Building Through Every Ad Tech Era: AI, Attribution, and What’s Next with Ionut Ciobotaru

Speaker A: In advertising, many people are passionate about the, uh, technical challenges, the scale, the big data, the algorithms. But I want somebody passionate about actual problems we're solving for businesses. We're getting them users and we're helping them reach the, uh, right. Users more efficiently and grow their business.

Speaker B: Welcome to How I Grew this. Where we dive into the world of digital marketing and connect with leaders in the space about how they're scaling, evolving and growing. Hey everybody. Welcome to How I. I Grew this. We are so excited to have Yona Chopotaru here with us today. Adam, I understand you two go way back. Uh, tell us about that.

Speaker C: Yeah, Yona and I met, what was it, back in 2015, 2016. Right. San Francisco. Both running edtech companies, partners in this space. And, uh, we've been fast friends ever since. I've watched your career with interest. I actually just looked at LinkedIn on the way to this recording and, um, I was amazed to see that you'd only spent two years at appleift before you founded your own company. So you only were in mobile for two years before you decided, hey, I'm going to go found my own ad tech company. Like, how did that work?

Speaker A: So I left Romania in 2012 and my plan was to do mobile actually from Romania. So my first company was in web. So I mean I was a bit of a, um, I would say tech geek, nerd guy, right. Trying new things. But I had an engage. I was building Symbian apps back in the days. My first job, I was a, uh, QA at Electronic Arts, right. So we were testing Java, uh, games on mobile. So that was my first work experience. And then I went with Symbian and with Nokia in Romania. I had some technology blogs and we were building websites. Back then 2000s, you would take offline business into online. Then E commerce came Pre Shopify, so OpenCart, open source tech, pre, uh, 20

Speaker C: to 15 era, that was online, not just mobile.

Speaker A: So not just mobile. Yeah, but then mobile came, right? The iPhone, Right, iPhone, it was hot, right? And people were buying iPhones in Romania, but we didn't have mobile plans, right? So there was nobody on Internet. You could not target them on Google.

Speaker C: Okay.

Speaker A: This is probably not the best place to start a mobile, so I should move a bit more west, right? And somehow the deadlines and the timelines didn't match. I was trying to go to Dublin because I don't speak German, um, but I ended up in Berlin. There was another startup, bootcamp, and ended up there with the plans to build a mobile SDK Back then for QA anyway, because that was kind of my background that didn't work out. And then I ended up at Applift leading their product for a couple of years before starting Paliti. So I knew already that I'm going to start a uh, mobile company. I pitched them. So the investors of Uplift, a couple of ideas, three and this one was the last one because in the end I got. That's the story why I spent two years at Uplift and how I started. So my elevator pitch for probability was that look, all the mobile apps are getting nicer and nicer in terms of designs and you have every new version of the OS improves, but the ads are the same, are just banners. Right. So we should make native ads in mobile apps feel native to the user experience. So that was the, the elevator pitch.

Speaker B: Okay, so do you ever still think about those other two pitches that didn't quite make the cut?

Speaker C: I do.

Speaker A: Uh, not too much. I mean basically one of the rejections I got was like, oh, it's not big enough. And I mean for VC model probably is not big enough, but it's still probably a solid business. And it was basically a web to app kind of thing. One of them. And I mean now web to app is pretty big, right? Of course for subscription and other things but. And it took 20 years. So you need to survive six years

Speaker C: running pub native from start to uh, selling and you focus on mobile the entire way. You had some changes along the way. I mean you started out as a native only ad network but you didn't stay there.

Speaker A: No, we didn't stay on there. When we sold, we started as a native only performance SSP and we pivoted a couple of times. The last one, the one the business we sold was I would argue mostly banners and programmatic. So why we didn't start with programmatic? Well, there was no programmatic for native mobile in app in 2014 or I know 13 early when we started. So that was one of the reasons. Second, we were going against banners and against SDKs and then we had an SDK and we were selling banners. So that became one of the largest revenue streams. So sometimes you cannot, at least as one company, push against the market itself. And what advertisers, specifically brands, I would say they want standardization right across all their channels. So the larger brands, which were most of our revenues back then. Yeah. And then door banners, many medium rectangles. So all the um, inherited ad formats from web traditionally. Right. Of course there was video and rewarded video. Those things came a bit later.

