[Music] [Tammy Pienknagura] Good morning, everyone. Last day of Summit. This is very loud. Thank you so much for being here. How has the event been for you? Thumbs up? Thumbs down? Yay! How was Bash yesterday? - Okay. - [Pete Choo] Okay. A subdued woo is appropriate for the day.
Well, thank you so much for being here with us. I'm Tammy Pienknagura. I lead Portfolio Strategy for our Digital Media Business for Enterprises, working very closely with our content supply chain teams. And I'm really thrilled to be here with Pete, my partner in crime from Accenture. We've been doing joint thought leadership for the last year or so. Right, Pete? Just trying to help our customers just navigate this changing world that we live in. So very excited to be here. Yeah. Thank you, Tammy. It's great to see everybody. My name is Pete Choo. I'm responsible for our Adobe partnership with Accenture Song here in the Americas.
And I know-- One, there's been a lot of discussion this week, but I think something that I'm really reflecting on from all the conversations we had with customers, with our partners at Adobe, with many in the venture community, is that it really feels like there's been a tone shift in the last 12 to 18 months. I feel like we were having a conversation about what we were capable of doing with AI. I mean, it really just feels like months ago. And now I feel like the conversation has shifted so dramatically away from whether or not we're capable of doing it and more about where we're going to derive the most value. And so we started this conversation really last year around how do we help brands navigate what we described as the age of content abundance. And I think that was really important in establishing some emerging trends around how the role of creative teams and content teams was necessarily evolving as AI became a bigger part of their workflows. But what we were constantly hearing from customers and leaders in marketing and brand was, I've done a million POCs. I need to understand where I can make some bets that are really going to derive scale value for my organization and that are really going to move us forward. And so for us, I think we're all likely in this room hearing many of the same questions from marketing leaders. How do I do more with less? How do I continue to maintain relevance at a time when it feels like attention and channels are only continuing to fragment and the battle to stay top of mind for folks is harder than ever? And then more and more importantly, marketing leaders under pressure to prove out the value of their investments. And so if we distill all that down, I think that actually end up coming down to three fundamental questions. We're keeping it pithy, I think, for the last day of Summit. The first was, why should we invest? What's the rationale? The second is, when is the right time? I think I was in three panels this week where the question was, when are the different various bubbles around AI and agentic gonna pop so I can really see what residual value exists? So is it now? Is it six months from now? And then the last one is, where do I focus? I think maybe show of hands, has everybody seen some statistic along the flavor of 50% of marketing activities will be automated by AI or something similar? Has anyone-- Yeah, I'm looking nodding heads, but heard a stat like that. I believe firmly with all credit to the groups including our own organizations that conduct that sort of research that we have hit a wall of fatigue from marketing leaders hearing that at the most gross level of their organizations. Great. 50% of these activities is going to be automated. Which ones? And to what degree? And what things will remain the domain of human ingenuity? And honestly if I have to invest in certain areas what technologies do I have to select? So I think there's a demand for greater precision that was really the backdrop for a lot of this research. - Any other thoughts, Tammy? - No. I think as we hope to be trusted advisers to our customers, we wanted to start answering those questions and really get specific into, "Hey, by industry or by phase of different processes, where are the differences? Where are the opportunities to just be a little bit more focused?" To Pete's point around the fact that there's all these scattered pilots and all these scattered experiments that may actually not be moving the needle, right? So that's the hope of this research.
And so I think just to do a little bit of context setting, with the group that we have, I think we can we'll do some Q&A at the end. But honestly, if you have a question or fundamentally disagree with us, please raise your hand. Let's make it a dialogue I think, for the last day. But I think to set a bit of context as we get into this, I honestly, y'all at Adobe, I feel like have had a really strong frame on this for some time. And one of our reflections as we are looking at synthesizing some of the research is that for the past probably 10 years or so, I think the belief was if we could get more robust and precise with our data, if we could get more programmatic in journeys and overall experience transformation, then we could turn the pipes on, the experiences would flow, and we'd realize the future state of real-time relevance, continuous personalization. I remember being embroiled in big debates as recently as five to seven years ago about is journey even the right term anymore because it implies a beginning and an end and we think actually that true personalization is about a continuous experience. And somewhat maybe intuitively or counter intuitively...
