How to Structure Your Organization in the Age of AI with Lauren Livak

How to Structure Your Organization in the Age of AI with Lauren Livak Speaker 2: Hello and welcome to The New Frontier. I am here as always with most of the time with Jo. How are you, Jo? Speaker 3: I'm really good. Thank you, Max. How are you? Speaker 2: Good. So today we are joined with an exciting guest as always. Lauren Livak-Gilbert is the Executive Director of the Digital Shelf Institute and co-host of the Unpacking the Digital Shelf podcast. The Digital Shelf Institute is actually putting on an event in London in October. It is free for brands and retailers. Yours truly will be speaking there on Answer Engine Optimization, all of that good stuff. So we'll link that in the notes, get a ticket and come along. But on Lauren, she has 10 years of experience in digital and ecommerce. She is ex-Johnson & Johnson. She is ex-Salesify, and she's worked with over 100 global brands to develop comprehensive commerce strategies, including staffing, technology investment, and performance-based processes, which is something we're going to be talking about today. The last plug I'll put in because we are hosting a Drinks at Amazon Accelerate on the Tuesday. So I'll put the link in the notes as well for that. So if you're in Seattle, come and join us there. But otherwise, Lauren, welcome. How are you? Speaker 1: Hi, how's it going? Thanks so much for having me. Nice to be here with you, Jo and Max. Speaker 3: Great to have you. Speaker 2: So, Lauren, you recently did a big piece of research. Maybe just starting there, what prompted you to go and talk to 90 brand leaders about organizational structure? Why did you feel this was the right time to think about this stuff? Speaker 1: It is the number one question I get asked from brands, from retailers, from anyone in the industry. It was a big question I had when I was on the brand side. It's a really challenging question, right? Because every organization is different and everything's always changing. So it was actually a paper I always wanted to write and I'm the most excited about this piece of research coming out. It really came from questions We're from the community and so I wanted to help them and bring that together and everybody was super willing to participate. So we had a lot of great turnout for interviews and for the survey and a lot of great data that came out of it. Speaker 3: Amazing, 90 people. That is a hefty take on. I can imagine that from speaking with all of these people, actually a lot of, as well, misconceptions have come up. And from an organizational point of view, people are facing a lot of challenges when it comes to AI. What are the main misconceptions that you've seen come up from the survey? And just generally, what are the main challenges that organizations have to face right now? Speaker 1: I love that question. And before I answer that specifically, I'm just going to start with one kind of takeaway that I think was a big theme that I think is connected to that. And that's the fact that this report specifically didn't just hone in on ecommerce professionals. I tried to make it omnichannel. And I think that's a big key in terms of success of AI. I actually met with legal, HR, procurement, supply chain. Like I actually went to many different functions. And I think that's one of the biggest problems with AI or misconceptions in organizations. A lot of people think AI, you can just slap this on here. It can be our strategy and then everything's going to work. But there needs to be a lot of fundamentals in place, whether that's your data, whether that's your governance, whether that's cross-functional alignments and understanding who owns it. There's so many of those fundamental elements that need to be in place. And they're not rocket science. It's things that we've talked about around organizational development for years. But if those are not in place, your AI initiatives, your AI test and learn, any project that you do will not really be successful. So I think that's the biggest misconception. And then also just not knowing what questions to ask. This is a new space, new technology and everybody has a different way of talking about it and there's a different set of language and there's different functions interpret things differently. So just knowing how to ask the questions, how to evaluate the technology and what makes the most sense. Speaker 2: Yeah, it is interesting. And it's something that is coming up with me a lot with my customers and potentially Jo as well. I don't know if you're consulting if people are asking this, but it is such a rapid change that for sure how teams are structured needs to change. So I guess, like, what do you see is in the specifics of that? Like, what do you think is changing? And how are you seeing these companies reorganize? Obviously, it's early days, but like, what is happening on the ground? Speaker 1: Yeah, there's definitely a lot of change that will happen. But the one thing I do want to start with here is that it's not doomsday, right? Like, it's not going to change tomorrow. We are not all going to be insignificant and robots are not going to take over, right? This is going to take time. But it's also going to take time to get ready for this change. So the really big changes are that AI is going to power a lot of things. So think about more of an AI-enabled organization. In the research, there's a diagram of what an AI-enabled organization could look