What Frogs And Socks Can Teach Us About AI Implementation

0:12Rebecca Sykes: Welcome to Ad Exchanger Talks, the podcast devoted to examining the issues and trends in advertising and marketing technology that matter most to you. Today's episode is sponsored by Basis, the leading intelligent operating system for autonomous advertising. Its enterprise AI solution transforms campaign briefs into strategies and media plans that integrate directly into omnichannel activation. You can learn more at basis.com.

0:49I'm Alison Schiff, Editor-in-Chief of AdExchanger, and you're listening to at Exchanger Talks. My guest this week is Rebecca Sykes, a partner at the Brand Tech Group, where she leads emerging technology, including generative AI. As brands rush to make more with AI, Rebecca is noodling a harder question. How do you scale the work without losing what makes a brand distinctive? We'll get into what it takes to make AI work beyond the demo and why the answer has as much to do with people and process as with the technology itself. But first, I'd like to take a moment to tell you about our upcoming Top Women in Media and Ad Tech event.

1:29Join us on the evening of November 3rd in New York City as we celebrate the leaders, innovators, and builders moving advertising and media technology forward. The Awards Gala will take place at the Lighthouse at Chelsea Pierce, honoring an extraordinary group of women across AI, commerce media, privacy, programmatic measurement and ctv i'll see you there hey becky welcome to the podcast thank you so much for having me well you know the drill what's one thing about you that not a lot of other people already know um so the thing that not a lot of people know certainly not from looking at my linkedin is that despite being working at and then and then leading the company uh mofilm for many years which is a um kind of decentralized content creator network i have missed most of the seminal movies of the late 80s and 90s it is literally like the black hole in my kind of cultural reference point so um it was it's kind of a challenge actually in in that job uh name any of the big ones star wars jurassic park the goonies pulp fiction none of them and i don't really know how this happened or where i was at that period of my life but it's it forms the sort of central references that you're that you'll encounter when you work with any kind of filmmakers so my my kind of film education happened in reverse I watched thousands of pieces of content made by young filmmakers and you know uh sort of aspiring climbing filmmakers uh and then I ended up watching the classics in in reverse um but yeah I've you can test me if you're like what's your favorite what's your favorite movie formative movie I bet you I haven't seen it have you seen E.T.

3:14okay you've got me okay I did see E.T. I saw E.T. but there are many many many that I haven't my husband Jurassic Park and Star Wars are the two that really get him he he can't believe that's uh you can rectify this this is possible grab some popcorn I know well actually as an adult I have absolutely rectified um and we love award season and even with two young kids kind of managed to get to the cinema a couple of times a week during in that period and go and go and watch everything. But there is a real chunk there that you wouldn't think given the role I've played and the jobs I've done. I will move on in a second, but please, I need to know, you have seen The Goonies, right?

3:52I have not. Still to this day, not seeing The Goonies. I haven't seen Breakfast Club. Honestly, there are so many. Oh, The Breakfast Club. Okay, we got to move on. I can't, I can't. John Hughes is just, I don't know, he's screaming from the yonder.

4:05Rebecca Sykes: Yeah. So you are a partner at The Brand Tech Group. lead emerging technology, including Brandtech's generative AI work. And you were named as a leading woman by AdAge in 2025, which is very cool. And we're going to talk about Brandtech stuff pretty much for the whole episode. But I do want to do a little more of a career rewind. You brought up MoFilm. You were there in 2015 when Brandtech acquired it, although that was back when Brand Tech was still called you and Mr. Jones. You were promoted to CEO of MoFilm in 2018. After three years, you joined the mothership, I guess you could say. But what did running a creator marketplace teach you about making creative decisions that maybe still comes in handy now, perhaps more handy than ever, because AI is such a big part of your job and you think about it strategically like on the daily yeah I mean it's a great question um I credit my friend with teaching me a lot of things actually um not about the 80s though nothing about the 80s but um but many things I think even uh you know 20 plus years into my career it's only now that I've kind of been introspective enough to look back and think what is it that I've liked about all of these jobs what have these roles all kind of brought to me and I've realized that all of my various roles have put me at the bottom of a very steep learning curve and I have learned that that curiosity is something I thrive on I love not knowing very much being beginnerful having to assimilate a whole lot of information having to make complex things more simple um and I and I've always kind of been in positions whether it was the start the advent of social working in kind of a decentralized creator network at the very beginning of the creator economy you you know, where that just wasn't a known entity.

