were managing ad campaigns, and we went and found a tool that could give those people the ability to, to be way more effective at managing ad campaigns and not have to go spend six months.
[00:11:50] Doing a bunch of training and get certification. And so like, that's another fundamental reason of why this stuff becomes so critical is many businesses. Don't have the luxury of having specialists who know this stuff cold. Now Mike does a lot of ad spend for us and, and Cathy, you know, our chief growth officer, she also manages a lot of the ad campaigns, but even now, like I'm constantly thinking like, man, there's gotta be better tools for what we're doing.
[00:12:13] Like there's better ad targeting tools, better scaling of creative. And I want to talk about these use cases, but one of the things I've thought about is, you know, we very early on with our trying to understand, explaining how we came with this five Ps of AI, this planning, production, personalization, promotion, and performance, and you can kind of think about AI for advertising that way, all the ways you can use AI to plan better, to better allocate your budget to.
[00:12:39] The content to personalize it and targeted at those levels then to like run the actual ads and management and then to manage the performance and find the insights. So it's like a macro level. If you just think about everything that goes into running an effective ad campaign and how much, like, you know, efficiency probably exists right now, or how much better you could do it, but we all don't have the staff to do it, especially now in the current economy.
[00:13:03] You're not going to be able to go out and spend all this money and hire all these people. So, yeah. And then, and then the other one is, you know, just the realm of where this can all go. And we're going to talk about some really practical use cases here, but just like some bigger ideas, like synthetic actors and.
[00:13:20] It's going to DALL-E. Like we looked at DALL-E and Google just came out with, imagine their version and essence of DALL-E to this week that can create images on the fly. It's like, well, we need stock photography in the future. Like if you're an ad creative and you're spending part of your time trying to find photos.
[00:13:35] Are you gonna be able to just tell the machine, Hey, can you create someone for me at the Cleveland Convention Center, uh, watching a conference, wearing an astronaut outfit? Like, I dunno, like, just get to say whatever you want and the thing's just going to create the image for you. So it's, it's just, it's such a fascinating space.
[00:13:51] And so much money obviously is spent in advertising, um, within the marketing realm, it's, it's just a massive part of the budget. So, um, yeah.
[00:14:00] Mike Kaput: Yeah. It's crazy to me. I, you know, I think there's like a really. A common quote in marketing that it's like a hundred years old. I'm pretty sure. But it's like, um, you know, it's something to the effect of, um, my advertising works.
[00:14:17] I just don't know which half of it or, right, right. So I'm actually pretty surprised that more brands that advertise. Even touching AI because honestly, I don't know at scale, like how you would do it. I assume they have very large pay paid or programmatic teams. Um, and maybe it works well enough, but when I look at some of the use cases like, you know, for instance, allocating advertising budget.
[00:14:47] Sure. I know how to do that. I'm sure that a team of experts knows how to do that far better than me, but artificial intelligence could literally. Allocate that budget in real time. I mean, how much waste is out there, how much performance is being left on the table, even when your ads do work and half the time, if not more, they
[00:15:08] Paul Roetzer: don't.
[00:15:09] I remember a couple of years ago. So Albert, I know as a tech company that we talk about in the book and they've since been acquired, um, I think sometime late last year, but I was in. New York offices. A few years back pre pandemics is by like 2018. And they were showing me the platform, you know, at a time, 2018, there wasn't actually that much really iTech.
[00:15:30] Like you, you were hearing AI from a lot of vendors. They were starting to talk about their AI capabilities, but we would like drill in or ask questions and you realize like, okay, there's not, not a hell of a lot to this product yet. They just have some ideas. And I sat down and looked at algebra. I was like, holy cow.
[00:15:44] You really started to understand. It's like, okay, you got, like, let's say you're a big enterprise and you're spending 50,000 a month. How in the world does a human manage that ad spend where you could be buying ads in thousands of places, tens of thousands of places you can be placing ads. So you've got 50 K a month and you're maybe across six different channels.
