Why you can't trust what AI tells you about your customers.

I'm the CEO and co-founder of Heatseeker, a customer insights platform, and I'm here with Patrick Stal, the CMO of HelloFresh, who has built his entire career around customer curiosity and his passion for building incredible brands. Before HelloFresh, he was at, Tomtom, at Uber, he spent time in fintech with WorldRemit and Zepz, and, we have an incredible conversation, about AI, about the growth of brands in the modern age. And where more marketing organizations are growing right now. Hello, Patrick.

Hey Kate, how you doing?

I'm great. So, Patrick, tell me a little bit about yourself, and tell me a little bit about, your origin story.

Oh, wow. Origin story. Right. Back to basics. All right. Origins. I think I am probably one of the most culturally confused people. I had the what I now call the pleasure of growing up all over the place. But when you're in it, it wasn't, didn't seem that easy. But definitely. I was born in Germany. We moved to Holland. We moved to South America, back to the Netherlands. I've always worked internationally. I would, not be a great sort of like one trick pony. National marketer. And I have great respect for people that have that depth and so I think I've grown up to be, really broad in curiosity global in character. I love getting out of an airplane, Kate. I love just what's new to here. What is different about this place? I love being in food now because it's such a passion. Across the world, anywhere you go, people have their own culinary traits and they're highly passionate about it. And so it's just this great area, around diversity in the world. But yeah, I think my origin story has made me this kind of strange, culturally curious, deeply confused individual. Maybe that's why I like marketing. So I sort of rolled into marketing. I started out as a consultant and, you know, like, like a lot of folks did that, didn't really know what they wanted to be when they grew up. So business, straight up business and strategy consulting and, I love marketing because it's such a diverse function, right? I mean, it's like we speak 20 languages a day. We speak data science and typography design and any, any one hour of the day. So I think that's shaped me in a way. And,

I love what you were just talking to then about food. You know, you're obviously central to the HelloFresh brand and the family of brands that you lead at the moment is this cultural moment for families. Talk to me a little bit about the privilege of touching people's families in that way.

It's huge. And I think just before we tap into that, it's it's, you know, I'm a bit of a strange beast in the sense that I've never been category specific. And across my career, I've always worked across a number of different categories. I've done everything from ridesharing to financial services. As you've said in my advisory years, I worked with everything from supermarkets to bumblebee breeders and museums and automotive brands. And so it's just been incredible to have the luxury to look across multiple different categories and how brands and people interact. And if you think about food is such a as I said before, sort of passionate space for people, but it's also so personal and so private. And today what we're seeing, we're seeing multiple large trends in the world of food. We're seeing food as nutrition really have a I don't call a revival. I think this is a new era where people are really starting to think more technically around, what is my food supposed to do for my body individually? I think that's the evolution of potentially some people that looked at food as fuel before. Very, very pragmatic, very functional, if you will. And so that continues to evolve. On the other end of the spectrum, we're seeing food as emotion and food as a practice and food almost as a call it a mental health aspect. Right? The aspect of preparing food and scratch cooking and de-stressing from the day. And then so importantly, the social aspect actually sitting down for dinner individually with a good book, with your partner, with a family, or even broader with neighbors and society and people coming together at the table. And there's a lot of forces right now in culture that are pulling people away from that specific moment. And I think the largest one being smartphones, distractions, televisions, gaming consoles, the amount of families that don't actually sit down for dinner and have a conversation is astounding. And in almost no instance do people feel good about that. So you've got this tension in culture where people know they want to have a conversation, they want to have a meal, they want to sit down for a family dinner. But at the same time, they're letting distractions come into play. And there are normal distractions. You know, kids are running off to different sports games. Partners come home at different times of the day, and then there's the smartphone problem. So I think that is an area where brands and food can start to play a material role. And particularly with our HelloFresh brand, that's that's an area where we want to make sure that we can not just help people answer the question of what's for dinner tonight, but also help them get back to the table together with their communities.

And how do you touch families? And in that, it's a deeply personal, culturally significant way. I hadn't even thought about, you know, the role of, you know, the family meal and the loneliness epidemic and these sorts of pieces. How do you touch across so many different markets? How is it that you, when that cultural moment means so many different things in so many different cultures, know how do you lead a marketing team to make an impact across so many fronts? At the same time?

