No Black Boxes

Speaker 1: So welcome to this edition of the Business Marketing, recorded live here at Cannes Lions '26 in the wonderful Dept Garden. Uh, I'm delighted to be joined today by Christoph Berg, who's the CEO and founder of Mint Square. Um, I think this is gonna be a conversation about sort of making programmatic useful again, um, for both kind of agencies and brands-

Speaker 2: Yeah

Speaker 1: ... who perhaps have been a bit sort of overwhelmed by the tools, the partners, the complexity. Christoph's built a business, uh, around simplifying, which I love, cleaning up and helping teams run programmatic with more confidence. So the interesting bit is where that clarity creates commercial advantage, and where complexity is still being tolerated by the industry, uh, is used to it. So without further ado, Christoph, why don't we kick things off?

Speaker 2: Sure.

Speaker 1: Let's talk about sort of why programmatic got so hard to manage. You built Mint Square around the idea that programmatic is too complex. Where do you think that complexity really comes from?

Speaker 2: I think programmatic advertising... Well, I used to work for an ad tech partner, uh, on the DSP side for AdForm long time ago. And, um, I was selling the technologies at this time, and it was already very, very complex at this time really to understand. And I think the expectation at that time from the technology was really having partners using it the way how it's supposed to be, but it was really never done at this stage, like, like how it was really intended. Because, I mean, background of such a technology is a very technology-driven background, and sometimes it's really challenging, like, to use it properly. And also, where the traders or other DSPs are coming from, I mean, they used to be like ad networks in the past. And then they got connected, and then they were just, uh, th- then they were, they were bought from Yahoo and from Google. And then it originated from a display perspective, which is, was very m- was, w- which was more known, like the, the remnant traffic, the cheap traffic. So programmatic advertising was kind of like the, the, the, the, the channel to go to, I would say. So, and from that moment on, more channels got connected, um, through, through this kind of DSPs, and then each channel has different ways on how to buy it. I mean, we have different types of buying processes in place, like open auctions, private deals, um, private auctions, programmatic guarantees. So when partners now try to get h- an really in-depth understanding, where do they start from? So it became more complex over time, and it's, I think, very challenging then for companies really have a, an, an really good insight how to, how to manage and maneuver through, through this complexity.

Speaker 1: And when people talk about sort of, you know, they want more transparency-

Speaker 2: Yeah

Speaker 1: ... in programmatic, what are they not seeing?

Speaker 2: So the word transparency, I think in terms of programmatic, has, like what the meaning of it has changed and pivoted over the years. So when I want, like, just to look back where when we started 10 years ago with Mint Square, I think the pricing was a major topic, but not on the technology side, more on the agency side. Because advertisers often challenged, like, what are the business models behind it? There were trading desk formed through technology, so it was very, very tough to see, okay, what is this actually at the end, what they are being bought, uh, what, or what agency was buying for them. So and nowadays, like advertisers have now higher demands. They really understand more about programmatic and what it is, like, really behind of it. So the expectations have shifted. And advertisers now, they don't really care about the technologies. They, they have, they have a understanding the big ones, what they really want to buy, but it's more about what's about the entire supply path. There are so many steps where, where technology is at the end, um, um, has to distribute all the budgets to. And I think with this it becomes very complex to understand, like, where is the money really flowing? And it's not only about this. There are so many different vendors out there, and it's about the, all, all the inventories that are connected. Who is really buying, um, then what kind of inventory? Because at, also at this stage, becomes very complex. Um, there are many ways, um, how to buy one certain inventory. Um, but is it the most efficient one? Do you get the best audience when you buy through this vendor? There are a couple of SSPs and bidding, uh, bid- bidding types at the end. So, um, transparency is not only about the pricing anymore, it's about, like, what kind of SSPs are you using? What targeting partners are you using? What verification partners are you using? That you have, as a brand, the, a very good feeling and understanding where, where it's shown and if it's safe as well.

Speaker 1: I mean, you've sort of started to paint a picture for me, but what are agencies and advertisers still doing that makes programmatic harder than it needs to be?