[00:16:04] And the human is supposed to somehow each day, look at the performance across all of those channels and allocate. Budget on the fly to the channel that is delivering the greatest value. So what Albert was showing was something like 30,000 iterations in a month that basically the AI, you would upload the creative here's sample headlines, sample images, Seipel body copy, sample offer.
[00:16:28] Mash it up, the AI would run hundreds or thousands of variations of those things. So the humans still did the creative, but the machine would mix and match the creative and find out what was working and in which channels and with which segments, which target audiences within those segments. And it would dynamically move the dollars around potentially while you were sleeping.
[00:16:49] And like now you come back in and you're down to two channels, but you're honed in on these very specific targets with these three versions of the ad. And. Humans can't even do this. It's not replacing the need for humans. It's just dramatically augmenting what we would do. We would spend the 50 K one way or the other, but now the machine is there helping us do this at astronomically hard to have them levels.
[00:17:11] And that's one of the ones where I was like, oh my gosh, like, it's just doing something we couldn't have done without the human and machine together. And so that's the budget ad budget alone to me was one of the early things that just convinced me of the potential with an advertising.
[00:17:26] Mike Kaput: Yeah, it's kind of interesting to Albert came up when we were researching the book because we talked a few years ago with a company I think was out of Australia.
[00:17:36] They basically sold like, imagine like Groupon for travel experiences. They were like, okay. Within Australia and New Zealand, um, here's a 50% off coupon for like an adventure or a hike or pain gliding, whatever. And the founder of that company. Did everything you just mentioned with Albert, but then there was also this really crazy story she had about how Albert, since it was just following data, saw that a huge segment of people were buying.
[00:18:08] So it allocated more ads towards them because the ads were working. But these were a segment of customers that the company didn't know it had. Um, it was someone that they had not even considered advertising to. So. That's kind of the crazy thing. We're not even just talking about like speed and scale of what we do already as humans.
[00:18:29] We're also talking about opening up. I think some really interesting capabilities that just completely were
[00:18:36] Paul Roetzer: impossible before. Yeah. What are some other use case? I know one that I always liked is a. Does it sell? Tra was that the company we did that in action with, so you create an ad in one variation, and then rather than the human having to now crop images and adjust it to fit the dimensions of a different channel, the AI, you just feed it, all the dimensions and it automatically goes in and makes all the adjustments, changes, fonts, changes, images, moves them around crops things.
[00:19:02] And so you can take one ad in theory and create a hundred variations on the fly. Like that's just a. It's like, oh, that's awesome. If that's what you do all day is sit around creating ad variations. You can immediately see the value of that, that one.
[00:19:19] Mike Kaput: Yeah, I think I'm also a pretty cool one. You know, that when you had mentioned it, obviously, and, uh, at the beginning of the podcast is, you know, our partners, AI advertising, um, one of their stories that I really liked that they had, um, They the sexually really interesting use of even advertising, I didn't even think of is that they basically, so they do something similar to like an Albert where they optimize, uh, use AI to optimize your ad campaigns.
[00:19:48] And they worked with like a equipment distributor to run ads. That were personalized to appeal to top candidates for open positions at the, uh, equipment manufacturer. And this was further back before there were so many problems with hiring people even, and they were actually able to like lower their cost per hire by about 20%.
[00:20:15] And they were able to. Influence like almost half the people that ended up signing on were influenced by the ads. And the reason they were is because AI was able to actually mix and match creative messaging. What have you to appeal to different types of people and to use whatever message would resonate most,
[00:20:35] Paul Roetzer: there is a.
[00:20:36] Lincoln the show notes, but Kevin Myers, the CEO did a really cool AI in action webinar, where he actually showed us the behind the scenes of the platform. And that's, what's interesting about their building is the vision Kevin had and what their team had is to build a platform across all these use cases.