Speaker B: What was most surprising about those days or what lessons popped up that you didn't expect?

Speaker A: So the most interesting piece was how international mobile was. Like you would build an app in Belarus, right? And then you launch it in the US or anywhere. Uh, so China would launch many, many apps back then. They could even acquire companies in the U.S. right? So that was the most surprising because if you think about traditional media, most of it was localized. Let's say you have Romanian media in web and TV or radio. Same for Germany, same for France. It was very, very localized. But with games and apps and mobile, basically you put it in the app store, you choose, your countries can be five, it can be 30 and you're everywhere. Of course if you're a large brand, that's not as easy. But for early brands, and I remember some of our early publishers like Voodoo, right, they was called Quiz Run, they had a small game and they launched it until they went into hyper casual, right? So but the idea was that hey, you're in France but you can launch a game, it goes viral internationally and then you scale, grow and iterate. And I think that's the one of the more surprising pieces. The other thing that I still really love and enjoy is the type of people you get in this industry. I would say some sort of gamers that a bit with ad tech and data, uh, hats, because you need those to actually perform and grow in the business. So actually smart people, fun to be around. So that's, I would say uh, the other piece also just the growth of the industry and now nobody talks about it. It's already, it's mostly mobile if you look at time spent and advertising spent, right. And I think the large brands took the longest to be there, right. So you had the mobile native brands and you had like Neo banks and all of these things, but you still have large brands. I would say more in the recent years. So maybe the last three, four, five years, McDonald's and the Starbucks that took crazy market share first, they executed well and they managed to get that many flops as well in terms of apps. But many, many of them managed to actually grow really nice audiences in mobile apps and grow their apps, grow their engagement, grow their user base with rewards, with um, location based loyalty programs. Ah, uh, it develops so much.

Speaker C: The data backs that up too. Because if you look at the growth over the last couple of years, the overall industry has grown about 5%. Games have grown 0, if not like maybe 1, 0. They're basically flat. And the non games have grown 10% so the only growth in mobile right now is from the non games. So it kind of makes sense that you only see the people adopting or at the large companies. Large corporations, yeah, exactly.

Speaker A: I think first you start with what's easy, right? Entertainment and that goes really fast. But then at some point, let's say number of devices and the time you spend there kind of plateaus. But then you need utility and functionality. So then that's uh, where the bigger players come into play as well.

Speaker C: I have an unpopular opinion about native ads and I'd like you to shoot me down if I'm wrong or say, I mean you've distanced yourself from the businesses so you can actually look at it from retroactively. But Facebook famously was the first large scale company that released native ads. Facebook native ads, as far as I can tell, they released them, um, they were somewhat popular. They never really got universally adopted. And then Facebook has basically walked away from that business.

Speaker A: So not he walked away, but mostly walked away. And that's also because of the idfa. Right. So story one of our pivots was actually because of Facebook audience network. So we were owning the market at the beginning, but when they came they were promising higher CPMs m although the fill rates were terrible because not everyone was in Facebook and they could only target behaviorally on the id. Publishers were not looking at the total revenues like, oh, but Facebook gives me 20 bucks CPM, you're giving me five or three. And everyone went Facebook back then. But that didn't last long, right? Because then at the end, end of the day, after one year they had to look at their bottom line. It's like, oh, how about that $2 CPM. I was like, yeah, we're still here. Where's your. Now you know how to look at the dashboard. That's nice. So our inspiration actually came from the fact that both Google in search and Facebook within their feed and now if you look at Instagram and Reels have very, very native units. And arguably I would also put up loving success in this reward the native unit within the game which is very native to the game gaming experience. And I think the native as a uh, display unit with just those elements that you put around maybe didn't succeed as much. Although by the way, when I was living this was actually growing much, much more than when we sold the company. Right. So that's the other piece of it. This is how, how it happens sometimes. But yeah, I wouldn't say it's the main format that's being traded right now, but if you look at the large platforms, they trade only native inside the platform. So for book, Pinterest, even Amazon if you look at their formats in shopping and so on.