Speaker 2: Agencies and advertisers? Um-

Speaker 1: I guess it's on both sides, right?

Speaker 2: Yeah, no. Yes, of course. But then it's when advertisers are talking about... It's harder to be... There's so many angles to look at it. Do you wanna dig into something very specific?

Speaker 1: I think, I mean, it's down t- to you. I can, I can sort of leave that question if you want. Um, I mean, it's an open question for you to interpret it how you like. I mean, ultimately, I guess it's like what's getting in the way?

Speaker 2: Yeah.

Speaker 1: You know, ultimately, I think is the kind of big question on, you know, what, you know, how do we m- you know, back to your point around simplification. It's just sort of like, you know, what, what barriers are kind of there that we want to kind of overcome on both sides in terms of the agency side and the advertiser side.

Speaker 2: Yes. I think expectation is also a very big, big topic right now. So we are-

Speaker 1: Tell you what. I'm gonna ask that question again.

Speaker 2: Yeah, sure.

Speaker 1: And then you can come in.

Speaker 2: Perfect.

Speaker 1: All right. Yeah?

Speaker 2: No, no.

Speaker 1: What... And by the way, always do this. This is good.

Speaker 2: Yeah, yeah. No, no, it's great.

Speaker 1: Yeah, yeah, it's good. Yeah.

Speaker 2: I'm gonna follow up-

Speaker 1: Yeah, of course

Speaker 2: ... and then we'll come to the perfect.

Speaker 1: Yeah, yeah, yeah, of course.

Speaker 2: Yeah.

Speaker 1: Yes, perfect. So what are agencies and advertisers, um, still doing that basically makes programmatic harder than it needs to be?

Speaker 2: Yeah. When it comes to this, I think expectations is a very important part of it. Um, at Mint Square, we are tackling the mid-market, so typically it's the indie agencies. Um, we have more than 80% indie agencies who are working, we are working with. So, and those partners typically start often with Google Ads, Facebook- Even the, the budgets they are spending for the advertisers are in this segment. So, um, and then there's the expectations, of course, that it needs to deliver the same output as the other walled gardens. So, and when you start off with this one, it becomes very hard to, to justify then programmatic advertising within this framework of the company. So it's, it's very tough then to really establish this, and this is the biggest challenge then for, for these partners. And then it's getting hard, and you need the patience, um, on, and, and management awareness as well that you need to take the time and to really understand and dig deeper.

Speaker 1: Let's talk about sort of like what, what kind of good sort of DSP access looks like. You've spent a lot of time helping teams choose technology partners.

Speaker 2: Yeah.

Speaker 1: What does a kind of good DSP decision look like to you?

Speaker 2: Yeah. Also here with the DSPs, I mean, there are so many. I think there is nowadays no bad DSP. I really think there are a lot of advantages. I think what... If, if we're talking to partners, I think our view is we really would love to understand, so what are the mechanism and the USPs for each DSP? We are not... When we are reselling three technologies at MINT Square, but that doesn't mean that we are handling other technologies as well. Our clients want a en- entire ecosystem and when, when they have licenses of other resellers, we are happy to consult and help them as well on managing these campaigns. So it's, it's not about, like, having the, a right or a bad DSP at the end. It's more does it really tackle the problem that you are facing for you or for your clients? And, um, and, and then there needs to be like a list of requirements and-

Speaker 1: Mm

Speaker 2: ... if they match, it's really totally fine. And if you're capable, just building this around your teams, I suppose, then it's fine. But then when, I think when, especially in the agencies want to establish like a couple of ad techs within their ecosystem, it becomes quite challenging because it's then more complex and you have to make every space available in order to make this happen.

Speaker 1: And let's talk a little bit about the sort of like the, the actual kind of selection process. You sort of said, you know, it's all about fit, right?

Speaker 2: Yes.

Speaker 1: Which is, which is great, but-

Speaker 2: Sure

Speaker 1: ... I'm going in, I'm going into this process. What are the kind of, what are the kind of the... Give me, give me the kind of the real top kind of like what are, what are those criteria? What are the kind of must, must do, must not do kind of hints, tips?