[00:20:51] So, so often with AI, you're buying a narrow feature. You know, somebody who's built something to do a very specific thing. Cause that's what AI is trained to do is very specific tasks. But if you have a vision to build a larger platform where AI is infused into all of these different stages, then all of a sudden you can actually do your planning, your budget allocation, your targeting of audiences, persona building, uh, D ad creative and testing performance and insights.
[00:21:15] You can look at totality and that's what they're focused on building is kind of this all-in-one solution. Um, but, uh, another one I know you've mentioned is Persado Mike, and they're another one of our partners, but, um, that was a Vanguard that they, that a case study with.
[00:21:32] Mike Kaput: Yep. Yeah, they did, um, Vanguard, which is, uh, if you don't know, like one of the world's largest investment firms, um, basically faced a really unique challenge to their industry.
[00:21:43] So obviously financial services is pretty regulated and they were really regulated about what they could and couldn't say, and. So basically, that's why I think a lot of financial services ads are really boring and all seem the same as because the can't differentiate because we can't do a lot of things that you would normally do.
[00:22:04] Now. Persado has really incredible AI technology that actually personalizes. Uh, messages in each ad or, you know, social media posts to resonate most with each individual consumer based on what, uh, the system knows about them. And so Vanguard was actually able to do that to just, you know, almost micro or hyper-personalized these ads and use that as a.
[00:22:30] To develop a competitive advantage, even though they couldn't really, you know, uh, think outside the box with the actual ad itself. And they actually raised their conversion rates by about 15% doing that. So it was pretty
[00:22:43] Paul Roetzer: incredible. Yeah. And I think, you know, a recurring theme people here, they listen to us or read our stuff is personalization and prediction.
[00:22:51] Those are words that just come out of our mouth all the time because of this. In the end, like that's largely what AI is doing. And especially in these cases, you're trying to personalize down to an individual level, whether it's the content you create or the ad you run, or the social media shares, uh, the, the sales outreach, whatever it is, you're trying to understand that individual and then tailored to them.
[00:23:09] And AI helps you do that at scale. And then the prediction is prediction of anything, anywhere where you can have inputs like data going in and a desired output, like a prediction you're trying to make like customer churn, customer acquisition. Um, conversion rates, click through rates, open rates, whatever it may be.
[00:23:26] AI helps you make much better predictions and then the machine learns from the data itself. So it can reallocate budget and move it across channels because it's learning in real time and evolving what it does. It doesn't need a human to write rules to tell it what to do. And that's the real key of the machine learning side as being.
[00:23:45] Better predictions and more predictions.
[00:23:48] Mike Kaput: Yeah. I think what's really interesting about that too, is that, you know, I I'm sure there are plenty of marketers out there that say, okay, this sounds really great. Uh, we definitely could use this, but you know, maybe next quarter, maybe we need a couple more people on staff or we, you know, we just don't have time for this right now.
[00:24:03] It's like, that's a valid concern, but it's like going back to that Persado example. How would you compete with somebody using this tool? If you are manually doing all of this, you can't, it's not even close. It's like, I almost think it's like, people need to realize that some of this marketing, AI technology out there, it's almost an arms race.
[00:24:27] Like you need to adopt this because if you do not, your competitors will. Yeah. There's just no contest by the end of the day, if they're able to hyper personalized ads to millions of people, and you're sitting there in a meeting trying to come up for a week with like three different creative variations.
[00:24:44] Like I just don't think it's going to work.
[00:24:47] Paul Roetzer: Yeah. I mean, no offense to, you know, the people who've made their career to an ad creative. Like it's an art form, I guess, but sometimes. You just need to test a lot of variations and to your point that you've spent a month or three months coming up with that perfect tagline and the perfect creative and the perfect visuals.
[00:25:09] And then what if it doesn't work when you know, the three months? So what it's doing, and if you listen to the lost and gained episode, we talk about what's lost as AI takes, you know, an increasing role in this knowledge worker economy, and what's gained like you're still can be an ad creative. You can still be a creative director.
