Episode 42
AI and Creative Disruption
This episode of The Backstory on Marketing and AI features John Gargiulo, co-founder of Airpost.ai, as he discusses the future of AI-driven advertising and the role of creativity in the age of automation. From user-generated content to instant video ad creation, John shares how Airpost is reshaping DTC growth through technology. Discover how AI can both replace and amplify human creativity—and what that means for marketers.
Click here to view the video: https://youtu.be/5Aqe1wPoelM
Transcript
Relevant. Dot com. So today I am interviewing John Gargiulo of Air post.ai. Fascinating story, especially if you're in the e-commerce space, it sounds like. But let me tell you a little bit about John and uh, and, and then we'll go from there. So he is the co-founder. And CEO of air post.ai and he's helped multiple consumer brands boost their growth through user generated content and ad strategies across social media channels.
Wow. This is gonna be [:John: Thank you guy. I'm looking forward to it.
Guy: Yeah, absolutely. So what is your backstory on getting into AI and marketing?
John: Uh, yeah, it's been an obsession since the early days. So I got into marketing long time ago, uh, in the New York Agency world as a creative.
I went to Miami Ad School. I got to work at my dream agency, cliff Freeman and partners, they did, where's the beef? Pizza, pizza, et cetera. And I got to work with Cliff and come up with funny ideas and show up at a set in your early twenties to see 50 people making it happen was really heady and amazing and fast forward 20 years, that whole sort of Mad men world of I.
ople, clients like DoorDash, [:Um, and in between there I was at Airbnb for a few years, uh, leading parts of marketing. There on the AI side since the G PT three days. I've been fooling around with that. Really curious about it. You know, obviously was one of many people that saw that it had huge implications for the future, particularly for marketing and started building air post inside of my agency, uh, a few years back and now we've got this foundation for this skyscraper that can, you know, ingest a link to your Amazon page or your.
PDP and make you a hundred ads to watch and make some money with.
Guy: Yeah, that's fantastic. Well, uh, you know what's interesting about, uh, your story is that transition from being a kind of a content guy and a creative guy and, uh, now getting over into the AI and using that with the support of AI and the automation of ai.
ore human creativity. Is sup [:John: That's my hope, but I don't share the opinion. I. You know, it was over a year ago, guy that a friend of mine, um, who's a very well known advertising creative legend, um, won, you know, can Grand Prix and everything.
He posted a five onion headlines and he said, guess which ones of these were made by ai? And spoiler alert, they all were. But I actually, I, I didn't see that coming because if you go to The Onion these days, I don't know, for me it's not as funny as it was 20 years ago. I just don't, you know, I used to think they're just so creative and hilarious and it seems to have faded a bit.
But these five lines, like three of them were laugh out loud, funny, and funny is hard to do. It's kind of the tippy top of creativity. Now, when we all go into Jet GBT and say, make me laugh, does it make me laugh? No, but you know, I'll give you the, uh, the latest Sam Altman posted recently. And this model's not even out yet.
mpressed with the creativity [:So. I, I think it's scary. I wish I could tell you, oh no. These robots are never gonna, you know, write a funny ad or do a Nike commercial, but I think they will. It just may be, maybe it'll be 12 years instead of two or two years, instead of two months. Um, but it's, it's coming. I.
Guy: Yeah. Yeah. Well, you know, I, I wonder though, uh, you know, especially where I, I don't know.
e that it's taking something [:Maybe, uh, you know, taking bits and pieces here, bits and pieces there, and throwing 'em together and putting, you know, together some response. And, um, uh, so on you describe
John: how I would answer you if you asked me a question to be fair.
Guy: Uh, well then I'm, then I'm, I must be close. Uh, but, uh, I don't know. But, so, but if that's the case and, um, so it's not really generating something new, what it's.
It is taking old stuff or existing stuff and kind of slamming it together.
John: I guess I, I'm being, I'm being flipped, but I mean. What is a new idea? Is it something completely out of thin air or is it taking a bunch of old stuff that's in my brain and as I come up with an add at Cliff Freeman, smushing something together.
Yep. You know what I mean?
duck? You know, I, uh, that [:John: There's something on paper, even the Aflac Duck. And that was, uh, Kaplan Taylor. Yeah. Linda Kaplan,
Guy: um, yep. Linda Kaplan Taylor. That's it.
John: Yep. And I knew the guys who came up with it. I, they freelanced at my agency once in a past life. Uh, and you know, on paper, I bet you a lot of people at the client and the agency even then before it aired, said, this is ridiculous.
