Loading summary
Transcript88 lines
- [00:00]
Jay Richman
Foreign.
- [00:13]
Alison Schiff
Welcome to Ad Exchanger Talks, the podcast devoted to examining the issues and trends in advertising and marketing technology that matter most to you.
- [00:25]
Sarah Sluice
This podcast is brought to you by adstrated. Adstra's leading identity resolution network enables precise targeting, scalable activation and richer customer understanding. Whether you're looking for person based accuracy or massive scale, only adstra has the transparency to give you both. Learn more@adstradata.com that's AD S T R A data.com.
- [00:56]
Jay Richman
Foreign.
- [00:59]
Alison Schiff
I'm Alison Schiff and you are listening to Ad Exchanger Talks and I can't believe we're already midway through December. It is crazy how time flies. But happy slightly early holidays and happy Hanukkah, which we're in the midst of right now. My guest this week is Jay Richman, VP of Products and Technology at Amazon, where he's working to transform how advertising content is created, personalized and scaled no less. We'll talk about what that actually means in practice, plus how to balance AI and human creativity. The new eternal question how AI is opening up new opportunities for smaller advertisers? Why sitting out the AI revolution isn't an option for marketers and lots of other good stuff. But first, please save the date for Convergent TV World. Taking place on March 5th and 6th at the Time center in New York City. Convergent TV World is the new name for our CTV Connect event. We'll bring together the worlds of linear TV streaming, CTV gaming, retail media and digital out of home to help you tackle the challenges of measurement, attribution and cross screen storytelling. Podcast listeners get 10% off the price of their ticket when they use the code POD10. Get your ticket today and see you there. Hey Jay. Welcome to the podcast.
- [02:19]
Jay Richman
Hello. Hello. Thanks for having me.
- [02:21]
Alison Schiff
A pleasure. Yeah. So what's one thing about you that not a lot of other people already know?
- [02:31]
Jay Richman
Well, I'm a, I'm a huge basketball fan. Which isn't maybe novel in its own right, but I, I played a fair amount as high school at a five and a half foot foot stature. So if you were to look at me, if we're meeting in person, you would not think, oh, that guy played ball. And I'm actually raising a daughter who is a whopping 5 foot flat that plays college ball. I'm actually going to watch her tonight. She's at Pratt, plays Division 3 ball, is a point guard. Yeah. So proud. Amazing. And yeah, a fun fact that you wouldn't know at least by looking.
- [03:14]
Alison Schiff
I'm trying to remember his name. There was a player in the NBA who was five foot something and he was really incredible.
- [03:23]
Jay Richman
Oh, Mugsy Boat.
- [03:24]
Alison Schiff
Yeah, it must be. And I just remember seeing clips of him playing looking just like diminutive next to these Amazons but basically like just an average heighted person. But standing next to someone who's seven foot tall with a seven foot wingspan just looks like crazy.
- [03:45]
Jay Richman
Oh, Mugsy was incredible. Yeah, no, he, he stood at Stan, he stands at 5, 3 and was not, didn't just make it but was a, a starting guard and phenomenal. So anyway, anything, anything's possible with the right my determination.
- [04:03]
Alison Schiff
So when you're not playing basketball or hanging with your daughter playing basketball, you're kind of a dyed in the wool ad tech guy. Before Amazon, you were at Spotify for nearly a decade, a bunch of different roles. VP of monetization, you were head of advertising and subscription tech. And before that you were at a tech company that focuses on contextual and native. And then at nbcu you were director of the digital media team that incubated Hulu and built advertising and analytics and video and social platforms for their whole network of sites and apps. And now you've been at Amazon for just over three years. And I'm curious how those experiences shaped how you think about the evolution of digital advertising platforms now that you're, you know, at the triopoly. Part of the triopoly.
- [04:55]
Jay Richman
Yeah, it's interesting to hear the journey put as such. I'll start by saying. So I'm a. We were talking a little bit before hopping on this call. So I am a native New Yorker, like born and raised, lower east side of Manhattan. I've spent my entire career based out of New York. And why that's interesting for Germaine is that there's really three main industries here. It's like finance, media, advertising. The latter two tend to go, you know, hand in glove. And so I spent a bunch of time like early in my career working on media properties, scaling them to the point of mass appeal and then turning around and monetizing through, you know, a combination of E Com at my time at Scholastic subscription and advertising, my time at Spotify, but like coming from New York, focused on media, you very quickly get asked to turn around and monetize. And I saw just like the power of being able to monetize as being ultimately accretive to the end consumer experience. The better we monetize, the more we can invest in things like content, the better the overall viewer experience became in the case of Hulu, now in the case of Amazon, especially as it relates to our prime properties. And so I've gotten very comfortable with the role that monetization plays to democratizing access and improving upon consumer experiences. Yeah, it's been the gravity of my career up until this point.