6:05The beginning of kind of the metaverse, the AI, I've always kind of had this same path. And I think just at a broad level, at a macro level, there's something about the curiosity that that, you know, you have to have to want to do that is something that I kind of bring with me and take on. But MoFelm very specifically showed me that kind of creativity lives everywhere. It isn't the privilege of a few. There are incredibly creative people from all walks of life who get to do this full time or not, who come from every part of the world. And that actually your lived experience is probably the best barometer you have in judgment for what is resonant and what works and what is meaningful and powerful.

6:48And I think in AI, it looks very similar, actually. um i don't know if you've had this experience but if you're working with ai on something you're not very expert in you can quite quickly be impressed with what your output looks like sounds great looks good okay and actually if you're very expert you're much quicker to cut through what is generic what doesn't work you know what sounds flattened um and you cut through and you get to good much much faster because your expertise is the is the difference there and I think um that is something you can pull across from working creative communities and working with AI you know that being unprecious about where the expertise comes from but navigating towards it as fast as you possibly can to that expertise and using that as the barometer and then kind of being able to work in whatever format uh makes sense so in mofilm this is all about opening up opportunities that otherwise weren't afforded to different types of creative from different walks of life, but reliant on their lived experience, their authenticity, etc.

7:53And in AI, that becomes about being able to present your thoughts and ideas in a million different ways, with the distance between something in your head and something in the world being so short now. But how good it is, is entirely dependent on the expertise you bring when you curate. I think it's been the same. I think being a beginner fool, and I love that phrase, is a superpower because it allows you to ask questions that are maybe considered foregone conclusions by people who are very experienced, who have stopped asking what should be obvious questions. They've become so accustomed to whatever the thing is.

8:28I find that with our new reporters, when they first come on board, they ask really great questions about how ad tech works because they don't understand. And a lot of ad tech doesn't make sense. So their questions are great because it's it's effectively cuts through a lot of BS. But then how do you become how do you become expert if AI is so impressive in things that you don't know about? and the way your brain learns is by hazing and having the experience. And so I think it's quite easy for junior people to not go a step deeper. You have to get to the point where you're able to cut through the BS.

9:10And that's kind of the preserve of, I don't know, people in their 40s who had to do all of this stuff without the tools. And I'm not saying that just because I'm in my 40s you and me both yeah I think um there isn't really a shortcut for the manual learning loop um and you know it's been proven over and over again people who write with AI can't recollect the thing they wrote in a way that you can when you wrote you wrote it yourself you brought a lot of yourself to it um so I think there's something really important in that we we talk a lot internally about the preserve of critical thinking and how you continue to think critically even when you're working in an assisted way I think uh that in itself is is really important um I also think it's really incumbent on us to have different learning paths and journeys so that you can be new in our industry keep all of what makes you know kept everybody to cut through the bs and ask good questions out of either naivety or just complete curiosity I agree with you they often are the better questions they challenge the accepted norms that actually when you go to answer the question yourself you go yeah I don't know actually that's a it's a great question we don't have a good answer it's just the way we've done it so I think you want to kind of keep that and protect that as much as possible um and that kind of way of working um but I think it's a it's a skill set we have to to develop we now look at training as a mixture of skills context and experience it's not just a learning environment it's not classroom based or you know, practical based learning, it has to actually come from context and experience.

10:48So I think it's incumbent on all of us as in industry to make sure we're offering that to people who are just beginning on that journey. Yeah, AI doesn't have lived experience. And I want to get into some news, though, because Brandtech just had an interesting announcement. You launched a global data and AI practice in early October. It's called Brandtech Data. It brings together 500 specialists from three companies under the brand tech umbrella. So your data and tech consultancy, 55, based in Paris, DP6, which is basically its Brazilian counterpart, or that's how I understand it, Jellyfish, your performance marketing agency, those three.