Speaker C: So I Google search in a lot of ways too.

Speaker A: Google Search, if you look at it, it's exactly like a uh, normal listing

Speaker C: and ADWords and all of the ad pmax right now. Yeah, I guess it's a good point.

Speaker A: It might not have win in Programmatic where we had the bet and also when Facebook tried it, but within, let's call them War Gardens or these larger platforms, it's actually the most successful format. And the other piece about it is that they innovate and change that format almost every couple of months.

Speaker C: Uh.

Speaker A: Right. Uh, in order to not get user fatigue. So there's a little tweak here. They add Legion form M there. You can submit the form directly there even if you look at YouTube or Meta.

Speaker C: Right.

Speaker A: So that's the other piece with the native format actually evolves over time maybe except Google, although even Google now and unfortunately here I cannot see them. But from my understanding they have ads in AI mode and AI overviews.

Speaker C: You mentioned selling the business or acquisition you spent after pub native. You spent five years, you spent four years ish with MGI and now named Verve Group, then acquired Bubnative and then you end up becoming the co CEO.

Speaker B: Correct.

Speaker C: Yeah, it's a good ride. Tell us about the acquisition which I'm sure was probably fraught and probably some of the best and most exciting memories of your life in terms of stress and accomplishment.

Speaker A: I uh, haven't had that level of stress before or after.

Speaker C: Tell us about it. Why did you pick mgi? I know they were active in the space during that time, but why did you pick MGI M and how did that go? Give us an anecdote story.

Speaker A: So we were actually fundraising. We were looking to do our series A around that time and we got some interest from another company in the space like strategic investor, another European company. And then we said with our investors to open up a process. So then we got another American company on board as well. And then we had MGI as well which came I think head towards the last ones. Right. But then they were more aggressive in the process, faster running the process in the end also the better outcome I would say for all involved in the sale. Yeah, they wanted to expand from uh, basically they wanted to in house advertising the ad tech stack into their portfolio of games because at the beginning they were media and games, but they were mostly games. They had Gamigo In Germany they had Trion Worlds and another one in the U.S. yeah, so they were quite diverse.

Speaker C: Right. They were an aggregator of games like stillfront or Embrace and they were relatively good at it.

Speaker A: I think they acquired before us 30 gaming studios, smaller and larger and then they wanted to do something similar in adtech and we were the first advertising or ad tech company they acquired.

Speaker B: Did your relationship with Risk shift quite a bit once you were in the Verve World or MGI?

Speaker A: Yes, 100%. So we went through a couple of pivots right, with pub native. So you always look at okay, what's my run rate and what's my burn and all of these things. Right. We were break even at the time but you know it's like um, it's a moving goal with startups and that changed with having like a ah, larger mothership. So then we were like okay, what do we need to acquire? Where do we need to invest? So the relationship with Risk. Yeah, I would think the relationship with stress changed more risk.

Speaker C: Well how up or down you were running your own company, VC backed. Some days you lost money, some days you made money. But then you become the co CEO of a public company. Was it more stressful or less stressful?

Speaker A: It's a different type of stress. I think you have more stress because there's many moving parts at the same time but you don't have um, this life threatening stress. Existential stress in terms of quantity I would say because it's really moving parts. There's investors stock here, then there's many people. You have to deal with many um, entrepreneurs which, how they are but you don't have the life threatening kind of existential stress that you had before. So this I didn't have and this was although you know I had to travel a lot because we were also acquiring business in the US and my co CEO was on the west coast. I'm here. There's not even a direct flight from Berlin to sf. So that's uh, another thing. I was traveling a lot. It's a different type of.

Speaker C: So it went from like live or die to grind like from 18 hours a day on, on.

Speaker A: Exactly. But it felt good coming out of that into this uh, I don't have to say it.

