Between Apple’s policy changes and growing government regulations, privacy concerns have sparked a significant shift in AdTech strategies. This episode investigates the potential for privacy-focused ads, exploring how companies are adapting to new regulations while still striving for effective targeting. Host Devin Becker is joined by Geeshan Willink, CEO of Nefta, to dig into the complexities of maintaining anonymity and its implications for the gaming industry.

Geeshan shares insights on Nefta's journey from its web3 roots through the struggles with building ad technology that meets growing privacy standards. We explore the balancing act between precision targeting and user privacy, comparing anonymous user profiling with traditional fingerprinting methods. Willink offers examples of how behavioral-based identification can lead to effective ad targeting without compromising user data and discusses the potential impact on player retention and ad inventory pricing. The discussion also touches on the challenges of implementing these new approaches in an entrenched industry, recent trends in AdTech, and predictions for the future of ad-supported gaming.

Lightspeed gaming

We’d also like to thank Lightspeed Venture Partners for making this episode possible! With its dedicated gaming practice, "Lightspeed Gaming," the firm is investing from over $7B in early- and growth-stage capital — the by far largest fund focused on gaming and interactive technology. If you’re interested in learning more, go to

This transcript is machine-generated, and we apologize for any errors.

Devin: Hello everyone. I'm your host, Devin Becker. And today I'm delighted to be joined by Geeshan Willink, CEO of Nefta. And Nefta is a privacy focused ad tech provider among other things it sounds like.

And today we're going to be talking about the future of games and ads and Sort of all this craziness around that and privacy. But just to like familiarize listeners, could you just quickly go over your background, the origin of Nefta and what you guys do today?

Geeshan: Yeah, absolutely. Hey, Devin, great to be on the show.

A lot to unpack there. I'm going to start with myself, my background. Originally I'm English Dutch, grew up in England to parents from the Netherlands has spent the last 12 years in business. First, the. I'm in consulting at banks and then i moved into startups and move to Barcelona for three beautiful years i built an early stage company there learn all about the tech and business sides of building a company and move to Berlin and co-founded my first company.

And FX trading was very technical built out all of the systems, the technology systems there sold the company in 2019 to a French company called Iband first, and then built another company in bond trading and machine learning. Combining a lot of data in to help insurance companies improve their bond strategies and that we merged with a company in 2021.

Yes. Moving on to the more important stuff, the more fun stuff Nefta. What is the origin of Nefta? What does it do today? We are trying to disrupt the ad industry. We think that gaming companies have been hugely underserved in the. In the ad tech market. And we're trying to do that with both our ad network and our pre mediation solutions to primarily, of course, deliver superior revenue performance in ads, as well as to rebuild some of this trust that has been lost over time in ad tech.

So Nefta started as an idea for a game. Basically the combination of monopoly pokemon go with a collectible cards of real world points of interest and those cards will be on the blockchain. So trade tradable and i did this with my brother who's been in media and gaming for more than a decade and i built out all of these technology systems and saw the potential to pivot into a sass model.

As my background is much better suited to working with business customers. And not on the consumer side. Pivoted that had all of the tech built, acquired early adopters, brought together a fantastic team and raised 1 million from Picos capital. A big shout out there in February, March, 2022.

And basically as part of that founding team, we hired someone from outfit seven who was instrumental in building all of the incredible ad tech there, which is the in house mediation service and their bidder, a lot more to unpack in these terms, et cetera. And so we came to market with this SaaS infrastructure platform, but ad tech was in our DNA and it was always the strategy to figure out how ad tech would would work.

So we scaled that web three infrastructure platform. We built the ad tech piece. We raised our seed round of 5 million led by Play Ventures. And they put their largest ticket from their play future fund. They're a fantastic team. Big shout out there. Also in the end that round was oversubscribed, but goodness, it didn't feel like it was oversubscribed going through it talk to a hundred investors over a three month period.

It was a real grind but we came out. And we we have the resources to have a great chance at actually disrupting this industry a little bit.

Devin: Cool. Yeah, definitely a lot there. A long history that definitely sounds like the destination is much different from the origin. I guess the question then is, and you touched on this a little bit, but I want to find out if there's a little more to unpack on this is what relationship, if any, there is.