like. And one of the biggest things is that, let's say, a brand strategist or brand manager might have five different agents supporting them. So, they might have an agent that's writing content. They might have an agent that's doing user testing with user groups using AI. They might have an agent that's reviewing. So, they can actually have AI agents supporting them in their role to do their job at scale and more quickly. But what that means is you need to have a very clear understanding of your brand voice A very clear understanding of your brand strategy. You need to have a connection with your IT, your technology departments to make sure you're training these AI agents to make the right decisions and to write the right content in the right tone for your brand. So that's where I think we're going in terms of organizational structures where some of those kind of hands-on keyboard types of tasks We'll be done by an agent, but it will free up time and existing rules to be more strategic. So I think about my days on the digital shelf on the brand side. We could only create so much content because we just didn't have the manpower to be able to do it. And I'm most excited now because you actually can with AI. You can actually meet We have hundreds of images at scale for your thousands of SKUs, and you can get it online to your retailers where you weren't able to do that before. But you just need to be very clear about your style guide, how you're training that AI. Make sure you have a human in the loop to review it. So it's not taking away, I think, a lot of jobs. It's freeing up time for those people who were hands on keyboard making the images, writing the copy, to now think more strategically or Go look at their competitors. See what's coming next. Come up with a new trend for the type of content that needs to be on your PDP, your product detail page. So I think it's exciting. Speaker 3: I couldn't agree more. And what I've been writing about as well in my newsletter is all about actually how AI is going to, in a way, redefine our job skill necessity, like how basically the worker of the future would reshape. Because at the moment, obviously, we are very focused on acquiring a lot of vertical skills. But actually now with having a team of AI agents, that kind of becomes less important. And now it's more about this kind of like strategic overall oversight of all of these agents and how they work together to get to the final outputs. Couldn't agree more. But I think this is indeed the future. And I'm sure like I already read about organizations that have adopted this to a certain extent, but probably there is quite a few as well that you've understood that are finding this challenging. So which are the organizations that are acing this and what is basically stopping the rest from your research? Speaker 1: Before I answer that question, Jo, I have to 100% agree with you about the talent piece. I think that this report and a lot of the research I did is a call to action to HR because the commerce leader of the future is very different from the leaders we have today. And if AI is going to take away some of the hands-on keyboard jobs, some of those entry-level analyst jobs, how are you then going to create a director of tomorrow? Because they're not getting that hands-on experience. They're not learning on the ground and then building their way up through their career. So, I really think that HR needs to rethink career planning, job descriptions, how we're training employees internally to get more of the kind of generalist mindset. And the leader of the future, to your point, Jo, I'm using the word orchestrator. So, thinking of that leader who has AI agents supporting them as the orchestrator across, let's say, the consumer journey or the holistic organization. So, they have visibility from start to finish and then can use their AI agents to do some of that hands-on keyboard work to be able to get there and tie everything together. So, I love that you said that. I'm super passionate about that point and I think that is a great thing to hone in on. In terms of organizations that are doing it, some that might be struggling with this, I think there's a level of risk. That's involved with adding AI agents. So the organizations that I've seen that are doing this really well are accepting a level of risk and they're willing to try this out. They're willing to test and learn. They have their legal department on board. They have their tech department on board. They have all of their functions working together to try this out. And they're looking at it as an accelerator and a test and learn, not total replacement, end all, be all. This is how we have to operate moving forward. For those that aren't really adopting it or trying to figure it out, I think they're hitting a lot of roadblocks with, one, what is the problem we're actually solving, rather than just, hey, we need to use AI. So I think identifying what that problem is that they're solving. The second piece is that level of risk and working with their organization to say, we're comfortable accepting this level of risk. This is how we want to do it. This is how we'll keep the human in the loop. And I think the third piece is really just making sure that they actually have all their ducks in a row internally with their cross-functional alignment with their data to be able to enable something like that. Because the old saying, bad data in, bad data out, I feel like everybody says that now, but