Speaker B: I'm pretty fascinated by the fact that you were brought in and then became co CEO. Was that pretty quick or was there some sort of like proving ground? How did you fall into that role?

Speaker A: So there was a bit of a proving ground. So I initially I was staying with pub native. I think a Year or two. Then I moved to cpo, which mean my responsibility was more on the tech and product and engineering, which meant for the acquisitions we were making. So we made around 10. During my time, I was in charge with the teams of doing the integration, the technical integrations. Right.

Speaker C: And did MGI go from like a holding company of mobile apps to an ad tech company?

Speaker A: Correct. Yeah. Now they still have the gaming studios, quite a few of them. But the main business and the main driver of all the growth and most of the M and A that happens currently is in ad tech companies. So Verve was one of the first acquisitions recently. They bought Captify and June Group. While I was there, we got Smarto and nextar Digital into ctv. So we acquired quite a few assets after Verve was the first US ad, uh, tech company. So after Palm Native, we acquired Verve and the name stayed as it was the best name from all the companies we.

Speaker B: I like it. Verve is a good name. And then Curveball. You started a bakery too, I see. In the mix of all this.

Speaker A: So after I left the, uh, M and a roller coaster, I took a bit of time off and I said, okay, I'm going to start again in, uh, ad tech or tech, and for sure with AI, because back then it was just starting. But it's the one moment in time where I can do something else for a period of time. And I wanted to do something non digital. Right. I said, okay, it's my first business in web, second in mobile. Now I'm going to do it again. I was like, wait, I should do something non digital. So physical actually. And I thought of a couple of ideas. It's a pretzel bakery. So it's made kind of like, uh, boiled, then baked. Yeah, boiled and then baked. So. And Berlin has, although it's a German city, has traditionally pretty low quality of, uh, bagels in general and then in particular. So I said, not anymore. I wanted something, but honestly, I'm not going to do it again.

Speaker B: Okay, so you're in there, you're doing something physical. You're out of the edtech world. Are you still keeping track of where the ad tech industry is going? Is your mind still there?

Speaker C: Wait, Amanda, uh, before we move on, I want to come because I talked to her about this a few months ago. I want to uncover some of the pain of the bakery here. So you say you're not going to do it again. Why not? Tell us about owning a bakery.

Speaker A: It's very physical and laborious. Takes a lot of time just going to one market. So we have different distribution channels. So we have markets, uh, we have deliveries and we have B2B and so on. And also if you don't have a brand, then you need to build a brand in a new field. So finding these people that have these markets, it's all you need to meet them in person and go there. There's no email, there's no digital anything, right? You just go there and so on. Then you need to pack, unpack your truck or your, your bike. Then you need to wait for the delivery. Delivery. Can't say I'm coming in four days, but no, actually it comes tomorrow. Right. And then at 8am you need to wake up out of your bed, go the other side of the city, take the delivery, put it back. Uh, same other things. Like our stall at the, at the Fe, it got burned after the fact, then we got flooded and then the wind took three days.

Speaker B: Oh my gosh. All the elements, literally nearly every element.

Speaker A: I don't even know where to start. I was like, I mean, it's fun to sell to people, to get to know them, tell them a little story about the products you have. But it's so laborious. There are so many times you had to pack, unpack so many weekends that it was coming back at 2:00am, you know, after. And then I had to work on my attic and other projects. So I think that's.

Speaker C: Well, so one thing I've always wondered about bakers is you have to be a morning person, right? Because you bake. Baking starts at like 4am and I am not.

Speaker A: That's the other problem. Yes. Sometimes we did shifts in the evening so we would prepare the dough the night before because I'm not a morning person. So I like this was. I did lots of mornings, you can imagine, but you know, I do it with very little enthusiasm.

Speaker C: That's pretty funny.

Speaker B: Okay, so going back, so you were also working on ad, uh, tech projects simultaneously or you were still thinking about it? Talk to me about the relationship there.

Speaker A: I was looking for a co founder cto. So now I was a bit picky with that.

Speaker B: That your bakery customers, any of them?

