Ostrich Report Featuring Malte Karsten

You know, like the store on the saloon, right?

It's just so typical. It's a simile for what's going on in the world, Marty. It's all the agents are taking over and we have no idea how to chase those agents.

Just pull the plug. That's my theory. Thanks so much for taking time out to speak with us as well. It's great to finally say hello and put a face to the name.

Yeah, we introduced already somehow— at least me, Hendrik, right? And then we somehow dropped off the introduction round and moved into the discussion already, right? I mean, if you want, we can do it again like first of all.

Yes, so I don't particularly care. So I'm Hendrik Loebeschar. I sit in South Africa. I work for NASDAQ Labs. Whoa!

Will we jump in and record and let's do it?

Okay, I could record this already. Very good, right?

Why not? Because we can edit it up top as well. We'll put in a proper introduction. So Hendrik, I'll lead in. I'll bring you in because you're so keen to get your message out there. And then Malte, we'll bring you in, say hello, and we'll jump straight into the discussion. Typically the way we run this We don't go too far into the background. We try and describe that in the post to go with this in terms of you, where you're from. We do a little bit about it. I think it's important. What I'd love to start out with as part of this is the global perspective. Like you're traveling all around the world and we're going to talk about LLMs and moats and all this kind of good stuff. But I would love you to maybe give us the worldview because we hear a very European view. We hear a North American view. But you're so well traveled in all of this, it would be great to hear what's the world saying, like what are the whispers out there. Hendrik, would you agree it's a nice, nice starting point for us?

Yeah, look, I, I, I think I have like 0.01 of Malte's international view, so I'm, I'm looking forward to hear what he has. Um, so yeah, let's go for it.

Amazing. Okay, give me 10 seconds, I'm gonna jump in and then we'll, we get recording. We'll chat for about 35 minutes. Or if we feel like it's gone longer, we just run with it. If it comes up short and you feel you need to go get a cocktail in Singapore, we let you go do that. Later, later, later. Okay, cool. Okay, count to 10 in your head and then I'm going to jump in.

All right.

Welcome to The Ostrich Report. I am your co-host today, Vinny O Brien. With me as always is my good friend in Cape Town who's getting ready for spring rugby season. Hendrik Loebcher. Hendrik, say hello to the world.

Uh, Vinny, very nice to see you, my friend. Um, I'm glad to see you because, uh, for the, for the listeners, uh, our dear friend, our Irish friend, has had massive technology issues. I'm telling you because it needs to be known. Uh, but Vinny, I don't know what you did, but you've got somebody on today that I've been trying to get to speak to me like for almost a year. So you clearly have a better connection with this gentleman from Europe. Martin Carsten, how are you, sir?

Hey guys, you mean me? Okay, Hendrik, I missed that point. Yeah, basically the problem is maybe that Vinny is a little bit closer, right? From a Cape Town perspective, that's maybe the thing. And that's why we made it a little bit earlier, right? Basically, I'm happy to be here and in a suspenseful mood what is going on here regarding LLMs, AI, and the world.

Well, could you tell people where you are? You've traveled so much and even trying to get this organized, we were thinking about time zones and places you're going to be, but maybe let people know where you are today.

Basically, today I'm based in Singapore and The thing is that I'm heading on to Europe already this night, like after the night a little bit and early morning, and because I have some appointments over there. So I'm basically coming from Mumbai where there's an engagement, actually my most important one, so to say, and trying to push things forward, right? And trying to engage as much as possible. Of course, Europe is, you know, like my base. And where my family is situated. And that's why, yeah, you know, I'm a little bit here and there everywhere, and that's my reality, and that's feeling quite normal for me, basically.

For the non-egalite amongst you, that means you're in stealth mode, right? I've seen that on your LinkedIn. You're obviously building something. It's dragging you around the world. One of the things I wanted to kind of just get your opinion on, or maybe you can just enlighten us a little bit, given all this travel, is something I'm calling the continental divide. Myself and Hendrik work a lot in the North American market. I'm coming back into Europe a lot more this year, so is Hendrik. But maybe you can tell us, we're going to talk a little bit about, as you said, the world of AI and everything that's going on, but maybe tell us about this continental divide. What is Europe, what is Southeast Asia thinking about LLMs? Is there a point of view that is strong? Are they reluctant, or do you see this as a world of opportunity? And you're going to press forward regardless?