Nature abhors a vacuum, there's always going to be a new bottleneck. And so what we started finding is we were setting really the scope of inquiry into the researches. It's actually content that's become the bottleneck again. We have so many clients now who have incredibly, incredibly mature capabilities in terms of targeting and precision and journey orchestration. And we are talking to brands every single day who are saying, "I have the ability to target down to an incredibly precise level but I don't necessarily have the message or the content needed to influence that piece of behavior to ultimately drive people closer to my brand." If we say that, I think then that this has resulted in a couple of things that I think we're finding a little bit interesting. The first is what I just talked about which is content scarcity. And this is the pipe is on, the tap is open but there's not enough water to flow through it. And I think the interesting thing though is now we've been talking about Generative AI for what seems like centuries, but it's probably been at least the last year or two. And I think the thing that has come through is, well, fantastic, Generative AI is now the solution. We can produce more content than we ever could have imagined. We're going way beyond like Modcon and some of the things we've talked about in the past. Now we're talking about effectively no theoretical limit to how much content we can produce. Imagination is effectively-- Imagination and Firefly tokens are the hard limit on how much content you can produce.
But the problem with that is that despite some of these rapidly maturing capabilities, I think I would hope to see some nodding heads here. But if I told you tomorrow that I could increase the volume of content you have in your DAM by 10X and we could just start pushing it through, I imagine what would be going on your head is something like more akin to this build, which is, so then we will run into some serious problems if we are just unleashing the raw potential of Generative AI. We talked a little bit about the idea of the sea of sameness, but the thing that has come up over and over and over again is-- Show of hands, does everybody remember the first time they used ChatGPT? Yeah. It's a seminal technology moment.
I remember how magical that felt. I was in Grand Army bar in Brooklyn in Boerum Hill and I was at a bar with a friend and we were just hammering ChatGPT and it felt like magic. - Sounds like a fun night. - Yeah.
The follow-up to this talk will be there.
But I feel like actually as a society and really and maybe like a highfalutin way, like as a human species, different show of hands, do you feel like you can tell when a piece of content, text, or image is AI generated nowadays? Can you recognize it? Yeah.
More images, less text. Text is getting quite good. Yeah. We're at that Turing test level.
Yeah. Exactly right. And so the models will continue to mature and obviously get better. We will pass through that. I think we've actually gone quite deep through the uncanny valley and get to a place of verisimilitude. But I think that the challenge that we have here with the insane pipes-- By the way, this was generated by Firefly. And to my friends on the Firefly side, we should probably consult with a plumber or something like that. But I think the big takeaway there is if we just unleash this flood of content potential that resides in Generative AI, then we may actually result in actually greater brand dilution. Even if we have the guardrails up that can make sure that the content is brand safe, we as humans are getting so savvy to what feels like generically created generative content that there's a great risk that we massively increase the volume of content, but that impact doesn't scale a pace.
And so I think to tie that up...