Speaker A: Well, fun enough. One of our bakers was a student at one of the computer, um, science universities here and she was asking if we can join the ad tech. Uh, it didn't happen. Uh, but that happened. Yeah, it was a bit, uh, hard to find somebody that's passionate about advertising and ad tech. And I was looking for something like that because in advertising many people are passionate about the technical challenges, the scale, the big Data, the algorithms. But I wanted somebody passionate about actual problems we're solving for businesses. Right? Hey, we're getting them users and we're helping them reach, uh, the right users more efficiently and grow their business. Right. And I want somebody that's passionate and energetic about this part of the business, not just uh, the building part.

Speaker C: And this is what you're one of the. I know you're working on a few things today, but this is hyped. That you're working on.

Speaker A: This is hyped. Yeah, this is the hyped AI agent. Actually now I'm focusing only on this one. So since the beginning of the year or, or bit before, I'm focusing on, I'll put bakery on the side. The other piece I had also on the side. So I'm focusing only on this one, 100%.

Speaker C: So this is HYPD AI. Tell us about this business.

Speaker A: So about this business. My thesis, when I left Verve, you also mentioned, uh, the growth of the advertising market, that games haven't been growing, but apps have been growing. Let's say one dimension to look at where time is spent, but the other dimension is growing in the open, or is it growing behind War Garden? And if you look at where most of the block list is actually Google and met, right? And TikTok and I was like, well, we have this new technology, generative AI and can we optimize the current workflows that marketers are doing, right? So one of them is you need to go into all these dashboards and you need to check the metrics, they need to compare it, then you need to make decisions based on them, then you need to make those updates, see the feedback from the platform, take it back, right? And then iterate on that. And that's very labor intensive. You can spend hours, you can spend days, and the more channels you have have, the more time you're going to spend. So my theory or my hypothesis was that, hey, with agents or with AI, we can automate some of these parts. That's one piece. The other hypothesis was, hey, we're going to have more ad, uh, channels isolated between each other. Like well, you have CTV now you have ChatGPT ads. I know, Doordash, instacarts, Walmart, everyone has their little ad network. But how are you going to manage all of those, right, as uh, a brand, even as a large brand, right? Again, you're going to need agents. And with those two hypothesis. So one optimizing workflows, the other one optimizing channels, we said we can help Cover some of those gaps, use Genai to do it. And that's uh, what we're doing, helping marketers be more efficient.

Speaker C: Are you focused on a certain type of customer? Is this like a one person shop? Is this like an agency that's servicing multiple customers? Like who would be an ideal customer for you?

Speaker A: So right now we're focusing on I would say small agencies and freelancers. So not the smallest one with the smallest budgets, but at least a couple of thousand per month. So those would be the couple of thousand per customer in advertising budget and ideally around 10 customers basically managing 10 advert. So small agencies, small freelancers, I would say that's the range that what we focus on. But we're building more and more features for this group, ideally expanding into larger budgets and larger advertiser as we go. We're currently bootstrapping so that's why we started, I would say the early stages of the market and we go up market step by step. We chose a different route.

Speaker C: Previously vc, then public, now you're bootstrapping. So you want either all three or you want to get away from having more investors.

Speaker A: That's a good question. I don't think I have the best answer for it. Uh, basically the founding team were a couple of founders. We decided to go this way to not spend time fundraising. So we have a bit of funds from other exits. So we said let's do it our way, figure out how we want to do it and when we see that we need more fuel in the business, then we can raise on our own terms rather than raising early. So that's how we're currently looking at it.

Speaker C: So out of curiosity, and I know less about this than I probably should, but why couldn't an ICP be an in house? Someone who's running a couple thousand dollars or tens of thousands of dollars for their own brand and we unuse your product.