11:29And the pitch is about helping brands get their data ready for AI and connecting marketing activity to business results, which are both very necessary and very timely things. I wanna start with the first part, getting data ready for AI. which everybody talks about. You need the foundation. You need the foundation. People don't really talk about the how so much, just the need. But what happens when a brand skips that step and feeds AI bad data or data that just isn't the right information? What comes out on the other end? I think for brands, the biggest risk here is that AI takes you towards convergence rather than divergence.

12:10And obviously, divergence is the lifeblood of brands. You have to be differentiated it's the purpose for being branded so I think um in that way um an absence of data is very hard to avoid convergence you're operating off of the same data set that anybody else would be that is the kind of purview of the of the the models you're working with whatever their training data set is so you definitely want to be infusing with your own data um what I love is that so many clients who previously considered themselves data poor are actually incredibly data rich because every kind of data works in this in this context um so you can you can take a lot more richness out of social interaction and lots of other things that can kind of really build community and perspective and and brand positioning points of view so i think uh that's exciting it means we've got lots more to play with as a business we're less worried about messy data um or bad data than no data because again the challenges that have always been fraught around taxonomy and the organization of data somewhat go away we can have lots more fuzzy kind of data matching we can have lots more you know a much broader approach to the use of data so what clients who come to me at the moment saying loads of our systems don't talk to each other things don't connect we've never been able to map this never been able to work this and there are a number of ways for us to solve for that today so i think that is a real boon that actually we can kind of get further even with messy data um and of course you want it because you want the you know the unique um history and um kind of experience you've had as a brand and interacting with consumers to guide everything that you that you do moving forward so i guess my my point of view on this would be there's no such thing as uh as bad or messy data now we can kind of solve for most challenges um and that's where you that's way you want to start from.

14:05So I guess there's no such thing as stupid questions, but applied to data. So once a brand gets its data in order and also realizes that a lot of the value comes from the frame that they put on it, that is just the first step though, right? Because you and I, we met in Cannes. We were drinking fancy juice on the beach, as I recall, as one does. We were talking about workflows because we're two nerds. And you said something like a workflow is what operationalizes the magic. And you were referring to the magic behind AI. And I took that to mean that having clean data is important. It's even like it's critical, but it's really just one part of this.

14:47You need a workflow that's connected and needs to be a repeatable sequence of steps so you can actually get value out of AI, use AI to automate, to analyze, to manage a business process, all the stuff that you want to do that is useful for day to day. Where would you say AI projects that look kind of cool in a demo stall once they hit the real day-to-day needs of a big company? I think it's because they don't have a good workflow in place. So I think running anything at pilot scale is very different to running it at enterprise scale. And what makes a pilot work? It being contained, you having chosen the data or the subject matter quite carefully.

15:32Actually, other things that trip you up when you come to enterprise scale, and you have to work with everything, the things that don't conform to the rule set you put in place, the places where half the context is missing, where the workflow has always been an exception to the rule. Those are the things that kind of need to be unraveled when you start to get to a full enterprise scale. And they're not just corner cases. They show up in loads of different ways so you know we run a process where we make sure that um we're not just looking at the ai element um that needs to be transformed but we're looking at the entire end-to-end process because we have clients who are looking at um a new ai way of doing something and saying well that piece is we're saving uh you know 70 of the time already in this piece but why does the entire task still take four days why is the sla for that tasks not changing and that will be because uh where the process itself takes you between six different systems, five rounds of cut and paste.

16:30That data doesn't work with this data set. That thing doesn't merge with that. And actually, there's a huge amount of time and effort being lost. So I think when you start to think about workflows and the application at massive scale, you have to look at the entire end-to-end efficiency of a work process, not just kind of the piece you can automate today inside of a single system. So in order to get to enterprise scale, you start to look at cross-system application, interoperability, how data flows between and how it can be tracked against, you know, from one set to another. So you start to unpack the things that are not sexy, but make all the difference in how AI actually scales.