Yeah, okay. You'll know this. You've worked globally across a lot of different businesses. You're working with businesses across the globe right now. And I think there's always an aspect of local, flavor insights, whether it's food or automotive. Right. There's something that people passionately believe in as theirs And in a specific culture. But then when you're able to look at things across multiple countries, you do find that a lot of those insights are human. And so the urge to have a meal with one's community is, is pretty universal. Think there isn't a place in the world that I've stepped out of an airplane or a train, or gone somewhere and have seen that the culture is so individualistic that eating is something you do by yourself. Different meals, different occasions and different practices, of course, but I think that there is a thread and I think across my career, I've always tried to be curious around those human insights and, and I think back to my Uber days, for example. So at Uber, when I joined, every country firmly believed that what customers wanted from an Uber trip was materially different because the ecosystem of taxis before that time was materially different, or because prices were materially different. But once you dig really deep into customer insights and you abstracted enough, what we found is that actually people wanted a safe ride within minutes that was affordable. What people want from their food is they want great healthy nutrition. They want it made easy, but they don't want to sacrifice on flavor or they don't want to sacrifice freshness. They don't want to sacrifice on nutrition, and they want to have a great meal together. And they want people to say thank you. That was a great meal, I enjoyed it. They want empty plates. They want people cooking. They want licking off the plates. Whether it's a porcelain plate or a paper plate, it doesn't matter, right? But we all want the end result from a great meal. And there are cheeky ways as a brand that you can then kind of pick up those human insights. That doesn't mean we talk about it in exactly the same way in every market. But I think that there are in almost every category, you will find that there are these human insights that I think, transgress countries and cultural barriers. Yeah.

I remember, when you and I first met, Patrick, we were talking quite a lot about, the importance of live market experiments. It's obviously a really big part of what Heatseeker does. And I remember you framing up almost like where in the funnel, you know, the team likes to test, you know, and where. And that the moment that you choose to run the experiment makes such a huge difference to what it costs us overall, you know, to pull our insights out this way and so finding, you know, light, early ways when it comes to things like product development to interrogate rather than, you know, experimenting in a place where we're going to be wrestling with fulfillment on some of these experiments for decades.

Yeah, 100% Kate. And I like that funnel in product development. Concept testing early on, the ability to particularly if you're mobilizing large, large parts of the organization before you mobilize, understanding whether slightly left or slightly right is the better direction to move in. And so that ability to quickly test, even at a conceptual stage or, at an early stage before we have real concepts to, to allow us to get direction quickly and predict what the right path for the organization would be is incredibly valuable, because I think a lot of organizations do commit to large strategic directions without deep testing, of which one of those directions might actually be stronger. So we and a lot of organizations test around is it AB asset AB like is it picture A or B or video A or B. Which one performs better on the channel. Or we have a user funnel and we test all these different routes around the funnel. Which organization has actually tested? Which of these five product ideas would work best? And FMCG businesses do a great job of that. But I think hyperscalers don't necessarily always have that muscle. Or if we have material strategic direction, are we actually sense-checking with our customers which direction would work best, and do we have a good understanding of where that would go. So I think upstream testing, even if it's more conceptual, can be of huge value to organizations.

Yeah, it's almost one of the fact that, marketing has done such a beautiful job of testing campaign creative. You know, AB testing, you know, the, the brand and messaging, almost means that when we talk about testing, people assume that's what we mean. And not that we're using experiments to interrogate, you know, what's the right job to be done for our customer? Using experiments to interrogate which feature how do we talk about each of these features? What matters more to our customer? Or, you know, this answer or that? And I think uncoupling that, so that we can, you know, experiment more broadly and be just as data driven about our products and about the promises we make to our customers as we are about, you know, is it magenta or lime for the fun?

100%. Right. And I think, unfortunately, a lot of marketing teams in their organizations are perceived as, oh, great, they know how to create assets. And, sure, they have this area around customer insights, but are those customer insights teams really embedded broadly across the organization outside the marketing organization? Because I think especially where we're going to find much flatter organizations. After, in an after the age of AI, where we're going to find much more individual contributors and builders with a lot more access to create. How do we embed that core muscle of customer insight generation? Understanding the customer and making customer insight led decisions into many, many more aspects of the organization. This is the era where insights can feed into practitioners in a way that they've never been able to before. I think you were working on some of the most critical tooling around this area, but this is tooling we've never had where an insights team can literally plug the MCP of insight generation and their target audience into a product team that's thinking about a product brief to build a feature or an app for audience ABC. Suddenly we can actually put insights at the heart of the organization. So for a CEO really wanting to build an insight led and an audience and a customer focused organization, this is the time. But picking the right players and the right tooling to get there quickly will be quite critical, because it will require a change in how we operate.