it's so true, right? If your AI is feeding off of bad data, you're not going to get a good result no matter how many people you have involved, right? So I think those are the three areas where I've seen those that are struggling to figure it out get stuck. Speaker 2: Yeah, so I have a question for you on the adoption side. So I have a friend who will be, I'll keep anonymous, but a kind of an owner of a large Amazon agency. And he was piloting our tool and a bunch of other tools for like the copywriting side. And he got the copywriters to do the pilot. And every single kind of feedback on all the tools were like, obviously, from the copywriters, super negative. Oh, this is rubbish. It doesn't do X, Y, Z. And you can emphasize, right, because these people are obviously very nervous. They're very nervous about adopting AI, because they know it could, in some ways, it can automate them out of work. And even if it doesn't automate the amount of work, does it mean that the traditional promotion of like, you're a copywriter, then you get a team of copywriters, and you become a senior copywriter or whatever, like that traditional path is broken, because there's no, you just use one of these tools, and you're just managing an AI system. So how, with the leaders that you spoke to, how do you, how do they get their organizations to adopt this stuff? Speaker 1: I think you hit on a really great point. Change is scary. And we don't actually, I wish I had a crystal ball and could tell you exactly what's going to happen. But the way I look at it, and I think a lot of the leaders that I spoke to were saying that we know that new jobs will come about. Right, this will create a new type of role and ways that we might need to pivot that we don't know about now. So there will be some traditional types of jobs, functions that might be highly augmented by reality by skipping by AI, right? That is happening today, and it will continue to happen. But I think the important part for the people in those types of roles is to identify how they're differentiated from the AI. And understand that career pathing might be different in the future. And I think that's that call to action to HR that I was talking about. Understanding that those traditional career paths are not going to exist in the future. So it's more of a generalist role. It's more of a zigzag than a straight line, right, in terms of your career path and being comfortable with that and knowing what that looks like and knowing how to navigate it. I think we're all trying to figure this out right now. There is no clear answer, even from the research that I did. It's not like I'm Prescriptively saying, this is the way your organization needs to be. But there are some very key tenets and principles that you should build upon, which is making sure that you're flexible, making sure your HR team is understanding of what's coming next and how to recruit for the future. There's 11 of them, so I won't go through them all. But there's some very key tenets that you just need to think through. And I would say for people in roles like copy editing or content creation, know that one, it will take time. For these things to change. Two, I don't think they'll ever be fully replaced. There's always going to be some element of human in the loop, but I would say, how can you differentiate yourself or how can you think about some of your strengths being moved to a different function or a different type of career path? Speaker 3: Yeah, and I think actually just to add to the point around, let's say, copywriters and work being replaced, from my point of view, I truly believe that, yes, AI can do a lot of heavy lifting and a lot of production, but when it comes to, like, true creativity and truly actually creating something of value, this is still very much a human remnant. And I don't really think that even with these systems becoming more and more intelligent, I don't think that they, at least in the short term, are going to be able to create something truly unique, something that really sparks that special thing when we see a really good copy or a really good ad, like a really human connection, I would say. And I think to an extent, it is really important for people, for example, in copywriting or in creative jobs. It's to realize that actually what they really need to focus on is like honing on that creativity and bringing that sort of unique point of view that humans can bring. I think to an extent AI is going to push actually the human value of creation much higher. And so I feel that it's really important to see that it's going to be almost like a loop where we are going to like devalue human creation and then value it back. I agree. Speaker 2: It's part of the definition of these models, Jo. These are prediction models. They're trained on a bunch of data and they predict the next word or pixel in an image based on the training data. So by definition, they're not creating anything new or unique. They're just reformatting what the model knows plus whatever kind of brand guidelines and whatever else you put in the rack. Speaker 1: And I'm not connecting the dots. I think that that is the orchestrator role that I was talking about, right? Like you still need someone to oversee everything and be able to say, Ooh, I don't think this is working. So let's tweak this or let's bring this together or this person's not talking to this person. AI can't do that. And we'll be able to do that. So I think honing that strategic