Speaker A: Actually that's an ICP we're aiming towards. But I will tell you why it's not an ICP now because most of these advertisers, I would say they need a couple more platforms right now we have Google Meta, but they might need either uh, Shopify or an Attribution solution. We will need to integrate an MMP or something like this side and they would have at least a couple of channels, right? They won't have just Google Meta, they might have Programmatic, they might have email, they might have different channels and we would need to integrate all of them. To actually create that value takes time. Each channel, it's a pain. So actually building with AI, which is a different story, but it's a pain to build. Well, to be considerably better than ChatGPT. Right? Because that's kind of the baseline. Took us a year to be better. But I think now we never hear we're not better. We're always better than just dropping files and data into ChatGPT for use cases. So I would say that's the answer. We need to build more capabilities to go ICP by icp. Direct advertisers, I would say they have more complex needs.

Speaker C: So um, if I'm running just meta, just Google today, I wouldn't get enough value out of the automation that you guys provide or the AI agentic help.

Speaker A: No, I think you would get. But you would spend up to 10k or 20k. I think once you go more than that per month, you're probably going to use different tools as well. Like even if you don't use attribution, you might use analytics, let's say Google Analytics. Right. Then you might have Shopify or some way, uh, I don't know to track real value.

Speaker B: Right.

Speaker A: Because there's over attribution. So even web and even in mobile. So you need a source of truth to be integrated as well. Of course you could just paste files, but that's not very agentic. Then we're not really solving the problem. You need a source of truth, I would say as a director of that. And then you need measurement. Right, okay. Which worked, which didn't work. These channels, right. They over attribute and optimize for themselves. That's also our thesis is that basically the advertiser needs to be in control and that uh, the platform actually optimize for themselves, if not the primary goal, but for sure the secondary goal.

Speaker C: So that's a uh, huge question. I've measurement and attribution today for AI specifically, what are people doing? What's working, what's not?

Speaker A: For AI specifically I think there are two parts to it. So one is AI is a channel and people are looking into AEO and ngo. Right. But if you look at most of the um, let's say the traffic that's associated with that, it's in the single digit percentage. Yes it grows 20% year on year but from a single digit to still a single digit. So that's one. So I don't know if you need to measure that much but what we've seen people saying again for small pieces and I think that's mostly applying for large brands. Right. If you're a very large brand, 2, 5%. It's actually a huge amount of traffic that might come to you. So in that case it makes sense to measure it. And what we've seen is that or heard those are not the brands we're talking to. Is that performance is better or seems to be better in terms of engagement. Like if a user comes from ChatGPT, they will engage more than a user coming on a website.

Speaker C: Yeah, that makes sense because it's got more higher intent.

Speaker B: Exactly.

Speaker A: It's higher intent than Google have a short query versus ChatGPT, which is usually a longer, longer query that you ended up at the end to.

Speaker B: Right. And they've gotten more to the bottom of what they're looking for. And at this point this feels like the best recommendation. So there's almost like a little bit of trust baked into like where you're sent as well perhaps.

Speaker A: Exactly. So, uh, it's quite interesting indeed.

Speaker C: What are people doing wrong? So it kind of sounds like people are focusing on when they shouldn't be.

Speaker A: I think they should be learning and I think if they are larger brands, they should be looking into it. In terms of what are people doing wrongly, I still see a lot of people not really using AI. There's two types of AI, right? There's gen AI and there's what I call traditional AI. And I think for measurement, attribution, incrementality, you actually don't need genai. Right. Like the two things I mentioned about gen AI was efficiency in operations and the other one is new channels. But if you think about the actual problem of advertising is mostly about measurement and attribution. Right. Like, hey, okay, the demand is there, which bucket is which, where I should spend more to gain more or more efficiently, so then I can reinvest. Right. And I think that that's been always the problem. Measurement, attribution, incrementality. And I would say small brands are not focusing on it enough. Uh, not at all largest brands as far as I know, but we're talking like really, really large brands. Millions is spent per month. They do look, uh, at these things and I think that's the right way to look at it. I don't know if agents will bring you much there or this SEO geo is just another channel. It's a small percentage, you should look at it, but more you should look overall at your channels. Uh, and maybe with the agents, add more channels. I don't think it's trivial right now, but that's one thing to look into and be able to experiment faster between where your user are and where your let's say value comes.