I mean, it's a little bit about noise, right? I mean, or the quality of noise, right? I have to say the noise in North America is the loudest, let's say, if you can say it like this, right? So everything that from an agentic AI perspective is changing is actually somehow happening obviously in North America, right? So everything who's engaging and reacting, and we come to that point maybe later, that's maybe a North American thing. It's maybe driven a little bit by the sentiment actually, right? I mean, you have those big players over there, they need to have that noise, you know, like to benefit a little bit from the sentiment, right? Basically, yeah, you've seen that maybe in the news. Um, Europe is more or less in some sort of, uh, fear of everything, um, situation, right? All this regulatory stuff and all that what's going on. If that's, um, if there's, um, what, what is happening, you know, we feel a little bit old, you know, like left behind. And there's the next change. We had that crazy internet coming over us and now there's the next big thing we do not understand. And funny enough, in Southeast Asia, a little bit also in China, I would say, where I've been recently, totally different. It's like the normal, usual, the logical next step is simply coming, right? I mean, it's so obvious, it's so clear. It's also a little bit— the sentiment is maybe a little bit different and more open. Yeah, maybe you can say it like this. Yeah, I mean, if you approach Shanghai actually and you enter a self-driving car or self-driving bus, you know, and then head somewhere, and it's a different perspective on those things, right? But no one in, for example, Singapore or Indonesia would say, "It's a crazy thing called a AI coming, what is that doing to me, or something, right? Not the kind of perspective we're seeing. That's much more open-minded perspective, and basically normally coming next stuff, next steps, right, one after another.

So you said something interesting. I think it was the first sentence you said, it's the quality of noise. So you, who is in build mode, I, I don't ask you about your, your idea. I think that would be just too too rude, I want to ask you about how, how does the— how do you distill that quality of noise? How do you figure out what's worth taking from the US market? And I, I fully get that. I think the sentiment is linked to the stock market, which is linked to 401k and how people are going to retire. And right now that's in flux, so people are really trying to figure out how do we, how do we do really well in the short term to make the long term a little bit easier to swallow. Um, in Europe, that reluctance which is probably shared in South Africa as well, Hendrik. I think it's a cultural thing, right? And now geopolitics is, you know, it's playing into it because there's going to be this standoff. So we've got to find ourselves somewhere in the middle. So how do you, as someone who's getting into build mode, who's saying, despite my age, despite my fear of becoming obsolete in this world, what's it doing for you and how you take quality of noise? And what do you say is good quality versus not so good. Loads of questions here, not just one.

Yeah, loads of questions. Um, I mean, like, like I talked before with Henrik already, right? I mean, the thing is, for me, for me, the perspective is already, you know, like skipping a little bit the actual noisy environment, right? I mean, the things that are actually going on for me are quite, you know, like clear and quite anticipatable. You know, there's nothing about it, right? It's too, too much logic in it that things will change now on the, you know, like short-term, midterm, um, significantly when it comes to, um, the intake of Agentic AI. And that's clear, that's coming. So, um, there are different reactions and there's different action actually taking place, but what I'm more interested in is what's coming beyond. What's next after— um, okay, wait, is everything okay?

Yes, okay, it was a little bit stuck, but continue.

Um, okay, um, what I'm interested in is what's next and what's, what's beyond, what's after that actual, you know, like noisy environment or that actual, um, agentic AI That's, that's what I'm interested in. So, so, I mean, am I—

there was a freeze, but it's okay. I'm gonna— I've bookmarked it here, 10, 10 minutes, so we're good.

And I wanted to say, because, because what I— you know, like, if you take the actual scenario and the actual situation and imagine the different kind of agenda guys are now stepping in and are changing the, the e-commerce landscape as we know today significantly. Okay, yeah, it's happening, right? And it will happen to a certain extent. What's in— what's, what's interesting for me is if that is— has happened, let's say, will have happened, then the big uplifts for those agentic AIs, those big advantages, and that they are representing and causing threats all across the world, let's say, actually in the e-commerce world as we know today, if that is equalizing, if everywhere is Agentic AI, if everyone is using Agentic AI, and that is equalizing totally. I mean, this is too normal then, right? It's like the invention of the internet, the invention of the car, and the first one with the car was driving faster than the ones with the horses in front of their carriages, right? But if everyone is driving a car, that's not the big advantage anymore. So that's why I'm looking a little bit beyond not skipping the actual step in terms of neglecting it or ignoring it. It's more like, but what's next when everything has become equal and the same and the big advantages are no longer there? That's what's driving me actually.

What time— just, Hendrik, I know you're keen to jump in, but what time frame do you have in your head for that, Malte? Because everyone you talk to is thinking in terms of weeks and months and dog years. In your— from your perspective— Hendrik loves dogs as well. Uh, so from your perspective, what kind of time frame is this equalization where the, the playing field's been a bit level and you figure that there's this new world that you can operate within? Like, is that 1 year from now? Is it 2 years from now? Or are you planning on being a digital nomad and traveling all around the world and this could be anytime?