It's like a doctor that just says our name. This is the cheesiest thing ever. It'll be gone in half a year. And it wasn't. And there is some juju there to like, how do you know, et cetera. Again, it's not today guy, but it's what, how long? Five years you're gonna blink and it'll be gone. Um, you know, you are gonna be able to say, Hey, I love that Aflac Duck and these other ideas, and I need something that's sort of hard to put your finger on why It's funny and it's kind of awkward or odd watching a lot of old Vine videos with my son, which are just so creative and funny and weird to me.
Come up with 10 ideas like that. Oh, these two, you're onto something. Just like talking to a creative.
Guy: Yeah. Yeah.
y headed there, man. Like if [:Guy: you know?
Yeah. Yeah. Well, even two months from now.
John: Seriously.
Guy: Yeah. I mean, every half
John: week there's new models coming out.
Guy: No, it's just, it is just crazy. Well, we kind of got a little bit off, but, um, you know, uh, so tell us a little bit about, um, you know, user generated content and how you're using that and how you see that kind of revolutionizing, uh, marketing that can really drive, uh, you know, DTC, uh, direct to consumer brand growth.
John: Yeah, I mean, I've met with. Hundreds of direct to consumer e-comm founders that are doing, you know, I don't know, 10 to 250 KA month in spend. And they are struggling for a number of reasons. A, they are not Facebook ad experts, nor do most of them want to be. They love skincare and she's developed this really cool packaging and she's having an issue with her three pl and the inventory's at a weird level for what?
s to wear and they just, you [:They can hire agencies, which mostly they churn through every seven months because it doesn't move the needle that much and it's a lot of work. So what we're doing is just jumping to the end with air post AI and saying, look, you know, you need video ads for Facebook, here's. A cell, put a link to your product.
We can use AI to scrape the page, download all the photos, download all the videos, look across your Instagram and your, you know, past ads you've run on Facebook. Ingest all of your footage. We have about 400,000 pieces of footage that we've quietly built up over the last several years. Um, you know, vertical UGC shot with a phone kind of stuff.
ince she turns her head. And [:Guy: That is just incredible. You know, when I, when I think about, you know, how hard that would've been able to to be done, you know, before, and especially I could imagine, you know, if you've got, you know, I don't know, 10, 50, a hundred different products that you're trying to sell Yeah. And you want to have something that is really just that edge to get, you know, that consumer over the hump, uh, doing something like what you're talking about can certainly be a diff uh, you know, a strong differentiator.
John: Yeah. Hope so.
ar or two, that that is just [:So how do you see AI. Uh, making even more in that transformation of the digital marketing process.
John: I think it's gotta be good enough, and it's not today. If you're listening to this, you've probably tried an AI tool for marketing that you may have even been initially excited about, or a friend told you about, and you may have opened it, you may have started to use it for three minutes and still been excited about it.
And then 13 minutes later you're like. Ah, I think this is not gonna work. Um, or, or God forbid, a week or three later, are you really still using it? That's because my personal opinion is if you can't get 95% the level of quality that people are with their old ways and manual ways, now it's worth zero. It's not worth 90.
w we do our evals. It's like [:And and so far, spoiler alert, people can't really tell and it's getting better and better every week.
Guy: Yeah, yeah, yeah. And I think you're right. You know, that first ad that Coca-Cola did, which got panned Personally, I didn't think it was, uh, I thought it was a good experiment. If you think about it as an experiment, it was a great way to, you know, to kind of get your toe in the door and, and see what it's gonna look like.
And then you get, obviously, you know, everybody, you know, and I mean, a lot of negative complaints, but nevertheless, uh, you know, it was something to gain some feedback. And so from that perspective, I think it, I think it worked. But now. You know, now that, uh, AI is getting there and, and, and, uh, is maybe on the, you know, on the threshold of really taking off that then has other impacts and all of a sudden, you know, the human touch and marketing is maybe gonna go.
y favorite ads, which is the [:At that point, I'm finding it hard to believe that AI is going to, uh, replace that. So tell me what you're thinking about.
John: Yeah, I think if you look backwards, it helps look forwards, so. I used to share your opinion more strongly, uh, when we started working on Air Post, when you went to make a script for a TikTok video ad.