thinking, also creativity, I agree with that completely, Jo. We can all tell when it's AI copy, right? Right now you can tell what it is. You can tell AI videos, AI imagery. I'm sure it will get better, but coming up with new provocative ideas and creative thoughts will always be valuable in my opinion. Speaker 3: Yeah, absolutely. And actually, like now, actually to the point around being an orchestrator of, let's say, a group of agents, that certainly will complicate goal setting and measurement, right? So how do you, as an organization, how do you evaluate the work of a team if part of the team is actually held by AI agents? So do you have any observations of what organizations are What are they doing right now, or how are they preparing for this, let's say, mind labyrinth? Speaker 1: I think from a goal perspective, the really big challenge right now is that most functions are goaled separately. And across the board, I find that as a challenge. Most don't have shared goals. So for example, like the sales team has a specific goal, the marketing team has a specific goal, and so does the IT team. They are set up to go against each other, right? So they sit in their box, but that is not going to move organizations forward and help them grow. What you really need is a shared goal. Let's take ecommerce, for example. If you say you want to get to, I'm making this up, 7% ecommerce in the next two years, then every single person who touches that, which would be Retail media, sales, marketing, IT, shopper marketing, category management, they should all have a goal that is connected to that. So all of the teams are working towards the same North Star instead of working separately. Because I've said this so many times, but what gets measured gets managed, right? Your bonus on a specific goal, that's what you're going to go for. So you need to have that shared goal across the board. And I don't think that changes if you have AI-enabled agents that are helping a brand strategist work through this. The brand strategist will still be accountable. For driving a specific goal, whether that's ecommerce sales, overall sales, growth, whatever that looks like, they still will be accountable to that. But if you're having AI agents on the brand side, maybe on the shopper marketing side or the retail media side, supplement those orchestrators, they all need to have a shared goal. And the one thing we found out of our research is those organizations that had shared goals had a higher percentage of ecommerce sales. So it does correlate to ecommerce sales and higher growth because you're working together. You're all working towards the same thing. You're not working against each other. So I think that fundamental shift needs to happen in organizations today before you can even think about measuring differently, having AI agents helping you and supporting you in your day-to-day. Speaker 3: Can I just ask you a follow-up question? Sorry, Max. Speaker 2: You always do this, Jo. Speaker 3: I don't always do it. I do it sometimes. You do it sometimes. Speaker 2: It's a joke. It's a joke. Speaker 3: Balanced. It's really interesting because I completely see how it would work in theory. In practice, would we need to redefine the way organizations are set up? Because it's like a, it's a, you know, I see it just working with not even like big organizations with like small sellers. You tell them, Hey, in order for this AI thing to work, you really need to redefine your processes. And then it becomes like a big head scratching exercise because that's the foundation of their business. And so it's the same thing when it comes to a big organization, but even more complex, which is. How does the organization work from what you say has to be completely redefined? And so what are your thoughts and do you think it's doable? Speaker 1: That's the complete point of the research. So you hit the nail on the head there, Jo. The way that I jumped into this research was the question I always get is where should ecommerce sit in the organization? And I think that's the wrong question to ask. I think it's how are we fundamentally changing the way we work to better match the way consumers shop? And that's a different question. And that requires a different result internally about how you're setting yourself up. So to your point, Jo, one, it's super complicated. Like I can talk about this and it could seem really simple on paper. It's very challenging. Org changes, people changes, the way you're structuring process, technology, it's all very challenging. But I think, to your point, it is possible, but you have to do it in stages, and you have to make sure that you're doing things in the right order. So in the report, I talked about kind of those key tenets of how you need to think about it. I think some of the big ones are having shared goals, right? So first, getting everybody on the same page. Then thinking about your leadership. Do you have a sales leader and a marketing leader that are reporting to separate leaders? You should have a growth kind of mindset, whatever the title of that looks like. So someone owns both sales and marketing and is looking at them being put together and being more cohesive. So thinking about that, thinking about a change management program, if you're going to change the way that you have worked fundamentally for the past 100 years, You really need to help people along that change, right? Change is hard. Change is scary. The model that I've talked