[Music] [Hannah Elsakr] Thank you guys for joining us. We are not going to make you stare at the screen with slides. So if that's what you wanted, you're not getting that in this session. We're just going to have a conversation about how GenAI is transforming how brands and agencies work. I'm Hannah Elsakr. I'm the Head of Global New Business Ventures at Adobe and proud Founder of the Firefly for Enterprise products that you saw on stage yesterday in Main Stage. And I'll let Treff introduce himself. [Michael Treff] Hey, all. I'm Mike Treff. I'm CEO of Code and Theory. - Thanks for coming to our talk today. - Terrific. So we're going to self-moderate each other, which could turn into a disaster. - I think it's going to be great. - It's going to fun. It's going to be great. So we have some questions that we'll just riff off of each other, and then we'll save plenty of time. It's a small group. If you have questions, we'll put a mic up and just start answering questions as we go. Great. So I think the first one was really about, hey, there's this thing called GenAI. It's changing really how we think, it's transforming creative and the creative craft. And maybe before we get into how agencies and brands work, how do you just see GenAI just changing how Code and Theory does approach things like creativity? Sure. Well, good question. First, let's start with, if you're a creative human being and you're not excited right now by having all of these new tools at your disposal and having all of these new ways to be creative and have all of this innovation in front of you where you can really do things that you could never do before, faster, better, in ways that you never imagined, you're probably not a creative, and you're probably in the wrong industry because as you know-- Yeah. From Firefly, we are at a moment where things that were not possible are now possible. And it's not just the generative AI piece where people look for efficiency and people look for performance, which is really important and part of the plan. But the way in which you can create and imagine it, like experiences, solutions, connections and relationships with your customers as a brand, that changes everything. So as an agency, we try to incorporate that. And it started for us with just curiosity and experimentation, which is I'm sure how it started for you. Yeah. I mean, for us, I mean, Adobe is obviously has a lot of surface area across marketing, creativity, and document productivity. And this conference, the heritage, many of you have been coming for many years, is really about MarTech transformation. And what we saw over the last couple of years is lots of companies have invested in laying pipe, CDP, Journey Optimizers, Analytics, all of that great stuff. But then they're realizing they have a bit of a problem filling the pipes with enough content and on-brand content that's engaging. You add to that the fact that the speed of social, you guys may have told me this too, but our assets pretty much go dead on arrival after a week and multiply that by the number of channels that you need to refresh. I've actually heard one of the brands tell me that they have analytics that say, if your consumer has seen the same creative three times, they get really upset with the brand and they disengage from the brand. So the worlds are melding together. So we spent 10 years on MarTech. I think we're on the early innings of what I would call like, there's no creative tech word, but like the stack, the creative stack is starting to get innovative. And agencies like yourself and other partners are super leaning in because to your point, taking the drudgery out is going to free you to do better creative work, better marketing, and be right there for your customer on time. And I think that's a critical piece to your question of how does it affect an agency like ours, like we have to start asking ourselves, I think all agencies should start asking themselves, "How do they want their humans to spend their time? What is the highest value for the humans for the clients of an agency?" And that's a really significant question. Because it's not just the obvious, like software can do things that it once took a lot of humans to, so it's faster and cheaper and it's better. And it solves the content supply issue that you're definitely demonstrating and talking about. But really having an understanding of where your skills and your people and your talents can create value for your clients, and then how do you want your people to spend their time doing so, that's the place you start if you're an agency and you're thinking about what to do about this. Because then you change the way you work. Then you change the way you engage your clients because you're-- Excuse me. You're getting out of the drudgery of production and delivery, and you're getting into how can I find high value opportunities to really drive your business as a client. And then those relationships start changing. And the market, if I can say, is going to force that. Because, look, we're here at Summit. And to your point, there's tons of technology and software to be bought and that people are buying that solve a lot of these problems. Clients aren't going want to pay twice for these types of things. They're not going to want to pay for the software and then pay the agency to go and operate the software to get the same output. So it even changes that, which we're seeing, which I'm sure you're seeing too, in the way you're engaged with your clients about what they want to buy and how they want to buy it and why they're buying it.