- [06:41]
Alison Schiff
I live in Peter Cooper Village now. I moved in August, so. So the Lower east side is kind of almost my backyard. I'm starting to explore it. It's really exciting.
- [06:52]
Jay Richman
Cool. Yeah. No, I spent a bunch of time in Stuy Town in my middle school years. I went to middle school right there. So all my BFFs were sty kids. So. Peter Cooper Stuy Town. Yeah. Holds a soft place in my childhood.
- [07:09]
Alison Schiff
I really love it. It's like almost. It's almost as if the East Village had a backyard, you know, like an actual lawn that you could hang out on. You know, it's very green, it's very quiet, it's very lovely. And you know what, this is going to be sort of a ham fisted kind of tie that I want to make. But when you were talking about advertising being accretive to the experience and how we can have multiple things at once, we can have green spaces and we can have a gritty city, we content and we can have advertising because there's this strange mentality, maybe not so strange because there are some pretty crappy ad experiences out there, but that it's one or the other. You know that when the monetization happens that the experience goes out the window. But it doesn't have to be that way.
- [07:59]
Jay Richman
No, I think the best, I think the best model is, are these mixed monetization models where you're offering both because that way you're able to extend a service to the broadest possible audience that's interested in it. You also have a clearing price for what you should charge on a subscription basis by virtue of what arpu you'd be able to generate in an ad supported world. So I really do think that they're complementary to one another. If everything's behind a paywall, you're going to limit your reach. If everything's behind ads, there are going to be experiences that folks would happily pay for to avoid interruption. And if you offer both, you understand how best to price while not limiting access. And I think if you started in this world as ad supporter, you're trying to figure out subcommerce. And if you start as a subscription base, Netflix, you're trying to figure out how do you grow your total addressable market through ad supported tiers. And so I feel like that's the direction of, of, of travel and everybody's trying to figure out what their right, right balance is.
- [09:31]
Alison Schiff
Well, you, you lead creative tech at Amazon and it's your job to develop agentic systems that do that basically. And I, I'll just read off of your LinkedIn page that transform how advertising content is created, personalized and scaled, which is, you know, kind of this overarching kind of thing. But you get to, let's say you get to the office in the morning, you pour yourself a cup of coffee. What does that look like? Like, what does that mean when you get to the office and want to transform how advertising content is created, personalized and scaled? Like, what do you do every day?
- [10:05]
Jay Richman
What do I do every day? Now you sound, you sound like my kids. What do you do? Well, they asked me that question and they tell me to stop talking about AI.
- [10:14]
Alison Schiff
Well, this is a safe space. Please talk as much as you like.
- [10:17]
Jay Richman
Okay, good. If I'll get them to listen. My indirect way of communicating with them. I don't know. I think I have a really cool job. I wouldn't pick a different one. I get to think about what is newly becoming possible by staying really close to the frontier technology. I specifically have been focused these last few years on generative media, how through a prompt, you can create rich images, turn those images into animations, and string those animations together into fully formed video and ads. I think it's just a great use case for the technology. The business model's baked right in the bar to create a 15 second spot is a lot lower than a full featured film. And so I get to play on the bleeding edge with a really cool set of technologies and techniques that has commercial appeal and like a market waiting to be born around it. One of the things I think you rattled off my LinkedIn was maybe something around democratizing access, which is probably an authentic used term. But for us at Amazon, most of our advertisers are small brands with little time, budget or skills required to tell their stories. And so they stick with sponsored ads, the type that appear inside of search results, kind of creative list experiences. Yet Amazon operates this really vast and sprawling network that extends beyond stores, to prime video, to screens in supermarket aisles, to Alexa devices and kitchens and living rooms. And for these brands to be able to reach their consumers wherever they are, they need tools that can help them venture beyond product shots on white backgrounds. And so to me, that's what democratizing access means. It's helping these brands grow their businesses on Amazon by providing tools that Let them reach their audience wherever they're at.