We'll get a little bit into the study in a moment, Tammy and I. But I actually wanted to lead with the number one blockers or the top three blockers that we heard from our respondents. A little over 1,100 marketing creative and content professionals, both leadership and practitioners. We'll get a bit into that. One is really the thing that's stopping me is I don't have the value case right now that will allow me to justify making the investments I need to unlock this value. The second, I think that this is honestly trending in the right direction, thanks to commercially safe models like Firefly, is I'm not 100% certain that we can do this in a responsible way that protects the ethos of the brand. We were with a gentleman from Delta Airlines last night and he was telling us this very interesting story which is they're not using Generative AI on airplanes and things like that right now because there's this really avid community of airplane nerds. And if they put the wrong tail number on an airplane of a particular model, Delta receives a material number of complaints. So they're not quite ready to do these generative planes. But I think that's going to go down a little bit as the precision of the models gets better. And then the last one, which is surprising to us, I think is, while ROI is obviously intrinsically tied to cost, cost by itself is no longer the number one barrier. It's actually the third by a fair margin. And so I think what we're seeing is brands are not saying, "We're not going to allocate meaningful investment to this." What they're saying is, "Help us understand where the value is and we are willing to scale the investment to meet that scale of value." I'm curious if this resonates with the audience here. We're going to not try to do too much participation the morning after Bash, but yes, no, are folks feeling willing to invest if the right case is made? Yep? - Awesome. - Okay. That's good because we're still pretty early in this presentation, so it'd be a little early for disagreement maybe. But I think if we come back to this piece then, I think the thing that are these three questions. I think where Tammy and I wanted to then anchor us as we jump into the data itself is that original question, why should we do this? Across industries, we generally found that there's an estimated 7X ROI on a 3-year time scale for investments in AI specific to content creation and production. And we'll get into how that data was derived. The second is almost unquestionably across survey respondents. This is not a 'wait until later' moment. This is very much a 'let's move forward with where we are now because we cannot fall behind'. And then the last piece, which I know everybody is hoping that where is a very finite answer, is the where is it depends a little bit. There are specific areas. But what we found is looking at the total value across content creation production across industries was insufficient. We had to go deeper to the industry sector level to actually start seeing some of the nuances and where that value shows up. And perhaps unsurprisingly, not every industry is likely to realize value the same way. And so remain seated to learn more. But I think the big piece here is we are all in our relative organizations as leaders under greater burden of proof, not just analysis and contemplation, but proof to justify these investments and very much it feels like we're in a particular moment. Yep. So with that, I think because we talked about taking an industry specific lens, Tammy, do want to talk about how we approach that? Yeah. So no need to read every word on this slide, but we did do, as Pete alluded to, a very deep bottoms up econometric analysis to talk about, hey, what is the ROI? What's the value across industries and different pieces of the content supply chain? But we also wanted to be very conscious that we're not just doing this for the sake of estimating the ROI in a silo. We know that across industries, you guys are facing some very deep changes, some very deep shifts, and we truly believe that unlocking...
Content can help you in different ways. One example is we were talking to a bunch of CPG leaders the other day. They're under threat from all these micro brands who are digital first. They're engaging directly with customers, not being disintermediated by channel partners, for example. I see a nodding head. Maybe you're in CPG. So how can taking a little bit more control of your content, working closely in a hybrid model with agencies, for example, help you have that direct relationship with customers? Pete, I think you focus on a handful of different industries. Any other specific trends that may be interesting here? No. I have a soft spot in my heart for CPG, and so you just triggered a thought as well, which is there's the competition from these upstart digitally vertical brands. And then on the consumer engagement side and retailer partnership side, it's also become quite fragmented. There are so many new channels to engage in.
Back in the day when I was working very closely with the CPG brand, we used to talk a lot about last mile, right? So how do we ultimately get the product to the shelf? And I was in a conversation with this consumer goods leader who was saying, "We're not even talking about last mile anymore. We're talking about last millisecond." And it's the last millisecond, not just at the shelf, but in that moment where somebody's going to make a decision. I make all my most impulsive purchasing decisions on Instagram.
I think Meta's clearly aware of that. But I've been-- People are going to start targeting you right now. 100%. Because I've said it, my phone is right here. I expect to receive an ad at any moment. But I think to your other point, we also are deeply appreciative of the fact that we can't simply lift and shift solutions from something like an unregulated consumer industry to far more regulated industries like financial services and pharma as well. I'm curious your thoughts on that. Everyone's facing a different reality. So hopefully, we won't get into the full research here. Our people were published in two, three weeks so happy to send it all to you but there's a lot of depth there. Why don't we get into some of the research? So just in terms of context setting, we looked at the creation and production of content, that value chain in five different ways. There's five different steps that all relate to one another, but it helped us have a framework to just break up the value in these areas. So first, you have creative ideation and development. So how do you empower your teams on the creative side to just generate more ideas, a wider breadth of ideas with GenAI and actually create those hero assets, those core concepts that are going to help you really scale over the organization? Pete was alluding to regulated industries. Across the board, it's important, but especially there, reviews and approvals, collaboration across teams is important. After that, you can scale all your content to power different channels, different geos, to truly personalize experiences.
