How Puma approaches emerging technology like generative AI

0:04Hello, everyone, and welcome to the Modern Retail Podcast.

0:30which I guess, I don't know if Puma's known for its hat, but I had that in high school. And it was a very big cultural object for me back at the time. But Ivan has a really interesting job. The head of emerging marketing tech means that he gets to play with new technologies, test things out. I know that this week, or actually when this airs, it'll be last week, the company announced a pretty fun collaboration, which we'll get into that uses JetAI. But also just going through the history of different announcements, Ivan's name is always tied to some pretty cool new technologies. There have been things with Web3.

1:00There's been a lot with augmented reality. I'm always interested just to talk on the marketing side of things of when you test it out, why you do that, what are the KPIs, all of that jazz, and especially when you're working for such a well-known brand like Puma. But we'll get into all of that. How are you doing, Ivan? I'm doing great. Thanks for having me. It's so good to be on this podcast. Yeah, thanks for joining. We're excited to have you. So let's just start with who you are. So you've been at Puma for quite a bit now. Six years, is that right? Yeah, six years and a couple of weeks here, which is crazy to think about.

1:29It's the longest I've been with any organization of any kind in my life. So it's really exciting. Obviously, I really love working there. I've been blessed to have worked on some really fun and interesting projects during my time at Puma. But yeah, six years now in January here. And so what were you doing before you joined Puma? Yeah, so I spent one year at a small tech startup. But before that, I actually spent five years at the National Basketball Association, the NBA, working on social there. That was like the early days of social. When I first started, my first job was literally sitting in a room with 15 different TVs, and each TV was on a different basketball game.

2:01And my job is to post on the official NBA Twitter account and the official NBA Facebook. And back then, social was very text-based, which is kind of crazy to think about. It was a lot of just constant updates on scores and stats and things like that. And throughout my five years, we really got to ride the evolution of social. Vine became a thing. We became the biggest brand of Vine with over 2 billion Vine loops. And then we watched Vine die. Then we saw Snapchat come in, become a thing. I remember talking with NBA executives and being like, we need to be on Snapchat. And the NBA executives looked at me and they were like, why would we join a sexting platform?

2:34But to their credit, they always wanted to push the envelope. They always wanted to be the first to do something and try something. And we were the first big brand to be on Snapchat. We had some really big successes there. And that's kind of really bolded my career. And that's why we do things now, like we'll talk about AI Creator. My mindset is always to kind of be the first to try something. I really get those early learnings and then reiterate on the process in the future. Got it. Yeah, I was looking at, you know, your LinkedIn, which is always helpful when you're doing research about a certain guest.

3:01But your evolution specifically at Puma is that you started with social media. I think it was head of social, yeah, the team head of social media. And then, you know, now you're on the emerging tech side. Do you think that is the natural evolution of where things are going on that marketing front, where five, ten years ago, the next frontier was social media, and now we're to these other platforms using these other technologies? Exactly. It's really funny. Like I said, I was there for the early days of social at the NBA, and I kind of feel like it's a very similar place now with these new emerging technologies like AI, like VR and XR, where we're kind of like brands that are kind of playing with them.

3:39There isn't exactly a playbook out there of how you should kind of do these things. So it's a very similar time comparing like the early 2000s of social media with where we are now, these new technologies. And it's kind of funny looking at social. Like social has now become like the media. It's like really the traditional channel now. That's where most brands kind of focus a lot of their attention on. And now you kind of see some of these new technologies trying to, you know, become like different ways, different channels that brands can use to interact with their community. And how did you end up being the person talking about these emerging technologies at Puma?

4:12Was it just you were playing around and you had an executive's ear and was like, we should really be doing more of this? What was the story behind that? Yeah, I think that's pretty close to it. Again, I always wanted to push the envelope. So when I was the head of social here at Puma, I pushed the brand to join TikTok. We were the first major sportswear brand to be on the platform. We had a lot of great successes there. And actually, TikTok named us one of the most innovative brands on the platform. I think that was in 2020. We had a lot of great successes on Instagram as well. The first to utilize Instagram Reels when that became a feature.