Yeah, it is one of the spaces that is about to be the most disrupted by so many parts of these agenetic organizations and agenetic enterprises that we're building. The AI is making most things faster, more efficient. But the amazing thing about customer understanding and customer insights is that it's not just the data. You know, it's not just how we apply it. It's not just how, you know, we unlock growth and build behavioral models that we can use across the entire enterprise. You know, it's all of those things at once, you know, through this opportunity of retooling. And it's an incredible cultural moment for the importance and relevance of the function of marketing within the organization. You know, is this is if AI is going to be about being more data driven. It's really a race to the bottom when it comes to marketing as opposed to, you know, this is really a beachhead moment for the for the for marketing to really be the custodians for customer understanding and that discipline from governance, you know, through to fulfillment and delivery, all of those pieces being touched by an opportunity to reframe how we connect with customers every day.

But I have two thoughts on that. So I think the first one, to your point on AI, there is this moment right now where there is this, huge opportunity for us to make the wrong decision and just trust the smart sounding, opinionated person in the room fallacy. Right. So I will go to ChatGPT and ask that ChatGPT about my audiences, and it's going to give me a really smart answer, and I'm going to believe it and sound really smart, and it's going to sound really smart. And it might be 80%, 70, or 50% true. I just fundamentally don't know. And I can feed it all of my insights. And I would say at that moment, synthetically, it would probably give me a largely correct answer. But over time I will start to depend on that. But that is just like as if in the real world. We were going to take the most smartest of our researchers, put them in a room as an oracle and say, tell me everything you know, and I'm just going to go to you with all my questions, and you just keep telling me what you think is the right answer. Whereas what we've done is we go to that smart researcher and we say, please tell us what we think we know, and then let's go validate that and research not doing that second step in this world will create real issues. And it's not just about the models, synthesizing and imagining things. It is actually about the fact that we need to remind ourselves that we've built a muscle to sense-check assumptions, and we need to build that into the fundamental ways of working. So now we're thinking about how do we use AI and GenAI, and then how do we build a shit-check layer on top of that. Right. So how do we say how do we create a research layer that is super agile and powerful to sense check quickly? Because I think that's one of the benefits of GenAI is we can get instant results. So now suddenly a research can't take three months anymore. We need a way to validate assumptions from those models within days, maximum weeks. And so I think the tooling to get us to that place is critically important. The second piece you said about marketing as custodians, just raising thought. Right. And this is important to me, I have seen organizations where marketing has positioned itself or has been traditionally seen as the owners of the customers. And, well, there's nothing wrong with that terminology is if you want to create a customer centric organization, you cannot have a monopoly on customer centricity within the business. I just fundamentally feel culturally, that's not the right approach. So I think we should see marketing as the enablers, as customer centricity. Right? We probably have tools alongside in tech businesses, UX and UI researchers and other parts of the business that also do an amazing job of surfacing insights in different ways. But working together to distribute insights and customer centric behavior more democratically across the organization. So yes, we can run the methodology to bring insights into the organization, but now the tooling will become available for us to actually implant that into decisions more broadly, without us having to send an insight team to work with market to market, or an insight team to work with... Here's product, here's technology, here's finance, here's distribution. Now we can actually implant a lot of those insights into the systems that they use. And that's really exciting. And I think that cultural aspect will come to life even more. But for those businesses that are saying there's a there's a monopoly on customer centricity, and at the same time, we're saying we want to be a customer centric business. I don't think those two things go hand in hand.