Either now or previously between what you guys were doing in web three and the ad tech stuff outside of just the ad tech being in your DNA.

Geeshan: So we built all of this web3 infrastructure stuff. The idea is to profile wallets on chain and look at the value that they have on chain. But what we saw in the web3 space was that it needs scale.

Web3 doesn't have scale. And where adtech has really thrived is in mobile and and using that wallet identifier to target users on mobile is not in compliance with the privacy frameworks. The scale element and the privacy element were the reasons why we. Pivoted away from web three ad tech, but then we got really into the privacy angle because that's what has caused so much issue in the ad tech space.

We hired expensive lawyers. We dug ourselves into this privacy angle. And what we came to conclusion is that the solution is first party data. The issue is that going from zero to one for gaming companies in first party data to actually increase ad monetization is extremely difficult.

You have to build all of this ad tech infra yourself or do an ad or ask an ad network to do that for you. So very difficult.

Devin: Can you explain what you mean by first party data and how that works in terms of like in relation to ads and data from users?

Geeshan: What we've done is we've standardized the way first party data is structured.

What that means is we look at all of the interactions that players have within games. We collect all of that data. We process it in a privacy compliant way. We'll go into more details there. And use that data inside real time machine learning pipelines and inference to modify the advertising bid price for a specific ad opportunity impression.

And basically, the idea is that we correlate the value of the user with the value of the advertising bid. With all of this in game behavioral data directly coming from the publisher.

Devin: So when you say value of the user, do you mean general value or value? Maybe to that specific advertiser, because obviously they may be advertising to different types of players and things like that.

Is that part of that? Or is it just value in terms of like, this is a player that spends money, things like that.

Geeshan: No, it's exactly like you described. The, all of that behavioral data is combined with advertising data and the algorithms. Correlate which type of behavioral archetypes or behavioral characteristics perform well with different types of ad creatives.

Certain users, certain cohorts will perform better with different types of ads and the algorithms will automatically improve click through rates with automatic training, retraining. Exactly. That's a, that's how the system works.

Devin: The obvious question, is advertising these days, especially post IDFA is really about targeting users.

So you're not spending money on people that just aren't your right customer. And that's generally what is a good idea for games to be doing. Cause they have limited budgets to try and acquire users and acquiring users that are just going to turn anyways, pointless. So the question is really why is taking this approach where you're Technically leaving yourself with a lot less information a better balance with privacy versus trying to actually learn more about the user and directly targeting people.

The Facebook style of things where they're very specific down to, age, location, all that stuff.

Geeshan: I think the key differentiating factor that you need to be aware of is is PII personable or personally identifiable information. And of course, Facebook have a huge amount of PII, name, email address, age, all of this you can use to identify a person.

What we are doing and what we believe is the future of balancing and tracking and targeting with privacy is using this in game behavioral data without collecting any email address, IP address, no IDFA. And just using only that data to track users. The big thing with AT& T, the Apple privacy framework is that you can't track users.

And the thing is the way users are defined. So users are defined in certain ways, but as long as you don't collect PII, you're able to track users and correlate with advertising value.

Devin: Can you give an example then of how behavioral. Actions lead to ad targeting, like what kinds of behaviors, how does that actually correlate at least at a high level to some kind of ad being targeted in a way based off your guys system or even just the theory in general.

Geeshan: What we ask from game publishers is to look at their games and we help with this to send us the data that differentiates the user player trajectories. And like I said earlier, we've standardized these game interactions, these game events, and basically publishers will look at these game events and send us this data and find as many data points that they can think about to help the models differentiate the users.

And what I mean by this is some users, they spend a lot of time completing levels, some focus on their inventory, some like to focus on cosmetic items, Some fail, many levels and so on and so forth. So using this standardization of in game behavioral interactions, the models are able to differentiate the users into certain clusters, correlate all of that data with advertising bids and improve click through rates. That's how how it works.

Devin: Is there any potential for the same sort of strategy to be used in with promoting in app purchases that are like targeted to the user based off their behavior? So obviously that's something that games try and do already, but they don't maybe have the same kind of technology or models to be able to understand if you're already consuming that data to try and understand what players are doing and what their, maybe intentions might be in the future.

Is there an opportunity there to promote those items?