I mean, basically we will definitely see different time frames for different regions, different kind of adoption rate. We will maybe not really recognize that it's common sense already in China because it's much more silent in a way and much more happening without talking too much. And we will definitely recognize when it's happening in the US. And that's maybe, um, also, let's say it like this, it will happen in waves, right? There will be a strong first wave and it will somehow fail, yeah, and then some sort of a pushback and a new, you know, you know, like new, um, let's say, arranging situation, yeah. The people will say, ah, it's not working, it's too dirty, and it's not really returning what the customer is expecting and blah, blah, blah and something. But when I look into the actual status quo, what's already there and what's already silently going on, if you think about Rufus and Amazon, yeah, what is Emilia doing in Amazon already for a long time? And we are still thinking that Amazon is, you know, like manually doing things. Right? They're simply not talking too much about it, right? Because they don't need governmental money for it. That's why, um, there's not so much, you know, like noise around it. It's simply happening. But, um, I don't know whether you've heard about ShopDirect and Project Starfish and so on. So that means, um, I mean, this is already happening. I mean, that's, that's completely the other way around. Outside of Amazon. That's complete agentic AI stuff. So it's happening. So are we concerned about— yes, it's not okay and blah blah something, but from a technical perspective, we're already there somehow because it does not require too much. Yeah, it's simply find the best stuff according to his KPI set. Yeah, that's basically what it is. So a little bit, um, you all remember, or I mean, you're all facing when you're using an AI today, okay, yeah, if your prompt is not right or if something is not really clear, then you are freaking out maybe sometimes because crazy replies are showing up. That might be the case, yes. But that's not a matter of years, right? That's not a matter now of more than 2 years until we come to a point where we say, yeah, okay, it's no part of everything, it's no part of the story. Main question, we should come to this maybe later. Main question is now who's benefiting from it the most going forward. How much is Amazon still benefiting it or platforms are still benefiting from the, let's say, order journey or from the e-commerce as such. But we will see arm length, we will see some sort of, uh, silent inflow, a silent intake, a silent change. And then we come to the point where we ask, okay, it was not that crazy as we thought, but what's next?

I want to bring you in here.

So I think what Malte is saying is, is what I'm seeing, in the sense that for me, the biggest news that no one's going to talk about in the West is Alibaba's Quinn basically being downloaded a billion times, and that, that was news 4 days ago in, in the West. I can tell you now, nobody's going to let an eyelid about that because it's from China. I think the one thing that Malte has said, which I think is the biggest, like, for me, from a second order impact situation, for me the big question is if you have to put, you know, the continents you know, together and say, okay, so who's going to be where? I think it's pretty clear that Africa and Latin America are going to be the last adopters of this. And I think what we will see is that the, the giant superpowers, you know, China and the US— this is essentially going to be like the, the Cold War in the sense that this is something that no one wants to lose or win. But I wanted to go back to something that Malty said, which for me is really interesting. What is after Magento Commerce? Because from a, at a macro level, the one thing that I'm thinking about by myself, and you know, I don't talk about it because I just, it's a, I think we have a security problem. I think we have a ginormous challenge coming to infrastructure for e-commerce because you have all this autonomous situations happening and you're going to need autonomous bots and all those other things to ensure that, you know, infrastructure doesn't get hammered. You know, for me, the one thing that happened that I, that I was thinking about while he was talking was, you know, the article, the big story that Claude said that, you know, they saw for the first time a bot-based attack on, on infrastructure from a hacking situation. The thing is If I have to summarize what I heard from Malte, it's— for me, it's, it's linear growth that we, that we knew in the last 15 years has gone. That's not going to be seen ever again. You're going to now have these massive platforms that have platform scale advantages growing faster because they have access to data, they have access to capital. And I think the only company that could potentially be like the, like the, the anti-platform is OpenAI. Everybody else just does not have enough money. This is, this is going to benefit the large. It's going to benefit the ones who have the most access to infrastructure and cloud and all that stuff. But that, that for me is what I'm hearing from, from, from somebody that's sitting in Singapore and has seen a bunch of things. I will also just make mention that We're in— at almost the end of January. I saw in the middle of December the Indian government doing something that no other nation has done, and that is said that, you know, let's test, um, you know, AI commerce in our, in our country with payment— with a payment gateway that we own so that we can see how this works. The thing is, is that you need guardrails, Vinny, and what Marty has explained is that the guardrails are available in certain places. Hello Europe. But in some places, the innovation, the speed of innovation, like China and the US, and I think certain parts of Europe, this is being hampered by regulation. It doesn't make it— it doesn't give it— it's right or wrong. I just think it's, it's a fact. And that is the challenge that, that is going to become a bigger challenge because you're going you're going to have people sitting in other parts of the world wanting access. For me, the best example of what he— what Marty explained is I wanted to see the new open— new ChatGPT go. I basically got a big, big fat message across my screen that says, this is only available to US customers, please sign up and, you know, we'll let you know when it's arrived. That for me is the code word that's never getting out of the US. Deal with it.

So just to pick up on a couple of points, because I want to bring our guests back in now and, and kind of talk through some of the, the other aspects of the topics that we want to get through. But, but there's an infrastructure layer here being created by Amazon, and they're enjoying the advantage they've created over 20-odd years right now. And you, you've kind of alluded to that up top, Malte. And the, the second part of that is I think it's been proven that the US version of e-commerce, that direct-to-consumer, free shipping, next-day delivery, the model that we have all aspired to has not been— it has not allowed global e-commerce to get beyond a certain level ever. What I'm interested in is, can we design country-specific ways to deliver the overall proposition of commerce, speed it up, through, uh, the agentic layer that's going to be put into all of this. But then you have Amazon on the other side with something like Project Starfish, and they're going to say, we want to control supply chain. And the supply chain is not necessarily the product supply chain, but where data starts and ends. And, you know, all of the news about AI and where the bottlenecks are going to be, everything is focused on trying to solve that bottleneck. Amazon are using their their scale to control that at source. And that is something that you've kind of come in with, something you'd like to talk to us about a little bit, Malte. Um, Frida Hendrik's dog clearly disagrees with me.