- [12:42]
Alison Schiff
By the way, I just realized that you started at Amazon the same month that ChatGPT was first released. Right? It was November 2022. That's exactly when you joined Amazon. And that was a very significant moment that started to change the world world in like a very rapid way. I mean, was that, I don't know, what was that like? I guess kind of being part of that wave. Because when you, when you started, were you already thinking like AI will shape reshape the advertising ecosystem?
- [13:18]
Jay Richman
Oh, I don't want to oversell my ability to stare into the crystal ball, but it was certainly a factor. You know, I joined Amazon about three years ago to lead this creative tech charter. It was a brand new charter. It was, it was, it was built when I joined. Teams that were kind of working in and around this space were kind of, you know, plucked from elsewhere. We created a central charter around the opportunity and because it was brand new, I didn't inherit a whole lot. So I wasn't saddled with legacy. And I got to build and invent from a clean slate right at the time when paradigm were shifting. And so I can look back and say how fortuitous the timing was. It did factor into my decision to join. Oh, this sounds like a, not just a great company, but a great charter and a great time. And so I hopped in with both feet. But if I were to say back then that we would be creating full on television commercials for Fortune 100 brands out of nothing more, a little more than a product detail page. Yeah, I would have been lying. I think that the pace of change and innovation over these three years has been nothing short of mind blowing. And I still feel that way. You were asking about day to day, maybe not day to day, but certainly month to month, if not week to week. I'm seeing measurable improvement in, you know, model quality and creative output in a way that I just don't remember at any other point in my career.
- [15:21]
Alison Schiff
The pace is wild and I want to talk about what that feels like as someone on the creative side or someone who's involved in ideating a campaign and launching a campaign. Just like the massive impact of AI on creative development, because agentic AI, and we weren't using the term agentic like as recently as a year ago. I mean, it's basically transforming workflows pretty much every stage of the process because, yeah, I mean, creative development used to follow a pretty linear path, right? Brief concept, storyboard production, maybe you'd measure it, hopefully you'd measure it and then you'd do it again. But AI is really collapsing those steps. And it also feels like it's becoming less of a tool that you prompt and more of, I guess a partner or. Yeah, a partner that you can use to research and come up with ideas and then produce that stuff across lots of formats. But with all of that potential, like from where you sit, our brands and advertisers already really doing that in practice, or is it still more potential than reality now? Like, has it really worked its way into workflows?
- [16:38]
Jay Richman
Oh, it most certainly has. I mean adoption is spiky. Like for some it's become an indispensable tool and has completely transformed how they compete in storytelling, the market. And at the other end of the spectrum, there are folks that are looking and seeing and watching. I would say fewer folks fit into that category than I would have assumed. We built Creative Agent and image generator and video generator before it for small brands effectively competing with non consumption, meaning that they just weren't doing this today. And so like now we're making it available. And so you're competing against nothing as opposed to something that was slaved over for weeks or months by professionals. Like that wasn't the bar, that wasn't the opportunity that we saw. We saw. Let's try to open up and expand the marketplace and who can participate in opportunities like television, you know, advertising. And to our surprise, we've seen some of the largest brands on the planet, like Procter and Gamble, be some of the earliest, most aggressive users of the tool and the capability because they see the opportunity to empower their brand managers, cut down on cycle time and often and ultimately gain a lot more efficiency and therefore performance out of their creative. And so kind of everywhere you look, we're finding these untapped pockets. Like the largest brands in the world have Long Tail brands in their portfolio. Not everything within P and G is tied. They have brands that F into Long Tail. They get starved for resources. This is a great tool for them. There's Long Tail markets within large brands that get starved for resources. So there's just like all of this like untapped, unmet need from brands big, medium and small. And what we're finding is that these tools are just starting to kind of fill in like those, those gaps, those, those white spaces. And so, so yeah, I've actually been pleasantly surprised how leaned in the, the, the industry has been.