Geeshan: Yeah, absolutely. And this kind of optimization has normally only been possible to the largest publishers out there who have built, large analytics teams and also add ad tech teams where they can. Optimize the user experience and user journey by showing less ads or less disruptive ads or optimizing for specific IAP purchases.

This is something which we think about, we talk about a lot internally. This to us is at a later stage the. The ad tech piece and launching the network, launching the mediation. All of that is a huge challenge in today's market. And we will focus primarily on that at the beginning and then these kinds of services we can add later on, but it's certainly in our.

Wheelhouse with the team that we have. And the knowledge to, to launch something like that on top of the solutions we have.

Devin: Then I guess the question that I have is too, is when you're talking about building up all this starting from zero with different games and things like that, are there kinds of games that this happens to suit better for a couple different aspects?

One is being able to collect player data how long people are playing those kinds of things, the different sort of nuances of data that might be coming from different types of games, as well as games that use ads more than others. For example, a hypercasual game is going to be very ad dependent, but at the same time, also not a lot of player like behavior differentiation or purchasing habits in, in those kinds of things.

What kind of games ideally or. Types of products, does this sort of solution suit?

Geeshan: It's very true that, the largest types of games that are driven by ads are more, these bite sized hypercasual games. And they have, more linear player trajectories and player progress paths that has, that is changing considering the privacy changes.

And this term hybrid casual has come about basically. Making these games that are just for advertising also available for IAP. And this is really where things start to become much better when differentiating users the more complexity and more opportunity to veer away from your standard path, the better the models will be.

Fundamentally, it is the volume depths and variability of events that are sent to us. That enable the models to differentiate the user trajectories. And lastly part of the privacy is that we build models per app and per publisher. And so it's not a question of trying to compare apples with pears.

We only have. One model per app or per publisher. And so the models will find the variability within all of that data that you send and not try to look at all of the different genres. So it's. Game by game with the model. And so that's how we improve accuracy improve performance on a model by model basis If you understand what i'm trying to say,

Devin: Yeah, definitely. I mean, would you say overall then player retention is something that improves accuracy over time so that like Better ads can even be served if a player, you retain them, but they don't spend money on ip But you might be able to serve them better ads the longer that you retain them in the game because of those deviations You mentioned

Geeshan: This is a publisher's dream.

A publisher's dream is to be able to have more control over what ads and how disruptive the ads are within specific placements. As a publisher, you don't have a lot of control over this. So this is something, again, we are thinking about and exploring to allow publishers to modify what is shown in specific placements for specific users.

Perhaps a high potential user doesn't or shouldn't be interrupted with an interstitial ad. But they should just continue on their user journey, but publishers don't have that level of control yet. Which is another area which we are exploring.

Devin: One of the things that I think is going to be the question in a lot of people's minds is how does this sort of thing compare in terms of pricing, ad inventory, all that kind of stuff?

Obviously you guys are starting a little bit from the ground up, but at the same time, like, how does this work compared to traditional style of advertising and bidding and things like that, especially on the ad inventory side, because you're targeting users differently.

Geeshan: At first it's cost per click and cost per install.

And these installs are attributed via MMP providers. How they attribute installs, the types of data that they collect. That is their business. We receive the data from them and that is how existing ad tech players are able to do post install event optimization campaigns. Campaigns and so on. And so we will start with cost per click and cost per install and move into using that externally attributed install data to post install event optimization and ROAS campaigns.

That is how we will be pricing it. Another key component here is in margin. Traditionally in the ad tech space margin is not transparent. Ad tech providers adjust margin at the beginning, they have a very low margin and then they increase margin after the trial period or proof that things work is over.

You have large incentives that are played, paid to publishers SDK custom deals, et cetera, et cetera. Which is also part of the reason. There is a lot of distrust in this place and the space we are trying to do something different. What we are doing is we're introducing a fixed margin of 15%.

And that is less than what I would call the average of about 30%. But we are also introducing a margin reduction system, which enables the margin to be reduced in a transparent way. Down to 5 percent based on the volume and contribution to the ad network that a specific publisher brings. So yeah, we're trying to do something quite different here when it comes to our actual margin and it being transparent.

Devin: And how does the scale affect all this in terms of when you're talking about like a game with a very few amount of players or a game with a huge, massive amount of players, obviously, for example, that might provide more training data, but also might provide like a variety of different ad targets.