The fourth one is a little bit about companywide creation. I think you heard David on the Keynote on the first day talk a little bit about how everyone can now be creative but with guardrails. So how do you empower your entire org to create content that's on brand, power that last mile, millisecond of content? And obviously we don't want to just create content for the sake of content, we want it to perform. We want to make sure we understand what is driving that performance so that's the last step. Good? Okay. What our research showed, if the slide wants to move, there we go, is that across industries if you think of a $30 billion revenue organization we can scale this up and down.
There's about $190 million of annual value once you reach steady state in a content transformation. That's the 7X ROI that Pete was alluding to. And that value is really concentrated in a few areas, right? You can see here that pretty much skipping each of those five steps you see a little bit more green.
What we did uncover that was actually a little bit surprising is that half of that value comes from higher productivity, quicker speed to market. A lot of customers tell us we need to get to market faster, right? So a lot of those efficiency productivity gains. The other five have actually can come from driving growth which is one of the things that CMOs are being challenged to do right now. How do you drive more growth with less resources? How do you actually prove that you're not just a cost center, you're actually a growth engine for the organization? So not sure, Pete, if you're seeing any reactions from customers around this or maybe the areas of value that are most impactful. Yeah. It's one of those things that I think when you and I were first looking at the data, was very surprised and then tempted to go make the team redo the analysis. But then I think as we talked about it and then we started vetting this against some specific customer conversations, I think the two things that were surprising to me on first blush that now feel intrinsically true are, one, it's unsurprising that revenue growth in this value model is disproportionately associated with scaled production of asset variations because we have so many brands right now who are saying again, "I put the plumbing in place. I simply don't have the water to go through the pipe." So there are moments every single day for so many of my audiences where I'm just missing them altogether. Or there's a really important retailer and I'm not able to get my digital shelf cleaned up and consistent sufficiently enough that I can make sure that I'm the one who's getting that buy now box. And so that intrinsically feels true. The other one I think maybe that surprised me more in our conversation last night was really the creative ideation and development. I don't think of that as a commodity or process intensive thing that I can strip lots of repeatability out of. But again, I think as we've talked to actual creatives, as we look at some of the exciting things, I see my friend Zach there, but you can see Zach with some of the folks at Droga5 who are starting to get into project concept. I think what we're finding is actually chemists call this activation energy, but there's actually a huge amount of inefficiency tied up in starting with a blank sheet of paper or an infinite canvas. And AI is actually proving to be where it's not the final draft, potentially sometimes the world's strongest first draft for a lot of creatives. And that's an enormous amount of trap value. But I think we're really gratified to see this. I see some people taking photos. So let us know what your executives and budget decision makers think about this. But I'm excited that we're heading to a frontier now where, thank goodness, we can stop talking about AI purely in the reduction of human labor and all of the potentially troubling implications of that, and we're talking a lot more about growth. And that's the conversation that we want to be in. I think our research from last year, we're happy to send it all, pointed to...
Much higher value in reinvesting time and energy in higher value creative work, for example, than in just reducing resources, right? The other thing we wanted to point out, and again, this analysis is available in a few weeks, the actual value in the ROI does change by industry. It's still pretty high, so it's still a pretty good case for change.
But maybe not too surprising, we are seeing that certain industries can probably extract more value, especially in different places of that value chain that I described. So again, pretty interesting...
Insight. Our analysis did focus only on marketing teams and related teams like field marketers, sales teams that need to get in front of customers. There's definitely more value across the org, but we just scoped it that way. Okay, so let's get into the secrets and the depths of the research. So we wanted to break it down for you and show you by industry, the six industries that we evaluated where's the money, where does the value lie.
We also know that it's the third day of Summit and many of us may not be fully awake yet so we did want to break it down a little bit for you. Happy to send these slides again to all of you. What we uncovered was that across industries, as I said earlier, there were definite areas of commonality and where the value lies, right? In four out of the six industries that we looked at, that creative ideation and production was actually the number one area of value.
The combination of revenue and inefficiencies as we said.