about in this research is the reinvented organization model. Because I think right now what we're doing is just tacking on ecommerce to organizational structures that have been built for in-store or have just not been built for the AI future, the digital future. And you really just need to rethink how you're doing business. And yes, that's a big change. I will totally acknowledge that. And it has to happen in stages. And every single organization is at a different level of maturity. So the one thing I was very clear about in this report is it's not prescriptive to say, you need to do this in this order, and then everything will be solved. It's absolutely different for every org, whether you're big, small, medium, in different categories, at different levels of maturity. But you do need to rethink and say, hey, are we built for brick and mortar? Are we built for an omnichannel world? And how do we start changing people, process and technology to be able to support that? And if you don't do that now, There's no way you're going to be competitive in the next five to 10 years because shoppers are now omnichannel. We know this. We know that they're shopping on social commerce, online, in-store. They're having so many different touch points until they get to that actual purchase. So organizations need to be able to adapt to that and it's going to take time. Speaker 2: I know the answer to this is probably it depends. Speaker 1: It usually is. Speaker 2: Yeah. But if we try and maybe give a specific example maybe of a company that you think has really nailed this in terms of structure. And obviously, it depends on personnel. And as a leader, you want to get your best people to do stuff, which they're best at. And that's obviously an important part of the equation. But is there an organization who you think, through this research, has really nailed a good structure for the AI world? And how does that look? Speaker 1: So in the research that we actually talk about Bayer and they were quoted in the research. So this is all important. You can read about it. And they have something called distributed share. Speaker 2: Who are Bayer, just for the listeners? Speaker 1: Yes, Bayer. Speaker 2: So what do they sell? Speaker 1: Oh, sorry, health care. I'm sorry. I'm cut out for a second. So they do global health care. So they have vitamins. I think they have over-the-counter. I feel like I now need to look at their products. Hold on. Speaker 2: No, no, no. I know Bayer. It's an accent difference. I think everyone will know who you're talking about. Global health care. Speaker 3: Guys, I'm sorry. Someone who lives in Germany, I have to correct your pronunciation. It's Bayer. Speaker 1: Thank you very much. I'm probably pronouncing it incorrectly. I apologize. So I actually interviewed them and they were quoted in the report and they have a very interesting structure. So I encourage everyone to actually look at the report and review it. There's actually a visual of it, but it's called Dynamic Shared Ownership. And you can actually look it up. It's a principle for organizational design. And the way to think about it is that You identify what are some of your key priorities that you're working through, whether it's, I'm making this up, right? We need to do better in walmart.com sales or hey, we need to have a better Amazon sales because our margin is getting eaten away. You create those priorities and then you have a team that works on that specific challenge for the next however many days. It can be 90 days. It can be a sprint. So you pull in the right people from the right teams and they go and they actually work on that specific challenge. So they might have a home team where they sit in a marketing function and an away team where they're actually working through that specific problem and trying to solve it. And what it enables the organization to do is pivot for the actual challenges that you're trying to solve and bring in all the right people to go and solve it. And then you might move on to the next thing. It's really interesting when you look at it. I like to describe it. This is not an official way of describing it. I like to describe it as like an amoeba. You're shifting and moving towards what you're trying to solve. And then you're deploying the right people and the right strike team to go and solve it. And then figuring out, okay, what do we have to do next? So rather than having like just a traditional hierarchy, It's more that you're distributed in these home and away teams so that you can work with the right people to go and help drive growth. And I think that's a really interesting model, especially when you think about how AI might influence that. We didn't specifically dive into that into their model, but I can think of a world where you would have AI agents help support you in that, whether that's through data analytics, content creation. Identifying what are some of the challenges to get to next. So they have a flexible model. They have shared goals. They can pivot on the fly. Those are really the kind of key principles that you need to have in place in order to be ready for any change, if it's AI or not. Speaker 2: Yeah, I'll throw in another one, which I think was the best thing that I saw at Amazon in terms of working in dynamic teams, which is a we used to have a monthly experiment review, where rather than separate teams doing separate experiments with AI, which of course would happen. You would have a shared review with