17:08It's all the same stuff that you had to do before, but it's AI now. Totally. Whose job is it to turn an AI pilot or just a good idea, a use case application of AI into something that the whole company uses? Because it's not quite marketing. It's not quite IT. It's not quite the data team, but it's actually kind of all of them. And I think that's part of the problem too, the human question. Absolutely. Very smart partner of ours, Paul Limbry at Google, wrote a brilliant book called Frogner Sock with Kelly Riley. And one of the lines I love the most is technology moves at the speed of technology and transformation moves at the speed of people and that is exactly true so there are new roles required um we look at this as there is a kind of operational core um needed around new technology it that needs to take on systems ownership and i think previously a lot of our focus time budget and role provision happened kind of at the edges of a system kind of where work got made in market close to culture close to customers etc and a lot of that is being pulled back because the nominal cost of making is it you know it has has dramatically decreased and actually the the value is in the systems ownership which is much more kind of core and central so i think there are new roles being created there to run ai governance to kind of take over systems ownership workflow maintenance um that actually kind of protect the system itself i think those and new roles that sit both in agency partner teams and in client teams to kind of co-lead and run the system that is being operationalized.

18:52Whatever happened to the chief AI officer role in brands? Because I feel like I read a lot about that, brands hiring CAIOs, which I hope that never becomes the abbreviation because it's so awkward. But I don't think I've ever met one. And I just can't recall ever encountering one in the wild. No. No, I think it's a really tough job in the same way that the role of a CMO right now is really difficult, too. It's all encompassing. You have to, you know, we said at the beginning, right, you need when you where you have your own expertise is where you can work fastest and be most agile and most assisted with new technology.

19:28If your expertise has to be everything end to end in an entirely liquid space, it's really, really hard to have the depth of knowledge you would need across everything. So I think there is a focus on transformation and I think that's really strong. I think you need that. People who are delivering in the business today aren't the same people who are devising, testing and pushing the frontier of what the business should look like and how it should operate tomorrow. So I think you see, I personally see a lot more transformation focused roles, whether that be called digital transformation or AI transformation, it's more of that.

20:00And then you'll see more technical roles existing inside of the organization and more operational kind of governance focused roles. existing in an organization, but less of that like catch all term of we've just got one figurehead who's going to just take us into the future, which I think is really hard. It's kind of like how we cover topics until they become infused into everything. Like we had a mobile reporter that was me for a while, how people cover AI, we have an AI reporter. And And I think that's going to change soon and everyone will just own that. And it makes sense that it has to start in a focused sort of way.

20:41And then as it becomes diffuse, then it becomes everyone's the air everyone breathes. Before before we take a break, though, I can you tell the story of the frog in the sock? I remember you told me when we were hanging out in the south of France, but I can't remember the reference. Yeah, no, it's a brilliant triumph, actually. They have written a book, a business book, as a fable. So it is an kind of easy read. You sit down in, you know, single sitting. And the concept is that the frog is the AI technology and your people empowered with it. so it is built to leap it is able to kind of you know clear a huge distance in a single bound the frog is the is the focus at the beginning of a transformation journey it's incredibly exciting we all focus on buying better frogs comparing frogs focus the pictures of your frog um but actually when you then go to implement inside of the business we kind of end up putting that frog inside a sock the sock being the existing operating model within that within the company and in general no one's sock is wrong or broken it was just built for a different time so you take this frontier technology and capability and you stick it inside the sock it will still try to leap but it absolutely cannot um it kind of it goes nowhere and shedding the sock is a real focused group task you know you don't you don't unravel it string you know thread by thread you have to shed it start again and build new operating systems and practice around the capability of the frog and so um they kind of uh suggest that the the delineation should be around 70 percent of the effort of transformation goes on the sock on shedding the sock on the people and processes um around the technological advancement and i think you know we that it absolutely mirrors our lived experience and the the parable itself has loads of other beautiful analogies around spectator sponsors and the dryer that is culture if you don't change that there's lots of brilliant And it gives us all great language for what is hard to explain, which is partly why I love it.

22:51But I think it gives real focus to why we spend so much time on those two P's of our three P's, which are the people in the process. And actually the platform, once you've got a good platform, you can kind of stop thinking about it to some degree. The change is going to happen because you put energy into the people in the process. Yes. And also just don't put your sock in the dryer if there's a frog in it. That's just like a terrible idea. It's a bad idea. Think of the carnage. Okay, we're going to take a quick break. And when we're back, we'll just keep talking about AI because that's what we do now.

23:25Great.