How does scale affect all this concept outside of just simply more training data?

Geeshan: Of course, things work better with more data, the models have more nuance to find those nuggets of players that perform well with specific creatives. It does work on small games as well. If you lean into the data that you send, like I mentioned earlier, that volume depth and variability the models can still improve ECPMs at small scale, but it's going to, because of the.

Model the model per app or per publisher but of course it's going to work with larger titles and have a higher impact.

Devin: So are you tracking anything on this stuff to verify if they go and they actually click through the ads and they do some kind of post action and actually, the ad essentially works.

Are you first off tracking that in some way to know the effectiveness of the ads in the system? And then secondly, feeding that back into the trading data at all to be like, Hey, here's positive feedback on that. We verified that you guessed correctly with the model, basically.

Geeshan: So we are not doing any of that attribution ourselves.

We are relying on third party attribution providers to tell us which campaigns have been able to acquire which users. And like I mentioned earlier, we will be introducing the post install event campaigns optimizing for a specific in game event and introducing the ROAS campaigns which combines.

Attributed data with the internal data.

Devin: Is there a risk at all with some of this, if you have say low ad inventory potentially skewing the data where, people are being shown the same ad over and over, and that might affect what the rates look like through that attribution and skew the model further where it's like, Oh, it thinks that ad's effective because it's just the only one being shown to that user.

Those kinds of things, those kinds of problems I imagine with low ad inventory at the beginning.

Geeshan: It touches upon, the pacing algorithms that are used inside the bidder. And basically there are levers which we control, which determine how often a user sees the same ad. What is the ad fatigue?

And that's something which we we control and we can, you can see when you launch a campaign the point at which ad fatigue comes into play and the effectiveness of that campaign. Starts to reduce.

Devin: Is there a specific type of ads that this is specializing in terms of like interstitials or rewarded and also playable versus non playable, any particular types that it's specializing in?

Geeshan: Rewarded video and playable ads being, the most engaging ads and the best performing ads out there. And we work with all of these ad formats as well. I haven't got enough concrete data to tell you. The system works the best with banner ads.

Devin: Overall, like now that you guys have been doing this for a little bit of time what have been the biggest challenges to really growing this idea and concept?

Because, as far as I know, there aren't a ton of people doing this just yet and you're entering a mature industry. Like what have been the biggest challenges overall that you've run into?

Geeshan: One of the key questions I get asked is why doesn't some of the big ad tech players out there do this?

Why don't they. Do what we've done. And it's a few things I think they've tried. It's very difficult. I underestimated how difficult it is to build such scalable and complex systems. So from a technology point of view, there is a lot of challenges from a business trust point of view. There is a lot of challenges as well.

I think there's a huge amount of distrust and dissatisfaction with large ad tech providers out there at the moment, but there are very few alternative solutions. Launching a marketplace business in itself is very difficult. You need buyers and sellers, publishers, and advertisers at the same time. And because the market is dominated by big players, we are being integrated through third party mediation.

And there is there's difficulties in that also, because you can only bid at fixed ECBM prices and you don't have access to fair ad auction dynamics, key areas. We've faced in launching this thing over the last six months, supply demand being integrated into other marketplaces, which don't give the possibility to bid high for high users and low for low users.

And of course that tech complexity. And so this has been a big struggle to launch this. But we're in a very exciting position now having faced all of these challenges. Where we are integrating directly into some of these large publishers into their in house mediation. And we're also launching our own SDK mediation, which is following the same principles of.

Trust and transparency and add revenue optimization. Exactly. So we went through a tough six months launching this, but now we are in a much better place with the next phase in go to market also in that time. Just quick point here. Is that we've we've done some case studies internally using internal data coming from publishers.

And we don't want to go to market and say, Hey, look at this case study because it's data that we have internally. And, there's not a lot of trust in ad tech. So people are going to be like, yeah, you just manipulate, manipulated the data. But what we have noticed is that we can improve eCPMs by about five to 10, 11%.

Using this behavioral data on iOS opt outs. So applying that at scale when we have, good feels then there is significant revenue to be had using the solution. And we are just building up to. Having all of that all of those pieces come together to actually deliver on that value.