Yeah. Look, I think this is some of the coolest stuff I've heard in a while. I've got to say, Patrick, I think the first piece that you framed up that like having the insight or the data or the synthetic version of your customer, but without an ability to validate in the real world that it's real without the ability to calibrate it is like the biggest, you know, it's the it's the biggest opportunity that we have. You know, we've coming from a world where, you know, a lot of our decisions would be made, you know, using surveys. And we know now that they're only, you know, surveys are only 20 to 30% correlated to real world action. So we're already doing better when we get to this next level of maturity that a lot of organizations are in now where they're using AI, they're using ChatGPT to create versions of their customer based on their historical data. So using their customer data or their CRM, their buying behavior, or the performance marketing they're responded to in the past to create synthetic versions of their customer that their teams can use in real time, which is great. Like, I think that's a beautiful way to mature and will be far more successful, no doubt than, you know, using surveys, for instance, you know, and it's much more available, you know, to your point about democratizing this, you know, everybody's got that available to them at the node. I think. I think the next piece, though, is that we got to be able to trust those synthetic answers. We got to be able to take them to the bank like we've got to validate regularly like regularly, that they are truly correlated to the real world. In order for us to feel confident in the huge decisions, like the more, the more the organization is counting on those numbers, or those level of correlation or those insights that they're using at the node. It's almost riskier, you know, than it was before. You know, like we as marketing organizations change, you know, a lot of them is central, centralizing. You know, a lot of the teams at the node are shrinking. And so in a lot of cases, we're coming away from, you know, a lot of the local customer understanding. So, you know, we really need to uplevel the tooling that we're using, you know, to scale, you know, that kind of customer understanding.

Yeah. And I think what we're both saying is that's actually not a point against, generative AI models. I think it's how to lean into them. Right. And so if you have a model, but you're able to feed it with a fresh layer of insights perpetually, it makes it more powerful. It makes it more trustworthy. If you, in the future. We're talking about this, as you and I are both not talking about. Great. We have an LLM and it's telling us this. Let's spend three weeks in focus groups to validate that. But no, like, there are really smart, quick ways to feed and validate that. In the future you might find your LLM training from agents. So I might have a personal agent. That agent might represent me in the world. And so now we have an opportunity where an LLM can learn from my agent. But it might also be able we might actually be able to build a layer of insight generation that asks the agent to validate a decision with me as a real person. Right? And so I might train an agent, but there might be a party out there that's training my agent to learn from me better so it can learn. So this is such an exciting space, and there's so much that's going to happen here. But fundamentally, that nucleus of much stronger predictive behavior insights, which is really what we need, right. We need to know is person A going to do X when I do Y when I put them in front of this. Marketing's all about behavior change, right. We're trying to get a person from not buying something to buying something, from leaving to being loyal, from not being aware to loving a proposition. So this is all about behavior change. So the closer, closer we can get to understanding which decisions we need to take as a business to drive behaviors in, in favor of the business that we're trying to build. That's the answer. And so the excitement around how to generate the engines that allow us to do that, I think is, is bigger than ever. And at the same time, the ability to now plug that into hugely, vastly diverse decisions across a business, I think is bigger than it ever has been.

Yeah, I love I love this, this opportunity that this moment creates to really sort of democratize. Like who owns the customer. You've talked about a pretty aggressively simple approach to that. You know, you love for everybody in the business to use the product as a customer. Talk to me about how this has shaped some of the decisions that have been made within your business.

Yeah, I mean, it starts with actually being a customer. And I know that sounds stupid, but, I've been in a lot of businesses and I've known a lot of businesses where even within some of the senior leadership teams and marketing leadership teams, you ask people straight out like, are you a customer? Do you get a box every week in our business? Like, did you enjoy a meal yesterday? And people don't wholeheartedly say, yes, that's a problem. And I think, I think you then need to ask yourself the question. You know what's wrong with the product? Because apparently you're missing a product if you're if you're actually employees don't consume. And look, I get that. That's different. If you're running a B2B SaaS business or you're running a large scale corporate or production enterprise, but if you're running a consumer business, everybody in your business is a consumer. They might not always be the perfect fit. But you need to ask yourself, start there. Why are people not using this product? What's wrong with it? But the ability of you to surface not just data and to look at the to stare at dashboards and to say, this is what the data is telling me. But the ability to then relate that back to actual customer experiences is absolutely critical. So I had I had the big pleasure of starting at HelloFresh when I was still living in the Netherlands, and we moved the family to Germany, where HQ is, and so I've experienced our product as a weekly customer and two markets now, and there were material differences and there were differences in quality and there were differences in menu composition, but it allowed me to surface and then escalate a number of topics to say quality of produce A is better in this market. What are we doing around our vendors? Can we make sure that the quality here is just as good as there? And it's a very specific example. And obviously the business solves those problems for other ways. But having gone through the personal experience, just allow me to add a personal layer, also storytelling to the insights that we were actually seeing in dashboards underneath. Getting back to your question, I need that from my people. I need my people to actually own the product. I need them to consume the product. I need them to be opinionated about the product. I need them to act and think as customers, and then I need them to understand who the target audiences and be able to put themselves in the shoes of that target audience, even if they are not. Look at that. So if we're looking at a target audience with families, children between the ages of six and 16, not everybody fits into that audience. Right? So how do we get to people that aren't already a part of that audience to empathize, to understand. And that's where, again, some of these I put them in front of those audiences, meet them, talk to them, do the safaris, do the things that FMCG businesses have been so good at training their marketers around, but some other businesses still fundamentally lack in the way that they try to uncover insights. They'll sit behind a mirror even in the age of AI There was nothing as exciting and interesting. And, you know, I make a habit of this when I visit our markets. I ask the team to please organize focus groups, and we go and spend time behind the mirror. And you'd be surprised for how many people. That has been a first, even later stage in their career. It's not something people do at scale anymore, but every single time it has connected a human story to a data point that we had already identified, which has then made the case for change so significantly more powerful within the business, and has developed often 12 to 18 months of physical and digital product roadmaps off the back of it. So do all the intelligent work. And then every few months, put yourself behind a mirror and listen to real customers.