23:37I'm Sarah Sluice, Editorial Director of AdExchanger. And with me today is Michael Olson, EVP of Client Development at Basis, which is supporting our podcast this month. We're going to talk about something that doesn't get enough attention, paying invoices. What is the cost of slow financial reconciliation and who is the most impacted? Hi, Sarah. Thanks for having me. The cost is time, cash flow, and trust. When financial reconciliation takes too long, payments get held up and finance teams, whether they're in agencies or ad brands, are just chasing down numbers. People are spending a ridiculous amount of time figuring out why two systems do not match.

24:22Agencies probably feel it the most because they're stuck in the middle carrying the float. They have the client asking questions on one side and media partners on the other. The longer it takes, the more time everyone spends fixing the past instead of focusing on performance and health and growth of the business. So you outlined a very frustrating problem. I would not want to be that person chasing down invoices. Is this primarily a technology problem, a workflow problem, or a people problem? It's all of the above, but I'd lean towards mostly a workflow problem. The industry definitely doesn't have a shortage of technology or data.

25:04We have a shortage of systems that actually work together. And when they don't, people end up filling in all the gaps manually. They end up downloading spreadsheets, comparing reports, sending emails, chasing down discrepancies. It's a mess. Technology is obviously part of that answer, but another piece of software by itself just simply doesn't fix it. It's imperative for technology to actualize omni-channel data monthly and push to reconciliation systems. As an industry, we have to connect the tech stacks interoperably, end to end, and then allow the technology to do the heavy lifting to align the data between the contracts, media plans, media transactions, and the payments.

25:49Okay, so it sounds like technology and workflow can make this just a bit easier. What does automated reconciliation look like? A lot of automation talk right now. And if I'm thinking about the future agentic state, ideally, reconciliation just happens. AI is constantly looking at delivery, spend, invoices, payments, and then flagging issues as they happen. Instead of somebody finding them three weeks later, and in a lot of cases, the AI should be able to fix those issues without a person ever touching them with the correct harnesses in place. People should only have to get involved when something unusual happens, which is the human oversight.

26:33That gets away from this huge month-end exercise that costs a lot of time and resources and much closer to knowing where you stand financially at any given time. This is the ideal state basis is going towards for our connection with Prisma by MediaOcean, a platform used by numerous agencies for financial reconciliation. Okay, so it sounds like by connecting those two different platforms we can and adding agentic AI, we can make it a lot easier and people won't have to chase down invoices all the time like they used to. So thank you so much for sharing with us what that would look like. And thanks to Basis for supporting our podcast.

27:11Thanks for having us, Sarah.

27:20All right, we're back. And let's talk about AI agents, because AI agents are starting to take on more marketing related tasks. I get pitched on this all the time. Planning, creative, media buying, the whole shebang. So Brand Tech has an AI marketing platform called Pencil. You just launched an AI agent called Scribble, which is a very cute name. Basically, a marketer hands it a brief and then Scribble works out which AI models and tools to use. It runs the job in the background. It does things like resizing ads, swapping products into images, remixing video, a lot of stuff that feels, I don't know, like grunt worky.

28:02But then the marketer reviews the plan, edits the work at the end. And your plan is to connect the creative to media buying later this year to make the whole thing a bit more end to end. In the announcement about Scribble, you talk about how marketers like don't really need to learn how to operate tools. Their role is about assessing and judging and guiding the outputs of technology. And then the technology works on its own. How do you get people to adopt something like this, even though it sounds great on paper, like you're not going to find a lot of people who say, oh, I want to resize, you know, ads all day.

28:40But if they're just assessing it, it also feels like they're kind of removed from the process, like maybe they won't actually be needed eventually. Yeah, no, I think so there are two things I would like love to comment on, I guess. The first is it's normally not the start point. So scribble is a full kind of agentic harness if that's something that's passing over your desk a lot. This notion of a single front door for all briefing, all briefing gets triaged by agents and then rooted to wherever the work needs to be done. That might be fully automated. So Scribble takes the first pass and makes the work.

29:18It might be rooted to any number of third party partners who craft that work in a combination of human and AI craft. So it doesn't take away all of the making, but it should, as you say be taking away the grunt worky things that could be automated could be highly accurate done with real fidelity without needing somebody to manage them i actually think that bringing in that layer of kind of orchestration wrapped around everything um is quite exciting for a lot of the marketers we um talk to ai transformation it's just too important not to own it themselves so being central to the system orchestration and the operation of how that works bringing the sovereignty of how work gets made and what good looks like and who gets to approve what back inside the business feels quite I want to say powerful but you know I don't mean that in an overarching way you know control but like there's something um about the ownership and the account empowering I guess empowering exactly that is empowering exactly that so it I think it feels exciting for that reason and the point about skills is that you know previously the idea of sovereignty and bringing things, bringing that work in to be owned by marketers meant you had to learn craft skills.