If you just asked ai, Hey, I'm making a script for a TikTok video ad for this brand, make me one. It will make you one. It'll tell you what should be in the scene, what's happening, or you could just, or it'll just give you a voiceover. And it was horrible. It, I, I was convinced that it's horrible because it was learning, you know, where are our actual ads scripts written down?
oks. At least they weren't in:Uh, that's what it sounded like. It just sounded like an ad. It sounded very corny. I was on a call with somebody four to six months ago. I. And I went to prove that point to them how our script writing, which uses so many microservices and ways to craft it, to make it actually good and educational and informational and interesting and specific.
Um, I said, yeah, like, whereas if you just go to Chatt BD I'll do it right now. And I, I, I pulled him Chatt BD and I said, make me a TikTok at, so it's like a couple years later. And it wasn't bad, and I, I, I even specifically said, make it colloquial and very Gen Z and how people talk there. And it did, did like an 80% good job.
TikTok now even multimodal, [:So they've really pulled in the corpus of like humanity, which mostly exists on the internet. Um. Things are getting more and more creative. My bet guy is that in two years max, you just go on and say, make me a funny TikTok ad for Aflac that incorporates the doc, or invent some new mascot or whatever.
You'll like it. You may not love it and say this, but I mean, if I showed you 200 ads, you wouldn't love them. You know, it's, it's, it's getting there.
Guy: Yeah. But, um, fair enough. Yeah, and I, I, I kind of agree with what you're saying. Um. And, uh, so that kind of means then you, you have then, let's say human managers or whatever it is, or executives.
e five concepts or something [:But one of the things you said right at the end was, is that, hey, these are pretty good, but they're not outstanding. I. You know, so when I think of, like, again, I think of the, uh, you know, the, the Aflac duck, I see that as being outstanding. Mm-hmm. And so, you know, the question then back to you is, is good enough good enough?
Or am I just gonna get, you know, after everybody's doing it, am I just gonna get lost in the sauce again? And, and I'm not gonna be standing out in a way that I really need to be? That is
John: fair. Like, if everyone could make an app, like level commercial, then. So we gotta go even higher than that. Again, I wouldn't overestimate personally, I hear that I wouldn't overestimate the quality level of humans today.
a lot of it's just gonna be [:But my bet personally, and I think I, I have a theory that the reason, and I'm a creative person, all my friends are creatives in advertising. The reason it's really hard to hear this or just to agree with it is 'cause it seems to be so counter to what we maybe would like the world to be and maybe like our own lives and jobs, but like.
It is gonna happen. We are gonna have outstanding, and I could be wrong. I says part of me that hopes I'm wrong, but I think we're gonna get there.
Guy: I hate to say it. I think you might be right. I mean, yeah, I'm an analytic kind of guy and um, I think creative guy today, now, or maybe you know, a couple years ago, but I think creatives are worth.
Every penny you pay him. Yeah. Because if you can pay somebody to get that Aflac quality commercial out and you know, then that right there is, it's worth it. I mean, the, the hockey stick that Aflac enjoyed after that commercial went on for, you know, I don't know, 6, 8, 10 years maybe.
John: Yeah.
Guy: Still going today.
. Yeah. And, and, um, so you [:And, and then those, I think you're right, those, at least in the short term, you could easily replace with an AI generated commercial. And you might actually do better than what you see there.
John: And they could be more smart. Uh, those AI commercials could be smarter around. What has been working on digital?
What has been working on these vertical TVs in our pockets? Oh, when we talk about this inside, even if it's a black box, the algorithms understand that this kind of vibe and this kind of person is working and so they make the TV commercials based on that. I've been a creative coming up with TV commercials.
omething last year, but like [:Guy: 'cause that's, uh, I feel, I think that's the, yeah, one of the, one of the big challenges, uh, you know, it's huge that challenge marketers and creatives are facing. Yeah, please. It's scary.
John: It's a scary feeling and I, I can almost feel people listening to this being like, oh, great, that's, it's all over. Um, I've learned about something semi recently called Jevons.
I dunno if I'm pronouncing it right. Jevons Paradox. Have you heard of this?
Guy: No.
John: Okay. I hadn't neither, so I'm gonna read it. So, 'cause I never get it right. Uh, the summary, the Gemini Google summary, Jevons Paradox, named after 19th century economist, uh, posits that increased efficiency and resource use. Uh, IE ads becoming just so easy to make can paradoxically lead to increased resource consumption rather than the expected decrease.
ling oil out of the ground in:Well, crap, there's gonna be no one. You know, they're gonna have to fire 90% of the oil industry, or the cotton farmers, or the advertising creatives. But actually the opposite happens. My favorite example of this, 'cause there's a visual, um, that goes with it, is there's a famous ish photo of an insurance company in Syracuse, New York in, I don't know, 1921 or something, and there's just lines of wooden and desks.