Totally.

I'd love to just— obviously, I'd love just to hear your take on that. Like, where are Amazon going with Starfish? What is this competitive advantage you're trying to gain? We all hear about moats all the time. I'd love to hear where you feel it's landing right now.

I mean, let's go one step back, right? I mean, Amazon, what— agentic, or let's say even one more step back, um, replying to what you— what Henrik said, and what, what I think it was about what Henrik said. Um, indeed, agentic AI has stepped in years ago. Let's say it like this. It's a, it's a development. The only thing that's new is the noise that OpenAI brought in to collect that tremendous amount of money, right? I mean, that's the only thing that's new. When ChatGPT went live, and I mean, that was a buzz, like, you know, like, I don't know, can't remember, like, like the Beatles, first Beatles album or whatsoever, right? So that is like, from a marketing perspective, that is the only change. Yeah, so many different companies across the globe were already working on something sophisticated, but when I said, okay, let's change the complete perspective on that bloody tech topic no one is interested in, and let's create some buzz, some noise, some fears, some whatsoever. That's the only thing that changed roughly one year ago, November '23 or something. I don't know, '24, November '24 or something, right? Or '23. So that was basically the only thing that was significantly changing. The rest is a constant flow of development since, I don't know, the late '90s, right? So, but okay, let's park that one. Next, um, regarding Amazon, let's also get one step back, right? I mean, what did Amazon do last year when this agentic story came up? What did they say? Okay, robots.txt, block everything that is agentic. Perplexity out. Google, okay, it's a different purpose. Yeah, partly allowed. But blocking off all those pure agentic AI, the integrated ones, the pure ones, everything. robots.txt said no entry. Everyone was freaking out. How stupid. Have you heard of the new cooperation of Shopify and Etsy, letting the Agentics in, this is smart and so much more advanced and something.

Hmm.

Hmm.

From my perspective, Amazon was simply buying time. Yeah, they're big enough to, you know, like don't care too much about the 0.2, 0.5, I don't know, percentage of Agentics AI checkouts actually that's taking place, right? And realistically, so they were buying time. For a simple, clear purpose. And you have maybe heard it, SAP News, not so important news, you know, they have those new fees for the APIs. And you know, like they are getting effective in April or something, but they dropped it around New Year's Eve. I don't know, can't remember precisely. I had the news already in November, I think. So basically, that's the core thing. Yeah, you change your API environment, you change the way someone from external can approach big data in your body, in your core body. It's not about the customer, the one who wants to take a look around in the deals page. It's about big data demand coming from outside in. This is totally changing now. API will cost some bucks for the guys who are using API. But the big thing behind is there's a cost structure, a cost system created for the agentic AI because the agentic AI players like Perplexity, you have heard all that, you know, like fighting and embars between Amazon and Perplexity. They need some sort of, they need Amazon. Amazon is too big. Imagine you, Hendrik, you are approaching your AI bot, saying, give me, search for me, I don't know, give me the best washing machine, running shoes, running shoes, you know, like for all that Cape Town brilliant views walking.

Yeah.

So you need those ones and then you get results, but there never is Amazon in between.. And that's definitely from a North American perspective, from a European perspective, and kind of new also from a South African perspective. That's an issue. Amazon is too big to be, let's say, skipped all the time by Agentic AI results. Amazon was betting on that one successfully. So now the Agentic guys are sitting around the table in the Amazon building and trying to negotiate something, right? How much do we have to pay per job, per ping, you know? And that needs to be arranged, that needs to be, you know, like, talked about. But Amazon is saying, A, we are limiting the access for the Agentic AI, because in the very beginning, without the robots.txt closing the doors entirely for Agentic AI— imagine that nightmare. I had this discussion in Seattle exactly like this. Agentica AI is capable of reading everything inside of Amazon within, I mean, some days at least, 2 weeks max. Everything is out. Reviews, history, sales history, everything. What if, you know, like, hand it over to someone, to all those Agenticas? That's why it was totally correct that Amazon was closing all the doors completely and not letting them in. Through certain dedicated doors with restricted, you know, like behind that door, restricted access. Plus the entry points have some sort of some, some, you know, like cash machines, uh, where you have to throw in a coin if you would like to enter that limited space. And right after, you know, like checking. And the funny thing is Amazon is also, um, has found a survival mode, let's say it like this, for its advertising unit. Because Agentic AI is killing advertising. It makes no sense for you to pay for advertising to get better visibility if the Agentic AI bot does not care about who paid for something, right? I mean The Agentic AI bot is capable of going through all 50 or 500 pages of the SERP, and that's not, you know, like a human being is only capable of the first page without scrolling, right? So that's why advertising is on the, you know, like somehow it's for me, it was the most important thing. Now if Amazon is saying this is the door you have to go through, Perplexity Bot, And after, behind the door, we are presenting you the best shoes. Amazon is presenting the best shoes behind that door. That is my selection. That is what I can offer you. And out of that selection, uh, make your choice, right? This is this limitation thing. So Amazon is still— can still, you know, like, ask the audience, the seller audience, if you would be part of selection I'm offering to the Agentic AI bot when it's coming, when it's showing up, then you have to pay. That's PPC, right? That's basically— that's like so many different things Amazon had to think about and had to arrange in their heads. And from my perspective, it looks quite, um, quite promising. And it's not really changed by a the big engagement of Walmart that they've, you know, like NRF, where they dropped that engagement regarding Agentic with Google and blah blah and something, and definitely for more to come. For me, as I said to Henrik before already, this is showing weakness. Yeah, what Etsy did, Shopify did, Walmart now has done, it's showing somehow to me, it's showing weakness. If you're not really negotiating, yeah, end to end and trying to keep as much as possible from that entire, you know, like order flow, order world in the e-commerce environment, keep it on your end, keep it on your side. Yeah.