Devin: You mentioned the opt outs thing. Is there anything you guys do when it comes to working with people that do opt in, or let's say the case of a game developer happening to have their own amount of data from some other source, right? Are they able to then combine that with your data to enrich it themselves?

Or you guys able to use that, like to leverage basically other sources of information that the developer or publisher might have? The, they're dealing with sort of the privacy implications of that on their own side.

Geeshan: If it comes from the publisher, we could use it because it's first party.

We can't link any of that data to any other source. We can't have any universal identifiers. Those two go hand in hand. So we could use it, but imagine working with, 50, a hundred publishers and collecting unique data from each one This is why the importance of standardization in the collection of that first party data is so key.

Devin: With you guys being in the trenches of this stuff and having seen like a, everyone kind of scrambling a bit to try and figure out what to do, what kind of other solutions you said, you've seen some other alternatives, what kind of other alternatives have you seen people trying outside of just hope for the best?

There's gotta be other tech solutions other approaches that people are taking.

Geeshan: Yeah, there's a couple of things. One is people are going to different ad formats, for example, offer wool. This is another ad format that people are experimenting with. The influencer marketing route has been Has been taking off, so there is a push for finding alternative solutions to this privacy induced problems.

Some are successful, some aren't, there is an interest to do that. There are also tech solutions out there that are trying to improve the actual ad tech space. It's just so hard. Launching a mediation launching an ad network, launching those things combined is just really difficult and require very specialized knowledge.

A fair amount of funding and a lot of determination and fighting uphills, so to say, but there are a couple of companies that are doing transparent mediation. Which have caught my eye and some other companies experimenting with alternative ways to acquire users at scale.

Devin: And when you mentioned like regulations and all that stuff earlier and involving lawyer, are you mostly talking about platform level regulations, country level regulations, like how are you looking at this from a, like a compliance standpoint because obviously these things can vary, right?

Geeshan: Yep, absolutely. The key thing to note is. The apple ATT framework like I mentioned, you're not supposed to track users, but how do you define users? It is this PII. And so as long as you don't connect PII, you don't try to replace the universal identifier. You don't mix the data between developers or link it with any third party data, then you are compliant with Apple's ATT framework.

Now, there's been the introduction of TCF 2. 2, especially in Europe with the e privacy directive. And there is a small caveat here. That if a user opts out of TCF 2.2, which is a small percentage, because the opt out rates for TCF 2.2 are very small, like five to 10% for example, then nobody can do anything on the European users that opt out of TCF 2.2 because.

No identifier can be stored on the device without that consent. This happens in on web with cookies nowadays. It still happens and it happens across mobile in all spaces. It's just that it's very difficult to regulate, but the regulators are coming and we've seen we've seen some cases already.

So there is a small caveat in opt outs where if you opt out of TCF 2. 2 in Europe, then nobody can do anything with those users. But we are focusing on iOS opt outs because we are compliant with the ATT framework. And that is the area which we can improve ad revenue the most. And now to your original question, we are not just focusing on iOS opt outs.

We are collecting this in game behavioral data from all users. It is just that we can improve the revenue the most for that cohort for that segment exactly.

Devin: If you're already doing machine learning or whatever form of modeling you're doing for these users when they're working is there any way to then also leverage this for stuff, things besides ads for the game developers?

Obviously you're standardizing the inputs, but the outputs could be, whatever in terms of looking at other things like behavior around like the first time user experience, for example, where being able to help, facilitate better improvements there. And things that aren't just directly related to ads, is there any, I mean, obviously AI is a big topic right now.

So everyone's looking for opportunities in this space. Is that something that if you're already building that technology, as you said, it was difficult to build up all this infrastructure. Is that something where there's some opportunity outside of just straight ads or even just around, like I was saying earlier around in app purchase offers and things like that to just better improve games overall outside of just showing better ads.

Geeshan: One thing which we are doing is building our churn and retention models. This. Reduces the training period for the models to be better at bidding in the advertising bid and these churn and retention models can be used for other solutions. If you think a user is high, highly likely to churn within that, those first few sessions.

Is there ways to modify the game to improve that, those, to improve the likelihood that user stays that the retains that the retain to be able to do that, to go from zero to one with that is quite difficult. And it also means you need to control, firstly, the ads because ads are part of that monetization strategy and the user experience at the beginning.