Yeah, we find that with the actual customer avatars that we build, you know, it's so easy, you know, to the point about great data. It's so easy to sort of go deep on, you know, what would the uplift on leveraging this job to be done versus this customer pain point? But we find that now that we, you know, ingest your qualitative research, we ingest, you know, what people are saying in the call center. And we listen to what people are saying at Reddit, It allows us to tell stories for those AI avatars to tell stories about why, you know, it's 4:30 and I don't have anything for the kids. And this is how it makes me feel. You know, I'm going to be late for pick up, and it's sporting games tonight, and it's not really until you hear that, that the job to be done really makes sense to you. And it starts to imply, you know, like solutions and headlines and ways to connect with the customer. In that way,

The why is absolutely critical, Kate. And, and so, you know, yeah, that's 100% the topic there. And exactly right. So factor customer is you've been working out of the gym. You've had a busy day. You want food that doesn't compromise. You still want it ready in a couple of minutes. But you don't want to feel the guilt of having a ready to eat meal. So as a brand, how do you solve it? How do you make a meal that is so fresh, so tasty, so crunchy that customer understands that? Like I'm not sacrificing, I'm having an amazing meal and I deserve that. And it's ready in minutes since I'm getting the best of both worlds. But that storytelling only exists in the real world. I need to talk to you about one thing though, because one thing that I find is super exciting is also how people behave in different situations and you'll know about this, but some marketers might not. We have an amazing insights team. And so one of the things that we're doing is we've really started to democratize customer interviews, I would say. So we have numerous customer interviews any time of the week. Anybody in the business can dial into those. So they're quite democratic in the way. And we can source questions from around the business. But what's really exciting is we're doing that with AI generated voice and interviewers. So we no longer have, like you and me, a person that we actually have a voice that introduces themselves as, hey, I'm a I'm an AI generated voice. And it's been fascinating to see that people are more honest. This is a small human insight. Yeah, because people know that they're speaking to a bot. Yeah. Their responses. We have found, are becoming less polished, less politically correct there. And it's not politically correct. I was like, sometimes it's, you know... Do you eat healthy? Yes. When you're asked by a human because it's a socially acceptable thing to say. Yeah. So when they know they're not speaking to a real human, we are getting more, unpolished answers and I would say better insights. So this is a one of these small nuggets of how new technology in the age of GenAI is, is really helping businesses, I think, become more customer centric and get to the real truths easier.

Yeah, I think that's great. There's like the ambition that I have to have a meal on the table with my family, and then there's a reality where like three out of four of my kids have got an iPad and everyone's fighting, and I just want everyone to be quiet and which is where the high conviction, tagline of Phones Down, Forks Up has come from recently.

I love that, I love that, yeah, yeah, I,

I did find when I was making a HelloFresh meal, you know, this week I was really impressed by I don't tend to use my oven for some reason when I'm cooking for my family, I don't know. It comes from, you know, in Australia, you know, there's a lot of Asian cooking, there's a lot of wok cooking. You know, everything goes into the fry pan. But what was amazing this week is that for the first time, you know, HelloFresh took me through a process of like baking, which was amazing because I don't have to watch it. I can, you know, leave 20 minutes, Wait, you didn’t watch it? And it was a great way to like, solve the customer job to be done, which is, you know, I'm busy, you know, I can't watch it. You know, I can't, you know, but I still get a great outcome. So a great way to kind of solve the customer job to be done. You know, it just in a different way, you know, than I would have, you know, one of the,