Huge room, and it's mostly women with abacuses and paper and they're calculating actuarial tables and their business cards say computer, they're called computers. Mm-hmm. People say, I work as a computer down at the insurance company and they compute things. You know where I'm going with this. Fast forward to where somebody someday said, oh, I could just put this Microsoft computer on one desk and press a button and it'll do all of that work.
Well, we're [:And so I think it's easy to be like, oh crap. Like, but that's what I do. I'm a computer and now I'm gone like. There's gonna be so much more opportunity for folks.
Guy: Yeah. I think there's a good chance that you're right. I mean, you know, you look back at, uh, like office automation and, I don't know, this is maybe, I don't know, 20 years ago and, uh, office automation was gonna just wipe out all of the office workers.
Well, now there, like you said, there's probably 10 times more office workers than, uh, than there were, than before. So, uh, certainly their, their positions. Are different and what they're doing is, is very different. And that's, or in some
John: [:Actually, the efficiency of a sales rep has not gone up. Mm-hmm. So, as weird as it sounds, it might be in a few years, like one way I could see it going is. You know, uh, Walmart has the same number of creatives actually. And yeah, they're all using these tools just like all the sales reps are using these tools.
But there's the same number of them. They're not getting that much more efficiency. They don't wanna go back to the old ways 'cause it does save them time. But as someone sort of cynically pointed out on this podcast, uh, what do they do with that time? Okay, now they can make 12 sales calls a day theoretically instead of six, because of this efficiency.
Do they or do they just fill it with, you know, having going on for another coffee? So it'll be really interesting to see how much that paradox comes through.
Yeah, yeah, yeah. Well, and, [:Although, you know, too, if the. Uh, on the flip side, just thinking out loud and, um, we're definitely way off on a tangent, but if the category size volume has not changed, then all of the sales folks in that industry have gotten, let's say, more efficient. But they're still chasing the same number of deals right now.
If one of them is more efficient than the other, then they are gonna have maybe a short term kind of an advantage. You know, at some point they're gonna all, even out again based on, you know, some share level because the category volume hasn't changed. Now in, in our example of office automation and in your example of insurance and the, the human computers were that were out there.
, it's may or may or may not [:And, um, you know, so how do you see AI driving and. Creating opportunities for brand growth. Uh, but before you get into that, what do you define as brand growth?
John: I was about to ask you the same question. Um, are you talking about top of the funnel, kind of just getting the word out? Well, that
Guy: is exactly where I was going because you got top of the funnel, middle of the funnel, bottom of the funnel.
And um, you know, and when I think about brand growth, uh, you know, certainly sales selling more grows your brand emotionally. No question about that. You got poor people engaged with it. They're using it more. Therefore, you got more brain cells that know what your brand is, but. Before they've even purchased, you know, brand growth could be considered.
more aware. Uh, or I've got [:To me, that's what I, I would define as brand growth. It's more the emotional side as opposed to the sales side, although there, hmm. They're tightly, uh, linked, but brand in my mind is emotion. And then, uh, you know, sales, so to speak is then, you know, conversion.
John: Yeah, I would say I, I got a masterclass, uh, in my years at Airbnb from our CMO and CEO on.
Building an incredible brand. It's funny, we've been talking a lot about what is it, what are the words and emotions we wanna elicit from an air post on our, on our page? And I, I thought we had a decent thing going, but talking with an external firm and some other people's, like, wow, yeah, this could really be meaningful.
guys at a company called Fer [:And that immediately said two things. Oh, this company helps design a whole funnel digital funnel. And, oh, I, I make ads every week. I, I don't, I'm, we haven't changed our funnel in seven months. I guess I'm doing something wrong. Uh, Airbnb, right. Belong anywhere. Talk about the emotional level. Nothing about, stay closer to the city center, you know, not just in the city center.
