Cutting Edge AI

#1 From Esports to Building Peec AI: Marius Meiners on Redefining the SEO Industry

• Podcasts from the Angel Invest Team • Season 1 • Episode 1

Share episode

Search has stopped behaving like a list of links. It’s turning into a stream of answers inside AI chat, and that changes everything from attribution to how brands meet customers. We sit down with Peec AI CEO Marius Meiners to unpack what visibility means when people ask ChatGPT or Perplexity for product advice, and how marketers can win in a category that went from a few players to 60-plus seemingly overnight.

Marius’ story sets the tone: from grinding to top ranks in League of Legends to jumping into venture and then founding Peec AI, he built a bias for execution and a “limitless” mindset. That lens helps decode a fast-moving market where differentiation comes from speed, narrative, and data moats. We dig into the practical differences between SEO and AI search: How prompts replace keywords, answers replace SERPs, and why sentiment, coverage, and structured information matter more than backlinks. We also explore the future of AI shopping: when do chat interfaces make buying easier, when do visual platforms win, and how will retailers react if platforms try to own the entire funnel?

You’ll hear the tactics behind Peec AI’s go-to-market, including an unconventional influencer play and a high-energy hackathon that produced real product ideas, like prompt clustering and coupon-aware plugins for ChatGPT. We end with a clear view on market size and fragmentation why a cross-platform analytics layer is inevitable as AI search spreads across ChatGPT, Perplexity, Claude, Google AI Overviews, and agentic search inside messaging apps.

If the future of search is answers, the winners will be the teams who learn faster than the platforms change. Listen, subscribe, and leave a review to help more people discover the Cutting Edge AI podcast.


Marius Meiners:

So we started with like two or three competitors, I think, in like November when we started looking at the market, and now it's just exploded. So as you said, I think now there's like 50 or 60 competitors, and like every three days we find a new one, even today, still.

Jens Lapinski:

This is Cutting Edge AI, brought to you by Angel Invest with your hosts Jens Lapinski and Robin Harbort.

Robin Harbort:

Our guest today is Marius Meiners. He's the CEO and co-founder of Peec AI, a platform for AI Search Analytics. Marius really had a random life. He started playing competitive League of Legends, and now he is building the analytics tool for AI Search. When we talked to him in July, he just raised his $7 million round led by 20 VC, and we are really happy to publish this episode now. We will talk about the future of Search, Marius' limitless mindset, and one of his secret marketing hacks. This is the Cutting Edge AI podcast by Angel Invest. Let's go. Hi Marius. Hi Jens. Great to have you here.

Marius Meiners:

Thank you so much for having me. Super excited. Um it's a beautiful sunny day here in Berlin. You know, the summer's been up and down, but I think today's a great day, so excited to sit down with you guys and talk a bit. Sounds great.

Robin Harbort:

Marius, you you often say that you have the most random life. Like what do you mean by it? And maybe you can tell us a little bit more about yourself.

Marius Meiners:

So I grew up in Bielefeld, pretty small town here in Germany. Essentially went to school, hated school, and then also as a teenager, I was pretty sick. So I actually didn't attend school for like five years in my life. And you know, I always needed something to do, so I tried to keep myself busy and then ended up playing video games um a lot, like just every day, like 10, 12, 14 hours. And then at some point was like, hey Marius, if you do this so much, like might as well become like good at it, right? So I, you know, started trying really hard to get better, and then you know, slowly realized that if you put in work, you can actually improve a lot and started grinding the letters. So I played League of Legends, um, which is one of the more commonly played video games, and then eventually, you know, played up to like the top 100 and played a lot of tournaments, won a couple, lost a couple, right? Just how it goes, and really enjoy that and realized through that experience kind of that if you just put a lot of work into something, you can really, really excel at that, you know, and and really get good at it. And then at some point had like an operation, like a surgery, so could live a normal life again, and then realized I don't really want to be an esports player for for the rest of my life. I actually want to have like a career. So started developing software for two years and then you know, worked in startups and then was never really happy with how the business people like led the startup, to be honest. And always thought, you know, in a way, like I could do a better job. And you know, if you're a developer, you're kind of like a really well-paid construction worker in a way. Like you typically get the task and you just deliver on the task, and that is cool, but never really what I wanted to do. I wanted to come up with the strategy and the execution myself. And then studied economics through that, really enjoyed it actually. Funny enough, um, because I you know I hated school, but studying was actually kind of something I actually quite enjoyed and learned about finance, about venture capital, then worked in a small venture capital fund for a while, and then after that started at PwC in the venture leads department, where I mostly did startup MA, but also helped limited partners invest in venture capital funds across Europe. So did a couple of like interesting transactions investing in, yeah, I don't know, like UVC Speed Invest, Project A, Lakestar. I think we did something too. So like a couple of interesting investments, uh, which was great because then you get to like look into all the you know data rooms of all the VCs and you get to understand that hey, like some of them you know aren't performing that well, even though that might be like very loud on LinkedIn. I think I read like every single uh pitch deck from VCs uh in Europe or something like that, I'm pretty sure. And you know, as a founder, you always like fundraising sucks, but from the VC side, it probably sucks even more because the deals are way longer, right? You fundraise for so long, and uh, you know, you don't really get to yes quickly, I think, except if you have like exceptional, exceptional returns. Yeah, lots of founders like to complain, but I think the VC side might be even worse from the fundraising perspective. And then, you know, always wanted to be a founder, so at some point said, you know, now is the time, I learned enough, now is really time to do it. So left in September last year and then joined the ENTA cohort, and then through that met my two co-founders, Daniel and Toby, and then really started working on PKI.

Jens Lapinski:

I actually have a question around that. When you were sick, right? And then you had this journey after that. At what point in life did you realize that this may have shaped you in unusual ways? How long did that take until that coin dropped?

Marius Meiners:

You always have these discussions with people where you realize you have a completely different mindset in a way. Like I really have a mindset where like you can do probably anything. And then maybe I'm a bit too extreme with this. I think if I wanted to be like now an NBA player, I think I could do it, which is a bit ridiculous because I never played, you know. But like I think if you just work hard enough, you can probably do anything you want. And then people would always like give me the counter-arguments, right? So, and and I, you know, through those discussions, you realize, like, hey, a lot of people don't really have this like growth mindset in a way. They feel really limited in what they can achieve in life. I was never like this because playing esports at this level, everybody also thinks that's like almost impossible, right? Because it's so hard and you you play against like incredible people and it's such a crazy grind. And then you just realize, like, hey, these are all also just people, like even like the biggest entrepreneurs that you know, you always have the names in the head, like Peter Thiel, Elon Musk, whatever, Jeff Bezos, these are also just people. And if they can do it, you know, you can also do it.

Jens Lapinski:

Yeah, it's fascinating. We're gonna get onto technology things. So I see it in like pretty much exactly the same way. I broke my back when I was 12, spent a long time in a hospital. Uh it was not fun. You develop a certain mindset of looking at the world, and it I didn't actually understand for decades that the way I looked at the world was not how most people look at it. Yeah, so anyway. So, how does that then lead to you know, podcast is called cutting edge. You started with cutting edge, with playing, but now you do tech cutting edge. Talk about what that is.

Marius Meiners:

Yeah, sure. So essentially, maybe quick intro to PKI. So, what we do is we help companies and marketing teams specifically understand how they're performing on AI search platforms like ChatGPT and Perplexity, and then we also give them the data that they can then in turn use to improve their visibility performance, and then hopefully drive traffic and drive conversions to their websites and increase their revenue. And the way we came up with it is basically by looking at kind of at the edge of what's happening in AI and consumer. And then when Search dropped into ChatGPT, we realized that this is actually a massive change in how people search. And search is obviously one of the biggest markets in the world, and definitely the biggest online market in the world. So there had to be something cool that can be built there. We then went ahead and talked to like a lot of marketing teams and realized that there wasn't really amazing solutions out there. A lot of them were kind of building their own like visual basics scripts and hacking around, and we immediately understood like, okay, this is something where you can really build something that people want to use and they will be like able to and happy to buy. So that's kind of how we got started on it. And I think it's always, you know, if you want to do something, it makes total sense to start on what's really changing in the world right now, and then go from there and see if if there are things that can be built. Uh so that's kind of how we landed on the idea.

Jens Lapinski:

There are dozens of these kind of tools in the world.

Marius Meiners:

Now, yes, now that's true. Yeah, but uh when we started, it wasn't true yet. Um, so we started with like two or three competitors, I think, in like November when we started looking at the market, and now it just exploded. So, as you said, I think now there's like 50 or 60 competitors, and like every three days we find a new one, even today, still. So that's super funny. In a way, I think it's because it's an obvious opportunity, like it's not something where you have to be a genius to discover this opportunity, right? Because it's just in a way SEO for AI. So SEO was a thing, so you know this is also going to be a thing. So it's kind of obvious. And I think the good thing about obvious cases is that it's clear that there is a market demand, right? So that's good around opportunities like this, but then it's also obvious to the point where you have a lot of competition, and that's obviously not amazing. So I think you you take both in a way if you do such an opportunity, and then it really comes down to execution. So given you are in such a competitive category, you just have to get the execution really right, and you know, you also want to be able to compete and want to be happy to compete, right? It's not a space where you can build for two years and then launch and then have like a big moment. It's really something you need to capture a lot of the market very quickly and early on.

Jens Lapinski:

Yeah, you just announced that you raised money from 20 VC and other lovely people. Yeah. Why do you think? What have they told you? Why have they picked you? They must have looked at all of this, right? Dozens of competitors, and they said, nah, these guys from Berlin, these Peec AI guys, those are the guys. We're gonna do that. What do you have, you think, that that made them pick you?

Marius Meiners:

I think different investors' pattern match to different things, and mostly people want to invest in people that are like themselves. And I think, you know, Harry and Kieran and Julian, also people that didn't really go the normal path to becoming a VC investor. So they look kind of like for weird people. We are really weird people in a way that neither Daniel, Toby or me have like the typical, you know, elite university, two years Goldman Sachs, VHU, LSE background. Really, none of us have that. Um, and then you know, they also don't have that. So they kind of pattern match like, okay, these are like outlier weird people. We kind of like these these guys, and then we also execute it super well, right? So I think from the growth we've shown, it was obvious we are able to deliver. And you know, that plus bringing great KPIs, I think was what convinced them at the end of the day.

Robin Harbort:

Did they mention your quite impressive brand visibility and and marketing strategies? Because I could also think about uh is PKI maybe winning because of like marketing and less because of the product?

Marius Meiners:

I think that's fair. So I think you know, as I described, I think you need to be very fast on go-to-market, so it is really what we focus on at the moment, right? So if you are in a land grab space, you have to be top of mind, right? You have to grab the attention, and that's what we're trying to do, and I think we've been able to do so far. So I think you know, if you think about the geo topic, we're definitely one of the like top three brands you think of. Uh, and that's really the goal. And to get from the top three to like the most dominant is now the next step that we really, really want to achieve.

Robin Harbort:

What is your most most crazy marketing strategy? I think you you mentioned one to me at the OMR.

Marius Meiners:

Yes, I think we found hacky ways to get people to engage with us. And I think one funny story is kind of about about Lily. Lily is one of our advisors, and she's great. Like, she's like the probably the most known influence in our space. She gives like all the keynotes, she does all the all the speeches on all the like main events, and we kind of intercepted her after her speech at the OMR. You know, we were like, hey, what is she like? Like, what can we, you know, how can we get in contact? So Daniel um found out that she was traveling to Berlin after the OMR, which is in Hamburg, and then he was like, Hey, do you want to like get like Berkheim guest list? And he was like, Yeah, sure, like I'm down. Uh so then we connected over that and and uh I got it done. And then she uh went to the club. I also went, and then we like drank shots all day and became friends, and now she's sporting us, so that's like fantastic, right? And I think you have to like find creative ways to to engage with people, especially if they're like you know high in demand, and get them to like uh you know work with you and be excited about working with you.

Jens Lapinski:

At Angel Invest, and and we we have a think tank called the Technology Briefing Center. We're currently thinking about um what the future of consumer shopping looks like in the age of AI. When you think about saying, oh, we do we're doing doing kind of like SEO for for AI or something like that, that's sort of like a neat shorthand for now. Let's let's talk a little bit about the differences, right? Because what you do is not really what people do, because SEO means I'm taking a shop website and I'm optimizing it for Google, which where people go and then they click and then they go to the shop site. But what you are doing is actually similar, but it's actually somewhat different. Can you can you give us some insight on the on the differences there? Absolutely.

Marius Meiners:

So essentially Google is in its core a list of links, and you click on a link to go to the website of the you know shopping provider or whatever, right? And that's kind of how the business model works. They, you know, take all your content, but then they give you some traffic, and the traffic in turn turns to revenue, and then both products are kind of happy, right? And then Google, of course, um was able to grab a very high percentage of the value creation, which is why it's like such a such a great company. This wave now will look very different because people now turn to LLMs and they expect not a list of links, but they expect an answer, right? So they want an output which they then can in in turn use and don't have to click through content anymore because the internet also has gotten not the best experience for consumers, right? If you Google something like, I don't know, opening a bank account or something, and then half the page is sponsored, the race is like affiliates, and then even if you click on one of these things, the content is not trustworthy, like you don't really want to purchase something on that site. So I think now that's really changing. And it remains to be seen how this wave will actually turn out. Like, will these platforms incorporate ads? How will the ads look like? And then also what's the experience on the platform like, right? ChatGPT now implemented shopping, so you can now see products in the chat response and then ask follow-up questions to those products, etc. And then the question is will they prefer their own like first-party sponsors of those products? Will there be like an affiliate model? Can you buy those placements? So I think you know this is really just starting to exist, but there are some fundamental key differences. That is answer versus list of links.

Jens Lapinski:

If you give the answers to what extent, let's just say you do this for, I don't know, practically all sites. I mean, strictly speaking, you don't even need the permission of these sites to at least to a certain extent to go in and take out products, I guess, right? I mean, you can we can just crawl like anybody else if you wanted to, although that's not quite so elegant, right? To what extent is is this kind of service? So if we look 10 years in the future, yeah, if you have an index of all the products that there are in the world, aren't you the the shopping side? Isn't that basically is that a possible path where where you say, okay, if you because it's not just SEO, you actually facilitate much more completely the purchasing of those uh items. Uh so you could say that you're actually an AI shopping engine, if you like. It's a pretty short step from saying I've got the shopping engine to I have a front end as well. And you know, I I directly facilitate that next to all of the other people who plug in. So I can have my own interface somewhere. Yeah, or I can be acquired by somebody who can provide me with a flow so that the the interface is just there, for example. What does that future look like that you're trying to build towards too?

Marius Meiners:

So I think this is kind of what Google's trying with AI mode. In their Google I.O. presentation, they kind of showed that that's the vision they're building towards. Yes. So, you know, by showing like lots of shopping pictures, like a whole shopping section on the side of the screen and stuff, and they really want to make it so that you can purchase like in that screen right away. Of course, the retailers will be extremely skeptical towards doing that because then they lose a lot of touch points with the customers, and I'm sure that they are not super happy about that. And also then the Google will probably like arbitrage away a lot of the profits because they can decide which product to play out and which not to play out, right? So I think there's a big tension in Google's ad business, which they now have with the retailers, and then doing that, which is kind of like against the retailers to some degree. So I think there's there's so many open questions. Um I don't dare to project how that will really look in 10 years. I think what's clear is that this like list of links will go away, and then that some shopping experiences will probably be in the chat interface. I think that's for sure true. Will it be, for example, the most visual interface? I don't know. Like I think, for example, if you talk about fashion, I think that's a segment where it's just much more likely to you know steer towards these platforms like TikTok or Instagram even more, and not really towards like these chat interfaces. However, I think for stuff that's very like information heavy, like an insurance product, a finance product, a pharma product, then it's super likely to me that those product purchases will move towards a chat interface.

Jens Lapinski:

If we took a search interface like that, or let's just say it's there's a there's a there's an AI chat interface and it sits on a side. Yeah. Because uh irrespective of how that looks, it could sit on the left, it could sit on the right, it could look like an intercom screen, right? For some weird reason, most of these interfaces are on the left, but then when you talk about customer support for some reason they sit at the bottom right. I have no idea why this is. It's a maybe a convention thing. If that is the interface that you have in ChatGPT, why is that not the interface you should have on, I don't know, Adidas.com or something like that. Isn't that in a sense gonna all converge onto the same point, if you like? How do you guys think about that?

Marius Meiners:

So I think the the one difference is the one starts blank and the other starts of information, right? If you go to a product detail page, they they choose to present certain information about the product proactively without you having to ask. So I think the the first like message in the chat, if you will, is kind of given by the site. I think it's it's clear that you will be able to ask follow-up questions on that on the page. So for example, let's say you have a complex product, like a drilling machine, and you're in like a web shop from like a drilling machine provider, and like there's this machine, and they give you like all the product details and like a description. I think it's clear that there will be like a chat interface where you can then ask follow-ups, like, hey, this is my exact use case. Does this drilling machine does it support this use case, right? So I think that makes total sense. And I think the more complex the product is, the more likely that is to be helpful. If it's a very simple product, I don't think it makes too much sense. For example, if it's a shoe, like what can you realistically ask about a shoe that you wouldn't already know by looking at the shoe? You know what I mean? So I think then it doesn't really make too much sense. I think it depends so much on the product and like how complex it is, how informed the buyer journey is, etc.

Robin Harbort:

Another topic is um you you hosted uh hackathon this weekend about um AI search. Please correct me if I'm wrong, but you invited a lot of great technical people and they they did different solutions. And I think there was a winner uh going home with 10,000 euros. Is that yeah? How was the weekend? What did you learn who won? Tell me more.

Marius Meiners:

It was so cool because essentially we did it in conjunction with Vela. Vela is um one of the guys that's really like pushing the European ecosystem with Tech Europe. So he's sourcing these events all over the place, and we did one with it together in Berlin and sponsored it. And you know, you're always a bit scared, like, will people show up? Is it gonna be empty? Is the vibe gonna be nice? And Vela asked me at the start, like, hey, do you want to do like a cash price? I was like, sure, what's what should we do? And he's like, maybe like a thousand. I was like, man, a thousand, no one takes a thousand seriously if you split it between like three people, that's irrelevant money. Uh so we pushed it at 10. We're like 10 is a cool number. I think it's gonna make people be excited and really push hard. So the opening time was like 10:30 or something, and I came at like 9.30, and there were already like a lot of people there, and they were all like trying to get an advantage. Some had booked demo calls with our sales team to get informed about what we do, which is hilarious, right? So people really tried really hard to win this money, which I thought was super cool, and the energy was amazing. People really built great solutions. We actually had two or three like really cool ideas come out of it that we also didn't think about from our team. And of course, it's like a cool like hiring event, right? So we want to find the best builders in Berlin to work with us and to build this product. So we now have like six or seven interviews scheduled with people from that event. So we hope that we will at least like hire one or two of them to come in the team and support the mission. And the team that won was actually a fantastic story because it was like five people. They all are not from Berlin, they all came here for this event. One um woke up at 3 a.m., took a plane from Barcelona to uh fly to Berlin, landed here, arrived at the event, and won. So that's super cool. The other guy took like a train from Mannheim and then also drove back with the train at night, like 10 10 p.m. So he arrived also like 4 a.m. in the morning or something back in Mannheim. So great team. I think they totally deserved. They built something fantastic. And yeah, I'm super excited to do more of these. I think it's a really cool thing to have in the ecosystem. We need more stuff like this.

Robin Harbort:

It was like a cr a crazy hackathon founding team, it seems like with the Yeah, yeah, for sure. What did they build? Is it something which can be part of PKI or is it a competitor to PKI? What did they build?

Marius Meiners:

So they built a browser plugin that gives you coupon codes for products that are shown in ChatGPT, which was like the first part of what they built, which is cool because then we get to, you know, of course, we get to then of course also see which prompts people are typing in through that browser plugin. And that's data that we can in turn use to like tell people what people are actually searching for. So that's really valuable. And then they also built a clustering algorithm for prompts, which is great because then you can, you know, aggregate them by meaning and by you know funnel stage, by commercial intent, etc. So they really went like through the whole flow. They had like two data science guys and then three like hackers. So they split up really well the two really cool products combined in one. And I think it's a really cool idea and great topic.

Robin Harbort:

Sounds fantastic. So you're you're looking for more people to hire now in Berlin. What kind of profiles are you looking for and what kind of roles?

Marius Meiners:

Essentially, we have place for at least three more full-sec developers, then also one front-end developer, one back-end developer, one AI researcher, and one data scientist. So those are like the engineering team roles we're looking for. And then we're also looking for like one or two male salespeople, so classical account executives who will continue to drive forward our agency sales, which is one of the customer groups we heavily serve, and they're a bit more high touch than typical brands. And for us, the most important thing is like energy. You have to be like super high energy, you want to work hard and you want to be here in the office. We have like an in-office culture, uh, which we're also excited about. So if you're listening and you are looking for a job and this applies to you, then feel free to reach out.

Jens Lapinski:

One of the questions that we always ask people before we hire them is we ask them about curiosity and we ask them about contrarian thinking. What are the kind of thoughts that you have on your space that where you think I the majority of all the folks that you talk to who are like competitors or in a similar space think the whole world goes like that, and you think, well, I'm not so sure.

Marius Meiners:

A lot of our competitors focus on content creation. And because it's one way to rank in GEO or SEO or any other like platform is to just create content at scale. Yes. We think that's pretty hard to do actually. So for example, there's like AirOps, this is like the whole focus. AirOps is really good. And competing with that product is probably too hard at the stage that we're in right now. I think there will come the point where we can do that, but I wouldn't want to do what our competitors are doing, which is like give out pretty mid content generation ideas. Because if you are a corporate and you buy a software and then they give you like content ideas that are like six out of ten, you lose a lot of trust, right? Because they also know that this is like not the best content. If you do this, you have to do it really well, and doing it really well is really hard, right? So I think there will be the point in the evolution of our company where we throw out content generation ideas, but it's definitely not right now. And I think that the competitors we have are doing this actually a big mistake, and I'm happy that they're doing it.

Jens Lapinski:

This is actually we have a portfolio company called Graswald, which is basically on the content generation side. And I think we will talk to uh to them at some in some future episode, but they focus 100% on content generation, and it's not easy to do this well for a corporate number one, number two is to stay within, have repeatability, follow corporate guidelines, la la la la, God knows what, right? And then stick to the product specification, have no hallucinations about what these things actually look like. It's not a trivial activity. So do you think there is the emergence of like a like a little stack around, okay? So there is the there is a system that displays these products. Do you think that this will get replaced? Because currently you have you have these systems that are optimized towards search, go to the website, purchase, or maybe go to the app and purchase, and then you have the the systems that power that. Do you think there will be a completely new stack around that? Or do you think there's just gonna be elements of the stack that are gonna get swapped out or gonna get added into it?

Marius Meiners:

The classical SEO stack is pretty diverse. So if you ask like a marketing team like what are you using for search, it's actually like five or six products. So it's like a Zambars Districts, HRS, but then it's also like Screaming Fork, and then five, six more tools. So it's a pretty diverse set of tools. And I think it's likely that this wave will play out similar because the companies really want to use the best in class solutions and they don't want like a whole platform as much as in other spaces. So I think this is gonna be similar where there's gonna be like one, two, three pretty big ones, but then also a pretty broad set of like diverse smaller solutions. And then, of course, we are trying very hard to become one of the like bigger ones and not one of the smaller ones.

Jens Lapinski:

And what do you think will go alongside with that? So is there is one of the differentiating aspects to say, okay, we tightly integrate with the three, four other categories within each of which is another three companies or something like that, and all of us are gonna become big together, or is is is that so you have the for example the content generation engines, let's just call it like that. Yeah, within that, you you basically say, Oh, there will be three or four of those, it will be big. And then you might have others. What are the possible categories that could could sit alongside of you?

Marius Meiners:

Sure. So I think e-comm is the like heavily underserved in what we're doing right now, because essentially what we're doing is we're taking a set of prompts, we're running it against the um models, but then for e-com, you have like a thousand SKUs or like 10,000 SKUs. It's just inefficient right now to run 10,000 prompts against the models every day, it's super expensive, so no one does it. So I think that part is very unsolved right now, and also we probably won't solve it in the next months because it would require a completely different business model. Then the whole part of like qualitative analysis is not really done yet. So, in terms of we tell you the sentiment of how the models speak about you, but then companies want something like a brand audit, like an actual crazy big brand audit feature where like for every product, for every target group, for every persona, in every market, they give you reports. Also, not really our focus right now. So, someone's probably gonna build that too. I can come up with like five, six different things there that I think could be like products that someone will probably build in the future, and then we will probably integrate some of them over time.

Jens Lapinski:

I agree. I think that there will be a completely new chain of sort of products that will work in in synchrony or you know, together with each other. And then that's actually gonna be is possible that this is gonna be one of the key differentiating characteristics because then as a as a brand, I know I can need to buy four products to get this job done. And here are the four that are come as a bundle, or not as a bundle, but they they they they are compatible to each other. Yeah. So if if there is a compatibility standard or something like that that everybody can adhere to, which is probably gonna get proposed by, well, I know who initially propose it, but that could be quite interesting.

Marius Meiners:

I agree. I mean, for example, you know, in Google you always had like your set of keywords. That was kind of like the core of what everybody, everything else was like centered around. So we kind of want to have the set of prompts you're tracking as a center, and then everything else can be built around those prompt sets, right? So for example, this qualitative analysis can also be done on that set of prompts. So if we kind of like have like an API for that or like an integration for that, then I think it's something that other companies could use to provide other services on top of.

Robin Harbort:

So the the way about how how you think about this industry, you have like a clear vision. Is this vision also shaped by a recent A16Z report? Because back in May, when when I contacted you for this podcast, you were mentioned in the um in the AI search market theory of A16Z. Did you talk with them about the future of AI search?

Marius Meiners:

We didn't talk to them directly. There were so many touch points now with different people that sometimes it's like hard to know how information goes where at the moment. Of course, like in the fundraising, we talk to almost all the big VCs. And I think a lot of them realize that this is a very interesting market for them and that there is the possibility to build very big companies on this like layer. So I'm not surprised that people are now coming out with like their thesis reports on this. Um, I wouldn't put too much weight on these. They have also been wrong in the past. Of course, you know, it's always super exciting. And then, especially the German ecosystem, I have to say, is like very driven by these. Like they read this stuff and then they're like super excited. Like, we initially talked with funds in Germany in like January, February. They were super uninterested in the topic, they hated the market. They were like, ah, this is too small, and like you probably can't build a unicorn there. Suddenly, this report comes out, and everybody's back in my email. It's like, Marius, we love the space, let's have another conversation. So I think Germany needs to a little bit more be on their own opinion and form their own sentiment than not always look to like these US reports and then like you know, jump on the train essentially. So I think that's a bit unfortunate. But it definitely helped us, right? So the hype is definitely exciting and helps us. Also, like, you know, Harry's uh post helps us and like the announcement video that they did and that we did. So having press is never bad. So we'll take uh everything we can get.

Robin Harbort:

Yeah, also this week in a pretty famous tech podcast, they talked about Peec AI, and one concern they mentioned is that this market is just too little. So you don't think it's too little?

Marius Meiners:

I don't think it's smart at all. Like basically on every major platform, you have at least one or two unicorns that exist as an analytics product there. And I think there's a lot of reason why this can be much bigger. So essentially, if you build for Google, that's a very monopolistic platform, they can always take certain services away from you, and that really cuts into your margin and value creation opportunity. This space will be very fragmented. There's already ChatGPT, perplexity, there's AI mode and Google AI overviews, then you have Claude. There will be like four or five agentic search solutions that will exist in WhatsApp. You now have like this you know, button to search in the bottom right, we might buy perplexity, right? I can go on and on and on. So I think there will be like 10 platforms that will be relevant. You do not want to log into 10 first party analytics solutions to get a sense of how you're performing. That's not what teams want to do, right? So I think it makes total sense that there's a big platform to be built across. All those platforms as an analytics product. And I think that can be really big.

Robin Harbort:

Okay, Mario is really impressive what you were building at Peec. The only question left for me is will you ever go back playing League of Legends?

Marius Meiners:

Yeah, so I have like, I, you know, even after stopping to play professionally, I always once a year installed it for like a month and just like grind it back to like uh like a certain rating, like diamond, and then I uninstalled the game again in like four weeks or something. And then in those four weeks, you could never talk to me. Like I was just like awake until 6 a.m., slept three hours, played again for 18 hours. And you know, if we ever exit this business at that point, that's definitely on the radar. So, you know, back to the grind at least a little bit, and then it's just fun to play. I I like competing, I like playing, so maybe in the future, but not professionally, but like definitely, definitely grinding.

Robin Harbort:

First step building PKI, exiting it, and then back to League of Legends, top 100.

Marius Meiners:

Definitely.

Robin Harbort:

That sounds awesome. Marius, thanks for being here. Thank you so much. Thank you. Thanks for being here. If you enjoyed this episode, support us by leaving a follow and share the cutting edge AI podcast. See you next time.