4:41So again, just like always had that innovation mindset. And then we saw some of these new technologies kind of popping up. And I always kind of pushed to use them and they kind of fell in the social realm. But then we also saw like other parts of the business also kind of try to use them in different ways. So at some point it kind of made sense for us to have more of like a centralized group focused on these technologies. And one of the first big projects we actually worked on was building a Roblox experience. So we've done the whole gamut of building gaming experiences, doing things in VR, this AI project we just launched.

5:09So it's been fun to try all these different technologies and play with them and learn how they work. Yeah, well, the Roblox thing, it seems like you guys were very early to that because we're still hearing about brands, you know, a month ago, you know, earlier this summer, we're testing out Roblox when it feels like you were like doing it before. Can you just talk a little about like, is it a hard sell to do these tests at a brand like Puma? Or like, how like how much bureaucracy do you have to go through in order to say we should go on Roblox? That's the beauty of Puma. I think we are a large organization.

5:38I always try to say, like if you show the Puma logo to 100 people, probably all 100 people know what the Puma logo is. So it's like this well-known brand. But the great thing about Puma is like, it understands that it needs to always kind of innovate and be in these places where culture is happening. And we saw so much culture happening. First, like in youth culture, obviously Roblox is very popular with younger people. But then we also saw like concerts happening in Roblox and other brands jumping into the platform. So we saw more and more culture happening there. A crazy thing that was happening with a lot of the executives of the company is they're asking their kids what they wanted for Christmas.

6:11And instead of saying like a toy or video game, they were actually asking for Robux to spend in Roblox. And I think that was like an aha moment for a lot of people internally. Like this is such a big part of culture right now. We need to kind of do something in there. And that's kind of what kind of kicked off the Robux conversation. Got it. And do you, because your title is head of emerging marketing tech. And usually at most companies, there's a head of emerging tech and it runs the gamut of maybe we'll test out Gen.AI to streamline our internal systems, and maybe we'll do something on the marketing side.

6:40Is there someone else at Puma who's thinking about these emerging tech on the inside of things, or how does that all work at the company? Yeah, I think it bubbles up in different ways. There's obviously the product side of Puma. People focus on building products, and they actually made an AI hire there who's done a great job of bringing different AI processes into the design process. And Puma just announced a shoe that was partially designed using AI tools, the Puma Inverse. So we definitely see it bubble up in different groups. And then I'm just trying to centralize it from a marketing standpoint.

7:11So really the point of view I try to take is that how can we take these technologies and have consumer-facing applications of them? That's really kind of my perspective. That said, there are projects that kind of come in where there are more internal projects that might help streamline the business that I'm also a part of. And how do you think about, on the marketing side, I guess it's not a one-size-fits-all for every new type of technology you're using. But when you talk about social media, ultimately the point is engagement, right? And you go from engagement, it goes down the funnel, you'll get conversion.

7:42But when you're doing something like, I don't know, virtual reality, or you're doing something like Web3, I feel like the KPIs are a little bit different in terms of what is deemed a success. So how do you think about that when you're doing your first foray? Yeah, it's kind of hard because I feel like every technology is a little bit different, has different KPIs. I think for us, a big part of this is just making sure that we are innovating. And as these technologies become a larger part of every kind of day-to-day life that we're not straggling behind, we just want to make sure that like, again, like we don't kind of miss the boat on these technologies that we're, we understand them, we know how they work, and then hopefully we can kind of be a leader.

8:20And I think we succeeded that with Puma AI Creator. I think we built something that was kind of first to market that is very innovative and is kind of pushing the envelope. Can you talk a little bit about that? I would love to. I've been waiting for this question this whole time. But yeah, so we've been working on AI Creator for probably over a year now, and then really closely with Manchester City, the football club, for the last 12 months on this project. Basically, what we created is a co-creation platform that allows anybody to come in and use Gen.AI to prompt our system. And honestly, you can just prompt it in a few words, and then I think we'd limit it up to 300 characters.