lots and lots of senior leaderships in there where you would go through experiment by experiment. And this is the experiment. This is the update on it. And everyone would discuss that. And that was cross teams, lots of senior directors. And there was a good process to anyone could basically submit an experiment to get it in there, but they would basically control and centralize everything so that there wasn't duplicates going across the business. And it was a really effective way of both enabling everyone to have visibility on what's going on, but also encouraging people to experiment and think about new things in their roles and then submit that to the team. And if the working group who wrote the review thought it was a useful thing to do, then you'd be plunked into this review with lots of senior visibility and work both for the individuals in the company and also the leadership. So that's another one that I've seen which works well. Speaker 1: I love that example and I actually had the opportunity to see Andy Jassen speak and he was talking about how you need to think about metricing people on growth and failure. How many times did you experiment? How many times did you fail? And that as a metric of success. Because if you think about traditional goals and objectives at most organizations, it's about achieving a sales target or achieving a specific growth target or doing something like going forward with a specific initiative and meeting those stage gates. But what if you said, hey, you tried 15 experiments. That's a really great goal for you. Or you failed three times and you learned something from it. It's just a different way of thinking. And if you want to test and learn and you want to move forward, you have to be comfortable with failure. And I know that large organizations can't be startups and startups can't be large organizations. Like that's just really hard to do. But there's ways to set things in place like joint leadership, like share goals. We're going to talk about just agile decision making and different ways of recruiting and talent development that can help set you up for success. And I think large organizations specifically need to start thinking about these things because they will take time to implement and it will take time to change a lot of these ways of working. Speaker 3: And this is really interesting because it then poses the question of who makes the tech stack decisions in this kind of like framework, right? So who is driving like the AI effort of, hey, we should buy this tool or we should build this tool. So what have you seen in terms of like how, basically, how is the ownership of the decision-making, this decision-making happen from an AI perspective in terms of how are we going to implement this whole AI thing? Speaker 1: Yeah, no, I think what's happening today is not going to be what is happening in the future. There's a couple of things I've seen today. I think one being heavily focused on the IT department, right, because they are making technology decisions. They have more of the expertise. So a lot of organizations are using CIOs and their tech organizations to make those decisions. A lot of them are partnering with legal. We're here to understand where is the data going? What do we have access to? Do we own it? So I'm seeing that a lot today. I have seen some organizations, not many, but some have created chief AI officer roles. My opinion on that is it's going to be similar to a chief digital officer role where it makes sense for a time, but then it will go away and it will get pushed back into the business and it will be more of the kind of traditional roles. When something like this comes out, whether it's ecommerce, that was shiny at one point. Now, AI is really shiny. You usually put in a role for someone to specialize and bring it up and then they go back into the business and it becomes a part of everyday work. Those are the ways I've seen people make those decisions today. But I think the challenge that a lot of organizations are having is their individual employees want to use AI, right? What are your guidelines? What are your protocols? How do you say you can use this tool or you can't? You can put the same in or you can't? A lot of organizations are creating their own LLMs so that they're closed and they're not actually sharing that information with the broader world and internet. So I've seen a bunch of different approaches, but today it's usually IT in partnership with the legal department. But I really see in the future that it'll be individual teams can make those decisions just based on kind of the rules and regulations that the organization has put in place. Speaker 2: So I want to unpick that a little bit. So from my own experience as well, I think at the beginning of this AI wave, small businesses were very eager to adopt AI and buy software. And larger businesses, I'm talking like back in 2022 when we started our business, Larger businesses are a lot more cautious. They may do some external experimenting. I know many of the big CPGs, for example, as you said, built their own kind of, let's do product listings with AI and we'll build our own thing. And now, at least from where I'm sitting, I'm seeing a bit of a reverse of that, which is the small businesses, small business owners, and many of whom are my personal friends, and Jo, many of these people as well, they love building these little connected GPTs that connects to this vibe-coded thing in Claude that connects