- Sorry, I'm having a coughing-- - Then let me keep going. I'm going to get-- So that's part of it. So then after that, if you're an agency today, and to the question of how is changing agency-client relationships, like our clients, most of our clients are at a place where they're looking for what we call like phase one. They're looking for efficiencies. They're looking for cost savings. They're looking for you to do more with either the same amount of money or less amount of money from just the deliverables perspective. So if you used to pay me X dollars to make Y things, now I want to spend either X or X minus some percentage to make Y plus some percentage of things. And everyone's shaking their head. This is very affirming. Thank you. So we understand that. There's different expectations on the relationship. And so everyone is in that phase one of efficiency and performance. And frankly, the software is going to solve most of that. That's what Adobe does best. Can we talk about, I mean, the elephant in the room and the early days, why some brands, agencies have leaned in but some brands haven't is, well, what about human creativity? - Yeah. - How does that change jobs? Am I going to lose my job? We also hear there's a tale of people who don't want to use it at all. Yeah. So how do you think about that and upskilling or process change in the work? It's a great question. And it's a real question.
Let's break it down into a couple components. So component one, just the resistance to it. I would say that's futile. Because just like all technical innovation, you can resist it as long as you want. But to quote my friend Dan Gardner over here, founder of Code and Theory right there, he says technology is undefeated and he's 100% right about that.
Show me the technology that lost over the arc of time. The app, we don't use the abacus anymore. - I use Excel. - Correct. Because it's been replaced by better technology. Exactly. So that's just futile. So the resistance, you either embrace it or yeah, you're going lose your job. So I believe, I'll speak for myself. This is not a Code and Theory position, but it's a Mike Treff position. If you don't embrace AI as someone who works in a creative organization, as an example, you're not going to be working in that creative organization for a long time for that much longer. The people who do embrace it are going to find new ways to do it and new value to unlock. So yeah, then there's the fear piece. So there's resistance and there's the fear. Fear is just about, I think, familiarity and curiosity and being interested in learning and interested in doing new things. And hopefully, you build an organization with a culture like that, and you have a legal team that lets you. That's also important. But a lot of that fear is also going to be, like the legal side fear things, that's going to just be removed because all of this technology is going to either be in your workflow or there's been no real legal issues anyway. And plus, if you're using Adobe, you're fine. - Plug. - Yeah.
So then it goes to the-- Are people going to lose their jobs? And I just touched on it. But yes, people are going to lose their jobs. And yes, people are going to have to find a different way to create value for themselves and their clients, and that's just true. Now the beautiful thing about new technology is that you don't have to be upskilled and retrained to just experiment with AI and to be someone who's on the cutting edge of it. Yes, if you want to do engineering and implement platforms and stitch the data together, yeah, they’re skilled there. But also, with any new technology, new jobs are created, right? So things that an agency like ours used to not do, we're now going to do. - And that's going to create opportunity. - Yeah. Our CEO, Shantanu, likes to say, "Humans won't be replaced by AI, they'll be replaced by humans who know how to use AI." So I think that's literally what you just said. Exactly. Yeah. I'll give you like a little story. I've talked to some creatives who were junior designers when Photoshop came out, and they told me that there was the same resistance. There were people who were like, "I'm sticking with my sketchbook. I'm not using that thing." And they raised their hand to learn Photoshop like they were in a bit of a minority. In the age of digital marketing, who do you thinks career maybe accelerated more or less? It's going to become ubiquitous at some point. People said that when the printing press came out, people said that when a camera came out. - This is all true, right? - Yeah. - So can I ask you a question? - Of course. So at the beginning of Firefly, you have experience with that. So you dealt firsthand with the first wave of fears around this. And when you were creating it, what was your guiding principle for what you were trying to solve for? So as a company, obviously, creativity is at our center, the creator community is at our center. And so we made a very conscious choice to come out with models that were only trained on licensed data. So the way we came out was we felt like we wanted to come out right to market. So I can say, and my legal team has my back, every piece of content that went in, we had license to, we know where it came from or a creator got paid for that content. That was really, really important to us. So that was just a philosophy that we had. The other thing is, I think the model, so commercial safety obviously then parlays into brands getting very comfortable, agency partners getting very comfortable. Depending on where you are in the workflow, if you're ideating and that thing will never see the light of day, use whatever model you want. But I have increasingly customers that come to me and say, "My legal team is asking me to rip and replace the thing I was using and use Adobe because at least I know where everything is. With Content Authenticity, I can also trace every piece of content." So that was the principle. The other principle is that we do have a large surface area of modalities as well. So image was the first set of models that came out. You've seen the video models that have been released in beta and playing with those. 