8:53And then you prompt it, it creates a pattern that then projects onto a 3D model. So in this case, you can create a pattern that will project onto a 3D model of a blank Man City kit. And then you can basically create your own Man City kit from there. There's some customization options. You can zoom in on the pattern, zoom out of the pattern. You can move it around a little bit, add different elements to it. But that's kind of the core functionality. And sorry, for those who don't know, and I think maybe I might be the one, but what is a Man City kit? Like what's a kit for Manchester City? Yeah, so Manchester City kit is basically their uniform, their jersey.

9:24Okay, got it. Yep. And a really fun part of this project is that we actually have a voting system. It's kind of similar to Tinder where you swipe left or swipe right. So you can actually see all the kits that everybody else is making. And the winning kit will actually be the 2026-2027 on-pitched third kit for Manchester City. And so this is pretty much like I could type in plaid with dragons for lack of, and like that would then be the Manchester, like, how do you imagine the use case of someone typing in something and then creating these uniforms? Yeah, I think, I mean, that basically is it. You can type in play out with dragons and hopefully the AI makes something that's really cool and interesting for you.

9:58Obviously, it's still a skill. So people that have experience prompting will probably have a little bit of an edge and create something interesting and captivating. But basically that's really it. What we want to do is kind of break down the barriers. I think we've done some of these competitions in the past and a big barrier is that people don't necessarily have illustrator experience or Photoshop experience or they're not an artist themselves. You know, they're not going to take watercolors and paint their own Man City jersey kit. So we really kind of broke that down. We created this easy-to-use prompt system that people could just type a few words in and literally create a kit with under two minutes.

10:32I actually had somebody on my team create 60 kits in 50 minutes, which is pretty incredible when you think about it. We used a bunch of other tools that are out there that are a little bit more manual, where you kind of select different options and create your own jersey or kit across different sports. We did a deep dive into these different tools, And there's some really incredible tools out there, but they're very extensive. Like I was spending 45 minutes, 60 minutes, some of these tools to make, you know, one kit or one jersey. And the beautiful thing about Puma AI Creator is that you can literally do it in under two minutes.

11:03Got it. And like, what is what AI is it built atop? Did you build your own or using ChatGPT or how does that work? Yeah, so there's a couple of different AI models that we use. But basically, we're using Comfy UI, which is a node-based system that's linking multiple different AI tools together. But the largest tool that we're using is Stable to Fusion XL. And basically that allows for a wide variety of results and a lot of freedom for people to kind of type in different things and get what they're expecting. Got it. And what type of, I guess, user, because you mentioned earlier about the Roblox where you knew that people you worked with their kids wanted Roblox for the holidays.

11:41And that is some good sort of demographic test of like, this is who we want to reach and this is how we're going to reach them. How do you do that for AI? Did you do surveys where people said, I'm dabbling with AI or I want to use this technology to make new uniforms? How did you think about that? No, I think the learning came from somewhere else. I think we had two really big key learnings that helped us come up with the concept of the Puma AI creator. The first is people really care about their favorite team's jersey or their favorite team's kit. That is just such a big part of sports fandom and really who you are when you wear your favorite team's baseball cap or shirt or what have you.

12:15you're really kind of showing everybody else like what community that you belong to. And people are very passionate about these kits and these jerseys. When my favorite team releases their new kit for the season or their new jersey for the season, I have a lot of feelings about that. And when you go on social media and you look on Twitter and you look on Reddit about how people feel about new jerseys and new kits, there's a lot and a lot of opinions out there, you know? And usually, especially when Puma makes them, usually they're very good, but sometimes you get some negative feedback too. But again, people are just like really passionate.

12:43So that's the first learning. And the second learning is we've done some of these smaller scale activations with some of our streetwear brand partners where we invite consumers to kind of come in and co-create with that partner in Puma. And we've done some of these that were open to the public. And you get a line out the door. People love kind of coming in, creating their own custom gear. And again, they like to express themselves and who they are. And they do that through this customization. So, but the hard thing there is like, you can't really scale that, right? When you're kind of doing these like in-person events that have like these little stickers or patches that you can kind of iron on to these garments.