one of the great things we hear from customers with, with Factor and ready-to-heat meals is around ingredient discovery. So actually, bringing in ingredients that people don't necessarily know. And they get excited about that. And they love the nutritional values. And, the same thing happens in HelloFresh. For all of fresh it's ingredient discovery. So we have markets where we've introduced things like panko or black sesame and we're sort of like known as. Yeah. Oh, like I discovered this ingredient for HelloFresh. But the other thing is cooking and cooking techniques I remember. So this group in Australia, in Sydney where our customers are saying, oh, you know, unbelievable. I didn't know I could make a risotto in the oven. And I don't need to steer it for 20 minutes. But there's a reason to do that. And, and that's where the sparkle in their eyes suddenly comes up. And so again, data will tell us that recipe got high pick up. But now I know why. Because it had just like unique satisfaction of, oh, I just learned a trick and that made me better. And people loved the meal. I got the satisfaction from preparing a great meal. So we've got the story there with the with the data.

Yeah, yeah, that's a great one. What's the link between, you know, these pieces and sort of loyalty and earning forgiveness actually loved, you know, the I've heard you make that build that bridge before. You know, we talk about brand loyalty, but we forget that loyalty means you know, in a marriage means you're going to forgive me when I when I mess up and I do the wrong thing, and I'm and I. And I'm late for your sister's wedding. You know, that's what real loyalty is about. So talk to me about, you know, what do you think the opportunity for us is? You know, when it comes to brand quality, it's a great example. Your partner might still speak to you, but, you know, there's this. The sister never.

Well, yeah. I think it's a great way in, Kate Right, so I think brand loyalty and strong brands forget about loyalty, because I think that's a very specific topic in and of itself. And it's often sort of like a has a loyalty programs and everything associated. But that's just great brands build the ability for the relationship to, to have tolerance. And I think that ability for the brand to try things, to step out of its boundaries, to try new things, but also in the worst case scenario where things do break and we're all human today and our businesses do have issues. If you have a strong brand with strong brand loyalty and people have a clear idea of what you're trying to do for them, and the fact that brand is absolutely fighting for that customer, if you stumble, if you have a mistake, if an ingredient isn't perfect, if, God forbid, you have a recall or other things. And we've seen this in history with big brands and businesses, I mean, take, take some of the, the phone recalls, remember when batteries were blowing up in phones? and, and, and some of those phone brands got over that issue very quickly because customers really believed that the brand was going to fix the problem. But then we've seen other problems with other businesses where the brand was weak or there wasn't a lot of trust, and customers really boycotted the brand afterwards right? And I think that is the unspoken about power, of brand strength is not just acquisition, it is not just the fact that you stand out from the crowd and you can grow faster. But strong brands have this incredible loyalty and active customer halo effect that I find a lot of businesses have been unable to properly quantify. And so for those of you out there that are not actively measuring for example, the impact of some of your media investment on active audience behavior and loyalty and churn prevention, I would hint that you start to do that because there are material effects there that you're probably not currently measuring as an ROI of your investment. But it goes back to this. No, no, your partner build a strong relationship with brands about different and then, you know, continue to build on that relationship. But when things go wrong, ask for forgiveness. Always do the right thing. But hopefully your customers will have some layer of tolerance if you if you've always fought for their, for their benefits in the past.

Yeah. Well, hopefully we're not, having to ask forgiveness for the AI we're using in marketing. It's been an interesting week this week in the EU, with some of the new guidelines about the use of AI and in marketing. You know, what do we think we need to be aware of at this moment when it comes to the promise of AI and the traps?

Yeah. So for those of your, listeners and viewers that are not aware, as of the 3rd of August in the EU, the EU AI act has gone live and article 30, paragraph 30 of the I think 30 or 30 of the AI act says that any AI generated content needs to be clearly marked as such.

34:02[Ed: the provision discussed is Article 50 of the EU AI Act, applicable from 2 August 2026.]