3D, vector, all of these modalities start to build on each other. And I think it actually democratizes for an individual creator who can now not just be a specialist in Illustrator and Vector, can actually work across modalities, which is pretty cool. Yeah. And so great. So I'll pick up on what you're saying there. To the question earlier about how is it changing the agency-client relationships, I believe, rightly or wrongly, clients are going to expect their agencies to do more. And I don't just mean in the delivery mode. I mean, when you used to go to me because I was not me. But when you're a social agency X, and you were my social agency and that's what you did, like channel agencies or type mode agencies to pull on multimodal, that's going to recede over time. Because the interconnectivity of all of the content, of understanding users, of where you want to distribute it, of versioning, of all of the things that AI can do really fast, if you want to take the real halo of AI, whether it's in the generative piece, or in the compute piece, or in the segmentation and distribution piece, whatever it may be, if you want to get the value out of that, there has to be a convergence across the client organization of what used to be discrete briefs. And anyone who's worked in an agency or worked with agencies knows what we're talking about. You're not going to have a future where it's like, this is the brief for the website redesign, and this is the brief for the social thing, and this is the brief for the big hero campaign. These all need to connect. And so if you're an agency, you have to think about, well, what are you doing with teams with skills that you're building to be able to respond to that? Are you an agency, like Code and Theory, where we work across the C-suite. We work across all of those different modes because we can't be successful in providing value to our client if our CMO and our CTO are not working together. It's just not possible anymore, especially because of the software and because of how horizontal that gets. So it does change the expectation, I think, that clients will have of agencies in their relationship with them. And we won't have single purpose player for that much longer. So maybe expand on that more. So how does that change the actual relationship between Code and Theory or the other agencies with the C-suite? So I'll just speak for Code and Theory. What we have done is we've brought in expertise and built real solutions and practices around doing this. Looking at you, Cory. So we built an enterprise experience transformation practice. What does that mean? It's a lot of words. Basically, it's to understand how organizations of today are looking at their MarTech and looking at their ways of working and looking at their capabilities and seeing what the gap is to where they need to be tomorrow given the revolution that's happening in software. And most of that, that's like 50%. We can argue about the percentages, but that's like 50%, the software and the MarTech, and it's 50% the way of working in change management in the organization. And we need to be able to do that. So we're building a practice to be able to do that so that we can partner with our clients at a strategic level to unlock value. Because ultimately, this is all about value. Our client organizations are spending more money than ever on software and on MarTech and all of the other pieces. They're also spending a lot more implementing all of that. They're also spending a lot more on the change management around that and changing their skills and their people and their ways of working. And at some point, there has to be an ROI on that. And so when there's an ROI on that, the partners who can go upstream, be consultative, lead that change effort, handle it through and through, and also, by the way, still do the things that agencies do really well, make creative, make experiences, build communications, launch brands, build websites, build apps. That being all connected is going to be what is expected. So if we want to be performing for our clients, we need to be able to do that. It's interesting. I'm just going to pick up on ROI. I think when GenAI first hit the scene, it was probably late '22, early 2023, and everyone was in play mode. Nobody was thinking about AI. They were just like experiments. Now what I'm finding at the end of '24 and '25 is people are trying to lean in, but now you're talking with your CFO. Right. And your CFO is like, what's the business case for the millions of dollars you want to invest? I'll just give you the Adobe example. We like to be customer zero on our own products. And last year in the summer, our CEO, who was our acting CMO said, we're going to implement Firefly Services for Black Friday. And interestingly, the team was a little resistant because there's a process and there's a flow they're used to. We went through that. At the end of that, we saw over 50% efficiency in time saved and the team was actually happier at the end. There was some pain, no doubt there was some pain in the middle. But when we start to see like 50% returns on time, people are like, spend the money. It's a massive unlock, right? And you saw it. You did it on yourself, and that's a really credible way to then take it to your customers. That it's a real thing. But your point on the CFO is another great point for agency change.
Very rarely in the history of agencies was there a direct conversation between like an agency lead and a CFO of a client about where their business is going and why it's going that way and what you can do to be value before. But we have to be fluent in CFO speak now. We have to build business cases around what we're doing and why we're doing it. It's not like, is this a great concept for the ad campaign, or do you like the UX of this PDP? It's like, show me the business case for why I need to implement this software, put the content supply chain in, retrain my team, and then show me where that return on investment is going to be. And like an agency, we didn't speak that fluently. If you'd hired Code and Theory in 2010, that's not how we talked, right? But we have to learn how to talk that now because it is all converging around what that investment looks like. And so if you have proof cases where in your organization, where you can say we did X, productivity increased by Y, cost was short-term and amortized, and here's where the return comes in. If you're an agency and you're able to have those types of conversations, you're going to win. Yeah. Maybe we can talk about specific use cases that we've implemented. Sure. I think we have one more sizzle that we'll roll and then we'll talk about it. - Yeah. Yeah. - Let's see. [Music] That's pretty dope. And I'm totally going to ignore all of our competitors that were in there. - That's fine. - I cleared this with you in advance. I know. I'm kidding. It was a good joke. It was a good joke. I know. It's for the crowd. It's for the crowd.