to that for their own system. And on the reverse, we now have a handful of billion-dollar CPG companies. And I know these CPG companies that are now our customers. We started off building their own stuff, and now they didn't really work, didn't work how we wanted. So how are you seeing, from your perspective, that kind of mix of adoption of AI technology between like SMBs to the likes of Bayer? Speaker 1: I think the challenge is a lot of people have built stuff, but they're not actually seeing the numbers come through. Like they're not seeing growth or they're not seeing anything change. I think a piece of that is what we talked about before, where it might not have been built on the right data set and it's not actually solving the problem. I think there's a piece of that. I think there's also a piece of this will take time and there will be a need for adoption. But I'm seeing the similar thing that you're saying, Max. A lot of organizations tried to build it themselves and now they're realizing, wait, We don't have the money or the time or the people to be able to keep up with it. So we need to figure out how we can partner. But the challenge is they are trying to ask the right questions and make sure they're having control of the data. And it's a little bit overwhelming. A lot of the IT professionals that I talk to, they get hit with We have 15 to 20 new AI platforms every week. And so they're trying to determine where do I spend my time, what's important, what's going to go away in the future because one of the bigger players is going to build something. So right now it's quite challenging to identify what makes sense and what they need to actually work through. I think the big hype that I'm hearing on my side around commerce is how AI agents are going to be showcasing products. And how they need to work with the retailer AI agents that exist like the Rufus and the Sparky of the world. Do they need to build their own brand, AI agent, that might potentially talk to the existing retailer agent? Those are a lot of the conversations that I'm having today because a lot of brands are saying, okay, we were pros at SEO. What is GEO? Speaker 2: We love this conversation. Speaker 1: Yeah. Speaker 2: We love this. Unknown Speaker: This is your bread and butter, right? Speaker 2: This is our bread. Yeah. Myself and Jo. Speaker 3: Yeah. Speaker 2: Go on. Take it away. What's your spin on this? Speaker 1: Oh, you're the pro. So you should answer this question. But I don't think that SEO is going away altogether. But I do think that the approach to GEO is different. And so organizations need to understand how they need to change their content. They need to understand what The LLMs are actually scraping. I actually saw something the other day, which I thought was really interesting, and I'd love to get your thoughts, Max and Jo, if you have any different data. But there was research that was done that said the top domains cited by LLMs, like ChatGPT and Perplexity, 40% was Reddit. 26% was Wikipedia and 23% was YouTube. And there was, there's a lot of other ones like Facebook and very small on LinkedIn. Speaker 2: We put some research out on this. We literally analyzed millions of prompt citations from all of our customers. Yeah. Let me just, let me just bring it up for the conversation. And this will be as part of my presentation that I give in October at the Digital Shelf. So we saw that in ChatGPT versus Google AI overviews, as you said, 43% Wikipedia, 12% Reddit, 5% YouTube in ChatGPT. In Google AI overviews, this is more skewed towards 20% Reddit, 19% YouTube. YouTube Obviously owned by Google but massive in Google AI overviews. And then interestingly, 10% on LinkedIn. So I don't know, Jo, you're big on LinkedIn. Are you finding yourself being pulled into Google AI overviews at all? Speaker 3: No, I think this is pretty, I would say fairly new. I don't really see LinkedIn as such an important source yet, but it might happen that it's bigger. I definitely see much bigger, like very big differences in terms of which AI search engine pulls data from where, but LinkedIn is just not one of them. Speaker 2: Statistically, looking at millions of responses, they do LinkedIn and Google AI views. Yes, it's like nothing on ChatGPT. Speaker 1: But what's interesting to me about this data is that ChatGPT, Google AI or Gemini, they are looking at what people are talking about. Right? That's Reddit. That's YouTube. That's LinkedIn. What are people saying? So it's heavily indexed on the loudest voices around specific products and what people like and what's the hot trend and what's coming up next. Like I was doing a search on planning a vacation and I remember like I had been to the place. So I was like testing ChatGPT and I like knew the good spot. But there was this influencer that had posted about this like smaller town. It was in France and they're like, this is the best place to go. And I was like, I found this in none of my research, but a lot of people were talking about it because there was an influencer talking about it. There was like a YouTube channel about it. So I just think that's very interesting, especially for brands, right? If people are talking about your brand, you're showing up. If people are not talking about your brand, they are not showing up. And now, I don't have the answer to how to solve that necessarily. I think social commerce influencers, there's a big piece of that, but that's a really interesting tell in terms of how