Yeah.

And all that. I mean, if you think about Amazon going forward and they are able basically take from the 5 e-commerce steps, they will basically take the intent step, maybe the first one.. But starting from the second step, and that's a little blurry, starting from the second step going forward, Amazon is deciding already and is in the, let's say, in the stewardship somehow. In the Walmart example— yeah, okay, Vinny, come on.

No, it's just a question because I have two of the biggest brains surrounding Amazon in front of me. So I want to try and make sure I get all the questions out that I can think of. But is this Amazon sweating an existing asset. Is this AWS flexing its muscle again, or is this going to be classified as new revenue? One of the things in the earnings calls that started happening late last year was all e-commerce GMV was now called Agentic GMV. And it's almost like you're feeling that this is the death of e-commerce as we know it, because it's now just going to be classified as AI revenue. But, but my fundamental question is, is this AWS sweating itself and saying there's this brilliant flywheel, there's this brilliant ecosystem we already have and we're getting so much revenue out of that, that when we renegotiate on API access and the cost associated with it, it's just flexing that muscle a little bit more. And the ads at the other end of it is going to be that extension. That's kind of my first part to it. And then how has the response been? You've said it's kind of shown weakness. How has the response been in the overall market? I mean, Shopify, they never do anything with Europe. I mean, If you're a European seller that's reliant on Shopify and their AI layer to grow your business, Hendrix's laugh says it all. But they neglect Europe all the time. They pick up revenue just because they can, I guess. But yeah, they can. So I just want to ask that first question. Is it flexing a muscle they know already works and that they're going to be able to squeeze loads out of it?

So what I heard from Olti was a few things. So let's start at the top. This is Amazon doing what it's always, always done, and that is that they, they manage supply and demand. What Marty is essentially— what he described is we will pick who we will be making kings and queens in our ecosystem. It's not going to be a third-party bot. Just get that. You're not going to do that. Number, number two, and I think this is one of the things that I've learned just by reading and listening to, you know, folks who know this, is that The infrastructure to handle all of this costs a lot of money, Vinny. So if you want the access to the entire ecosystem, you're going to pay for this. One of the things that I continue to believe that people misunderstood is, and I've watched and followed Amazon for close on 20 years, Amazon is not a first mover. It's never been one. It's a fast follower. What does that mean? It means exactly what Marty said. It said we are going to put a big door in front of our ecosystem and we are going to tell you, you are going to pay a quarter, a 35 cents or 40 cents a call. It's your call to take it or leave it. It sucks for the sellers. It— but it also, once again, for me, what Marty communicated to me validates a thought that I've had for almost a year. And that thought is, is that the new flywheel that Amazon has runs on AWS advertising and logistics. Everything else is secondary. Who cares? If it works, it works. If it doesn't, it's not the end of the world.

Can I just pause you there? Because there's just something I want the two of you to answer here. We're talking at a level now that is above barroom chat, okay? Like, this is visionary type stuff. It's an incredible brain dump that you're giving us, Malte. It's incredible to listen to. From a seller perspective, where they get their news is going to dictate the sorts of decisions that they make this year. Who are going to be the net beneficiaries here? Are we going to see the return of the legacy brand? Are they going to start using commerce to their benefit now, or is it going to be nimble, um, nimbler direct-to-consumer type brands who have kind of hit a wall for the last while. What way do we see this playing out in the real world for sellers? If we can take it down to trading and merchants right now for a minute, where do you guys see the winners and losers?

So for me, the winners is the, is the legacy brand that have cash, that can literally go to any platform and say you know, we will give you a million dollars, please make sure that we get listed and found via whatever AI bot. For me, the best example, what we are talking about now, is exactly why PayPal bought Symbio last week. Symbio got bought because they needed access to inventory data, they needed access to things that they didn't have. As a seller, And you know this better than most, Mr. O'Brien. I love it when I call him Mr. O'Brien. The reality is, is that the message, the message will be owned and managed by Amazon. It's literally, you can play inside our ecosystem. If you look at Reddit, I, I read Reddit, I do read Reddit every day, and the biggest thing that I can tell you inside the Amazon channel is that everybody and their mother, sister, and father is moaning about the fact that Rufus is being forced on them. So as a brand, what I'm telling you is if you're a small nimble brand or DTC brand or a legacy brand, you're gonna have to kiss the ring, and the ring's name is Rufus. Malty, can I have your thoughts please, sir?