So how. How quickly do you start showing ads? And is that driven by data? How quickly do you start offering IAP opportunities? Is there perhaps an offer that you can give to the user if they are likely, if they are perceived to be likely to churn? These are all questions which game developers are asking.

There are some solutions out there to combat this, but also publishers need to be able To control the ad stack and be data driven enough to use third party solutions to to to control that user experience. So for us, Like I said, we've explored this, we will probably do something in that space, but not for in the near future.

And I guess there are a few platforms out there that game studios can look at to to improve that game design. Just wanted to add another point here in, in, in the way the models work. Right now we're collecting all of this behavioral data using and correlating that with ad creative data to improve click through rates.

Where we get really excited is more in player trajectories and sequence embeddings, similar to how these large language models work. When you hear they say, yeah, this model is 70 billion parameters, 400 billion parameters. I don't know how many GPUs that uses. Basically, this is the same idea where we take all of these events and put them into these large matrixes, if that's the right way to pronounce that word and look at the sequence of events and correlate the sequence of events with the ad creatives.

And I think that is, an area which will hugely or no, it will be the next iteration in improving ECPMs from our side is actually improving the modeling techniques based upon those large language models.

Devin: Is there any chance then to also feed this? I mean, We're talking about the game developer, but also towards add creatives and other things around that advertising space when it comes to it.

Filling that ad inventory when you're, I imagine talking to advertisers, is there also a way then you can use what you're doing to also help improve the quality or the, obviously you're improving the targeting, but just anything you do to improve the ad space in general on the ad side, on the creative side.

Geeshan: I think there is a lot of dissatisfaction again with the quality of ads having, many end cards, the X button is not being shown very long videos tap on off until the ad is finally gone. So the actual user experience is pretty poor. There are a lot of complaints. There is something you can do there when it comes to configuring ad placements from the publisher side to give them control over what kind of videos are shown, how long they are how many end cards are shown.

So there is certain controls that can be given to publishers from that angle. That's certainly something we would we would like to do from an advertiser's perspective, getting the data into which creatives are performing well in which jurisdictions, in which segments. That is key in iterating with creatives.

And I guess I touched upon a point that iteration of creatives and there's been discussions, on generative AI and advertising creatives. What is the way those two things work hand in hand? And does it increase that iteration in creatives creative production? I think yes. And over time, as these models get better, it will be creative.

Teams will be better assisted to produce larger amount of creatives and let these models loose on comparing a larger number of creatives more variations of creatives with all of the other data to. Squeeze the ECPMs and squeeze the performance in a from a new perspective exactly.

Devin: Cool. I guess just in general, then you were talking about a little bit right there is just looking towards the future of the space. Obviously there's a bunch of different solutions being tried. I imagine you have a lot of faith in your particular solution, but just overall in the space out, assuming privacy doesn't go back the other direction.

What do you kind of see as the future of ad supported gaming? And this could be outside of just what you're doing specifically, but I mean, different business models, we've got subscription models like the Xbox game pass, you've got the apple arcade and Google plays one as well. There's all these different ways to approach monetization of gaming.

And we're kind of at a tough spot in the industry. So having, especially having pivoted a bit, you guys. Yourselves, what do you kind of see as the future? And is there any relation to web three in that or any other, the sort of technologies obviously AI as well.

Geeshan: I think, and here, a much larger shift to IAP.

So ads that are only monetizing through, sorry, games that are only monetizing through ads are struggling clearly. So they are introducing IAP and, testing whether or not IAP is viable with that user base. To diversify away from just purely ads. I think there is a large dissatisfaction with from gaming companies, with large ad tech providers who are, you mentioned it, publishers, game developers are struggling posting their worst quarters yet, whereas some of these large ad tech providers are posting quarter over quarter growth.

And so to me, that is a. A trend which won't last and some, at some point it will break. And I think so there is certainly a need in the attic space for subscriptions for console to be honest it's difficult for me to say I'm so focused on the ad tech space. There are certainly other people who are better equipped to answer that question.

Devin: You did actually think of a bit of a thing I don't think we completely addressed, which is obviously there's a lot of non personal information that you're getting, right? So one of the things that I wonder is about you definitely want to kind of understand Maybe some idea of the demographic in terms of like where the user's located.