The number two area of value almost across the board was that scale production of variance for personalization, localization, refreshing content more frequently. And then the last one was that optimized content per campaign. So how do you actually finally get insights into what's performing and why to actually create that firewall of content and performance from your insights. Pete, anything to add here or-- No. I think this is pretty consistent with what I'm seeing from clients. Certainly, I think the content optimization piece, I think we'll dig into a bit as well. But I think the idea of not just having volumes of content, but very carefully taxonomized content, having content that can effectively be dynamically tagged or where we can take content and derive its principal attributes and composition as we start seeing in the Adobe products is incredibly powerful because it's taking us from what, how do I get that piece of content into the market into a why. Why is this piece of content performing differently in AB or multivariate? Why is this performing better for different audiences? And I think that there's enormous promise there. Yeah. So let's give you a few examples. You may be familiar with some of them, but for that creative scaling and creative ideation, there's a really cool case study with Coach. I'm going to try to get this video to work. Here we go. Where they leverage a lot of our creative capabilities and slightly more advanced ones like custom Firefly models. So models you train with your own brands assets in a very safe way to create what's called digital twins. Through this, in fashion, apparently, like the creation of all these physical samples to test with customers, to test with leadership, it's incredibly time-consuming and incredibly expensive, right? So by leveraging technology and AI in this way, it was much easier for these design teams to make new ideas come to life, to share those with a lot of...
Teams that needed a way in, and to translate those ideas into something that would be much more realistic. Not only that, but they leverage these same models to do things like focused groups with customers to get feedback and then downstream to create content for social media campaigns and in-store campaigns. So we really see that role of creatives as not just being pixel pushers anymore but really creative technologies that help the rest of the organization scale. So this is one of my favorites.
I think that for the asset variants, Pete, you've been working closely with Gatorade. - They'd have a few sessions here. - Yeah. I hope some folks were able to catch the Gatorade session. I think on day one with Gatorade and the agency work and code, this has been out, I think, since around MAX last year. If you haven't had a chance to do it, I strongly encourage you to take a look. This is my dream as a kid, honestly, this idea that I could always get a product in my hands that was made by me for me. I'm an avid custom ink consumer, and so if you become buddies with me and make some joke or something, there's actually pretty even odds I send you a T-shirt with that joke on it in a 30 or 40 days later. And I think maybe the takeaway for this is really, we spend so much time as a brand or as brands hoping to maintain a certain degree of control, right? Control of expression, control of experience. How do we anticipate, predict, and produce content in all of the myriad variations that our customers will love? And I feel like what this really brings to life actually is that increasingly, I think the reverse is true. I think the truest form of personalization in certain brands and markets is putting that control into the hands of your customers themselves. And that's a very scary proposition. And this is only possible not only with commercially safe AI like Firefly, but this is possible through a lot of really deep and thoughtful technical work, close partnering with compliance and legal. And I think that the big takeaway here is when we talk about N of One as we have for many years now, there's a big question I think for a lot of brands, which is how much of that control do we want to relinquish and put into the hands of our customers so that it is a more co-creative process between brand and customer. And I think we're going to see a lot more of that in the coming years. I think what was really cool to see about this Gatorade campaign is there's a lot of bottles being sold. And then as I mentioned earlier, CPG companies like this one are really trying to have a much closer relationship with their customers. So again, there's a financial case for this change, but there's also a strategic case for this change. The last example you've probably been a little bit...
Aware of our GenStudio for Performance Marketing application, but at Adobe, we've actually tested this ability to not just create content, but really measure it, test it, experiment on it, activate it in almost real-time and get insights that allow you to understand, hey, this is what works for this audience. So this example here is from our own growth marketing teams that we're able to-- And I think this is last year's stat. There's probably higher now. Get something close to a 57% increase in click-through rates for targeted emails, through much better targeted tested content and things like that. So exciting things to learn about and again if this research is valuable another area to start testing.
I think we have a little time before Q&A, Pete.
Do you want to cover maybe some insights that are different across industries that may be relevant? Yeah, so if these were the three that were consistent, we found held generally true, I think there are two patterns we wanted to share with this group.