I mean, basically what, what you are saying is right. Maybe my takeaway is slightly different. And how is that coming together? Um, basically, yes, it's who's been benefiting from the exit situation going forward. It's the one who has the deepest pockets. Whether it is a brand, I don't know. I mean, going forward, imagine you are— and that's hard to imagine— let's imagine we three are Agenti KI bots. Do we really care about brands? Do we really care about legacy, about that was always good, or my mother already loved Miele or something, you know? And AI, yeah, it's great brand, it's a KPI set, yeah, has some sort of, you know, like, um, has to rely on something, on some, on certain information, but basically does not care too much about the brand. So what we are seeing is, you've maybe seen the data, not 45% in 2030, I think, in 2030, not in the US, 275 million customers. And I can't repeat the number exactly, but it was a little bit less than 50% will be AI, agentic AI checkouts, let's say, right? Or customers appearing., right? So that's the one thing. On the other, on the other end, I've dropped yesterday this post regarding the percentage in the different, in the US, in the different categories. What's still brand search, branded search, and what's non-brand search, right? So, and you've seen, you've seen categories like, was it like baby or something with 97% non-branded or something? If I would be Procter Gamble, I would try to worry. So, but it's not— maybe I would, I would not start to worry. Yeah, I would, I would definitely— I mean, it's too late to start worrying, and this, that should have happened much earlier. So the thing is, still the best-selling brands are those ones, and that's what's right in, in, in the, in the stuff that you said, Henrik, right? It's still the biggest brands who can adopt and arrange and, you know, like, have the deepest pockets and can pay for the best Sports on top, you know, when it comes to advertising and so on and so on. So they can arrange, but it's no longer because it's a Procter Gamble brand, the offer. You have to play by the rules of the marketplace, you have to play by the rules of the advertising system to stay on top. And they have hopefully the best guys, otherwise they would do something wrong. Yeah, the best consultancy, best, best in-house stuff, uh, stuff and so on and so on. That's basically— but it's no longer about brands. So I think we, we will see an exchange of the brand going forward more by the best players maybe, or something. I don't have the right narrative going forward for the future, but that will significantly change the entire brand perception. Brand definition perception will definitely change.

Amen.

It's long overdue though. I mean, we've been in this false reality for such a long time as to how commerce should work, um, and we have an over-selection problem. Like, everywhere you go, there's just too many, too many ways to buy the same thing. And, and as a result, I could see the, the kind of consolidation of the market, and that has to happen if search is to work and discovery is to work the way that we're told it's going to work because what I call it like the rule of 3%, it's 3% conversion rate for the last 20 years has got us in this shitty position where no one's making any money in commerce. So I think you're going to have consolidation of tech, consolidation of product, and what that results in is a much narrower window for us to have to think. And CPG categories completely get that. In terms then of how we start talking about this better, like this is a conversation I don't think I've heard in our world, not maybe in Europe. Malte, maybe you hear it. I'd love to figure out how do we distribute this kind of message into operators and how do we get them to pause. As Hendrik was talking there the last time, it feels like we're after meeting a parent who's telling us it's okay to sleep at night and not worry about tomorrow because there's a long way to go in this whole journey and we can just take our time and That's kind of what I get from listening to you and listening to how Hendrik describes it as well, that we can just kind of pause for breath a little bit. So how do we make it easier for people to figure out what's quality noise and what's just noise?

So before Marty answers the question, and I think this is a big deal, I think as a brand you need to find the people inside your organization that know where the buck is going to. Uh, Vinny O Brien, you and I talk about this often. The, the, the really good talent is mostly in places where there's lots of money. I want to also just add to this— Multi mentioned this, and I think this is an important part. If you— in most markets not called the US, this thing is going to happen, and it's going to happen on a platform level, in a sense that some of these platforms are moving faster than others. And I think that's important. But I think the thing that he— the point that he made right at the beginning is the point that I'm getting more and more happy with, to just say this is what it is, is that scale isn't going to be communicated to many people. It's going to be literally just— you're going to be told, this is what— if you want to play, this is where you're going to have to be, otherwise you're going to not be anywhere. And this is the challenge that North America is creating for the rest of the world, because in markets like China in India, in Africa, and Latin America, this is going to be in most cases be run by governments. Let's just face the facts because this goes against demand. But I want to hear what Malte says about, you know, what do brands do, because I want to know the answer as well, because I think nobody knows, to be honest.