13:18But with AI, like you can really scale that. Like anybody can kind of go in, use this tool. It doesn't matter where you are. You don't need to be near New York City or Paris. You don't need to be near a Puma event. Anybody can kind of come in and use this tool and use it pretty quick. So we took those kind of two learnings. First is like people are very passionate about their kits. And second, people like love to customize and kind of create their own gear. that was kind of we had these big two learnings and then we saw AI technology coming in and it was very natural for us to be like all right we can't use AI for people to tap into these two passions.

13:47And do you view this as sort of a step one and if successful then you'll partner with other teams and do other similar things? Yeah I you know I think this is the first time we're doing this so we definitely want to see how it's going and we're only a day and a half in but the results have been incredible so far. And I think there is already conversations of what this could kind of look like in the future. That said, I think we still have a couple of days here to go. I think this is going to run until December 20th. People will have the ability to create kits and submit them into the contest. We really want to see what all the learnings kind of look like before we decide what the second iteration of this will look like.

14:21But there's a number of different ways that we can go. What are your metrics for success specifically for this? And especially you're a marketer. And so you need to sort of show there's been some type of an impact. And you're also only doing this with one team. So it's a little, you know, it's a beta test. And it's very, very circumscribed. What are the thresholds you need to hit in order to know that you can have a step two on this? Yeah, so we looked at some of the past Jersey design competitions that we've done. And again, those were a little bit more manual people had to design using either a platform like, like Illustrator or Photoshop or actually hand draw them themselves.

14:58We looked at how many submissions those types of competitions got and that helps us set a benchmark so i think that's part of it i think a second part is just working with the club i think manchester city's been such a club a partner for us and we love working with them because they are also pride themselves in being an innovator try and be a first in sports to kind of do things so a part of it was also kind of getting their feedback like are they happy with this are they excited about this and i think so far there's been nothing but green flags for them i think they're been very happy so that's been a really good KPI for us.

15:27And then we're also looking at what people are saying about this on social media, both from a media standpoint and also individual users. And again, I think we're getting a lot of positive signals. Yeah, I wanted to ask, especially when you're doing these types of collaborations with other teams, has that ever been a hard sell? Not necessarily for this one, but for past things where you're using some new buzzy technology that not a lot of people know about? I imagine maybe the execs don't know about Roblox. Were they a little trepidatious about going in on that with you? Yeah, obviously, we work with a lot of football clubs, but we also work with a number of other different partners.

16:01And we've obviously pitched these types of projects to other partners, not necessarily this AI creator project, but other types of projects. And, you know, there's partners that want to avoid these types of technologies and really want to kind of stay in the physical space. And there's other partners who are like, yes, love innovation, love trying things. So it really depends on the partner. And I think, you know, everybody has their own brand and how they kind of want to position their brand. I think working with Manchester City, they want to position their brand as a tech innovator, somebody that's trying new technologies and pushing the envelope.

16:29And that really aligns with what my group's working on. Got it. And can you just talk a little bit about, I'd be really interested just to know, because you've worked with a bunch of different technologies, some that have sort of had a lot of buzz and then fell. So like I think about like Web3, you've done stuff like that. Is that just part of your job? And when you're working with new technology that some are successes, some aren't? Do you still think that they all still do have legs? Just talk about, especially when you're in such a sinusoidal industry where some things rise, some things fall, maybe they'll rise again.

16:58What's your thought about that? Yeah, I think if I'm not failing, I'm probably not doing my job right. I think it's a little scary at times because you do invest a lot of time and sometimes resources into a lot of these projects, and you're not sure how people are going to react to them. But I think that's a part of this journey. That was a part of how we started in social media 10, 15 years ago. and I think as marketers and as innovators, you kind of have to do that. You have to kind of try new things. I definitely don't want to have a job that is very cyclical where I'm just doing the same thing over and over every 12 months.