30:31Even if you were learning how to complete those tasks with AI, you had to know how to make things. You had to have kind of creative judgment, not just for the end work, but how to originate the work. And I think that was, that felt very, very far away, you know, teaching people those kind of depth of craft skills and actually what an agentic platform and a harness like Scribble is able to deliver is it can take away the need for learning all of those technical skills. And you can ask in natural language for the things that you want. You can assess the outputs. And we have a great framework for assessment as well and run our own kind of public evals platform.

31:06So there's just so much more in that that a marketer can wrap their arms around and be empowered by. That leads to the question that I have about the sea of sameness and avoiding it. Because if everyone has access to the same AI tools, a lot of ads could start looking the same. And we're already seeing that bland plastic look that a lot of AI creative has. And you wrote this piece in the drum last year about that risk of the sea of sameness. And your advice is to make brand, like brand with a capital B, machine readable. So like spelling out what a brand stands for, the values, the colors, the sounds.

31:48And AI can understand that and then reinforce the uniqueness. So it's not just pulling generic things into an answer or into an output. But brand guidelines are usually written like for humans. There's like kind of fuzzy language, like bold but approachable or whatever. And I don't know that a machine can do much with that. But yeah, so there's that. I'm going to put a pin in that for a second. And then also, I think some of the best brand work happens when brands break their own rules. And so if you've written all the rules down for a machine, like how does it know when to break them? So you probably saw this campaign Skittles.

32:35Their whole identity is the rainbow, but they took the rainbow off their packaging in the UK for Pride. And it was great for their brand. Like most brands that show up at Pride, they adopt the rainbow. But Skittles whole thing is the rainbow. So they gave it away so that Pride would be the only thing that mattered. it would be the like giving up their most recognizable asset and it made the gesture feel sincere I think rather than opportunistic it got a lot of attention it even helped with the sales decline I just don't think a machine would have thought of that yeah yeah no it's true I think there are there are two things in my mind um the first is that the the brand guidelines as you say the kind of rules they have to be uh machine readable as you say AI is interpreting brand guardrails in in many cases now um but you can only get there if you had really consistent strong brand identity to begin with that even the people who work in your building would all answer the same way to the same questions and just so many brands don't exist in with that level of consistency and fidelity to a single identity so it really matters that you know who you are that everybody in the organization would articulate in the same way and then you can turn that into machine readable kind of content so that when you're speaking to AI audiences or human audiences or when you're working with AI partners or human partners everybody's interpreting the guardrails in the same way so I'd say that's the first barrier is just getting to that level of consistency and there are some brands right there are game there are whole games that you'll play at Christmas with your family where you see one piece of a brand or you hear one tone can you name that brand yeah or can and it's instantly it's it's that's the power right so that that is if you don't have that if you're not a brand that operates in that space if you couldn't be in that board game that's your first problem because you know making uh tons more content in in lots more channels you know in many other formats isn't going to solve that problem it's just tonnage at that point exactly it's tonnage it just absolutely uh continues to dilute and kind of extrapolate the challenge so that is your first problem the second problem then is how do you work with models that get um that zig when everybody else with zag how do you get not how do you walk away from convergence and you and you start working in that way I think this is fascinating um so there's there's uh and there are lots of things I could talk about in this space kind of partnerships we have with UCL where we're kind of building out um approaches and research around exactly this um there's uh you know kind of whole frameworks we're bringing into to pencil in terms of how to look at that the most creative approach and application um inside of a brand world but one of the things that um i love the most is a colleague of mine james dow um wrote a brilliant presentation um around brand islands and the way he framed this thought was that for years we have um written strong brand guardrails because experimentation was really expensive if you left the island and went off exploring you had you ran a lot of risk you know you you spend a lot of money you might come back with nothing and so with the kind of guardrails are there to protect increasingly as their guardrails can kind of serve that purpose and exploration becomes cheaper and easier than actually the the value of going off to leave your island and work out what happens when your brand bumps up against absurdist humor or sneaker heads or some other moment in culture or time or you know a different audience with a different flavor of how they might use your brand or your product or or see you as a business you start to bump into opportunity and And actually, that is the unlock around innovation that AI can bring.