I mean, basically, as you, as you already anticipated maybe a little bit in the question Right? I mean, AgentX AI will simply clean the backyards of all those guys, right? If it's today okay to have 20,000 times the same toothbrush, yeah, as a CPG giant, if this is okay, then going forward they will quickly find out, okay, it doesn't make sense too much because the AgentX always picked the same one, the best one with the best reviews, with the best, I don't know, you know, like, uh, legacy whatsoever. Yeah, the the beachfront real estate will basically always sell. It's not about today, you can, you know, trick a little bit around. And that's maybe the question that Amazon might play a little bit different, preserving this Amazon Advertising world. But basically, if it comes to the entire e-commerce environment, I would say for you, it makes totally sense to have this one great item that is definitely picked by those bots, but not like, you know, like, I mean, today it makes sense. You have different items, yeah, you have maybe cheaper Prime costs or whatsoever, and then you play this item up with advertising for a certain deal scenario in Black Friday or something, and you will sell it anyway. So, but it makes long-term Makes no sense, but it's still okay. Yeah, you buy a little bit more, you purchase it, you know, obtain a little bit more, bring it over the containers from China, and you have it and you sell it and something. But going forward, focusing, focusing on the best items that have a chance to compete, you know, like external. Because an OpenAI whatsoever, whether it's ChatGPT or something else, will definitely pick the best ones, the best 5, and then decide for one and then offer this one. To the prompter or even buy it for him. Yeah, um, so that's basically— it's a true demand to, you know, like melt down this over, um, let's say, yeah, this over-selection, let's say, like this, right? That's basically that. But that's coming automatically. I think this is coming automatically if you see where it flows, where the river goes, right? That it's basically this one item that's selling all the time why should you invest too much engagement into the rest? Yeah, into the long tail definition will change.

Do you know what? I was looking at your post yesterday and for branded and non-branded search, and I just had a glance at it again just now, but something occurred to me looking at the categories. It kind of tells its own picture. If you look at pantry staples and things that you'd expect in your kitchen or in your pantry or whatever. It's coming down to non-branded search and it's dominating at about 80, 90% in certain cases, which makes sense because people are value shopping right now. And value shopping is on a week-to-week basis. Where am I going to get the best in this category? But the one category I think it can transform is baby, right? And I've thought about this loads because I've got 3 kids and I remember each pregnancy being different. And your requirement after having a child is so unique in that experience. You might be getting more pharmaceutical products, or you might need formula because baby's not breastfeeding. And there's all these variables that go in. And typically the mother in that instance will go to friends, she'll talk to her mom or sisters or cousins or aunts, all these different things. And it becomes this really personal experience that is almost impossible to replicate from an e-commerce perspective, which is why I think there's been loads of category killers and lots of money invested in the category. I can see now how a conversation with an LLM becomes part of that discussion. So the products that start surfacing in there and the baby care brands and the post-pregnancy brands that start catering to that, I think it could be transformative in that category to start taking a much larger chunk of the spend that goes into it. And when I look through the categories that you put in there, I think it's quite transformative. In its nature. And, and then the other one, and you've alluded to it again in a later post, was around the, the change in the nature of luxury. I, I think we are really looking at some very different days of how luxury is. And there's luxury at the very top level, and then there's aspirational luxury. Aspirational luxury, I think, is that's that category that's in real threat. I think the layer above that is something that just doesn't exist in the world of commerce the way we think about it. The way I can describe that for people, anyone listening, is you go in to buy a new Porsche and they give you a free holiday to the Seychelles just for buying the new Porsche. Like, sorry, it's a stupid example, but maybe it's a Maserati or something you're buying. But that's the kind of level that you're dealing at, and they're the sort of things that are happening. So that's not touchable. But as I look through those categories, going back to my original point, I do see that there's room for transformation in there. Beverage being another one. It's one of those—

Surprise, right? Yeah, it's also a little bit counterintuitive, right? I mean, you would say, but that's already what's happening. The categories that normally you would say require the highest level of trust are the ones that are exposed to non-branded search the most. And that's totally counterintuitive, but it's not because I have 3 kids. They don't care about, you know, in the past, It was like, it's, uh, as you said, it's a, it's a Miele, it's an, it's an, it's an whatsoever. It's a, no, Porsche is different, but you know, like those commodity stuff. It's a Miele, it must be working, it must be brilliant, it must be okay. I don't need to read too much, you know, like stuff, um, reviews or whatsoever, like 30 years ago. Yeah, it's a Miele, it's okay. Because they build up some brand-level general trust thing. In Amazon, Amazon is the most important trust layer. It's on Amazon, it's existing, it's buyable, it's, uh, shipped immediately, it's, it's in my hands tomorrow because it's Prime, blah blah blah. That's overarching trust layer underneath. For my son, for my daughter. It does not matter too much how the brand is named. I'm looking like an agent, okay? I already— I'm looking already for the best thing, and reviews already replace the brand. That's what you said already, Vinny. Reviews already replaced like personal relationship. I would also, you know, like compound into reviews, yeah, and asking your friends or whatsoever. So The review layer within an overarching Amazon trust layer, let's say, the review layer has become the most important layer and not the brand name. And that's why non-branded search might end up most of the time with P&G products. Okay, because they're doing a great job, but not because of the P&G narrative on top, just because of, yeah, they represented the best enhanced brand content, pictures, videos, explanation, exactly, you know, like meeting the customer, you know, where the questions are, blah blah blah, you know, they have experience, they have money, they have, you know, like, they can do it. Deepest pockets. And now we are, you know, like, closing the circle. It's all about the deepest pockets if you allocate that deep pocket, um, you know, like, um, content or like, um, ingredients perfectly into right decisions, right? I mean, that's not natural.