So if you have visual content, it needs, caveat if you have audio content, you need to tell people and it's really intended to prevent deepfake, type of thing. So if you are mimicking a real person or a real product, you need to call that out to customers. So I think we're entering a fascinating stage in the world of GenAI, particularly Europe, is how our customer is going to react to that? Will they prioritize, vast volumes of entertaining content over potential, trust and, and validity and realism? And I think we'll see brands taking bets on both sides of that spectrum. I think most brands will continue to have a bit of both, but it'll be an interesting, I would say, 24 months as we see how customers react to that. And we are going to see other, other, other countries follow suit, we'll have certain American states follow suit quickly. I think we'll see other countries outside of Europe go down a similar direction quickly. So we need to prepare for that. As marketers, we I mean, obviously we're ready to label, but there's a lot of, infrastructure work that goes into actually knowing which assets are generated, which are not, how do we label that? How do we maintain that? So we track those assets. How do we make sure the metadata is correct? So there's a lot of homework that businesses today need to do. Around the GenAI piece there specifically. But I think the dynamic of entertainment versus trust and authenticity and how customers interact with that in the years to come will be really fascinating for us to see as marketers.

And do you think the role of AI, you know, is more about sort of scaling or sameness or. I mean, I like to think at least, you know, with Heatseeker, you know, our ability to bring data around bold choices, bold decisions, you know, allows brands to swing for the fences a little more. You know, often, you know, the bigger, bolder ideas are harder, you know, to get support across an enterprise for, so talk to me about, you know, I and the risk of averaging, you know, when it comes to marketing conversations,

it's the core risk, right. So anybody can create an average good outcome with AI in minutes today. And I'm give you one I think that's great. But if you want to have amazing outcomes, you actually need to know what you're doing. And you know, actually work with AI and using AI to get the best of AI, and you know how to bar-raise in AI, right? So how do you get to a place where you quickly get to MVP's, you quickly get to outcomes? But I would say almost in the way that you would have managed a team in the past. You can get any person to create an average outcome, probably. How do you bar raise that? And traditionally you'd raise that in craft by sitting next to them with experience and raising the bar and illustrating, showing best practices and otherwise. With AI, I think we're going to have to go through the process. How do we how do we maintain critical thinking in our teams? How do we put them in the place where they are curators and bar raisers, in a place where I would say building becomes highly democratized? Anybody? Now, we don't need people to have 20 layers of teams to execute. They can do that through the tooling. But how do we actually start to resharpen in the minds? And I think we're starting to see a bit of a crisis in that space where because you can create anything. The volume aspect has been the first chapter. How do we just create a lot of volume? Great. That volume is easily average. And so now the question becomes how do we create volume that stands out. That is not average. That is absolutely fantastic. Whether it's content, whether it's analysis, whether it's, strategy. How do we get back to real critical thinking in our employees and in our leadership to make sure that the outputs that we raise the bar so high for the outputs to continue to, to raise themselves, that'll be the challenge because it's really easy to do average stuff. And I fundamentally believe that part of the tools to do that are not just people's brains and critical thinking. But again, know the target audience. Per business, we are a conglomerate. We have 6 to 7 brands. We use know per brand, per region, who exactly is the target audience and what makes them tick? I can't bar raise output if I don't know who I'm doing it for and what the jobs to be done. Yeah. Have access to the right insights and rapid testing and turnaround. Right. I can bar-raise, but what's even better is if within a couple of hours I can get signal from audiences to then bar-raise whenever the output is. Well that's incredible. So insights proposition clarity target audience clarity. And then strategic direction because you can go left or you can go right. As a business you need to know what you're what it is that you're trying to build. So you need those ingredients and you need those ingredients to be democratized across the business. For anybody to really elevate the game. And don't get me wrong, in the right areas, real creative thinking will be absolutely business critical because if we have people that just accept average and bar-raise for quality, but don't bar-raise for stand out quality and think around creatively and prompt AI in a way that actually delivers creative outcomes? I think that will be of incremental added value in the future.

Yeah. Well, look, in terms of some of where you've been, you know, bolder and bar raising recently, you know, the involvement of celebrities and some of, you know, your big, you know, brand bets recently has been really interesting to watch. Like what's behind some of those decisions? Take me through those stories.