36:18And I think the efficiency, the kind of consolidation and consistency around brand is almost like a step one, extract a bunch of efficiency from the model, right? Stop having things that are off-brand with no purpose in order for you to go off and explore being off-brand on purpose, because that actually is where you sort of start to generate a whole load more value. So I think that's really exciting. And that is definitely an AI plus human endeavor. There's also, though, the question of what AI, quote unquote, thinks of a brand. Like if you ask ChatGPT or Gemini or whatever to recommend a shampoo or laundry detergent, like which brand does it bring up and then what does it say about that brand?

37:01Jellyfish has a tool for that called share of model, which is like share of voice, but for AI. So Jellyfish will ask the models about a brand over and over again, track how the answers change, and then a brand can use that information. But there are a lot of companies that are doing that right now or that say they're doing that right now. Startups, agencies, everybody, they're claiming they can tell brands what AI thinks of them, again, quote unquote, because AI doesn't think. And then help them show up in AI answers in a better way. I mean, if you could see my inbox, it's hilarious. So many companies are like, we crack this.

37:40I'm like, I don't think that's true. What makes share of model like different? What are you doing that maybe these other companies aren't doing? And what sets it apart? So I think one thing we're incredibly proud of is share of model existed before the industry had a name for this whole category, which has become GEO or generative engine optimization. um uh and so it was something we were pioneering um very early on um and we're very quick to understand that uh the way that a model understood the brand and i i will in you know understood i'm i i know we don't we're not thinking we're not understanding it's shorthand it's shorthand exactly the anthropomorphization is helpful in some ways um but you know how a model uh interpreted a brand um and then made a recommendation off the back of that um was going to be a huge driver for engagement, for awareness, but ultimately for commerce.

38:36So I think we've been in that space for a long time. We actually think of it in a broader sense, more than just optimization. We think of it as gem, as generative engine marketing and how it impacts every kind of approach you might take. So we've been looking, share of model data is underpinning where the volume of source data comes from for recommendations. A lot of that comes from video. and then looking at like you know which influences are creating which kind of content and which might lead you to a more sort of lucrative rich scene of partnership so there's lots of ways in which we're taking this much further than kind of a search equivalent and into looking at the entire ecosystem of how you go to market as a brand and all the ways in which you might need to understand both your AI consumers or your AI recommenders as well as your human audiences.

39:29Creators, which we've touched on here and there, so important. And we're trying to figure out how to cover the creator economy more at AdExchanger because historically it hasn't been something we've written about much, but it feels like it's being infused into everything. And AI search pulls in so much from YouTube and from other Reddit and other places, but particularly particularly YouTube and a brand does really need to know where people are talking about them or where there are alignments it's not always easy to find out. Yeah exactly you do need to have a kind of a way to reverse engineer where that source material is coming from what does that look like and then to look at the work you're making and understand whether it hits those beats and and you know can help you assimilate into a conversation um in that way so yeah your experience is exactly the same as ours we have uh two big kind of pillars two big things that that brands seem to be very focused on right now which i agree with think are right um the first is how do we live at the absolute forefront of culture how are we being resonant in everything that we do and the second is how do we unlock ai transformation to do that at a scale that the market demands um i I think just for most brands, it's just become so clear.

40:52Growth used to live in a handful of moments, you know, seasons, products, you know, cultural moments and sponsorships that you could resource them with time and people and a good agency partner. And there are now so many moments blown up. It's in every single step of the customer journey. They all require insight and data and craft at a level that you just can't fulfill with a human only model. So AI transformation and that ability to live in culture has become the key to whether you can grow as a brand. And we know many aren't. Right. So like this is a critical decision for them in terms of what they can do.