Speaker B: But really interesting to hear you say kind of the difference between how important this is for a smaller versus a larger company. It seems like for the smaller companies, like it really doesn't need to be a focus but for the big ones, if they're not looking at it, they're going to fall behind or if they're going to miss out on a big chunk of data or how are you seeing that?

Speaker A: The larger brands, the way I see them, are usually at the top of their market. Right. So they look at competitors, spend, hey, this guy is spending more on my keywords than I need to spend as well and things like that. While a small brand, it didn't even exhaust one channel even Google or Meta or even two. Right. So it means that not really a problem they have between channels. But once you are different channels and you maximize more or less the value and actually competing with others, then measurement is probably the most important thing uh, you can do. But early on is more making each channel efficient and making sure you can maximize the value of each channel.

Speaker B: What feels harder about performance marketing now than it used to?

Speaker A: That's a tricky one. I think we have loss of signal. Right. That's the first one. I mean now it's recovered a bit. I mean everyone readapted to it but we had that one and that one that turned around the whole ecosystem for a couple of years. Right. So I would say that was for sure harder. Now platforms are getting more complex and also more black box. Like Applovin is a full black box. You don't get any level of detail of the um, placement id but not uh, publisher level. Right. Google with PMAX more or less the same now they have channel distribution but yeah, you don't see much Meta wants to automate everything. So where is your control? Right. And if the platform has the control, you know you're losing it. Like the control must go somewhere and uh, it's not like they will always deliver to your goals and then leave

Speaker B: the budget unspent and the variety of the output too surely is going to decrease.

Speaker A: So I think that's becoming harder. Yeah, at least those are.

Speaker C: Yeah, well question on that. So if you are optimizing towards your roas and your business is less focused on the details of the targeting and the administrating, uh, and you just give it to the black box and they spend down to your roas. Shouldn't you be experimenting with what you said earlier which is making sure you set the appropriate Targets and then reinvest, assessing where you are achieving your targets.

Speaker A: I would say so, but again, not many. So for example, if you're a small brand, you might be able to afford one agency, right? So then you're going to run more channel or if you're a big brand then okay, you can have that, you can talk to your agency or whoever runs your campaigns, maybe it's in house and say every monthly report or quarterly report, we should reinvest. And as far as I understood, or at least what I've seen is that this reporting, and even for large brands like media reporting, so channel reporting has moved from one year to maybe quarterly. So they do do make these adjustments. And look intra quarter, if a new channel is emerging or if a new channel is under or an old channel is underperforming and so on. So this is happening, I would even argue, at least in mobile app people are looking for new channels, right? Because they already know the channels are kind of saturated, there's competition, prices are high. So yes, we had that deep, but with idfa. But now as much so people are looking outside of. Okay, mobile app. App. I have an app, but I want to advertise somewhere else. And where, you know where, uh, I do it work. Gardens? Yes. Programmatic. Yes. And network. Yes. What else is there that I can go and engage with my audiences? Uh, but again, it's only if you're big enough that you can afford to ask yourself these questions and go out there and uh, answer them.

Speaker C: So looking ahead, and I'm reading this back to you, so correct me if I'm wrong, if today the focus of the business should be on making things measurable and investing in the appropriate channel, the right channel that's profitable. Looking ahead, what is the future of the performance marketer with AI is giving them guidelines?

Speaker A: So good question. One dear to me, and I think I was chatting with some people on LinkedIn about it the other day. Um, and the question about allocation, so what you can do with agents, and this is, uh, adcp, Ad Context protocol, which is an open source protocol, part of agentic advertising organization, which we're also part of. But the thesis there is that, okay, if you get your measurement relatively well in place, what you should be doing right as a performance marketer. Well, allocation. Right. Find the new channels, specifically the channels that are not easily available with your current workflows. Right? So think about out of home or even traditional TV or I don't know, other channels that are not easily available now that are not traded on our

Speaker C: Impression offline billboards, like that's coming back to billboards, huh?