[00:25:27] You made. Have the AI create variations for you rather than you sitting in a room for three months and developing these things. Maybe the eye does it for three minutes and spits out 50 variations, but you then curate that stuff and you use your intuition and experience to say what's the most interesting thing here, which one might resonate, which one might motivate, which one triggers the emotions I'm looking for.
[00:25:50] And they eyes your assistant that may enable you to do way more. But actually maybe enjoy it more. Like I don't, I don't know, like I'm not a creative director, I'm not an ad person. I took media buying in college and enjoyed it, but that's about the extent of my advertising world. It's not a world I've lived in extensively, but you know, we touched a lot on it and we work with partners that specialized in it.
[00:26:12] And, um, you know, I do, I do think that the advertising world has probably been as affected by AI as any category. Going back to your point about programmatic advertising, buy side, sell side, like the stuff's been infused in for probably the last 10 years in a growing way. It's the creative side. To me, that is the new realm where there's a lot of unknowns.
[00:26:34] There aren't that many people using it on the creative side yet, but I think it's going to quickly work its way into, um, ad creative. And I don't, I don't, I don't know that people are really ready for that. I don't know that there's too many people who are embracing it yet and finding the new opportunities within it.
[00:26:50] But I do think that. It's going to change things. There there's a video. I'll put a link to this, but, uh, Perry Nightingale, who's the head of creative AI at WPP. He has this like 20 minute talk. It's like sitting in his bedroom like this. Isn't some crazy Ted talk or anything. He's just sitting there talking about the future of AI.
[00:27:08] And it's a really powerful, like 20 minute talk because he talks not only about. Like the, uh, it talks about the environment to start. He like tells the story of how Hey to do a photo shoot, like a big photo shoot for a 62nd or 32nd TV spot. You would send like 15 or 20 people on site. You would scour the locations, you would do all this stuff.
[00:27:29] And so you're, you're flying there. You're taking cars to places. So now. Send a drunk so that a single person with a drone, they use the drone and maybe they use LIDAR and they're like map the thing. And then they bring that back and they simulate the environment and, and then they can actually like film some of the stuff in front of a green screen and then use the simulated environment.
[00:27:46] It was just like, Ooh, like I wasn't even thinking of this stuff. And here they are doing things that are at this whole other level. So I do think that the big ad agencies like the big dollars and certainly in the media entertainment industry. They're using AI in ways that would blow the minds of most businesspeople and like the average B2B ad person.
[00:28:07] Who's maybe just like, you know, we're not doing this high scale production value stuff, but it's playing an increasing role on the creative. And I think that's a really fascinating area to pay attention.
[00:28:18] Mike Kaput: Yeah, for sure. I mean, it's just, it's getting really crazy out there and yeah, I don't mean to scare people about jobs and things.
[00:28:26] It's just, you really do need to start paying attention because it is moving so, so
[00:28:31] Paul Roetzer: fast. So to kind of, well let's real quick. Any other. Tools. We want to recommend people check out. I know you did a on LinkedIn. I'm looking at it. Now you did a couple of days ago where you had 26, you made a list of like, Hey, here's a bunch of AI or ad tools, or are there any on that list that jumped out to you or any responses you got from people that were like ones we should check out?
[00:28:53] Mike Kaput: Yeah. A couple jumped out to me, um, as worth kind of exploring, um, I would say, so company called next role, which I assume people have heard of because they own ad role, which a ton of marketers use. I think that's really interesting example. Um, because it literally says on their homepage, you know, here machine learning is essentially baked into every part.
[00:29:16] Of that tool. And that can kind of, you know, they're not going to say, Hey, we have AI for this. They say, look, we are a machine learning, powered, uh, ad retargeting platform. So I think that was an interesting one in terms of seeing how this stuff is infused in everything. Um, there's another, uh, company that looked really cool called blip, R B L I P P a R maybe blip AR um, Basically they create, they helped you really quickly create augmented reality, um, objects, advertisements, campaigns.