- [19:11]
Alison Schiff
So there's something I really wanted to ask you about. The Coca Cola holidays are coming ads which you've probably seen they used AI for the past couple of years, and it is pretty amazing. But it also sparked totally understandable backlash from artists. And there's definitely some uncanny valley going on there. But what I think is really interesting is that Coke experienced that backlash last year, but then they decided to use AI again this year because apparently it turned out that that ad was one of Coke's most effective campaigns with consumers. And Kantar raked it as the top performing Christmas ad globally last year across all categories. And that obviously encouraged Coke to keep on experimenting with generative AI. But I went and I watched the video again before hopping on this podcast with you and I was reading some of the comments and they're really vicious. One is like, nothing says Christmas quite like dad got fired because AI replaced him to do animation. The unintended irony of using real magic as the tagline is hilarious. And this AI ad made me want to drink Pepsi. So I won't make you comment on a specific brand, but what's your take on this whole vibe, like the tension between these emotional responses and the drive to just push AI creative forward. And it is getting better and better. And when you look at that Kantar result, like, that's really interesting.
- [20:43]
Jay Richman
Yeah, well, first, I'm very familiar with the campaign because my dad stopped talking about AI. Kids were very quick to forward it to me. I hope they weren't any of those commenters. But familiar with it. And look, I don't think it's probably too different than how CGI in the film industry was originally greeted. And if you look back, you could see how primitive those original versions are, but how if you zoom forward and look at the latest Marvel releases, how it's become an art form onto its own. And so I just look at it as evolution that you're going to get these kinds of comments in the beginning because it's new, because it's seen as potentially disruptive, and because it's imperfect. I mean, these experiences are very spiky. We're not that far removed from Six Fingered hand and we've moved past that. But that doesn't mean that we don't have. I forget what the defect was on the Coca Cola, I think was it.
- [21:56]
Alison Schiff
The wheels on the trucks were like all different sizes and some of the animals that were in the ad wouldn't be in North America. Stuff like that.
- [22:05]
Rick Irwin
Okay.
- [22:06]
Jay Richman
Yeah, you have to have a good. I mean, some of the stuff that I've seen. So I was doing a bunch of stuff around Thanksgiving for, for personal and professional oh, you were asking also earlier about some of the day to day stuff. So I've incorporated generative media in a fair amount of my day to day work. I tried to decorate. I had a Thanksgiving, my dining room table and it created an abundance of small pumpkins and fall foliage such that the table left no room for anyone to sit around and and dine. And so that was just the model as being a little too aggressive. I've seen when you try to do a setting that includes wines or beverages, it'll tend to put three or four different glasses for a single setting as opposed to. So these things, they still exist. And that's why we've tried to develop our tools such that fault tolerance is built in, that they're not designed to be one shot and done that the human is a feature in the in the loop, not a bug and allows for them to spot these things and iterate. And as they get better, more and more of it will be automated for sure. But I look at it as just like an extension, the new canvas or paintbrush. My daughter is studying art and design in college and my son is going to go on in all likelihood to become a scientist. So I got the left brain, right brain on there and I can't even imagine what their jobs are going to look like once they graduate. But the one thing I encourage them to do is to kind of run towards a new medium, run towards the disruption test and tinker. And you're going to make a bunch of mistakes along the way, but you're not going to learn as much as you would by being on the sidelines. So anyway, that's kind of my advice when folks ask.
- [24:14]
Alison Schiff
I mean, it's solid advice. You just have to roll up your sleeves and understand it, even if you are skeptical or don't like it philosophically, for example. This is not a perfect comparison, but if I'm reading a book and I don't like it, I will finish it. A completist. So I'll finish the entire book. And one of my reasons is because I don't feel like I can have an informed opinion about how bad the book is unless I finish the book. I want to be able to shit upon it with authority. So I'm not saying that AI tools are all problematic and blah blah, blah, it's not a perfect comparison, but you really shouldn't have a knee jerk reaction. It's okay to want to protect yourself and to be worried about your job and the future and, and all of those things. I mean, you should be reasonable and rational, but the knee jerk reaction isn't useful.
- [25:10]
Jay Richman
Yeah, I personally don't think AI is a bad book and I'm quite keen to see how the story plays out. But it's here, it's moving very fast and I really do encourage folks to be hands on.
- [25:29]
Alison Schiff
So we're going to take a quick break and when we're back we'll talk a little bit more about keeping the human in the loop because this is about enhancing human creativity. And we'll get into a few real examples of how this is playing out for some of the brands that Amazon works with. So stick with us.
- [25:54]
Rick Irwin
Foreign.