AI agents are going to service results. Speaker 2: I can talk about that quickly. So there is on Reddit, obviously, is a massive one. AI reviews, ChatGPT, whatever it is. And you do want to be very careful on Reddit. You don't want to break terms of service. But what you can absolutely do is stuff like what one of our employees has done here, which is basically declare upfront, hi, I work for Ozoma, so full transparency in the comment, and then go on and add value. And we've seen this was a highly cited Reddit post in the topics that we want to be visible for. Basically, how do you, for those listening, are there tools that show which queries, keywords on my website are ranking on popular LLMs? GPT, Gemini, et cetera, is a post that someone wrote. And obviously, we wanted Zoma to be cited for that. So yeah, you can absolutely have like employee generated Reddit to go on and say, hey, full transparency, I work in this company, but we're great at doing this. And like, we think about it a lot every day. And we yeah, we've seen our share of mentions from ChatGPT increase because of that kind of thing. Speaker 1: If I bring this back, Max and Jo, to org structures, if you now know where the AI agents are recommending content from, you need to pull in the right teams who own those channels or think about those channels on a daily basis, right? So a lot of times at a brand, it's probably the social team that's working on YouTube or maybe someone on the marketing team, maybe from a retail media perspective. Do you have any ads that are on YouTube or some of the other platforms? Do you have ads that are on LinkedIn? So it's pulling in the right team members to say, hey, this is what's happening. Here's where our content is. Here's where our consumer is shopping. We all need to work together to have full visibility of our consumer journey from start to finish on every single platform that they're working on. That's really hard to do today because most of those functions are siloed. I often hear that the retail media team doesn't even talk to the digital shelf team, so they're running ads on PDPs that don't have images, right? That's the connective tissue that needs to be solved first before you can even talk about some of this AI kind of shiny object feature that we've all been talking about. I hate to be a kind of a broken record about the fundamentals, but there's so many of those fundamentals that need to happen first before you can move forward and have an AI agent that's talking about your brand and trained on all of your data and kind of thinking about the future. Speaker 3: Yeah, absolutely. And actually talking about the future, then how do you see the AI native organization in two, three years time? Like where, like how do you see the entire paradigm changing and developing into that? Speaker 1: I think we touched on a lot of those elements already, but the orchestrator, the person who's orchestrating having those agents support them, I think each function will have more agents that are doing the work for them. So think of having five agents that are doing the work of 12 people. So I think it means taking away some of those hands-on keyboard types of work that needs to get done and being more strategic. So you might have fewer people on a team enabled by more agents. And you might be able to move faster and do things differently. I think that's not going to be a full reality for the next, I don't know, I don't have a crystal ball, a number of years, right? But I think you're going to start to slowly see that work supplemented by AI. I also think that is going to change the way that brands show up to their consumer. There might be a world where agent-to-agent conversation is happening, right? And the consumer is not even interacting with the brand. How do you create an authentic experience for that? And I think a lot of the conversation today we talked about digital and online and online commerce, but the in-store elements still exist. People still want to shop in-store. If you think about Gen Z and even Gen Alpha, Gen Z actually wants to go in-store. They want to have experiences. They want to have multiple different touch points to get them to their end goal. So I think we also need to think about how the teams that are enabled by AI are also working with the in-store shopping team and creating a holistic experience. There's a lot of things we need to figure out, a lot of things we all don't know, but I think there is a world where we can do more, more effectively, it can be more strategic, it can free up time, but it will change the way organizations look and it will change the way we work. Speaker 2: Lauren, thank you. And I'm sure customers, I'm sure listeners have many more questions. And if they do, they can enter their brand or retailer and in London, or not in London and want to travel to London, they can go to the Digital Shelf Institute in, when is it? Speaker 1: October 16th. Speaker 2: October 16th. It's a free ticket if you're a retailer or brand. So make sure you go there. We will link the research in the show notes so you can read the research as well. Speaker 1: And you can also just go to digitalshelfinstitute.org and click on resources and it's like the second thing that comes up. Speaker 2: And you can do that as well and we'll link that as well. So yeah, thank you so much for your time. We appreciate you coming on and look forward to seeing you in person. Speaker 1: Perfect. Thanks for having me. This was fun.