41:28And it's the reason that AI transformation has become the center of our business and how we help them. So penultimate question. And all of that sounds super tiring. Like that's a lot. So going back to that piece in the drum, it was a really good piece. People should check it out. there was this other point you made that I like which is that AI means moving from fixed point change management to being a business in perpetual transformation which is the phrase you use and that that's just really hard what is perpetual transformation feel like for an employee like how do you keep people excited and keep them from burning out or checking out when you have to be on all the time as a brand totally um so i think perpetual transformation needs to become the feeling of being comfortable with maybe even good at change which is not where most businesses start where most people start you're exactly right not when most people start if there is a an ambiguity um to what the future looks like and how quickly it will arrive that we have to get really comfortable with so um it's something that occurred to uh to me to us a long a long while ago which was just that this fixed a to b transformation um wasn't going to work and people kept confusing transformation with implementation so is that six months where you just install a new system for us no no no because that system will be changing even within the six months people will need to continue to learn so this has become a much kind of it's not transformation is not a project this is a capability it's a muscle you have to exercise and you have to build that out um it's part of why we believe it's so important it lives inside the client's organization too otherwise you're absolutely dependent on other people coming in to change you but from an employee point of view in order to not be tired in order to not be you know done with change I think people need to understand uh you know I think there's less concern about like is my job going away or will I be interested in and more what is my job becoming like where am I going what does this mean if I unlock this skill you know and I reach the next level and I'm playing the game where do I go what what is interesting for me and I think historically we give people a woeful lack of clarity around what that could look like I work with a brilliant client who sat down every member of their team in the beginning of a transformation and wrote a personal development plan for them plus AI where do you go what gets unlocked what's interesting to you.

43:56And I have never engaged with a more up for it, brilliant, inspired, motivated team on day one, because they knew exactly where they were going with AI and what that was going to look like. And there was no fear for them in the change from A to B, because B felt really great. And they were pretty sure C was going to be even better. So I think that's a cultural dynamic to change and how we make sure that people don't feel tired, they feel energized by the notion of continuing to change. I'd love a personal development plan just for me that would be really helpful because yeah I mean we also get directives use AI just use AI we got you a Claude Enterprise subscription go for it I'm like I don't know what to do with this um I wanted to end on a more personal note what what do you use AI what do you use AI for in in your own life but not work related like what have you outsourced to that makes your personal life easier?

44:54And what's something you wouldn't use it for in your personal life that you want to keep? So this is a great question. And I feel like it really cuts to the chase of how much adoption is there because work is one thing, home is something completely different, isn't it? But so we always frame AI as there being like a A moment of efficiency extraction where you reduce waste, whether that's creative waste, media waste, etc. And it gives you a little bit more liquidity to go and reinvest in brand growth in some new and exciting ways. That's how I think about it at work. And at home, I have kind of the same idea.

45:30So I do it. I use it for the really boring stuff. Meal planning, shopping, whatever it might be. Birthday present ideas, blah, blah, blah. I use it for all of these kind of, you know, efficiency extraction things. But the thing I've loved so much is the reinvestment. I have two young boys and we've been writing books together and what they love is when they uh I don't do it with voice at the moment I I type they talk I type um and they explain characters and things to me we can then watch their movies right of the books they're writing of the characters they developed there is so much additional time that I get to spend being creative with them and I always had this thing before I had children would I be a crafty mom or would I be like an Amazon mom and I always thought I'd be a crafty mom that's what I really wanted to be with the the scissors and the glue and making things and the reality is if you work full-time you're often an Amazon mom and this has given me back the chance to have this like creative moment with them because they ask me you know they'll wake up at 6 a.m and say can we write our book now can we write our book you know they just want to they want to do that kind of thing with me and I get to like not worry about anything else but just capture their stories and be with them and be present and the delight on their faces when things appear that you know it came straight out of their imaginations is great so I think it's for me it's for both it's there's a little bit of the time back and the the friction out of just the organization of life and then every 10 minutes it gives me to do something that feels like fun like that it's just a benefit what's their book about oh it's it's about birds of prey they're very into birds of prey um then And there's a school, a bird of prey school and all of the characters who go there.

47:06So they're loving it. It's great. Well, when this is published, I want a signed copy. Absolutely. Send it my way.

47:26This episode is sponsored by Basis, the leading intelligent operating system for autonomous advertising. Its enterprise AI solution transforms campaign briefs into strategies and media plans that integrate directly into omni-channel activation. Learn more at basis.com.