- [25:59]
Sarah Sluice
Hello, I'm Sarah Sluice, Editorial director of Ad Exchanger and I have with me here Rick Irwin, CEO of adstra. Hi Rick.
- [26:07]
Rick Irwin
Hi Sarah. Great to see you.
- [26:09]
Sarah Sluice
So in a market dominated by very large identity providers, how do you see adstra disrupting this space?
- [26:18]
Rick Irwin
Well, really simply we set out and every day we do the things that the bigger players in the industry either cannot do or will not do. And when I say cannot do, it's because we have technology that we architected from day one that allows brands and their partners to do things that they can't do with the larger partners, like manage all of their identity resolution in their cloud environment rather than in the provider's environment. And then we do things that that larger players perhaps can do but won't do because it might threaten their business model that they've very successfully built up over time. A good example of that is challenger brands often need the ability to scale their business using identity solutions. And volume based pricing methods like CPM really don't allow for scale because they they tend to charge commensurate prices as the client grows. And so we work with clients oftentimes to provide them subscription based pricing that's flat and level and they can plan around that for a year or 18 months or two years at a time. Those are two examples of how we disrupt.
- [27:32]
Sarah Sluice
So more flexibility in terms of pricing and in how you use the identity product.
- [27:38]
Rick Irwin
Exactly.
- [27:39]
Sarah Sluice
Let's talk a bit about transparency and drill into that. Ad Exchanger listeners tend to love transparency. And where do you stand on that issue?
- [27:49]
Rick Irwin
Well, we think the two values that matter most in identity are transparency and control. One goes with the other in both directions. Transparency allows the brand to actually see how identity resolution is being performed because it can be adjusted to be more precise or less precise depending on the appropriate customer use case. And so we give our clients the ability to see how matches are made and Then control rides along with it. They can change the level of precision for the use case that they are pursuing that particular day. So if they're pursuing an awareness campaign, they might actually want a little bit less precise identity resolution because they want to cast a wide net and reach a lot of customers. And in an application where they're offering best customers a special promotion, they would be more interested in having a very, very precise match so that they don't offer a best customer promotion to someone who perhaps accidentally isn't the best customer. So transparency and control are both extremely important.
- [28:58]
Sarah Sluice
Okay, the opposite of a black box is what you get with Ad Strat. And let's talk about identity. It's such a broad term in the industry now. It can mean a lot of different things. How do you think people should be thinking about identity? And how is identity sometimes misunderstood?
- [29:16]
Rick Irwin
Well, the biggest thing I would, if I could wave a magic wand, I would love for people to understand the real and important difference between identity graphs and identity resolution platforms. Astra is an identity resolution platform, and every identity resolution platform produces a graph, which is a snapshot. I use an example. I use a metaphor of a movie. What's your favorite movie? Can I ask you your favorite movie, Sarah?
- [29:42]
Sarah Sluice
Right now I'm feeling Melanie Griffith in Working Girl.
- [29:45]
Rick Irwin
Okay, Working Girl. So imagine if you were a production assistant or a camera person on the making of that movie when that movie was being shot, and you made a scrapbook of everything that happened during the movie shoot. You'd have this big scrapbook and page by page it would sort of tell a story, but it's not the same as watching the movie. And that's the difference between an identity graph, which is like a scrapbook, it's just a point in time, and an identity resolution platform, which is actually like a motion picture in that it is running matches perpetually all the time updates to make sure that the association of one form of identity to another is accurate. And that requires software, hence why it's a platform. And a graph can just be a bucket of data. But an identity resolution platform is actually a working, breathing piece of machinery, if you will, that performs the work of identity 24 7.
- [30:45]
Sarah Sluice
I love that distinction of identity resolution and identity graph. It feels like an ad exploitation article. I'll file that away. Thank you, Rick, for educating us on identity and for supporting our podcast.
- [30:59]
Rick Irwin
Great to be with you, Sarah.
- [31:16]
Alison Schiff
All right, we're back and yeah, let's talk about human beings and keeping the good old human in the loop. Talk to me a little bit about how you see AI helping humans be more creative instead of AI replacing human creativity. Especially when it comes to just combining speed and scale with intuition and judgment, the kinds of things that only people can really bring to the process.