17:31I want to be having a job that's pushing the envelope, working with our legal team and having our legal team being like, you want to do what? That means that I'm probably on the right path when the legal team is a little shocked. Yeah, you're one of the only people I've ever met that said they enjoy it when they have to go to the legal team. To be honest, we have a great legal team here that is very open to a lot of these ideas and they've been such a great partner for us to work with. So, you know, it's been a joy working with them and it's been a joy kind of getting their reaction on a lot of these and working through a lot of these projects.

18:00There's, you know, when you first like come up with the concept, you're like, oh, this is pretty simple. But then you start going through the legal checklist and other checklists that you have to follow through. You're like, oh, wow, there's a lot of things that we have to kind of really think through for these types of projects. Do you have a full-on sort of game plan in mind when you're like especially like let's talk with the ai for example what's going on now you're just going to try it out maybe it'll work with manchester maybe it won't maybe you'll if it does you'll work with another team or do you see like a whole multi-pronged plan where ultimately the technology gets refined people will design their own puma shirts using your gen ai that then they can buy or something you know like something along those lines like how how far out do you plan for this technology to potentially be a hit i think right now we have probably three different decks that depict like three different ways that this technology can go basically three different paths and maybe the solution is that we do all three paths or maybe we do two or three or just one of the three i think there's a lot of ways that this technology can be used in different ways of puma even like from an eternal standpoint it's like we can give this technology to our design team and when they're working across all our different clubs and designing kits, they could potentially use that tool to kind of give it to the club and be like, hey, for the first round, use this tool, come up with what you're thinking in your head, make it happen in this tool, send it to us, and then we can iterate off of it.

19:21So there's this whole kind of work stream that we can do that's just an internal work stream where this tool can really be helpful for our design teams. And then from an external standpoint, I think obviously working with other clubs or other partners that we have, I think that's an obvious one. But there's other things that you can do, like you talked about potentially creating a system that lets you create your own t-shirt. We could do that at events, so we could do that on the website. And I think there's other kind of big ideas that we can kind of go down with. So I think, again, a lot of different paths here.

19:48But the first time that we're doing this, we really need to be disciplined here, see what the signals are back from the audience, see what really worked here. And also be honest about the things that didn't work or the things that need improvement. And then kind of come up with a game plan for version two. Yeah. And what was the interaction? because you mentioned that you could be working with other teams internally. How much of the other teams worked on this specific project? Or was all of this built externally? How did this work? Talk about the overall background of building this. Honestly, the amount of people that have touched this project in one way or another is probably in the hundreds.

20:20Working with Manchester, we've been on daily calls with them, I think, for the last month. And then before that, we were talking with them two or three times a week for a few months. And then before that, at least once a week. So we've collaborated really closely with the Manchester City team there. And they've been such a joy to work with. I really enjoyed that relationship. And I'm excited to kind of see where we can kind of go with them in the future as well. And then internally, obviously, we work really closely with our legal team, marketing team, product team, our e-com team as well. And then as far as third parties, there's also kind of developers.

20:52Google Cloud was a big part of this too. Deep Objects helped power a lot of this technology. So, again, there's just hundreds of people that have kind of worked with this. I have, I guess, the honor of kind of talking with you and kind of being the face of this. But there's just been so many people that I have to thank and that have put in a lot of effort, a lot of sleepless nights to make this work. And how long has this been in the works? Like, did this take a year to set up? Yeah, I think we had the concept probably 18 months ago, maybe a little bit less, maybe 14 or 16 months ago. We pushed Man City maybe about a year ago, maybe a little bit more than that.

21:25Then we got rolling from there. Luckily, that first conversation we had with the club, they were like, yes. I was like, hey, I have this crazy idea. We want to create kits using AI. They're like, yes, let's do it. And then I was like, can this actually be an on-pitch kit? And they're like, yes, let's do that. So luckily, immediately, they were just giving us a thumbs up to do this, and we were able to hit the ground running pretty quick. Got it. And what would you say, given that you've now been a part of so many different projects and new types of technologies, I guess, A, what sort of surprises have you gotten from any of these projects in terms of like long-term marketing tools or way people are adopting things, but also like have certain projects had more legs than others, or there are certain lessons that you learned from doing an AR thing that might inform future marketing projects down the line?