But I think it's a great point to bring it all back around because we talked a little bit about talent, not much, but something that's on my mind loads. But the large retailers have been busy bringing talent in-house to make use of how AI is going to benefit them. They're not trying to buy agency talent for this, which I think is a bit of a shift. That's kind of like early 2000s mentality, which I think is interesting. And I'll be watching to see the brands benefit there. I'm also looking at the clock and seeing that we've gone well over 50 minutes, which is incredible. I think we could talk about this for hours and hours. Maybe what we might do is get back together in 3 months or 4 months and just see how this is shifting and see what patterns are emerging. Because Amazon has been the narrative for commerce all around the world for over 20 years now. And they've created ecosystems, economies. They've built careers for people. So I think it's going to be interesting how it plays out. The other one I'm going to be— I'll ask us to watch is Shopify on the basis of your trust comment. Shopify now has a decision to make, I think, Hendrik, with Shop Pay and Shop itself as an app. If they're going to be competing on trust, they've got to put their balls on the block and say Shop is going to be this counterpoint to Amazon because we now feel that we can give you something that they can. And we've said before, you're either trading on trust or price. And the days of trading on price are long gone. So now it's going to be trust, and that's where consumers are going to end up. Uh, Hendrik, anything you'd like to ask or kind of comment on before we try and bring this to some sort of natural close? This is a brilliant, an unbelievable discussion.

So yes, it's a great discussion. Um, I, I think for me, the, the one of the comments of the year thus far that Malte mentioned, which I think is a big deal is I think brands are gonna have to do their homework about how many agents are available in a specific platform. And I, I think the one thing that, that Malte mentioned to me today is, is that Shopify and all these other platforms are partnering with others because they have to, because they don't have the ecosystem. The fact for me remains is that these, you know, these protocols that they are building essentially makes APIs for me almost redundant, which, which is going to scare quite a few people. But the one thing that I'll add is, is that we, we're in the beginning of a massive transformation, and there's lots and lots of change coming. I know I'm stating the obvious here, but Malti has just confirmed to me today that I don't think we understand the change that is coming. That for me is the takeaway from today's conversation.

Vinny, there's one thing I'd love to get you guys to contribute to. Have you ever heard of the Cluetrain Manifesto?

No.

So there's this kind of like a guide to how to treat the internet when it first came out, and this is when it was meant to be this layer, and there's 95 theses in there of how we should be, how we should treat this new utility for us all. And it was about access and democratizing knowledge and all those kind of things. It, it kind of strikes me now that there's, there's an opportunity to write a Q-Train manifesto for the age of agentic commerce right now because it is fundamentally changing the internet. I see these things like UCP as being that first step in it, um, almost having these protocol layers which fundamentally define how information is transferred and shared right across this new ecosystem. And then there's got to be a brand layer as well. And there's also going to be some consumer layers. I'm going to share some information with you guys to see what you think of it. But defining some sort of new— it's not a playbook, it's a way of thinking about the world in this new environment, I think would be really useful for anyone who's engaging in this. But look, I just want to say thanks, Malte. I know it's late. You're in Southeast East Asia now. Um, it's probably cocktail time for you. For, for me, it's going to be a cup of tea and potatoes. Hendrik, probably some fish given it's height of summer for you.

Uh, yes. Um, I'm actually gonna go make dinner now right after this because, um, we've spoken for a while. But, um, Vinny, this is, this was a mind-blowing conversation. And, and it, and I think what, for sure, what I particularly like is that Malte brought a global point of view to this. Which I don't think there's enough of for this conversation. I genuinely—

yeah, look, I, I, it's the sort of thing I, I think should go on news channels, and, and you should be putting it out there a little bit more. And as I said midway through the conversation, I think we need to try and bring more access to this kind of information. So that's something we're on a mission to do, bring more of it throughout the year on the Ostrich Report. But Malta, cannot thank you enough for giving us your time, your wisdom. And hopefully you'll join us again in a few months and we can continue the conversation.

I enjoyed it a lot. Thank you guys. And it was indeed a global round here, right, if we take our positions on the map actually. It was great and very, very interesting. And yeah, I hope to see you again in the same round here.

Sounds good. Thanks for watching the Ostrich Report. Tune in next week and we will have more guests, more information, and more of Hendrik, unfortunately. See you.

Oh jeez.

Oh jeez, guys, thank you so much for that. That was amazing. Um, I'll get the edits done and we'll be releasing this in a few weeks. We, we'll give you the raw footage as well, and, um, I'll give you the edited version as well. I've noted where we need a few, uh, a few edits. I'll send it on to our production manager and we will take it on from there. But thanks so much for giving your time so freely and And yeah, it was mind-blowing. It's just such a huge wealth of knowledge. Thank you both.

Thank you, guys, and take care. And we talk again, right? Just drop me a message and we talk again. Bye, guys. Enjoy.

Take care. Have a good evening, guys. Bye-bye.