Speaker A: Uh, anything that's not traded on impression, by impression, but can be bought in bulk. But you don't need to buy like before. You would need to do upfront of, I don't know, uh, tens of thousands or hundreds of thousand. Now you could with agents because you simplify both the buyer and uh, the seller part. You could buy in small amounts of course, based on the cost of the media, but you could open up to retail media, all of these many, many smaller retail medias, big term, you could try them out with smaller budgets and decrease the barrier of entry both on the sell side but also on the on the demand side to access those. So I think that's one of the first order effects that I can think of for AI to help. Right. You get your measurement in place, you want to experiment, you want to grow, these are the options, go from there.

Speaker C: I love that. So reading that back to you today, the most important thing is measurement. Once you have measurement nailed, then then apply experimentation. In the old days, experimentation was maybe new ad units, different things you could control within a channel. Now that's handled by AI within the channel. So instead focus on experimentation outside of the known given kind of scaled channels.

Speaker A: Correct. At least that's where I see. Because if you look, the channels are going, they want to use AI within the platform to maximize the value of the platform. But what you can do as a marketer is go outside of the that. I mean that would happen with or without you and find out where your users, your valuable users are.

Speaker B: So you felt your career really focused on solving the challenges for performance marketers. That's what's kept you passionate? Seemingly.

Speaker A: Yeah. So I didn't say it, but one of my first customers in my first business was um, a math teacher that had an online shop that kick in at a Google Ads in AdSense ads. Right. So that's performance, the CPC and it was a customer for a couple of years. I was surprised like who would click on that banner on Google on leak and then converts.

Speaker B: All right, so what's the next challenge that you're looking out for that you're hoping to solve?

Speaker A: We're looking to expand our icp. I think as Adam mentioned, now we're moving towards mid sized agencies and we try to cover more talents and agencies. But the challenge after that would be direct advertisers. Right. I think that's the end. Even connecting agencies with advertisers. Right. Because we use the same tool and Anyway, you need the humans in the loop. I think for now what I've learned is that just letting the AI on

Speaker B: its own, it won't work out. If it's slop feeding slop feeding slop, then the whole world becomes slop. We need the strategic input from the human.

Speaker A: Exactly. Direct advertisers would be the next one after we scale a bit and we add more capabilities. And personally I'm very excited about ads in LLMs. Right. ChatGPT will be the first one this year already. But I do think a new channel is super interesting for me. Right. So we had mobile full explosion. I think LLM ads, chatbots a bit more uh, nuanced like it's not a channel specific like just mobile app or web. You have different use cases where you find those ads. I've seen ads in an ID amp, um, code. I don't know if you know this one. It's not an open source or is it. It's a small ID in the terminal and they have an ad on top of the chat, the agent box. Right, so. Exactly. So having ads for in developers, you don't expect, expect that.

Speaker C: So you have an ad in like a development environment. That is crazy.

Speaker A: That's crazy. Exactly. But it's happened, it's live now.

Speaker B: Right.

Speaker A: So that's the, the one thing I'm more excited is about shaking a bit the, the current status quo in terms of channels. That was called the mid to long, long term excitement and that's actually literally

Speaker C: follows through exactly what you've done your entire career is kind of find new channels and take them on. I get it.

Speaker B: Um, awesome. Well, where can people follow your work or stay in touch?

Speaker A: I'm on LinkedIn. Uh, LinkedIn. Yonotrobotaru. Um, and that's where you can find me the most. Or email yonothype AI. That's um, another easy way to find me.

Speaker C: Wow.

Speaker B: Thanks for sharing.

Speaker A: Yeah, thanks for having me.

Speaker B: Oh, thank you. This was awesome. Thank you so much for sharing all the insights and congrats on a um, really lovely career and I'm glad that you're fix the bagel problem in Berlin.

Speaker A: Well, somebody had to do it, right?

Speaker B: Amongst many other things for marketers. Yeah. Uh, awesome. Thank you. Bye everyone.

Speaker A: It was a pleasure. Thank you.

Speaker C: Thank you. Anat.

Speaker B: Thank you so much for listening. If you like the show, please leave a review wherever you listen to this and share with someone trying to grow their career or their business. Until next time. Time keep growing.