Yeah. And so potentially one of the major ones that you're thinking about is our work with Serena Williams, on Factor. So we introduced that partnership recently. That's, a long term partnership. And yeah, I think it, it, it comes down to a couple things. Again, this is a brand where we're fighting, to really introduce the fact that people don't have trade-offs. They can have real food with amazing nutritional value, the highest possible bar within minutes. So they're not trading off convenience for nutritional value, finding somebody whose criteria is so high, who has perpetually raised the stakes, who can be the voice and the face of that type of no compromise attitude. Serena just perfectly set the bill right. And so it is about that fit. And I don't think every brand will find the right individual and so I don't think every brand should work with celebrities necessarily, but where you can and where you have a relevant space and culture and where there is somebody that is such a perfect fit, you should absolutely fight to make that happen. But I also firmly believe that it's a it's a core partnership. So if you're going to work with 1 or 2 major ambassadors, make sure it's not a one off, make it structural, invest it in each other, build that partnership over time. Because like any relationship in the beginning too both parties, you're going to have to get used to what works and how you work with each other. And I think that only grows over time. Besides the larger sort of celebrity endorsements, we work with thousands of creators every month, literally. We have one of the largest creator programs in the world, and we across all of our brands. And so even there, we see that those creator programs take time to materialize. They're much smaller in nature, but a first post does something a second or third post or something else. Why? Also because their audiences are starting to see that it's authentic. It's not just a one off. Hey, you should try this. I've tried it. Here's a promo code getting out of the discount spiral. The actual storytelling. This is amazing for my family. This is why we cook this meal. This is why I love these Factor meals. So I'm back to Factor. I'm like, I just tried the, Chilean Seabass and I'm trying the filet mignon. I'm building up those stories. I'll send you creator programs really need to fit with what it is that you're trying to do with your friends.

Yeah. It's like, it's back to the loyalty. But loyalty in your relationships as well as, like, your loyalty with one customer at a time.

It is. But it's also about cutting through. Right? So there is a lot of noise. There's a lot of slop. There will be ever more in the world. And so I think one of the things that creators and celebrities still allow brands to do when the partnership is real and deep and authentic, is cut through the clutter and bring authenticity to conversation, where you where you need it most.

Amazing. Well, Patrick, we're almost that time. I like to kind of finish these conversations by thinking about, you know, for the modern marketer, the world is changing so fast, you know, no other craft is going through as much transition and transformation as the craft of marketing is right now. With all of this context, you know, what do you think marketers should do on Monday? You know, someone's in their commute listening along to this podcast, you know, like, what's one example of something they could pick up and use in their business on Monday?

Well, good morning, fellow marketers. And that's the number. Happy Monday. I would say take a deep breath and, realize you'll be fine. Marketing will continue to exist, because I know that there are a lot of people going through existential crises right now. I think there's a lot of teams that are think they'll be eliminated. There's a lot of people that see AI and the tooling around AI as a threat. And, I don't believe that. I, I think marketing has a lot of his most exciting days ahead of it. As a function, I think. Sure, the way we do marketing will be completely revolutionized. That's already happened. I it's not that we are at the beginning of this thing that people are talking about Will in be GenAI? We are in it right now. Teams are organizing differently. And I think some of the things that we're going to see, we're going to see flatter organizations, we're going to see flatter organizations. Because with AI at your fingertips, everybody starts to become a builder. And we are going to organize, see organizations that can move much faster. Execution is faster. Rather than writing an extensive brief, I can generate an MVP of a campaign in the next 20 minutes, and I can take that to a creative team. That creative team is using GenAI to create, scripts to create, you know, and storyboards to create basic versions. Our feedback loops become much more condensed. The ability to get quick work out the door at an incredibly high level starts to change materially. And so I think marketing has become more important. Everything that we've just discussed around customer centricity, democratization of insights, marketing, having a broader role within the organization, all of those topics, I think is exciting. But you need to grasp it. So I think the thing I need every marketer to do on the Monday morning after taking a deep breath is go build. If you are not currently playing in these systems, if you are not already developing MVPs of campaigns, if you are not building micro sites and funnels for your business yourself, even as a CMO, and even if that never sees the light of day, if you are not using Claude to build assets, if you're not using Gemini to build assets, if you are not working with your teams, if you are not, trying to program synthetic audiences, if you are not taking historical, research databases and trying to get insights from them and sharing them around in your leadership team, you will fail. So the time is still ripe for you to try things. And if you have, if you're in a mental space where you're objecting against, GenAI you will fail. And so after taking a deep breath, go build, go try things. You're going to fail. That's perfectly fine. But you're going to need to lead teams that are much better at all of these tools. And to lead them, you need to understand how those tools work. So teach yourself, go build pick up a pet project, build a business in the weekend, whatever it is. But get on top of your tooling ecosystem and understand what GenA I can really do.

Well, thank you so much, Patrick, for that incredible conversation. It's been amazing listening to you share everything you've learned about curiosity, customer understanding, and driving growth in the era of AI.

You're welcome.