So let's talk use cases and an actual customer example. You go first. Those were some of ours, but you please-- Well, yeah. No. Please, please. So in our case, I think what we're seeing working directly with enterprises is they definitely feel like the bar for creative is going up and up and up back to-- CMOs are asked for 5x the content they were asked for last year. Their friends, the CFOs are saying flat to down budgets. That's a tough math problem to figure out. And so we've started working with everyone from Mattel. You could see 4,000 toys a year. The name of the game for them is decreasing the time-to-market. They're using it in packaging, innovation. They're using it in ideation for dolls and shrinking that supply train down tremendously. You can also see Paramount. And that example, if anyone has kids, maybe you saw the movie IF...
It was the one key point of IF is like imaginary friends. And so they wanted to engage the customer and created a unique experience where you could generate your own imaginary friends in the style of the movie IF. That was a very high social engagement campaign for them. So those are some of the examples that we're using. Gatorade's another one that we've done with the custom AI bottles. But people are definitely leaning in and unlocking the power of AI. 100%. So from our perspective, all of those use cases apply. Anywhere personalization can be applied in a piece of content, distributed, owned channels, paid channels, doesn't really matter. Certainly, anything that affects supply chain and you can do like, on bottle as you mentioned-- - Yeah. - That's a huge use case. I mean, even Coke talked about that yesterday, right? Where we think it's going, however, is different. The content, volume, and personalization at massive scale and even anticipatory content that's different. If you think about it, personalization is reactive. I know a data point, so now I can do something and push it out. But AI is pretty good at compute and thinking about anticipatory models, so you can be ahead of even what a customer might think or even know or even indicate that they might want. That's a huge piece of it that will play out in content. But we really think the next phase of use cases are in experiences, are in actually generative experiences, like generative UX, generative UI, where your experience, I'm just pointing at you because you're right here, your experience on a PDP, for example, that PDP might look completely different and nothing like my experience on a PDP because you came in from a different place. It knows so much about you because of you've been there before, what you've looked at, etcetera, your triggers to buy, whatever it may be.
Generative experiences are the next piece. Because then that's the use case where all of a sudden, it's not just the content still in a push or a pull, but it's everything surrounding the relationship with that customer. Is that generative or really great at the edge assembly? Great question. It's both. So it has to be both. It's generative. - I agree with that. - Yeah. Because generative still takes like nine seconds to generate. Right. Depending on what it is. Yeah. Exactly. But it has to be both. So in the future, where we're building it and where the use cases are going, there are these systems that you're going to have to have. You're going to have to have a brand system that understands your brand, a model trained on your brand. It has all your input. It knows how to behave. It knows how to speak. It knows your look and feel. It knows your tone. It knows all of these things.
It generates your content ultimately. So the brand and the content live in a system. Then there's a design system. And I don't mean it in the traditional sense of what we have today where it's a set of modules. But there's atomization of what experiences get built out of, to your point, where it's the components of UX and UI, the fundamental components of it, the understanding of how it will orient, is it on a mobile or is it on a desktop, what are we trying to do on a page like this, etcetera? So then there's your design system. And then there's your data system, right? - And your data system is the triggers. - Yeah. What am I looking at in order to compute and compose all of these things? And those are the use cases that we're starting to build out now, which for some reason...
Some clients are more open to that exploration than the content exploration singularly. Because the content exploration, it's like, there's still a sense of control on it that maybe they still want. Whereas generating or composing an experience in real-time, that's a natural thing that technology makes sense that it should do. Yeah. So I'd like to draw this differentiation because people want to talk about generative, but sometimes what they really want to talk about is smart assembly, which means that in the DAM somewhere, you need to have enough atoms to mix. Yeah. And the secret is most consumer brands, they don't really personalize against more than a couple of segments. - They're not doing N equals 100 segments. - Right.
So the idea that you could actually pre-generate and I think on-brand enough to remix is part of the unlock. I could see why you then would retain control because you probably do have a review and approval cycle for the 5,000 assets that you're going to have. But it will feel very unique to me when I show up. - Correct. - Yeah. Correct. So for the use case piece and, again, back to the agency-client relationships...