Speaker 2: So we are talking about different markets, I suppose. Uh, we are, um... And then it becomes inventory, which is the first one you have to look at. I give you a real-life example. Um, we are operating with three technologies, one a local one of Germany, which is then Active Agent, we have Adform, and we have The Trade Desk. Completely three different vendors in the market. So if a client comes and wants to, wants to have Netflix, for instance, you won't be able to find Netflix either on Adform or on, uh, Active Agent, so you have to go for The Trade Desk. So inventory is the must-have you have to, you have to look into. If it's just about display, it really doesn't really matter, but this is one specific. Then API connections. Um, especially talking about indies or, or mid-sized agencies, they want, uh, they want an API connection with the reportings. Some agencies also require kind of connections to a, uh, to a, um, to a production API in order to have like an, a UX on top of it. So these are the mechanism that you would then say like, "Choose this then, choose this technology against another one," I suppose.

Speaker 1: And of course, that's just the beginning in a, in a sense, right? Choosing, choosing the DSP is just the, that's the, that's the end of the beginning maybe.

Speaker 2: Yes, of course.

Speaker 1: Then, of course, you know, you've picked the DSP, like then what kind of, what tends to kind of make the setup clean rather than messy, you know, as you, as you onboard, as you integrate?

Speaker 2: Yes. Um, here I need just to reflect a little bit, uh-

Speaker 1: Yeah

Speaker 2: ... deeper. So there is... So, um, you wanna ask again?

Speaker 1: Yeah, sure. Yeah, yeah. So of course, you know, choosing, choosing the DSP is a sense the end of the beginning, right?

Speaker 2: Yes.

Speaker 1: Then the, the, the real work begins.

Speaker 2: Yes.

Speaker 1: Um, what, how do you kind of make that kind of setup clean rather than messy?

Speaker 2: Y- when you start a very first initial campaign with any DSP out there, uh, you can't assume everything works like the other one. Um, I think one of the biggest, um, struggles of any trader setup in an agency or advertisers, the expectation is either you come from Google or Meta, and then it should work and it, it should work a- once you've switch it on, the performance is gonna drive to the top and it's like it's gonna be amazing.

Speaker 2: But this is not the way how it is. So-

Speaker 1: We dream.

Speaker 2: Yes. There's, I, me, like a lot of dreams, and then the expectations. So what, what we see, I think you have to have like, first of all, a budget allocated really to test any DSP inside out, and you will not lose at the end because you will have some insights if this makes sense at the end for you or if it doesn't. But at least you have a learning that makes sense at the end to understand if this is like something you want to go for. However, um, how would you do it ideally? Um, if you, if you set up a campaign on any DSPs, there are four levels essentially where you should have a focus on, um, and it's also on the level of skill the agency brings with you. Uh, first of all, you have to translate, like if you do traditional IO bookings from the past, it's a different way of creating a media plan programmatically. You're talking with audiences, different target vendors, uh, sales houses which are needed to be orchestrated, which are more dynamic. It's not that you like really allocated certain budget anymore to one vendor. One vendor can really go well for one campaign, and for the other it doesn't really work then anymore. So- Media plans are completely different than the way how you implement it. You need to have a deep understanding how this is working. Um, for instance, you have technologies, they're working with ad groups, technologies, they're working with line items, so it's a deeper understanding how you really set it up and which kind of algorithms are there. And then you can compare also. You don't need to have, like, just one campaign, like, running through one this- DSP, but if you have a setup of couple of DSPs, then just run this parallel and seeing to compare what are the advantages. The entire reporting part is very essential to understand how would you report at the end of any campaign and how can you communicate this or integrate this into this entire reporting. So, and this takes approximately, I would say, three months really to understand this entire process. And the way how we solved it is a little bit different than typical DSPs would do it. Um, when you onboard any client from the technology side, it's mostly you get all, everything presented. They will, there will be- the technology will be shown, every single detail, but then it's a workshop of one or two days, and then it's very hard to follow and then to remember how the real work is actually done. So, um, and over time we realized we need to take the clients really with them, with them together, sit in the boat, and create a blueprint for the future. And they sh- they shadow us and see how, uh, things are done. And then we can be sure that the entire capabilities of every these technologies are really seen and we worked with. And, and then after these three months, I'm 100% sure then these clients, they really have a very good understanding if this is working well for them and if they move forward with the technology or not.