[00:29:54] So some of the videos on their site are really crazy. Um,
[00:29:58] Paul Roetzer: augment. Yeah, probably this blip, they are then
[00:30:05] talking about the branding. I think that's
[00:30:07] Mike Kaput: a whole other podcast, but it basically, this is really cool because look, I mean, I guess AR is no nothing new though. It probably is advancing at a really fast pace, but it just makes it easier. For you to do something like AR it's lowering the cost and democratizing this really interesting, uh, next phase of what you can
[00:30:29] Paul Roetzer: do with, it's a really important, important point.
[00:30:32] Like if you're an Adobe, if you're, if you're in this world, like Adobe is, you know, one of the leaders in being able to do creative, using AI to power creative. Um, but because of all these rapid innovations, this stuff is becoming accessible to the rest of us. And so to your point, like two, three years ago, Could I have gone in and tried to experiment with AI to power creative, probably not like unless I had the most powerful tools out there.
[00:30:56] And now all of a sudden things like Dolly too, and imagine from Google and whatever else spins off from that, like sometime in 2022, We as marketers may be able to go create our own photos and illustrations and designs by just prompting with text and it could get infused into ads. It could get infused into copy.
[00:31:16] It could be used anywhere. And it's like, think about that. Like that's real or, or giving texts to create video. Like that's the next step. Okay. Now let's go from text to image to text, to video. And now I just say something and it's created in video. Like that is not that far off. I mean, that may sound crazy to people.
[00:31:33] But I would be shocked if people aren't working on that right now. Like it it's, it's probably achievable in the next couple of years.
[00:31:39] Mike Kaput: Yeah. I think we'll probably have to have like a AI agency's podcast at some point, because that's the immediate based on our background. I'm like, whoa, if a client could just do that,
[00:31:54] Um, just to wrap this part up here, there's one other I want to mention, I will preface this by saying, I literally have just gone on the website. I do not have any affiliation or know any of these at these people, but it's called real eyes. And what it does is it uses AI. To measure human attention and emotion in response to advertising, which sounds just kind of wild to me, but basically, you know, facial recognition is able to test out, are your ads working?
[00:32:22] I, I would fully assume that, you know, this is more in like a focus group setting on everyone's computer. Um, but what's also really cool is as I was on their website. So that's a really crazy use case where you're optimizing ads by actually registering in real time where my eyes are tracking what my emotions are.
[00:32:45] Uh, that's just so futuristic to me, but like, again, it's one of those things, as long as it's being. Used ethically and a focus group. It's like, if your competition is using something like that, like how do you even come close manually? Um, but one other thing they mentioned, which is cool is that they actually advertised, uh, on their homepage that they have.
[00:33:08] The world's most accurate emotion classification system enabled by the largest culturally sensitive AI training set, which kind of speaks to the importance of bias and data and AI, because I know there's been. Litany of stories about, and complaints about facial recognition, especially because it is, uh, sometimes bias towards different people of different races and registers them as a angrier or more upset than others.
[00:33:37] So I just thought that was really interesting that they kind of get in front of that, but yeah,
[00:33:41] Paul Roetzer: it's just some wild stuff. Yeah. And that's know we can do multiple episodes on the bias stuff, but in essence, the thing to understand, if you're not familiar is AI is trained on datasets. And if it's fed data of largely one race, then it, it's not going to know.
[00:33:57] How the other race reacts or looks or things like that. So it's, you have to know how it's trained to do what it does. You may buy a tool that claims to do something, but if you dig in and realize like, oh, it was trained on a really limited data set, and there's no way this is inclusive of the kinds of audience that we reach and influence with our brand.
[00:34:15] And that it's probably not gonna be very valuable to you, but most. Wouldn't have the first clue to, to even ask those questions or like, you know, drill into the data set. But yeah, and we, I mean, we have plenty of experts that we lean on for that kind of stuff. I mean, that's not me and Mike, that's not our expertise, you know, in terms of our domain.