Go to a neighborhood outside of Tokyo or something practical. It was much deeper. 'cause everyone wants to feel that they belong and. No one wants to be a tourist. Uh, so you know, Brian and Jonathan landed on that in the back of a cab in Asia somewhere, but two words, very powerful. So I think brand definitely still matters.
w. I think it will be better [:Com consumer. Um, it had a B, the B word was a dirty word for a long time. Oh, what is this? B2B company wasting money on some brand agency to talk seven months about changing their logo. Now people have realized this one guy put it, if you and I are starting a company and we need some CRM software, Salesforce, it's probably already in your brain.
If we need some inbound marketing or some lightweight CRM, or let me, sorry, let's skip to, to mail some Fisher-Price, uh, email solution MailChimp. When those things jumping to our brain are really powerful, right. Uh, and it really makes it harder for other people to get in. It's a moat. Uh, it's one of, uh, the seven powers, um, that, uh, the consultant wrote the book that ne, that Reed Hastings loved.
or don't know how to do, but [:Guy: Yeah. Yeah, yeah. Well, I, and I, like I said, you know, in my definition that you know that a brand, you know, is, is gotta be the, you know, how many brain cells or synapses do you own, you know, as opposed to, uh, another brand. And it certainly has to be also competitive. You gotta look at it and relationship, you know, between Dove and Olay or you know, Vaseline or whatever the other brands are in there.
Um, anything else? I've got one more good question for you, but anything else, uh, on, uh, that we haven't talked about or,
John: no, I, I hate to be pitchy on these things, but just for anyone who's sitting there thinking, man, I need more video ads and then I'll only feel like churning through another agency or figuring out how to use some tool, just go to air post.ai.
Put your link in. Watch Magic happen. Uh, if you're not happy, let me know. Just john@airpost.ai, but it's a new thing in the world, so, uh, if I think it could be helpful to folks, check it out.
you and, uh, definitely air [:So what advice would you give to an up and coming new marketer?
John: Oh boy. I think there's two kinds of marketers, uh, one people who embrace AI and are constantly trying to figure out where the bleeding edge is and, and where the useful bleeding edge is. 'cause that's not the same thing. You know, what, what stuff is actually practical to use.
For example, uh, content stuff like, uh, SEO, uh, optimization writing. Those tools have crossed, I think the 90, 95%. A lot of 'em still have humans in the loop, but there are areas that you could be using AI right now to help your business. Although there are quite few, in my opinion, that are, that at that level don't be one of the people that doesn't, that that just comes up with an excuse for why it's not gonna work or point out things that are wrong with it.
. AI generally, as you know, [:And you heard about him. You heard about another guy. You could go a few doors down and mix tables. And we are walking into town with our modern clothes being like, we have this idea for Crate and Barrel. We're just gonna stamp out the same tables, try to make it look like that one and that one at a factory.
And what do you think people are gonna do? They're all gonna scoff. And you know who's really gonna hate it is people who make tables and they're gonna go, that's ridiculous. Well, that one, they just came off the line. Did you notice how that's wrong? And that's wrong. And, but eventually there's a place for that.
And by the way, Arjun table was made by a person, like by hand. So there'll always be craftsmen. But don't, don't be the, we, we had five categories of personas at Airbnb. The, the last was called the heel dragger. You know, you have to like, come on grandma. I promised the Airbnb. No one's gonna watch you with cameras.
ssion, you do not want to be [:Guy: I don't know about you, but I aspire to be a heel dragger.
John: That's funny. I have a friend who's always. He loves that, that, that that was a thing. 'cause he's like, he's definitely that with tech.
Guy: Yeah. Well, wonderful. John, thank you so much. That's a great answer. I really appreciate it. And, uh, thank you for being on the podcast and, uh, great discussion.
I, you know, I think there's, I. You know, we are, I mean, I think we all realize that, but we are at the cusp of something that's gonna be crazy. Uh, crazy fascinating. So, uh, thanks for being here. So, where again, can, uh, folks reach out
John: Yeah. To you
Guy: and, uh, get ahold of your company and, uh, what have you.
John: Yeah, first of all, guy, this was super fun.
Uh, I'm on LinkedIn, so John [:Um, so either place.
Guy: Fantastic. Well, thank you John. So air, post.ai, and certainly, uh, John Gargiulo at, uh, at, on LinkedIn. Yep. Um, so for the audience, uh, definitely stay tuned for many other videos in this series on marketing and ai. And if you get a chance, please go to marketing machine.pro relevant.com and find out more about my upcoming book, the, uh, AI and marketing.
So, uh, John, thanks again and uh, we'll hope. Maybe do this again at some point.
John: That'll be fun. Thanks guy.
Guy: Thank you, [: