Masters of SearchPresented byRadyant
← All episodes
Masters of Search episode 1: SEO is DEAD | Malte Landwehr, CMO & CPO @ Peec AI | #1 cover art
EP 01·Nov 19, 2025

SEO is DEAD | Malte Landwehr, CMO & CPO @ Peec AI | #1

SpotifyApple PodcastsYouTube

Show notes

Malte Landwehr just quit his 6-figure corporate job (VP SEO @ idealo) to bet everything on AI search. So I went live with him on the 1st episode of "Masters of Search", to ask him everything about it.

Malte spent 5 years leading one of Europe's strongest SEO machines. And he walked away from it all to join a 10-person AI startup.

Why?

Because he believes we're about to witness the biggest shift in digital marketing history.

When someone with Malte's track record bets his career on it, you should pay attention.

We dove deep into: → How he thinks about the "Google is dead" claims → What he believes really drives visibility in AI search → How he thinks about the shift from SEO to GEO → Why he left his VP role for a startup → What this means for your marketing strategy

▶ Let's connect! 🔗 Niklas on LinkedIn: https://www.linkedin.com/in/niklas-buschner/ Radyant on LinkedIn: https://www.linkedin.com/company/radyant/ Malte on LinkedIn: https://www.linkedin.com/in/landwehr/ Peec AI on LinkedIn: https://www.linkedin.com/company/peec-ai/

Transcript

Full conversation

via podigee
  • 00:00:02So I think maybe we can just casually start in and everybody that pours in still I will just let them in along the way. Um so I know everyone has come from Malta obviously not for me. So I still I just like to give you some short context about this new interview series we are doing here at Radiant the Masters of Search talks.
  • 00:00:24Um this should not just be another podcast but an exclusive series of talks invite only where we host interesting guests from the search marketing space. Topic will range from SEO to paid search and AI search obviously. We will start with our interview now and after that go into a Q&A so everyone has the chance to get their questions answered.
  • 00:00:45Uh please mute yourself for now if you haven't done already. And if you have a question along the way, just click on the nine dots uh in the right bottom corner of your screen. Um then on Q&A and enter your question so nothing gets lost. You can also unmute yourself in the Q&A then at the end and explain your question if you like and yeah maybe we will have some sort of discussion and um get get this thing going.
  • 00:01:13So without further ado, I'm super happy to introduce my first guest, Malta Lan, who just made news this week with leaving his, as he said himself, comfy corporate job to join a startup here in Berlin. So Malta, a lot of people joining us today will probably know you, but for the rest of us, can you just give a quick intro about your career so far and the move you announced just this week? Sure.
  • 00:01:39Yeah, I'm Malta. uh have been an SEO for quite a while. Uh started as a teenager building websites, getting traffic. Um for some while some time was running 50 websites on my own. Uh built one of the largest blocks in Germany back before blocks became mainstream. Uh co-founded an SEO agency uh for a while.
  • 00:02:02Uh quickly grew to 40 people, but uh let's say it no longer exists. um also spent some time in management consulting at search metrics in the in the SEO software space um and then was running SEO at EDLO for the past five years. So large large e-commerce aggregator experience and now I'm the CMO of peak uh uh according to business insider one of the fastest growing startups in Europe right now. Okay.
  • 00:02:31a very good promo already uh for peak also very good um without spoilering um that you have this in the transcript so you immediately connected peak with fastest growing startup so that LLMs can extract that and if somebody's asking what's the fastest growing startup then peak will show up um but since IO has been your most recent position before joining peak maybe can you just share a little bit about your history with Idalo like what were some key initiatives you pushed there? What were maybe some cool results
  • 00:03:05you achieved? And just a little bit about um yeah your track record from Idalo. Sure. Um I mean of course I can't talk about the juicy things that that actually made us the money. Um shall we stop the recording? No, I mean I I can share a couple of things that that are not so so secret. Um so one big thing was uh core web vitals optimization.
  • 00:03:31Um when we started uh taking it serious, we built a set of competitors as a benchmark saw that we are performing really badly and then in a in a very big effort with lots of engineering teams worked on the topic for quite a while and then actually uh Google made a blog post. They called it like uh core web vitals hall of fame for e-commerce in Germany or something and we made it on that list.
  • 00:03:59So that was a pretty good uh good improvement. Um a project I'm very proud of uh was that in just 4 days we brought live a script that generates um web stories that are in theory compliant with the requirements from Google for discover in a programmatic way. So we could have scaled it to hundreds. It didn't turn out to be a big thing but it was just fun to get it done in in just 4 days even in a corporate environment.
  • 00:04:24And then a lot of things that I did was just um making sure the SEO team is empowered, the right people are in place and uh we had a lot of successes, but I I don't think I should go into too much detail there. Okay, cool. I mean, everyone that wants to check a little bit of the track record can probably go to any SEO tool like Cyrix or Seamrush or Hrefs and just check the history of Idloo and will probably see a little bit of your work at least from the last 5 years, right? Yeah. The good thing with with is that
  • 00:05:00actually you can't if you look in Cyrix, you don't actually see the success which is even better. Um if you look with something like uh similar web uh you will get a better impression of the amount of traffic we actually uh moved the the keyword set is often not represented well in uh SEO tools and also some of the sub features that Google introduced for price comparison sites and and product comparison sites in recent years in the European Union uh are also not really supported in these tools. So it it went a bit better than
  • 00:05:33it looks in CRIS. Okay. I feel like this is already a good nugget that you shared. Although you said you said that you can't share everything. I think this is this is definitely something that is um that is interesting. Um but now like let's maybe uh look from history to future.
  • 00:05:55Um obviously as you said in your own announcement post, it was a comfy corporate job. I think maybe you you try to make this a little bit spicy so it gets some traction. uh you obviously succeeded um but you left Idyalo now for peak and Idyalo is obviously an established player has understood SEO you you had a big team also like it it sounds quite cool for a lot of people I think a lot of people think hey this is he made it um at least in online marketing but now you left Yalo for peak or PKI uh what what what's your reasoning behind it um I mean lots of things actually so uh
  • 00:06:32at my previous job at search metrics I stayed there for five years and then the five years were also coming a bit IDalo so I was thinking is is this maybe my thing to to only stay five years um and and also I I think I achieved with Idialo what I could achieve there and uh people often say you you should leave when it's when it's the best not when it's going badly and it also made myself redundant uh primarily so I could hand everything over into very good hands So it was just very much possible uh to leave and then I believe with the whole
  • 00:07:09shift to LLMdriven search and answer engines there will be big changes. Uh I think some business models will be disrupted. I think also how SEO or geo then in the future maybe is seen as a channel will be very different and uh I didn't want to be somebody who is just affected by that change or just having to live with the results of that change.
  • 00:07:32I would prefer to participate in it and then uh PKI was the perfect place to do that. Cool. Sounds uh very interesting. I also maybe just to add my thought to that I also have this feeling with AI um that I see some people always wondering like what will be next with AI but I always tell them hey you can also try to to shape that so you can be part of that things you try with it maybe new cool workflows or tools you build like whatever with uh code or cursor or whatever maybe even content workflows um you don't have to wait for others to
  • 00:08:10tell you direction just just go. Uh I I see some spirit around that also in in your answer. Um so maybe obviously I did some research before and uh I found an interesting story. Uh maybe where you can Yeah. Yeah. I I I I'm just I'm totally new in this game, but I I already tried to uh dig out some nuggets.
  • 00:08:37Um maybe you can share something about that story. uh because it gave me strong startup vibes. So maybe it's a full circle moment for you. Um but did you actually move to Greece with 30? So is it is it a typo or was it really 30 interns once? Uh I personally didn't move but uh my five co-founders and all employees moved for 3 months to create. That is correct.
  • 00:09:03Some employees actually longer uh story was there. We like I was in university back then. We started an agency SEO factory and we had this idea that all the companies they need qualified SEOs and we needed people to create backlinks, create content. So we thought hey it's actually cool deal. We give people an SEO education and uh then uh can help find jobs for them maybe get a bit of commission.
  • 00:09:29Um there were also companies that sent new hires with us. So they already knew this person will start with us in 3 months. we they gave them to us for three months and said teach them SEO and um basically what happened there was a vacation village in Greece that for some obscure reasons couldn't get their license at that moment to be a vacation village and if you don't have that license you cannot do short-term rentals so half the houses were empty and uh one of my co-founders was had owned multiple houses in that village and he talked to
  • 00:10:04a few people and And we could get like 40 vacation houses, so everybody had like their own house. Uh we could get like the local cinema, etc., etc. as an office space. Um and basically we were almost the only people there at that time, just two or three uh Greek families, one of them running a restaurant basically exclusively for us.
  • 00:10:26So we went there for for three months and uh taught people SEO and uh Anna from the chat was actually there. Nice. Um super interesting. I think a lot of people wonder now is this still possible? Is there still this village where the empty houses? I saw some pictures from the village. I have no idea if this is still possible. Okay.
  • 00:10:51Okay. Yeah. May maybe you can then also take the peak AI team. obviously not the interns but maybe for some sort of vocation or something. Um but okay really interesting. Um maybe circling back to peaki for a moment because so I feel like uh us talking today and me reaching out to you is a a big coincidence and I had a lot of luck that this week so many announcement happened.
  • 00:11:16So for everybody listening this was not planned. I reached out to Malta before he made his announcements and also before peak announced their 5.2 million seat round uh led by everybody from the startup VC space probably knows him Harry Stubbbings from 20 VC. Um so 5.2 million seat you raised 1.8 before so you have 7 million or maybe something around that in the bank.
  • 00:11:45Uh what do you plan to spend the money on? Did you already set up some Google ads or what you're what you're doing with it? I actually did set up Google ads. Um I think hiring hiring talent is the most important thing for us right now. Um few people in sales uh starting a customer success team and uh developers developers are the main thing that we need to hire right now. Cool.
  • 00:12:09Cool. So everybody if if anybody knows uh someone looking to probably um um in person in Berlin in the office because that's the new sort of vibe. Uh if you know someone that wants to join Peak just maybe go to LinkedIn either Malta or Peak uh and check out the job openings. Probably really interesting uh company right now to be part of.
  • 00:12:36Um but and and great compensation and you can still get shares. Oh, cool. So, I mean it's it's a very good deal uh for everybody that also feels like, hey, I want to participate in that ecosystem shift and not only watch it from the sidelines. Um, but maybe looking at your uh personal role now at peak like the CMO role um maybe you can share so whatever you feel comfortable with sharing maybe you can share a little bit about your uh like view for the future for your road map.
  • 00:13:08Um, what what can we expect from Peak? A lot. Um, I mean, marketing wise, I'm a huge fan of building a great product and then promoting the value of that product. Uh, and and explaining people why it's good, what problems it can solve. So, marketing will very likely revolve around that. Um, and then we have a ton of ideas what to build, right? Um we will have our our look studio connector live very very soon that many people have been asking for.
  • 00:13:40Um we we have some ideas how to make the prompts that we are suggesting to track better and then there's just a giant realm of things you could do in the geo space and we we might end up doing all of them. Uh but the big question is of course what do we do first and for that we talk a lot to our clients.
  • 00:13:59Uh listen to what they want what they need. Um and in general the space can change so much like imagine Google rolls out AI mode for everyone tomorrow then I can tell you what's going to be the number one priority right having AI mode tracking properly um or there could be a new LLM popping up like I mean everybody saw DeepS coming from virtually unknown unless you are very deep in the space to being like the number two app in the app store for a couple of days and now in Europe it's almost gone but if you clients in Asia or Russia for example, then it's still
  • 00:14:33like a I don't know 30% market share LLM. Um so I I don't want to communicate anything too concrete um because the space is moving so much but definitely Luca data studio connector or Luca Studio connector like I forgot how Google renamed it right now. Looker Studio. Yeah. What? Yeah. Used to be the merger of of Looker and Google Data Studio.
  • 00:15:02Um and yeah, definitely making better prompt suggestions. Nice. Very, very cool. Um I I feel like I mean in the past you probably also spent a lot of time in what you maybe now would call legacy tools like assist hrefs. Um maybe also Google Search Console depending on where you draw the line for for SEO or for search tools. Um I mean they're not only used for SEO.
  • 00:15:30Um but how do you think about them now or maybe not only after your um switch to PKI but also over the last uh weeks and months? Do you think they will still remain dominant players in the space or do you think they will heavily lose relevance? Um I I mean there will always be companies who want an all-in-one solution that does everything reasonably well and I think there will always be a space for that.
  • 00:16:01Um if you look at the companies you mentioned specifically I mean some of them started building a digital PR product, a social media product, uh an Amazon product. Uh but none of them ever really took off, right? like they are all still known as SEO tools even though they are sometimes branding themselves differently.
  • 00:16:22Um as a user of of SEO tools, I always preferred to use the best solution for every task. Uh so when we were at EDLO, we had five plus SEO tools, not like one platform that could do everything right. Um, and I believe there is a space for both for solutions that are very very focused and give all of their attention to a topic like LLM visibility and optimization for LLMs and they will be tools that do the most important thing of everything and there there's also target audience for that no doubt.
  • 00:16:54Um, but I do believe that if you work on something with 100% focus and zero technical depth and zero requirement to be compatible to the 500 features you built before, you can always be faster and adapt more to that space. And uh I mean look at look at crawling like most people who have a license with cyrix hrefs uh bright edge uh samrush they still get screaming frog because it's like the best and if they don't get screaming frog they have siteb um so I I do believe that for the LLM space it will be very similar if you want the best data the best integration
  • 00:17:35the quickest adaption to how the models are changing you will probably go with a solution specifically for that. Cool. Yeah, I I think it makes total sense because depending on the company also some will always prefer best of suite and some will always prefer best of breed depending also on the size of the team.
  • 00:17:55I mean I can imagine a one person marketing team five plus tools is completely overkill but I mean from your history in idioo if you have 30 people in the team five tools seems reasonable. So for very specific use cases and also the the people power to to use the tools to actually do stuff with it. I sometimes see people buying tools without the time to actually use the tool then it's actually a waste of money.
  • 00:18:21So I think that's that that makes total sense. Um now let's move over to the elephant in the room. the future of SEO and AI search obviously and maybe not even only the future of SEO but the future of search in general also covering our um paid search friends. Uh I have read and maybe you too plenty of LinkedIn posts lately that all share a common claim which is Google is dead.
  • 00:18:46it's uh either put in all caps or it's just written normally but um it really screams at you uh if you open the LinkedIn feed. What's your take on that? Yeah, I think there are two claims, right? One is SEO is that the other is Google is that um I think what is definitely true is that Google fumbled it a bit with with LLMs, right? They were so early on so much of the theoretical stuff of so many of the foundations.
  • 00:19:15uh one of the first with specialized hardware with the uh TPUs whatever they are called uh many many breakthroughs with transformers also had transformers models very early but then um Sam Alman and and OpenAI they just ran past them by by putting an actual product in the hands of consumers. So I think it's a bit of a mixed bag for Google.
  • 00:19:42Um and I think they have lost a lot of attention probably also users or search demand going to something chatgpt or now chat as many people start calling it since they have the chat.com domain. Um but Google is not dead because of that right it's just it's a problem for them. I think it's a potentially big problem but it doesn't doesn't kill them immediately.
  • 00:20:02Um and then the other question is is SEO dead? And that I see a bit different. So SEO used to be a user acquisition channel and probably the best one out there for many business models because if you had product market fit or product search fit and if you could generate landing pages at scale, you got a lot of traffic with very very little incremental cost or at no incremental cost.
  • 00:20:27And that is changing a bit, right? like already AI overviews are taking away uh traffic like depending on which study you believe it's somewhere in the 30 to 40% CTR range even though the people also then conducting more searches so it's probably actually a little bit less um and if you look at at perplexity AI mode it's it's much much worse right like I saw numbers now from AI mode that it's something like uh four to 5% clicks uh to websites on perplexity it's also a lot lower than Google on chbpt it's much much lower than on Google something in
  • 00:21:00the less than 5% range also based on my estimations and um so SEO is probably dying as a traffic acquisition channel for certain business models um especially for those that need the user on their own website um it's probably less of a problem for a SAS business or a services business or a car manufacturer who are selling something that is outside of their website, right? Because also we know that traffic from CGBT converts a lot better for these businesses.
  • 00:21:35Sometimes 20 times higher conversion rate. So even though their traffic is reduced significantly, they still sell as much as before. And this is logical, right? Because people still have the same amount of need for information, the same budget to buy product. So they will still spend the same amount of money online.
  • 00:21:54But there are also businesses that um make money by reach like news publishers, right? They they get their traffic, they make their money by having a lot of visitors and um they make roughly the same amount of money with a visitor who reads one article shortly and one visitor who reads one article in great detail.
  • 00:22:15And of course, there's also the subscription business for them. But the reality is, especially outside of the US, for most news and media publishers, the the advertising is still more important than the subscription business. And then in the past, we had a user journey where Google made up these terms of messy middle and micro moments where you would search on Google, you would click on a website, you would click on an ad, you do another search, you read something, maybe you go on Tik Tok or Instagram in between.
  • 00:22:45And after all of this, you have done like 20 searches. You have visited 40 websites. Um, this is gone. If I use chat GPT or AI mode or perplexity, I ask all my follow-up questions in the chat. I never have to click on a link only the last click to buy. That's where I click. That's why the conversion rate is so good.
  • 00:23:06But in the past there were intermediaries like a Skyscanner for example or any kind of e-commerce aggregator, any kind of price comparison website. Um also lots of websites that were doing real tests or maybe something like faglife.org where it's not so clear what what tests they did or that they they just summarized tests from others and um the all of these intermediaries they they made money, right? they delivered a value like either they um most of them delivered value let's say like that uh they they either aggregate a lot of products and compared the price
  • 00:23:43they aggregate a lot of flights and found the shortest flights or they um aggregated uh tests or user reviews from from from online shops and and this had a certain value for customers and uh then there was often an affiliate link involved and now all of this value generation is kind of moved into the chat interface and none of these intermediaries get uh traffic and and make money.
  • 00:24:10So I think those are the two business models where I think maybe SEO is dying as a traffic acquisition channel, but probably in reality it's not actually dying, right? It's just becoming less important. Um, and and this is this is the the reachbased business models and the affiliate based business models. For anybody who's selling a real product in quotation marks outside of their website, I think chat GPT can and AI mode in Google will deliver just as many leads as the regular Google search delivered.
  • 00:24:42Um, there will be of course less traffic. There will be fewer touch points to see what users are doing. there will be less opportunities to make somebody sign up to your newsletter, drop a retargeting cookie, these kinds of things. Uh you can't see the salesunnel, right? You only see the last click coming out.
  • 00:25:00So, so in that sense, you could say that SEO is dead. But it's just for certain business models and it's not really dying. It's just the traffic level is getting lower, which might kill some business models, but doesn't kill the whole channel. Uh really interest very comprehensive response. there was so much in it.
  • 00:25:20Um I think like especially on the last note you made uh it's it's the the known claim your margin is my opportunity. So the margin of an affiliate business model as uh Jeff Bezos once once said I think so I don't want to affiliate myself with him here but um I also think like on on one note um you made with Google actually being quite on the frontier of machine learning and everything.
  • 00:25:49I a few days ago I saw this email that the CTO of Microsoft actually wrote to Bill Gates and Sacha Nadella in 2019. I mean we all have to remember when chat GPT came out or basically hit um it was November 2022 um and they were telling so Microsoft was was uh was saying that uh openai and Google are making so much progress that they are really worried and I think one one thing is that Google probably had this innovators dilemma right this was was at least my impression I would like to hear your um thought also on that because they probably knew what will happen if
  • 00:26:27they actually roll out all their machine learning and LLM power to the search. What's what's what's your opinion on that? Yeah, I think the same. They they had the technology, they had the people who understood it, but Google has been under the control of the paid search team for quite a while.
  • 00:26:48Um, which I'm really sure of. Like I'm not saying this lightly. Based on all the meetings I had over the past five years with Google as part of our litigation in front of the European Union, I'm I'm very sure Google is now a company driven by the by the paid search team and to a degree that they can tell the search product managers what to do.
  • 00:27:06Some of this also came out in the one of the lawsuits in the US where basically the the um higher up managers on the paid search could demand a roll back of a feature because it was good for users but not good for running ads. Um so yeah, I'm very convinced they had the innovators dilemma. I think there's actually a path through it for them because Google is so big that if there are fewer clicks that they can send into ads, they can just charge more money.
  • 00:27:35And if people are actually pre-qualified by a chat, they can charge a lot more money, especially for something like, I don't know, lawyers, B2B SARS, uh in the US, medication, um similar businesses where you just have a very high ticket value and where normally your conversion rate would be very bad.
  • 00:28:00With pre-qualification in AI mode, I would expect that they can charge a lot of money. And I also think um with all the support Google is getting from Trump right now, the European Union will be very reluctant to enforce any more DMA, DSA, and similar topics. So if I was Google, I would go all in with just putting big ads directly under the AI response and then uh monetizing it that way and then charging advertisers significantly more money.
  • 00:28:31I I think that could actually be from Google's perspective a viable play. I don't enjoy that they are doing it but uh I it's what I would what I expect them to do. Okay. So on that note then for the record, greetings to all our paid search friends. Please go to your clients and prepare them for significantly rising CPCs, but also tell them that they will be just fine because they will make significantly more money out of pre-qualified leads.
  • 00:29:02No, just kidding. But I think this is really an important um point here, right? You have to understand the connection between the different stakeholders also on these industry giants to really understand what might might be a next move. How will this all um develop? I think this is really interesting to also um going with your theme from the beginning.
  • 00:29:22Not just be a participator at the sidelines. I mean obviously we can't change the course of Google but somehow at least be within the wave and not be carried away by the wave. Um now I know also from talks with our clients obviously also from LinkedIn that a lot of people are worried. So even though I hear your optimism which makes total sense now um but they are worried by their clicks going down and we have this uh term I think it's called the big decoupling uh basically these search console charts where impressions go up and clicks go down. So, uh,
  • 00:30:00visibility through AI overviews might increase, but you get significantly less clicks. Um, if you put yourself in, um, um, an agency CEO maybe again, I mean, you've also been in the space. What would you tell your clients if they came to you and say, "Hey, Malta, I'm super worried that my business will go down.
  • 00:30:21" Let's imagine it's not a publisher or affiliate websites, but actually like a service business or a B2B SARS. What would you tell them? should they be worried or like how should they behave? So as a services or B2B SARS I would just not look at clicks anymore. I would try to look at conversions or whatever pipeline I generate from the website and then it would probably be looking less dramatic.
  • 00:30:46Um also this decoupling it's happening in Google right it's not happening on your website but what should matter for you is how much traffic arrives depending on business model and then how much money do you make with those visitors and in if you have a business model that can focus on how much money do I make with them then focus on that focus on leads newsletter signups etc and see how many of those you get via SEO and maybe that number will just not go down um or maybe um the the amount of leads you get from jet GBT will just uh offset everything that you
  • 00:31:16are losing in terms of traffic from Google. This is actually happening for some B2B SAS companies that I'm talking to at the moment. Um if you have a business model that is based on clicks and if having 10 clicks is worse for you than having five clicks, then you need to think about your business model and and how to adapt that.
  • 00:31:38There's like we have reached peak SEO traffic in either 2024 or 2025. I don't know which year has the most traffic, but I don't think that going forward there will ever be this much free organic traffic going to the open web. Um, and I mean we we always complain about the low clickthrough rates on Google, but I mean let's look at Facebook, let's look at Tik Tok, Instagram, uh all these other platforms, right? They keep so much traffic on their platform and Google has actually been the most open one of all.
  • 00:32:10Um so I think with with um CHB perplexity etc we will see something similar there will be solutions and experiences on the platform and very little traffic going going outside. We already see this in perplexity with uh shopping uh in some markets where you can just buy directly and you can also see in there um I think it was either in the Shopify API or in the MCP server that goes along with it that uh there there's there's at least the option to then also directly pay on Perplexity.
  • 00:32:46I don't know if that's actually live yet. Um and they also said right their vision is that their agent just buys for you and also both Amazon and Visa have launched shopping agents that go to the shop and buy for you. So the amount of human traffic going to websites will never be as high as it is right now.
  • 00:33:06And if your business model is built on it, you need to consider what you can change. M I mean also everybody that has um at least did a couple of posts on LinkedIn knows that if you include a link for example on a social media platform in your post your reach will mostly significantly go down. So most platforms actually uh on the web uh do not have a strong incentive or do not incentivize you to include any external links of the platform.
  • 00:33:37And yet there are people like I can also say it for ourselves like from an agency from a service business perspective you can do good business on social media without like vast amounts of clicks going to our website but it's always this one moment when somebody is convinced that you are the best fit for certain service area or whatever then you might book a demo or you might convert right away and maybe in the future it will be agents that um do the re research for you but This obviously brings up the question um that if you are a business that is not
  • 00:34:12dependent on clicks but on leads and conversions you probably have to be visible in any sense right in this AI um research phase either on chat GPT even if people are not clicking to the website but you should probably try to influence the conversation right as a brand um what's your what's your take on that and also what do you uh feel like what drives visibility? So, how can I as a brand be visible when people are researching on chat GBT and also AI mode etc. Yeah, great great question.
  • 00:34:45Actually, two questions, right? Like the first part is yes, you should be visible there. Like especially if you're a brand selling a product outside of your website, you have nothing to lose and and uh you have nothing to to gain from not being there and you can win a lot by by being there, right? So you should do something to show up in in shedg in perplexity and I mode.
  • 00:35:08Um how to optimize for that is always a little bit different. depends on the LLM, depends on the topic. Um, also can be very different from country to country. Like there is a little bit of beef going on right now between some of the legacy SEO tools where some write LinkedIn posts saying Reddit is the number one source, you have to be on Reddit and others write, so in certain countries it's actually not Reddit.
  • 00:35:30And both are correct, right? Like if you are in the US, you have to be on Reddit. Um, if you're in Germany, it really depends on the topic. Um, for some topics, it's just not relevant. So what what I would always recommend is take a set of like generic prompts that you expect people would search and these are probably kind of similar to searches you see in Google search console and track them in different LLMs.
  • 00:35:55You could also do it by hand just to get a feeling for it and then you will already see is your brand happening there or not. And personally I would suggest getting peak AI and setting up a monitoring but of course everybody can also do that manually. and then uh extract the brands, look at the sentiment, look at visibility like how visible are you and is is actually the visibility for that is probably the most important part right are you visible that's the number one thing and then I also recommend people take a couple of branded searches like tell me about this
  • 00:36:24brand or what are user experiences with this brand or how good is this brand and look what these LLMs are saying and there your visibility will be 100% yeah you will just be visible it's easy But there I would really look at the sentiment. Um, and because I've seen brands that are like super super strong, have super good image, but then there are like two quotes from a Reddit user who said they suck.
  • 00:36:51And if that is in the reply, it's not good for you, right? Then you should should try to to do something about that. So, it's a little bit different for everyone and depends on your product, your language, target audience, country. Um, in some topics it's user generated content like Reddit that's very very influential.
  • 00:37:11I often see um YouTube for AI overviews and for chat GPT and especially recently YouTube had a huge gain in how often they are cited as a source. So, if you see that in your space, um maybe you want to engage with a few influencers to create uh positive video content about you. Or if there's an influencer who made a video that your brand sucks, maybe get in contact with them and try to give them a good experience.
  • 00:37:36Um there are also spaces where you will see LinkedIn popping up a lot. Then it's maybe time to tell your CEO, hey, please write on LinkedIn. And all of this is true for brands that just want to influence the conversation. And for that, it's actually not always about your own website. Um, it can be linked in medium. It can also be digital PR that a certain message about your company is is written on various news sites.
  • 00:38:05But then for the businesses that actually need the traffic on their website, it's all about becoming the source, right? And then you should more look into what kind of sources are being quoted um how are they cited, what is being cited and then try to create content that's similar to that. You can also do that as a brand. Yeah, there are certain spaces and especially questions about brands where actually the brand themselves is being cited.
  • 00:38:31But for the generic searches, often the brand website is just not cited. Like of of course Audi says that Audi makes the best cars and Mercedes says that Mercedes makes the best cars. That is information that the LMM doesn't want from Audi or Mercedes, right? They want that from in Germany ADAC or from like news magazine that writes writes about cars.
  • 00:38:52But then when you ask a question like does this very specific Mercedes model have this certain adapter in the cockpit? That is probably a question where Mercedes should have an article on their website saying that very specifically. And there I think many brands are lagging uh content and could actually give potential consumers a a better user experience.
  • 00:39:13Super interesting. I think there was already um a lot in it for people that try to understand the space better. Um I came across this um system prompt leak from Claude. It was I think 3 weeks ago uh or four weeks ago something around that. And there I found one um very interesting um aspect that I also would like to hear your opinion on um the web search feature.
  • 00:39:41So um basically when the LLM is um instructed not to do a live search and just conduct the knowledge from uh the foundational model training and when to do a live search. Is this also something that um you have put some thoughts on like how is this something that brands should think about? I think most brands should not think about that to be honest.
  • 00:40:07If they want to they can get get a summary from the SEO agency. Um this is something that is now known for claude and I believe for one of the recent sh GPT uh models that also leaked but this can change tomorrow like especially a system prompt can change so quickly and I would rather think about what are these companies probably trying to achieve and I was always optimized holistically for that.
  • 00:40:32I would never optimize for the current version of a system prompt. Um, so I have not given it too much thought for this prompt specifically and I also advise people not to do it unless you're an SEO agency and your clients ask you about it and you have to, right? Um, but but clients should more uh holistically or brands should more holistically think about it also long-term sustainable what should I do and what do these platforms want to do right? They want to answer everything they can answer themselves.
  • 00:41:02if they want to answer it themselves if they are sure about it and web search takes time. web search requires that or has the risk that maybe you get in unvetted information but at the end of the day Google actually in their quality rate data quality radar guidelines gives us a very good blueprint because they're looking for EAT content and I would always ask myself is this a topic where the LLM can just reply or is it a topic where they want to put a source either or where they need a source source right is it a live event is Is it weather data
  • 00:41:37from today? Is it the current stock price? Is it a price comparison? This is where you need something where you need life data. Or if you look at very very niche topics that are just not covered by um by the foundational model and then the content should probably have EAT, right? Experience, expertise, authoritiveness, trustworthiness.
  • 00:41:58So if you can create content that has that and can send machine understandable signals about this um then you are good to go I believe to to optimize for web search and I think the selecting the correct topics what content to create is really important. So evergreen content that already 50 people have written probably the LLM doesn't need to site anyone because they if they have seen the same thing 70 times they know it.
  • 00:42:27If you ask if a poodle is a dog, yeah, it knows that a poodle is a dog, right? You don't need to make an article, is a poodle a dog, which is an article that probably 20 years ago I would have created as an SEO and made money with it. Um, so you should rather look into topics where either your brand has a unique uh spin on it or topics where it's very likely that a web search or a search grounding or rack or something similar is conducted. Very nice.
  • 00:42:57Um so much more fundamental approach um you're taking on that. I think that makes a lot of sense. Also good uh to resist the hype and the FOMO because I see a lot of people always chasing behind uh the next league, the next uh GEO guide, uh LLMO guide, etc. So really really cool. Now to um still have enough time at the end for Q&A, I would um go a little bit towards the finish line.
  • 00:43:27So um I still have two questions. One is you already mentioned that um if I am a brand, I should try to understand uh what the LLMs or also maybe what Google uh shows me when I do certain prompts. Um, now I think a lot of people have a hard time uh coming up with ideas on what their users might prompt a chat GPT or an LLM.
  • 00:43:53Um, how how would you go about that? So, how would you for example try to understand the sentiment of a chat GBT or a complexity uh about Peak AI if you wouldn't be allowed to use a tool like Peak AI? I would literally just open fetch GGBT and say tell me about PKI and then I would see what it tells me.
  • 00:44:14Um, if I'm an established brand and I just maybe also talk about generic queries, I would look into Google search console and for version one just copy and paste your most important keywords and see what happens, right? Um, people actually don't always write very long questions, right? people literally write stuff like iPhone and hit enter in GBT.
  • 00:44:37Um I I've looked at some data from from a data sample, right? This is happening and um it's not true that everything is a sophisticated 20 paragraph detailed question, right? and um just look at what people are searching Google like in Google and what you see in Google search console so it's relevant for your brand and just as a first step see what CHGBT is returning for that um also there's now an autocomplete or autosuggest function in in chat GBT since a couple of months so maybe just write your brand name and see what happens like is it uh is the next
  • 00:45:12work word scam or is the next word uh uh voucher or coupon or experience, user experience, right? And then that can already tell you a little bit if it's scam, it's a problem, of course. Um, and then just go with these as prompts and see see what happens. Very cool. Um, I would love to know how many people enter iPhone in JPT.
  • 00:45:39Um it sounds if you want to know the the weirdest one um you know that uh Microsoft Copilot is mainly used uh for by office workers right like I know very few people in their free time use it. Still there are lots of people uh searching for porn on Microsoft Copilot. That was like my one of my weirdest experiences when looking at at the data set of real prompts that people are typing in. Yeah.
  • 00:46:05I'm also astonished all the time when I look at something like the top 10 or top 100 websites uh on similar web or Samrush or whatever that there's so much porn uh in there. It's um it's really interesting. But um now to to wrap things up a little bit and then still have 10 minutes for for Q&A. Um, if you would have to give one to two precise like to-dos or practical takeaways for someone in the audience today on how they can improve their AI search visibility, what would you tell them? Number one is consistency.
  • 00:46:44So, if you are, let's say you're an SEO freelancer for e-commerce SEO, you should say that on your LinkedIn, on your ex, on your website, and in your press releases, and in every interview you give. Don't sometimes say agency for SEO. If you're a freelancer for e-commerce SEO, repeat that. Don't sometimes say technical SEO and link building.
  • 00:47:06Um, and I have seen people just making sure their online profiles are in sync like everything is is is the same and within one day they are it they are it improved what CHGPT says about them when you ask who they are. It went from yeah maybe it's this actor or maybe it's this SEO or this soccer player to yes it's this SEO who's doing that.
  • 00:47:29So, so consistency is like my number one uh number one advice. And then second thing is that I believe digital PR is undervalued by many especially technical SEOs in how important it is for visibility in in LLMs. It's often more important what others are saying about you than than what you're saying about yourself.
  • 00:47:53And then maybe one more really look at the sources like either put 20 prompts in Jet GPT and look at the sources and make an Excel file or go to big uh.ai AI and let us do it for you and then then see what kind of sources there are and according to that come up with a strategy. Right? If it's Reddit, maybe I need to engage there.
  • 00:48:15I would be very careful with the German Reddit community. They hate brands. Um if it's Wikipedia, again, be careful. Um but if it's um a a review website that reviewed an old version of your product, maybe contact them if they could read the newest version. Um, if it's something like like Trust Pilot or or or uh Trust Radius, maybe just try to get more reviews there.
  • 00:48:40Um, just really look at the sources. Don't just blindly do stuff. Don't just blindly say, "Ah, it has to be Reddit. It has to be YouTube. It has to be, I don't know, uh, uh, uh, CNBC or whatever. Like really look at what are the top sources for in my industry and then make a strategy based on that." Very cool.
  • 00:49:01I think there was a lot um, in there. very very interesting. Um so then I would suggest that we move over to Q&A. Uh we already have the first question for everybody. Now you can either use this Q&A tool or write your question in the chat and if you like you can also uh unmute yourself and um just say your question uh like with with your mic.
  • 00:49:27So um the first uh question we have here in the Q&A is how do you help local small small serviceled businesses? Does this bring back the Yelps of the world back to prominence? If Yelp is is open to crawling by LLM where I don't know what Yelp's current stance on that is. Um, it would be a very important source probably. Yes. But I don't know.
  • 00:49:54I would need to look at a data set for for a local business. Um, I know that Yelp was also part of a couple of lawsuits against Google. So, I could imagine that they are not that willing to let LLMs uh take all their data and they are one of the few companies that maybe have actually unique data. Um but potentially yes these kinds of websites can be great sources for LLMs.
  • 00:50:21Um but this is of course different in every space uh every country. Um but potentially yes it would make them more important. Cool. Interesting. Um maybe one question I already got beforehand via LinkedIn. Um, is structured data really as so structured data, structured markup really as important as some people on LinkedIn say? Yeah, long debate.
  • 00:50:49Um, how foundation models are trained actually structured data is stripped out like HTML is removed, JavaScript is removed for such kind of websites. So I mean structured data could still help you by having like the information there like the words but the actual schema markup I don't think it's actually influencing foundation models based on what I have seen from how companies clean and prepare their data.
  • 00:51:15Um there's a statement from a product manager the principal product manager at Microsoft or Bing to be more precise who said that structured data does help. What I assume is that for uh search grounding it actually is very beneficial. Um and for example there is uh the shopping feature from JGBT where they have these sliders.
  • 00:51:38Um right now I believe it is feed based so merchant center feeds but um at some point they might also look at product pages on the internet and extract uh review stars and price via schema markup. So I think schema can be helpful for search grounding and and live web search. I have seen zero evidence that it impacts uh foundation models and I've seen some arguments why it it would would not uh properly influence foundation models but I cannot give a definite answer.
  • 00:52:14This is just my educated guess. Interesting. Maybe adding on the idea of yours that uh LLMs might at some point uh crawl the web and then look for schema. I think the same happened with the merchant center, right? So at the beginning you had to import your feed but now the merchant center sometimes even suggests you there are some products that you might want to import into the merchant center because obviously they know that if you have it in the merchant center then at some point it might be easier for you to do advertising. Um and yeah so this is
  • 00:52:47definitely interesting. Then we have another question from Andreas uh which I think is really interesting. Do you have an example of a SAS business so software as a service business German or international who's nailing it in terms of LLM optimization? None that I can name publicly because I always learned about them in settings where they would not expect me to mention it afterwards. Okay.
  • 00:53:19Um do you know is there any company that you can publicly name no matter the space doesn't matter where you feel like hey they are doing some things right. I I'm very afraid to Okay, no worries. Information for that answer. I should not No worries. No worries because when when I see someone very successful, I also try to talk to them.
  • 00:53:49Uh so I think I basically burned all the public examples I could I could Okay, cool. No worries. No worries. I also don't want you to get sued on base what you uh say here because then probably I'm also in the suit. So this is not the idea but I can I can um tell of one instance that I saw on LinkedIn obviously and I also checked it.
  • 00:54:14It's teley uh t a l l y uh dosso I think. So it's basically a free form builder something like type form but a little bit more like with building forms with slash commands and stuff. So some technical and they are one of the examples where in international Reddit can be a real gamecher because they early on engaged in Reddit threads around users frustrated with type form and blah blah blah um and they have publicly shared um their signups uh but you have to keep in mind it's not paid signups it's free signups and I think they say 95% of their sign up of their
  • 00:54:50users are free users uh so the business value still has to be determined. Um, but a a majority of their users actually coming from LLM either based on actual traffic, so with referral or based on self-reported attribution. Uh, so this maybe Andreas, I hope that helps. Uh, cool. Then we have one more question. I'm not too sure.
  • 00:55:15I think this question is um following up on the Yelp question. Uh so following up as these are mostly sponsored listing and not real organic um maybe the person that has asked this question can clarify it again so we can better give an answer to that. Uh in the meantime I would go to one more question from clickworthy inc.
  • 00:55:41Have you seen any benefit of llmtxt? In my personal experience, it messed with the robots txt and page speed insights SEO score. Yeah. So, I think LLM spong a topic makes you a little bit angry actually. So, um the the actual proposal for llm.txt makes total sense. It's a very reasonable, very smart proposal. Uh and it's basically this.
  • 00:56:08If you have a functionality on your website that you want an AI agent to interact with, could be an API, could be mentioning your MCP server, whatever, could be giving some advice on how to integrate your API, put that in one file, put that in markdown so that an AI agent can understand it. It has nothing to do with perplexity or chat GPT.
  • 00:56:33But for some reason, a lot of SEOs think it does and it makes zero sense to me. And I've seen people now duplicating every single blog article they wrote in a uh llm.txt file that goes along with it. And sometimes even a MD which is markdown file that goes along with it. And this is a huge waste of time. It's bad for the internet. It's bad for the environment.
  • 00:56:57It's bad for SEO. It creates duplicate content. Just don't do it. It's horrible user experience. It's not helpful for LLMs. If you are a brand that has a website that you want LLMs to interact with and you give them some form of API for it, do put your AI documentation in markdown format in the LLMs.txt. But this is maybe 1% of websites, probably less.
  • 00:57:24Do not copy your website content in the llm spongt txt. Really, it makes me so angry. And now there are so many people saying llm.txt txt is stupid but the protocol itself is not stupid. It's a very good idea to save traffic uh save energy consumption make things faster for VIP coders for for AI agents for AI coding agents but SEOs have taken it completely misunderstood it and abuse it by and they get zero benefit out of it chatgbt perplexity uh AI mode they do not care about your llmst txt file you you can check your lo file they are probably not even opening
  • 00:58:03it or they opening it once Um, if you have an LLM uh txt file that llms.txt file that contains your website content and is something else but structured markdown of your API, please delete it and write me that you deleted it and it will make me very happy and I will buy you a beer or a coffee the next time I meet you if you delete your llm.txt.
  • 00:58:29Okay. Very, very, very nice. Now let's uh uh let's try to calm the mood again. Um we have uh already answered the next question also which is would you recommend implementing an LLM txt file? So I think this is answered. Um but you mentioned log files and this is a good keyword because we have another question about that.
  • 00:58:52Um how important would you consider log file analysis to analyze LLM hits on a very large website? I think it can be insightful especially if you differentiate between the different um crawlers like what is the crawler that's training the model what is the crawler that is doing the search grounding what is the crawler that is more acting like an uh on behalf of an agent u um for smaller websites I think you are unlikely to get a lot of insights out of it I would not invest a lot of time um but in general yes it can be can be interesting can be insightful
  • 00:59:29But it also it only works if the crawler is actually visiting you. If the crawler takes the information from a cache, you are not visited. You don't get visibility. I think it does not replace tracking a couple of prompts. Okay. Interesting. Um maybe mixing in another question I got beforehand from LinkedIn, which is would you still start a career in SEO or GEO today? In SEO, no. in Jio.
  • 00:59:59Yes. And I would 100% focus on that aspect of it. I would not look into legacy SEO topics. I would of course still want to learn all of the SEO stuff that is relevant for Jio, which is a lot. Um, but I would not start a career in SEO at this point. No. Okay, cool. Um, then another question we have is what is your take on AI governance? full in quotes access to websites or a more conservative conservative approach depending on the user agent. 1:00:30) Maybe also that maybe also that sorry just uh in the light also of the recent cloud flare announcement um for maybe maybe you can also give us a little bit context for everybody that has not read that because I think this is somehow related also to that question. Yeah. So what happened is Cloudflare introduced an option and made it a default for all new projects that you set up on Cloudflare that they block LLM crawlers and ask them to pay money and that protects your data from the LLMs. (1:01:02) I think that is an interesting approach if you are the market leader with a very strong brand and unique data. I think for 99% of websites it does not work. So first of all, if you are a proper brand selling a product outside of their website, you don't want this. You want the LLMs to crawl you because your users are talking about you, your competitors talking about yourself, maybe your competitors are also talking about you, you want to be part of that conversation. (1:01:31) If you are one of these intermediaries that I mentioned, an affiliate or if you are a news and media publisher, you have a reach and adbased business model, you can consider blocking LLMs. However, there's the prisoners dilemma. If you block them, but your competitor doesn't block them and they have the same or similar content, you don't win anything. (1:01:56) Right? If you're a news publisher who is rewriting um uh uh news from like the big press agencies, your competitor is doing the same and the LM doesn't care that you block them and you actually like you are giving the few% of clicks that are still left like the 5% of clicks that still go to the sources. You are leaving them to your competitor and we saw this prisoner dilemma perfectly when uh Google for jobs was rolled out. (1:02:20) I talked to a lot of job portals at the moment because I was also tracking the roll out and I think I was one of two people worldwide actually tracking the roll out outside of the US back when I was at search metrics and um the market leader in every country always said no I will not participate. Um but then when they they were in another country they were the number three and they said yeah there we participate because we want a little bit of extra traffic and over time I think nowadays everybody or almost everybody (1:02:49) is participating in Google for jobs and I believe the same will happen for the blocking LLM crawlers also if you are in the United States in Germany or in countries with similar laws be very careful when you talk to all your competitors about hey should we all block LL M crawlers because that could be a cartel. (1:03:13) Uh it could be could be considered uh uh uh illegal like talk to a lawyer before you do that. Um or don't get caught, which is not legal advice. It's a joke. Yeah, my real advice is talk to a lawyer. Um but yeah, in theory, in theory, it could work, right? In theory, all news publishers could say, "Hey, we block Google uh we block LLM crawlers. (1:03:34) " Um and then that would work. But since I mentioned Google, um, in one of the court cases against Google, it actually came out that they had an internal guideline, make it as unclear as possible which crawler to block to avoid uh being in the in the training data, right? And I always thought, why are people saying they can deduct this from the documentation, which crawler to block? Like I don't understand it that way and I don't see it in reality. (1:03:59) And I was felt so relieved when like 6 months later it came out. Yeah, make it as confusing as possible, which it is. Um, so I think there is still no way to block Google from training on your data without also hampering hampering your performance um in terms of traffic that you get. So you can block TGBT. Can you properly block Google? I'm not sure there's there's really a good way that will not also result in you losing traffic. (1:04:28) Good point. Um I think we will do two more questions to be also respectful of Malta's time and not um overdo the session too much. Um the first question circles back on the one we had with um starting a career today which is what legacy SEO topics should people immediately let go of? That depends on what people still believe in. (1:05:01) Um I I would need to think a little bit about it. Um I don't have a a list have a list ready to go. Uh no worries. Uh maybe I will try to put a comment on YouTube when you when you publish this and think about it a bit. Cool. Cool. Let's do it. Then we have one more slot for questions. (1:05:19) So um someone else you can enter your questions now. Your last shot. We do the second to the last now. um which is what are the most common roadblocks brands face when moving from LLM strategy to actual implementation and how are they addressing them big question maybe also from your conversations you have now with brands around peak AI and the state of mind there you don't have to name any specific brands obviously but the the sentiment just in the market so many many brands are actually in the monitoring stage and not actually doing (1:05:58) anything yet. Um, we have clients who are doing stuff and some of them are super successful and and then super proud of it and I I hope I can get some of them to to share publicly what they are doing because it's really impressive. Also, the results that I've seen are really impressive. Um, I mean in enterprises the the roadblockers are always the same, right? you want to do something, you need legal approval, you need approval from 20 people to publish content. (1:06:26) Um, you need maybe a place to publish content, you need people to write it. Um, and then if you have technical stuff like you want to track the crawlers, you need uh buy in from the from the engineering team, there are probably data privacy and security concerns as soon as you want to touch lock files. um this is what I know from the from the enterprise space what what hampers you and for smaller brands it's often lack of budget like there's an online marketing manager who's doing paid ads and SEO and LinkedIn ads and now it's asked to also do chat GPT um (1:06:57) what is that person supposed to do right you can probably set up a monitoring um and then get some insights from out of that but will you actually have the time to do something for it I'm not sure but I mean at least with the monitoring you can go to the CEO or CMO and say Hey, here's a list of five recommendations and here's how we actually standing against the competition. (1:07:17) Uh, but I would say lack of resources for smaller companies is definitely the the problem and also sometimes lack of knowhow. Cool. Interesting. Then since I don't have another question in here, the last question will be from me and this will be a personal question and I would like to know how often if you search for something or if you research something online, how often do you use chat GPT perplexity or the new school tools and how often do you still use Google? I use 90% chipt um I recently got a new laptop and I never bothered to change a default (1:07:57) search engine from Bing to Google because I only use it for spellchecking or to find a specific website. Um I was on vacation recently with my wife in South Korea for two weeks and I set myself the challenge to not use Google for search. I still used Google maps. Um, and I tbt I use it for everything and it worked very very well. (1:08:20) Like I I mean navigation of course you need a maps but recommendations where to go, what to eat when going to restaurant, uh what to visit in a certain city, where to book my train, how to book a train, a lot of translation work. Um had a bit of a medical incident. So went to a pharmacy where I couldn't read anything. (1:08:40) So just took pictures and ask what do I buy? what do I buy? Um, everything worked very very well. Um, much much much much better than than uh searching on Google. And I personally will never go back, especially on a mobile device, to clicking on websites with unknown user experience, often shitty user experience. (1:09:01) I I I'm very much in love with this like I have an LLM to chat with and that answers most of my questions. Strong. Maltto, what a great session. Yes, Andreas. Uh I second that great session. Um this was really really insightful. Thanks so much uh for taking the time also for going 12 uh minutes beyond time. Uh now that you are in a startup again the the claim uh time is money uh is more relevant uh I think than in the last 5 years maybe. (1:09:36) So uh everyone joining live thanks so much for participating. Thanks so much for your cool questions. Um, just so you know, we will have a recording of that going live probably the latest on Monday. Uh, it will go live on YouTube. I will also share it on LinkedIn. Um, if you want to follow Malta, Malta, where um can we or where can when the people in the audience uh hear more of your thoughts about the whole ecosystem, where is it best to follow you? LinkedIn is by far the best place um where I share like insightful stuff. If you also want to (1:10:12) hear random things, you can follow me on X, but um I might not post anything for a couple of months and then I might post 10 things in one day if I'm interested in it. Uh if you want to follow me, get insights, learn together with me, learn what I'm learning, then follow me on LinkedIn. Cool. Then thanks again so much. (1:10:33) Um have a great weekend everyone. Malta, have a great weekend. Thanks so much for um taking the time. Was a really cool talk. I enjoyed it a lot. Um and for everyone else, we will try to do this format obviously with um other guests over time. Um every Friday now, uh 400 p.m. German Berlin time. So, um yeah, maybe put a blocker in your calendar if you liked it. Um and join from time to time. (1:11:00) And if you have a cool guest that you want to propose that you liked me to squeeze out all the good nuggets as I did today with Malta, just send it over on LinkedIn. And besides that, all the best to everyone. Um, and have a happy weekend. Thanks for having me and thanks for asking great questions. (1:11:20) Yeah, thanks so much. Bye-bye everyone. Bye-bye.