These are not conversations that you were having or even having to understand when the brief was which was make my PDP look better and move the button here or here or here. What does the analytics say about where is the best button-- No, no. The best button needs to be a little rounder. - Yeah, exactly. - That's usually the feedback. Exactly. So these are totally different types of conversations. Yeah. And to the point we were having earlier, this is a conversation where the CFO has to be involved because it totally changes the way they appropriate technology and CapEx spend and all of these other pieces. The CMO has to be involved because they have a huge oversight on it. Probably the product team has to be involved because generally sites can be owned, especially if they're commerce based, can be owned by product teams. Right. The analytics piece in the CIO has to be involved in that because how do you get the data model to be right unless you're plugging in all of the right things to it and making sure that that works? And so the future relationships between clients and agencies, even unless you're fluent in these things, I don't know how you don't keep getting squeezed on the dollar, like if the agency keeps getting squeezed on the dollar side because people expect this stuff to be cheaper and cheaper if you're doing it the old way. That's fair. So we've talked about experiences, web experiences. We've talked about physical products, we've talked about content.
Are there other use cases that you're seeing in motion? Yeah. I mean, for sure. I mean, we use GenAI in research more than we've ever used any other dataset historically. So we do things like we will build personas of the key segments of our clients. So if our clients say, "Hey, these are the six segments we're going after in this conquest campaign, etcetera, etcetera." We'll build models for all of them. We will then run our research and insights against them to get to a brief. We will then run the creative against the models, against the personas before we even deploy any of it. So we have pre-optimized creative before we spent $1. And we found that the AI personas are generally more accurate in predicting performance of the creative than human feedback and traditional research. So we find applications in research all of the time. We find applications in just using it as compute all of the time, to just be able to synthesize things much faster than we ever could. So for example, you win a client, you got to onboard the team that used to take weeks, sometimes months, to get everyone through it. Now there's a brain right here that can onboard you immediately and be your copilot on that account. Yep. I'm excited about the video use cases and the 3D use cases. So if you've used our beta of the video model that we have, you can do Text to Video, which I think is a great use case mostly for media, but for agency and for brand, I love the Image to Video use case because you can take a still of something that's approved, use it as a reference in the model and get a five-second essentially video back. We all know video is like nine times more engaging than a still. And usually, you don't have a budget to shoot video. So I love that application and then just take it one step further with our 3D technology that we're now playing with. If you give me a still or two stills, I can create a complete turntable of that product. We also know in PDPs for commerce, like if you've got a turntable...
People love that stuff. Localization is another one that I think this opens up lots of vectors for. We usually only support the top couple of markets. Yeah. Yeah. And I think with generative, great translation, we have avatar, lip-sync technology, dubbing, you can start to reach markets that you probably would not have natively reached. There's no question that's true. - Can I ask you a question? - Yeah. Have you found...
It easier to engage end client-user organizations with additional use cases and expanding deeper into their AI journey or agencies? Oh, I'm going to say yes. No, I'm kidding. So I think early in the journey, there are a couple of brands and enterprises that are very lean forward. And therefore, they were interested in experimenting at a serious level with us going in. What I would say for all of our agency partners Is like, you grokked it.
This was a tectonic shift for the agency landscape and you all grokked it pretty quick. I would say Stagwell, the Stagwell family was one of the earliest people to come and say, "We got to figure out how to use this." It often helps if we both go in to talk to the customers because there's still a pretty big education gap. Probably everyone in this room has used many tools, understands lots, but if you haven't used it, there's an education gap. And there's a gap on understanding the ROIs and the use cases where it applies. So I would say you guys get it easier. You will also find just generally the creatives in your teams are embracing new technology and more and new tools. That makes sense.
That makes sense. So then I guess another question would be on top of that is...
So as the agent-- You just made a great point about it's easier to pitch if you go in together with the change management. Yeah. - It's like we went in together to-- - I mean, not literally together. - But like if we're both in same-- - Yeah. Coordinated on it and we're going in together on it, it's much easier to get an end result. That's another big change that you pointed out in the agency-client relationship is, we're supposed to show up with technology and partners and our partner ecosystem in technology that is so much deeper than it ever was historically. Yeah. That's totally true. We would just be like, "Hey. Buy some Creative Cloud. - We'll see you later." - Right. - Yeah. - And now we have to go in together. So a huge change in the agency relationship with the client is now they expect us to validate the technology and show up with the partners and bring the right ones to bear and act in an advisory capacity, but also having the partner relationships where if they're going to implement a new piece of technology on the client side or we're to use it for them, there's real liability between-- I don't mean legal liability but relationship liability between the client, the agency and the tech partner, and they're in that together. - I think it's changed our relationship. - 100%. Where we're developing to get-- You're teaching us, but though leadership wise, I think brands are asking us, "Where do I start? How do I start?" And that has probably changed our relationship. - Definitely. - You're definitely inside under NDA. You know our roadmaps now, whereas maybe before you wouldn't. I understand your business model and your client base and how you serve them so much better. - All right. - Thank you all. - Anyway, thank you so much, guys. - Yeah. Thanks for your time. [Music]