[00:34:33] Um, but that's where we turn to people for our may con conference, the marketing ag conference. You know the book and for online education, like we'll just turn to the experts. Like, what do we need to know? Like how can we prepare people, um, for the, for the things that may be coming around the corner that they need to be thinking about as brands and as advertisers, the market.
[00:34:51] For sure. Um, the one other one that I just, because we've done an AI in action, that was one screen. So they're there they're a sponsor of ours. Um, one of our other partners, um, and again, like Mike said, sometimes people would talk about our partners and we always disclose this. Like we just tend to know them better because we, we usually see the platforms.
[00:35:09] And so. Brands, we're familiar with vendors we've seen. So, um, we, we don't have any preference to any of them. It's just like one screen does outdoor advertising. And so they're trying to build it. So it's way easier to connect through a marketplace. Uh, and actually I get into predictive modeling around like traffic patterns of the specific types of buyers you're looking to reach.
[00:35:27] And then they can actually target ad campaigns. Based on those traffic patterns. So there's, there's just such fascinating stuff. There's so many areas. And now it kind of leads me to wrap this up with how to get started. And the thing I would say is know what you're currently doing in advertising and start by saying, okay, how can we be more efficient in the things we already do and maybe improve our performance, our conversion rates, our cost of, you know, cost per lead or whatever it may be.
[00:35:54] And then the second step is. What aren't we doing? What, what could AI on lock for us that we're not even thinking about? And that's where, you know, you just, you really need to understand what AI is and what is capable of doing to ask that second question, because until you understand it, you may not even have the frame of reference to, to know, like, if you didn't know Dolly to, and imagine we're a thing and that AI could create images from text prompts, you would never know to even factor that into what you're doing.
[00:36:21] So that would be my kind of two tips. What would you say. How to take this podcast episode, maybe go do something with it.
[00:36:29] Mike Kaput: Yeah, I think, um, I think I would check out that a good place to start, you know, shameless plug. I actually think it is the best place to start as our pillar page, which will be in the show notes.
[00:36:38] I'm an AI for advertising. The reason I say that is because it actually lists out probably 10 different bullets of what AI can do today. Like allocate budget, write ad, copy, whatever. I actually think it would be really smart to just go take a quick look at that list. And if you're doing one of those things, then start looking into some of the tools we mentioned later in that post, or just start doing your own research on, you know, AI for ad budget allocation or whatever.
[00:37:06] Um, I really think that that could be a good way for you to start seeing, oh my gosh, there's 10 tools that do some crazy stuff for this use case that I already.
[00:37:16] Paul Roetzer: Yeah, I guess if it's just ad budget, like it, it's hard to manage ad spend across different channels. And so in other words, it's, it's pretty easy to get started in the ad for advertising because there's so many use cases and so many that are very practical.
[00:37:29] Um, so yeah, I mean, I, hopefully this has been valuable to you. This is the similar format we're going to do for the, uh, the rest of them. So we would have. As I open my book to see our list. So we're going to have, um, next week will be analytics and AI. Then we'll do communications PR and AI content marketing, customer service and experience economy.
[00:37:52] Email marketing sales, SEO, social media. Those are gonna be the 10 episodes we'll do tied to this. And in the meantime, go grab a copy of the book. Um, you can kind of follow along as we go through these. And, um, yeah. So I hope it's been valuable, Mike. Thanks as always. All right. Well until next time, um, be curious, explore AI and, uh, check out the book marketingaibook.com.
[00:38:18] We'll talk to you next time.
[00:38:19] Thanks so much.
[00:38:20] Thanks for listening to The Marketing AI Show. If you like what you heard, you can subscribe on your favorite podcast app. And if you're ready to continue your learning head over to marketingaiinstitute.com. Be sure to subscribe to our weekly newsletter. Check out our free monthly webinars and explore dozens of online courses and professional certifications until next time, stay curious and explore AI.