- [31:46]
Jay Richman
Well, it's, it's interesting. I mean, in my. I'll just step back and say in my own day to day, in my own personal life, I find AI to be most helpful on subjects that I'm already well informed about.
- [32:03]
Alison Schiff
Yes, by the way. Absolutely. It's really about surfacing something I already know, but I don't have time to think about it.
- [32:11]
Jay Richman
Yes, that, I mean, I don't, I mean that's, I was going to ask but like if I know nothing about a topic, I feel like I actually get less. And if I'm a power user of something, I get to learn more, I get to go deeper and I understand where it's off and ultimately why it's off. And I think that that's true. In other domains like the creative arts, we see the best output being generated by, you know, creative professionals themselves or, you know, folks that have some amount of skill or have some amount of vision or picture and are able to manifest, you know, through the tool. And I think that that's going to remain. I also think about like photography, which I guess for some remains like affinity art, but back in the day was incredibly complex in needing to, you know, operate, you know, camera equipment and lighting and chemicals.
- [33:12]
Alison Schiff
Plates.
- [33:12]
Jay Richman
Chemicals. Yeah, my, my pop, I remember I grew up with a dark room in our, could you believe it? So a Manhattan apartment. And my, my pop dedicated one of our closets to a dark room. But yeah, priorities. He used to brag that he had more value in camera equipment than the, than his car, but I actually came to learn that it spoke more about his car than I do. But the point I was trying to land is that the best photos still, even in the day of smartphones, come out of folks that have taste, that understand balance, that can focus with all of the AI that's built into today's digital cameras, you know, bad photos are as prevalent as, you know, ever in our, in our history. So I think there's like something to having experience, prior knowledge and some degree of experience or expertise and being able to get the most out of the tools. And that's what I see in my day to day. And then that's what I see as far as, you know, early adoption of our creative tools.
- [34:25]
Alison Schiff
Then to go back to what you were saying during the first half about democratization which you're right, that's a word that gets tossed around. But I mean AI generated creative and other agentic tools without question. Those things are making it way easier for brands to run full funnel campaigns that look consistent across multiple channels and formats. It's really good for multi format ad production. And that's particularly important for smaller advertisers because they just don't have a lot of resources and taken together they're quite mighty. I mean small advertisers. That's why meta is meta, that's why Google is Google, right? I mean that's why Amazon is Amazon. So it removes the barriers to scale for them. And yeah, I mean that just makes sense. But looking ahead, where do you see the biggest opportunities for AI to open up more creative possibilities or business growth or whatever for smaller brands that weren't really possible as recently as a year or two ago?
- [35:31]
Jay Richman
I'm really excited about the TV space. TV and advertising on television has long been the domain of Fortune 500. Doing large upfront deals, buying large tranches in bulk and seeing that like open up to many more brands, resulting in a much more personalized viewer experience because it's been closed off, because it's been the domain of the Fortune 500 almost exclusively up until this point means that you wind up seeing the same ads on repeat. And I, I wonder now that we're at a place where we've got hundreds of millions of monthly active streaming TV viewers on Amazon alone, why the ads aren't as personal as the ones that I've grown to expect in my feed or in my search results. And I think in large part it's because the creative barriers are too high. The other barriers have been largely removed. TV advertising has become way more targetable, way more measurable through self serve minimum spends have, have gone all but away. And so the last remaining hurdle is then the creative and having something that is of quality, such that it can play in non skippable sound on full screen living room environment. And I think that once that last barrier comes out, which we're well on our way towards achieving, you open the floodgates to brands that have always wanted to make the leap onto the big screen, but, but, but couldn't. And so, and so, and so my head just like spins like thinking about that world and how much better like a viewer experience that becomes. I mean I don't, I don't know about like, I think, you know, the ads that I get on, on Instagram are like phenomenal. Like if I could go into Kind of ad only mode. I probably, you know, would it at times because it's really like just part of the overall experience.
- [38:05]
Alison Schiff
It's kind of like Vogue, right? When you're waiting in a waiting room and you look through a glossy magazine and you see ad after ad after ad for beautiful things and they're high production and they're quite beautiful and the.
- [38:21]
Jay Richman
Magazine is better for it. And I would love to see television get into that same, get to that same place. And I don't think that they can without the ads being much more tailored, much more personalized, more interactive, moving it from a one way to two way communication, which is some of what we're doing with remote controlled ads. And so I see all of the conditions are ripe and I fully expect over the next couple of years you're going to see a dramatic expansion of that marketplace, which will be great for creators, for viewers alike.