At Accenture this week, we were hosting a lot of industry roundtables. And one of the big takeaways actually is we had more cross pollination than I think we've ever seen. We had people from the health care industry coming to retail sessions. We had people in CPG going into high-tech. And I take that in a very collaborative spirit. I think that we're starting to view AI really as a team sport, but I think that there's actually a lot of hope with the leaders that I talked to this week that there's some uncovered truth or some insight based on the nuances of another industry or an adjacency that can serve them in theirs. So if you're outside of retail and CPG, I think something that you can be aware of or that we're seeing very consistently in that segment is far, far, far and away, largely because of SKU and channel proliferation. The highest area of impact overall in CPG and retail and consumer industries ultimately is about addressing content insufficiency. This isn't even about the most ingenious content in many cases. This isn't even about having infinite permutations. This is, "I need to serve every channel. I need to serve every audience segment. I need to serve every geography. I need to be able to get to market faster and localize a lot faster." And this is just really, I think, the hard reality for a lot of these brands. And maybe that's not the sexiest data point, but I think it gives us a really fruitful place to start because while we can do a lot of really interesting things across content creation and production, I think we want to make sure that we're surfacing the areas where we know there's concrete value and particularly a lot of top line value. And then I think the second one we'd call out is really, in some of these more regulated industries, I'd say both regulated and industries where the value chain is profoundly B2B, what we're seeing is that companywide creation is actually a huge theme, right? For the folks that are branding content in here, I assume that all the people in your organization follow your brand and design guidelines to the letter, that they never go outside. Yep. All right. They're perfectly reinforced from the PDF that's on the SharePoint that they are supposed to reference every time they create some bottom funnel thing that's just horrible.
But joking aside, I think in these industries, this pattern is very logical to us, which is, it's not-- I love David's point, which is, "Let's all think of ourselves collectively as creatives." But I think we just have to live in this reality that an increasing percentage of content in organizations is being created by non-creatives and without the oversight of brand. And so huge amount of trapped value here, but I think the idea is what's the role that marketing and brand would love to serve all these content requests. We simply don't have the capacity. But aligned to the previous research, what kind of role can we play as creatives, designers, and brand to use technology to help establish the necessary guardrails so that a greater percentage of content created by sales or even HR is the best expression of our brand as we would like it to be seen. And honestly even calls out or brings out some of those guardrails to say, "Hey, you might want to look at this. This is what it needs to be in order to be really great." And I think it's unsurprising that's happening in industries that are heavily regulated or B2B because that content pool represents such a significant part of their total content. And I think the last thing we won't touch on this today is we can get a little bit into summary and Q&A. I don't want to leave behind any of my friends from the health care or pharma industries. We do understand some of the broader implications of the MLR process. Unsurprisingly, I think review and approval is very much still a bottleneck today. I am encouraged with every six months, years or so that we're seeing an increasing number of players who are seeking to streamline and drive better automation and tooling in highly regulated approval and review processes. I think that'll be a big breakthrough in the next, hopefully, couple of years. But in the meantime, that's a big bottleneck for those brands. And so part of what the research shows is there's a lot of amazing content ideation that's basically getting trapped upstream because the funnel is just too narrow when we get to the regulatory approval process.
So yeah, I think that maybe I'll hand back to you as we think about how to summarize. Yeah. So no, I think we wanted to end with bringing it back to these five areas of value. We hope that this research gives you and your leaders a good place to at least start thinking about what to prioritize, right? And obviously, we have products that line up to this. We have expertise in transformation, change management jointly with Accenture.
But the answer to that question of where that Pete asked about a little bit earlier...
Is it depends, but I think there is value in being focused as you try to continue doing experiments but in the right place and then scale them across your organization. So as I said earlier, we will be publishing this research in-- What, about three weeks there, Alyssa? Yeah? Don't hold her to it. That's Alyssa and my team. But if you want to, please, please, please scan this. There's a lot of depth by industry into both the strategic and the financial estimates that we collaborated on with Accenture, and we'd love for this to be a really helpful resource for you to make decisions and keep progressing.
Last thing before the Q&A, you've probably seen this in a few other sessions, but we'd love to hear your feedback, what's resonating, what isn't, where actually would you like to dig more. We intend to do a lot more research in different spaces, so we'd love to hear your feedback and hopefully you win one of these wonderful prizes.
Thank you.
[Music]