Speaker 1: You're famous for saying, um, "No bl- black boxes."

Speaker 2: Hmm.

Speaker 1: Um, in the real world, what does a black box setup look like?

Speaker 2: Well, a typical black box would be a walled garden, I suppose. It has advantages. I'm not saying it's really bad, it's just something what do you expect and how do you wanna... W- how do you see your budgets going somewhere? So when we're talking about black boxes, there are advantages to it. It's very comfortable. You don't need to understand where the money is flowing. You don't need to understand how the mechanism behind this bidding is. You don't need to understand how things are done because sometimes it's done, it's done by managed service for free. Um, and then it's more about the output that's been delivered. So if your expectation is you give and throw money in and the output is that what you want, then it's fine. It's fair deal. So there's no wrong expectations on both sides. So the thing is then just go with it if it's fine for you. If your expectations is kind of you really have, you want to understand where the ad is shown, um, what kind of targeting vendor is being used for, is it first-party data that you're using in order for any purpose within your campaign? Is it, uh, second-party data where you really understand who is actually providing your, um, this, these data segments? So then I think it becomes more tricky and this kind of... If you ask this kind of question, sometimes there's just no answer to it. So, um, walled gardens have advantages, and we also work with walled gardens together because we need to tackle this kind of audience at some point. But if you want to have a holistic understanding of this entire programmatic ecosystem and if your standard is, like, really tackling this, you need to open up to, uh, to the open internet in the world.

Speaker 1: Going back to kind of, you know, first principles, I suppose, in a way. When a team's looking for more clarity-

Speaker 2: Mm-hmm

Speaker 1: ... where do you start first? Do you start with the data? Do you start with the partners? Do you start with the workflow or do you start with the people?

Speaker 2: Well, it depends also w- who you talk to. Um, typic- so there are two ways I would suppose, like, how this, this programmatic advertising, like, kicks in in, in institutions such as agencies or advertisers. So it's more about, so either it's the management, so they hear from programmatic advertising and they think it's a very good thing to tackle, and then they keep p- uh, then, then they bring it to the team and then they're asking, um, the technologies or, and us resellers, like, to, to support and onboard. And then it's a lot about people. Um, it's, I think, very challenging to explain all this pricing nowadays for people who are not with programmatic advertising. Um, you have technology pricings, you have data, data providers that need to be paid for, your verification partners, they have some, um, some CPMs on top of it. So it becomes really, really challenging then to understand why s- such a high percentage of the media is going to the ti- technology, uh, is going to the technology. And then if you add up all the targeting partners, then there's just a fractions going to the, to the, to the publisher and to the sales houses. So also there it's very, then it's a lot about education in order to give them a sense of so what's the added value to it and why to bring it into the business? Because if you are then going back to the walled gardens, you don't see how much they are earning at the end. You spend your money on a platform, you charge it with your credit card, or it's going just withdrawn from, from your bank account, but you don't... I mean, you know it's going to this platform directly straight, making a lot of margins. But when, when it comes, like, to have access to the entire programmatic ecosystem, there are so many players involved to get the most out of it. So then it becomes challenging, I would say.

Speaker 1: Let's talk a little bit sort of about teams. Um, you know, a big part of what you do is-

Speaker 2: Yeah

Speaker 1: ... all about helping teams, right? Sort of, you know, how do you operate programmatic confidently and independently?

Speaker 2: Yes.

Speaker 1: What does that sort of independence look like in real terms?

Speaker 2: Um, so when you are... This comes a little bit back what I said before.

Speaker 1: Yep.

Speaker 2: Um, having a very good idea about the entire journey on setting up the campaigns. Um, programmatic advertising is not only, like, defining five parameters about geo and IP targeting and then your audience and you let it run. It is a different way of buying, and it also depends on what kind of outcome you expect. If you're talking about performance, you need to have constant and regular optimization circles within this entire process. So for new traders, this is very unusual because typically how you buy media, and it's still, still going on, you brief some kind of an insertion order to publishers and sales houses, and then when they, um... And, and then it becomes more of, um, more technology. I think we have to rephrase this once more again.

Speaker 1: That's, that's cool.

Speaker 2: Yeah, yeah.

Speaker 1: That's all right.

Speaker 2: Yeah.

Speaker 1: So let's talk about building teams that can really kind of operate independently. You know, you do a lot of work helping them operate programmatic confidently, but what does that sort of independence look like in real terms?

Speaker 2: Yeah. Well, traders, I think it's very unusual for most, for, for, for new programmatic tra- as a programmatic traders and when you're new to this topic, that you really follow this optimization circles on a regular basis. We are trying to educate all of our clients to really stick on circles more times in a week, um, to optimize the campaigns of the c- of the clients because there are so... I think the biggest challenge is, is like having a great overview about all the capabilities of the platforms. It's not about inventory anymore, it's about choosing the right SSP, for instance, as well. It's choosing the right targeting DMP partners, choosing your first-party data, connect first party to the DSP, for instance. You have, if, if you wanna, for instance, connect or match your email addresses in order to, to enhance and enrich your campaign's capabilities, it becomes very, very complex. And I think it needs also a little bit of not only, um, not only the- theory on how it really works, how it's being done, but it's more about like having real-life situations. And that's where we also support along this process, that, uh, the traders really have a clear understanding how to onboard from the beginning until the end to the reporting for, along this entire process, if it makes sense.

Speaker 1: What are the skills that teams most often lack when they're trying to take more control themselves?

Speaker 2: So the biggest struggle of teams are about expectations, right, regarding programmatic advertising.

Speaker 1: Mm.

Speaker 2: We are talking about, um, machines where you have, like, all these different options that you can, can work on your campaigns. You can do the upper funnel, which is then the branding campaigns, and then you go down at the lower funnel with all the performance aspects to it. And, um, when you're, when, when you're working... So can you please read?

Speaker 1: Yeah, of course. Yeah, yeah, yeah.

Speaker 2: Thank you so much.

Speaker 1: Yeah.

Speaker 2: I need a good thing of water with it.

Speaker 1: Go for it.

Speaker 2: Mm. Oh, that looks nice with the-

Speaker 1: Yeah? You okay?

Speaker 2: No. Good testing.

Speaker 1: Yeah. What are, what are the skills that teams most often lack when they're trying to ma- take con- more control themselves?

Speaker 2: So I think the, the, I suppose it's expectations. Um, when you run campaigns originally, it's very about, like, output-driven, and the expectation is very, very often that teams are just, like, have the expectation you switch on the campaign and then it just do everything for you. And especially nowadays when AI comes so prominent, I think they're, we are not there yet. Just to give you a little bit. I, I'm, I'm the ... There will be, there will be a moment when we get there, but now currently there are, the algorithms of the technology that take care about a lot. But the human being aspect to it is so important, and you need to intervene. You cannot control the money of your clients or the advertiser budgets when they do it by themselves and let the machine do their work because the complexity just, just doesn't justify this. You need to have control if the spend is, if the pacing is fine, if, uh, if the, if the intended target group has been, has been, um, has been hit then through the campaign. And this becomes if, if you just let it, if you have the, if you have the expectations this machine is doing this for you, then it's, it becomes very difficult and then the decision going for programmatic or for not is going to be, um, definitely against the technologies end of the day.

Speaker 1: And, you know, if I'm a leader and I'm hiring in this area, what should I look for, you know, beyond people who just know the platforms?

Speaker 2: Can, can say again? Can I-

Speaker 1: Yeah, sure. Yeah. So basically it's like, it's like, uh, what's be- beyond the kind of the, what, what... It's beyond just kno- the knowledge of the platform. What, what are the kind of actual skills that you need to look for as, as a, as a leader?

Speaker 2: Oh, as a leader?

Speaker 1: Yeah.

Speaker 2: Yeah.

Speaker 1: Like, if you're, if you're hiring, if you're hiring a team.

Speaker 2: Yes. Yes, yes, yes, yes.

Speaker 1: Yeah. What, what are the skills, what are the skills that I need to look for beyond just they know these platforms?

Speaker 2: Yeah. I, I can tell you also what I'm looking for.

Speaker 1: Yeah. That's great.

Speaker 2: And the, so since we want to get rid of all these, um, repetitive tasks, which would be the reporting one, we are not, we are, it's, it doesn't mean that we are not reporting to our clients. We have all the, the dashboards to it, but the interpretation of the results doesn't need to be necessarily done, um, through someone, through a campaign management person. It's highly technical. So when the, the process has switch- has, has changed from a typical media planning position to something more technical, so I think the, the person definitely needs to be not afraid about numbers, about analytics, and to understand to interpret this. And AI becomes more important also on this aspect. Mm, we also work with two LLMs, so Antropic becomes one of our major LLMs now currently, where we're just not... We- we don't want to substitute the people, but we want them that they really properly understand how they can optimize the day-to-day business. And when comple- when it comes on the complex, uh, technologies, you also find answers to this. There are always academies with all the technologies, and you don't need to dig into yourself too mu- uh, too, too much anymore. But it's very interesting if you ask your LLM to support you with the question, you find your answers. And I was very struck by the fact that it's still, we are in an industry where I think it has been es- embraced with AI and all this technology which comes with it, and all the models and everything is building around it. But still, I'm surprised that also very young people, uh, in- along this interview process are kind of hesitant using this. And I was al- I'm always wondering, so, so why you... I mean, the way how it's been used very often lately is like asking and probing questions, and then you get something back. But utilizing all this project and coworker capabilities can really help you to become more efficient for these tasks which are really not interesting anymore, which, like, takes a lot of energy and headspace to it. And then you need some... You- I think you need people, like, really that they can focus then on the relevant ones to dig into more complex campaigns and to understand the mechanisms about all the capabilities that tech- technology is, like, offering because this is changing all the time. I think if you like, if we are talking now and in one month from now, there are different new providers on the market, uh, which are connected to, to the DSPs, new DMPs, new verification partners and contextual partners along this process, and you need to keep up with this. And not to know about it, really to experiment and be brave enough to include this and to test it out so that they can, uh, include this for future campaigns and successes.

Speaker 1: What's really interesting about you is you've, you've kind of worked in publisher sales, private market players, SS- SSP relationships. What did sort of all of that experience teach you that maybe sort of many buyers do not see from their side of the table?

Speaker 2: Also, the motivations about all these, um, of the publishers. Of course, the, the, the, the aim for publishers is to have the highest revenue and get most of the bid. Um, when I was working for TravelAd Network, which is, like, 15 years ago now, um, and when we started, that was programmatic really at the very beginning. So it's all about kind of, um, where do you put the ad, and, um, and, and also when. And the more relevant it becomes, the more the publishers earn. So now with programmatic advertising, it becomes also for the publishers, they have, like, a portfolio of different ways how to connect inventory to the programmatic ecosystem. And for the buyers, it's very hard to understand, so which way, where would I get the most out of it? Because if you're, if you're buying Huffington Post or if you go in Germany for, for the Bild site or in Austria for The Standard and Le Monde on- and they have different connections to all of this, and there are different ways how to buy it. But even though you see then the media at the end, it doesn't mean that you buy it in the same quality. It's a different, it's a different position, and it depends do you want some, do, do you want some branding? Do you... Or do you expect some performance out of it? So do you need to go to a one-to-one relationship with the publishers, or do you need to go through the open auction because it's sufficient enough? So also there, I think, um, if you also have a little bit of more understanding on the SSP side, it helps, um, to get more out of it when you also have, like, a direct relationship to publisher sales houses to get special deals and different, um, different options that you might not get if you just click on, uh, something on the DSP.

Speaker 1: When a team says, "Programmatic's not working"-

Speaker 2: Yeah

Speaker 1: ... um, what's usually the real problem underneath?

Speaker 2: Mm. So here it's also about a lot of expectations again. Um, programmatic works for some brands. For some, they don't, um, because then it becomes about measurement. We have it very often that clients come, especially now we're talking about, like, advertisers. So they have, like, they've run through the Google Ad space. They know exactly all the social media campaigns. They're very into, um, building their brand and using influencers. And then at some point when it's, when you have, when you, when you reached your threshold, so you're really putting more money in and it becomes more expensive, whatever you want to achieve with your, with your media campaigns, then typically the marketing responsibilities and the teams are looking for an answer for this problem. And then they come to the programmatic channels. So, and this needs a little bit more than just, like, switching on again, uh, the machine and, and having then some output coming in. You need to measure it, the success at the end. So how do, would you do it? With attribution modeling. And it's nothing that is just, like, implemented easily. You need to, you need to integrate it in, in all the campaigns and to have really bet- uh, better understanding how to utilize it, and then also to, to measure the success and, and interpret it accordingly that it makes sense then for this individual, for this individual brand. And this also needs some dedication and time to it. It's not like that you use programmatic advertising, it, it will have all the answers that you are expecting. It's more about, like, having, having also the setup in place that this is working, otherwise, otherwise it will fail. To give, like, a real example to it, um, if you're buying Google Ads, the users are in the mindset of, like, looking for something very specific. So the likelihood that something is being clicked and then, let's say, bought, when purchased, it's very high. So it's a last-click attribution. Whereas programmatic advertising, it's more like a lean back situation. So you know your audience Through different identifiers and targetings that you are then tackling or using through the DSP. And, and then at some point, maybe you consciously or not, or unconsciously as, as an end con- consumer see the ad or don't see the ad, but you have a feeling that something is tackling and then it's being marked. And along the customer journey, through these identifiers, there is an, there is an ad coming up on a screen on the street or on CTV, on a screen at home. And then, um, and people do not buy because they see this ad then, but then at some point they go to the website, but they need this proper tracking in order to really follow this entire customer journey. And then it becomes quite tricky because if you think, like, just having a click onto it, it's not really working. It's also the post-view conversions, and this needs to be, this needs, this understanding needs to be well, needs to be there in order to make it a successful programmatic advertising because it's proven that it will contribute to your, to the entire success, but you need to measure it.

Speaker 1: So many things are changing right now. Everything is going fast.

Speaker 2: Oh, yeah.

Speaker 1: Um, what are you seeing in the market now that sort of tells you where this is heading next or that you're really excited about?

Speaker 2: Yes. So I'm super excited about AI. Um, I, I, my background, I'm not a, I'm not a programmer. I'm not a developer. I, um, I f- I'm... I love new things, new evolutions in my industry, and I embrace it. Every change that comes with it is, um, is for me adventurous, and I just wanna try it out, and this is the same with AI. Um, I think there will be a big change on how to trade generally. There are initiatives with all the DSPs out there. Some claim that they are there yet already, and, uh, it's being done through AI. I think it's, it's easy to frame AI. I think there were algorithms already long time in place. It's just the way how it's being used is now differently. But it's not there yet that you just allocate budget and then it's working. It's just, it would be just too simple. Uh, starting from liabilities that are in place. So it, this, this will take time, but trading itself will be automated, and then it's more about, a lot about relationship management with all these vendors that they are out there, to find the best one, to get the best deals for the clients, to understand, like, which one is helps you to achieve the goals rather than, like, just using the time of trading. I'm not sure if this is by the end of the year, um, but the next five years I think we will have a solution for it.

Speaker 1: Christoph, it's been an absolute pleasure talking to you today. Thank you for joining us-

Speaker 2: Thank you, Tristan

Speaker 1: ... on the Business Marketing.

Speaker 2: Yes.