- [39:04]
Alison Schiff
Well, we haven't really talked that much yet about Amazon's tech and magentic. AI is this huge, huge investment for you guys and it's what you focus on every day. And you've released a lot of things. So just like no jargon, no self promotion, but give me the rundown on Creative Agent for people who aren't familiar and the broader creative Studio and yeah, like how it works, what it is, and maybe an example or two of this stuff in action.
- [39:34]
Jay Richman
Awesome. Yeah. Well, thanks for the opportunity. All right, where to start? So, Creative Agents. So we conceived Creative Agents. I don't know, maybe about like 18 months ago. I'm actually going to do an internal talk on the making of. But the mental model when we started on this journey was to try and recreate the agency model of old, the 1960s. We had all kinds of slides with Mad Men figures on it. Internally we call it Creative Agents Peggy, named after Peggy Olson kind of, but of course Don Draper 2.0. And we're like, okay, what if we could train a large language model to act, I think you used the term creative partner earlier, but to act as a creative partner, to act effectively as a, your own personal creative director, someone who can do market research, come up with new novel creative concepts, turn those concepts into storyboards and ultimately bring them to life by being able to access its own, you know, art, radio and television departments in the form of, you know, image, audio and video models. And so with that, so hopefully that, that, that meets the, the no jargon rule. But that's, that's what we try to do is like, we wanted to see, like, can we like rebuild that experience, do it in such a way that everybody had access, not just Madison Avenue types. And through now what amounts to like 50 some odd sub agents, we've got creative, we have this multi agent system with a creative director like at the helm that can reason about a whole host of requests in order to deliver ready to serve ads, including ones that are good enough to run on tv. And so this was thought of from the ground up. There wasn't a creative tool that existed that we were trying to figure out how do we bolt AI onto? We started with AI, we started with okay, let's build a conversational experience with no barriers to entry and then let's have that experience effectively generate the right user interface to go along with the job to be done. And so if you're asking to build an audio spot for Alexa, the UI will morph itself into an audio project. If you're asking to create a television, you know, commercial, it'll, it'll build you a video project. And by building a video project, I mean it'll create kind of scene by scene assets, scripts, voiceovers, you know, background music and allow you to have, you know, full control over the end, you know, experience by iterating at each layer at each scene or by editing it at the, at the, at the, at the very end. And so that's the approach and that was like our, our mental model. And I think I mentioned, you know, earlier, we've got, we originally designed this, I should say for, for mid market brands. Like that was the, the target audience that we're going after. Not the, the smallest of, of brands, not the, the largest, but ones that were kind of in the middle that are, you know, advertise doing like at a home advertising podcast, you know, advertising that like valued brand and investing in creative but maybe didn't have the means to go all the way up to video and, and tv. So that was kind of like the market that we had, you know, in mind. And I think we've been successful, you know, thus far in reaching that market. You're asking for examples of, I don't know if you're familiar with the company. Bird Buddy, is it?
- [43:49]
Alison Schiff
The Birdhouse Company.
- [43:50]
Jay Richman
The Birdhouse Company. Really, really cool set of products. Smart Birdhouse with a camera with AI built in that makes for the perfect gift. And they were going to advertise their product during father's day using image ads, was an early adopter of the tool along with their independent agency partner BTR Media, together worked on producing video spots that was served in Sponsored brand ads. And so they basically oh sorry about that. They went from an image ad planned to a video ad plan and they saw a 338% increase in performance by virtue of going from static to full motion. And they did it in time for a Father's day promotion they just wouldn't have otherwise. So this was kind of the point I was making around non consumption or expanding the market. In this example it would have been image and it turned out to be video. Video outperformed as its known to do for all of the obvious reasons. And so we opened up that opportunity by making this tool available. And then I mentioned I think the PNG example. They did an ad for dreft that has run on prime video that would have otherwise cost way more than free and taken way longer than days and involved way more than two people. So those are just a couple of examples from brands kind of medium and large and we're just getting started. So many more stories. If you invite me back to come for sure.
- [45:49]
Alison Schiff
How do you balance AI driven efficiency in creative production, which you've just outlined some great examples, with the risk of creative commoditization where the ads become a little too formulaic or they lose distinctiveness. I know they're pulling on assets that are brand specific, but there's this like an AI look that creeps into things and maybe that will change over time but it's a kind of like too perfect sameness that you see.
- [46:23]
Jay Richman
It's interesting. I do get this, this question quite a a bit but I don't see it. I don't see it in any of the results. And maybe that's because of our somewhat unique approach which doesn't start with a prompt but rather starts with the product being promoted. And from there we do where the agent does custom research around it. So it's, it looks at not just the details but all of the customer reviews. It couples that with world knowledge gained from the open web. It compares the product to other similar products in the category. It tries to identify what's stand out about the product and where there are needs or gaps in the market. Amazon sits on so much unique data. Like we can look at things such as, you know, category top return reasons and positive customer reviews, comparing the two to generate unique insights which goes into the creative concept which then carries through to the creative output. And because we're starting from a place of insight, every creative should be its own snowflake. I think we're going away from a world of sameness where every insurance ad has got its own mascot to one that truly tailored.
- [48:14]
Alison Schiff
My dog totally agrees with what you're saying, by the way.
- [48:18]
Jay Richman
Well, how big? How big a dog?
- [48:21]
Alison Schiff
I can't even hear you over his barking.
- [48:23]
Jay Richman
Wow, that's big.
- [48:24]
Alison Schiff
Should we leave this in the episode?
- [48:26]
Jay Richman
Oh, yeah, Oliver.
- [48:28]
Alison Schiff
There we go.
- [48:29]
Jay Richman
Oliver.
- [48:30]
Alison Schiff
He's an Australian sheepdog. He's made a few appearances on the podcast before, no doubt. Someone just walked past my door and he went into protection mode, and now he's lying down next to the dining room table as if nothing happened. Thank you, Oliver. Thank you for that. You were saying?
- [48:47]
Jay Richman
I love that you went with a. With a. With a people first name. My dog's Pablo. They should hang, but it sounds like they'd get on.
- [49:00]
Alison Schiff
What were you saying?
- [49:02]
Jay Richman
I don't know.
- [49:03]
Alison Schiff
I've totally lost my train of thought. I think you were talking about how you haven't really noticed that sameness that other people bring up to you because every ad is its own snowflake. I think that's where you were.
- [49:16]
Jay Richman
Yeah, that had a ring to it. I think I'll leave it there.
- [49:21]
Alison Schiff
Okay, well, last one. Since we are nearly out of time, I'm going to spring a version of my time machine question on you, because I usually ask guests something like, if you could travel to one specific point in history and make one strategic change that would alter the online advertising industry for the better, when would you travel to and what would you change? But I'm going to tweak that question slightly because AI is changing everything, like, everything so rapidly. What's one thing that the ad industry needs to make sure that it gets correct, that it gets right now, so that looking back five years from now, or whatever it is, we won't regret that we didn't do it or we didn't address it?
- [50:12]
Jay Richman
Oh, my. Well, that first question, it was incredibly nerdy.
- [50:20]
Alison Schiff
You can answer both if you want. You are more than welcome.
- [50:23]
Jay Richman
Oh, man. What are we doing? I mean, I think I'm going to go back to the point I made earlier, which is run towards the change, run towards the disruption. I think the biggest risk is sitting it out or waiting in late. I think that your competitors most certainly are going to take advantage, and the edge will go to the challengers if the incumbents aren't out there like Coca Cola, pushing the envelope on what's possible. And these things that start off as toys that are met with sneers that come with flaws, you know, today become the things that change the world tomorrow. And that just seems so incredibly clear from where I sit. To sit it out just seems like a grave potential mistake.
- [51:43]
Alison Schiff
Well, you know who doesn't sit it out? Oliver doesn't sit it out. He made himself heard. And Jay, thank you. And we should get Oliver and Pablo together.
- [51:53]
Jay Richman
Yeah.
- [51:54]
Rick Irwin
Today.
- [51:55]
Jay Richman
Well, thanks so much. This was fun.
- [52:10]
Sarah Sluice
This podcast was brought to you by Adstra. Adstra's leading identity resolution network enables precise targeting, scalable activation, and richer customer understanding. Whether you're looking for person based accuracy or massive scale, only adstra has the transparency to give you both. Learn more@adstradata.com that's A-S T R A data.com Sam.