Is there still some way to target the ads based off geolocation to some extent obviously, a tier three country and a tier one country might have pretty different ads in terms of what you're providing, but the gameplay might be exactly the same.

Geeshan: Yeah, absolutely.

And this touches upon the IP address component. And the storage of the IP address and whether or not that is PII which is yeah, it's it's certainly a topic for discussion. What we said is we are not going to. Expose ourselves to any of that peace. And therefore we don't collect IP address on any account.

But what you need to be able to do is differentiate the country. So we use a service to. Determine the country from the IP address and just store the country, and then we don't save the IP address. And this is critical to have that country level information, because like you said, you're going to have hugely different ECPMs in a tier one versus a tier three country.

And our focus is. Tier one iOS, and that's where we can improve the eCPMs the most but there is a whole debate on the IP address component and we've we've said we're not going to be exposed to that. It's also, working with great gaming companies easier to go through the legal side of what we're doing.

When we say, we are not doing, we don't do any of that stuff and it's in the contracts.

Devin: If that situation changes, say the laws change or an apple changes its stance for some reason goes back the other direction Are you guys willing to incorporate more data that whatever basically whatever is normally allowed and seen as legal and abiding by that to enrich the data or are you going to just Stick with the kind of data that you have because you're already building up around those models

Geeshan: well, our opinion is that this is The right way from the letter and spirit of these privacy laws is you are not trying to identify a user, you are just grouping users into segments and cohorts and understanding what those users intent is.

So we believe that this first party data, not IDFA, et cetera, et cetera, will be the. The best way and the long way forward for targeting in ad tech. And you'll see this across multiple industries where, the Revolut, the Neo bank is now an ad tech provider where PayPal have said they are now an ad tech provider.

And there are lots of instances where companies are realizing that first party data is. The key and is the value to unlock amazon is another one and of course these great companies they sit upon a wealth of first party data and what we are trying to do is. Enable publishers to leverage that first party data.

But not need to do it all themselves. That's a really key point.

Devin: Is there a particular size of company that you're generally targeting? Like maybe the double a size or something in that space where, they don't have enough resources to do this kind of thing themselves, but they're big enough to need it.

Geeshan: Yeah, it's a, it's an interesting question. Certainly, as we've been launching this thing, working with smaller to midsize studios that have some. Risk appetite to try new innovative solutions. That's how we've been going. But it's moved up the chain to be talking and working with some very large reputable gaming companies.

Even the largest publishers out there struggle, I think, to really leverage all of that data. And I say this because. I know King, for example, have a huge analytics department and I have been squeezing every percentage out of candy crush for the last X amount of years. And I've done extremely well there.

But I think, the companies that are able to actually pull that level of optimization off and that it makes sense at that scale are few. And so I think that there is a large. Area of the market that is underserved in having the technical and so on ability to be able to properly leverage.

All of that data that is generated from in game.

Devin: Is there any case studies or games like live right now using this kind of thing where people could see some of this in action or at least read the case studies and see how it works just because right now it's a little abstract and I think it would help people to get some concrete details.

Geeshan: Yeah, no, absolutely. And this is, really what we've been working towards is releasing Business value case studies that have publisher name, real percentages and that are not using our data, but data that is from the publisher. So it's verifiable. And that yeah people can see that it really works.

It goes back to the issues that I've described earlier. In launching the network in acquiring the scale to have statistical significance to to really show how much this works at a large scale for a specific or for multiple publishers. We've got this technical case study internally, and we will release these case studies.

We will release, hope to release one very good one in the next month or so. Two months i would say and that is because we are able to integrate through a large publishers in house mediation and not go through these other mediation providers which is where we can really bid high and bid low and we have access to these.

Fair auction dynamics where the auctions aren't controlled by by other people. They are just fair. So we're working on it and we're getting there. We've proven internally. We're very committed. We know that first party data. Is the future in the longterm and we will continue iterating until to improve eCPMs the most using this data over time.

That's just a couple of thoughts.

Devin: Yeah. And where would be the best place for people to find you're talking about these potential future case studies, things like that, would that just be on the website or is there going to be a good place where people could just be ready to get this information,

Geeshan: LinkedIn's the best place follow the company.

Interact with the team. That's definitely the best place. Of course it will be posted to the website, but LinkedIn is the best.

Devin: Cool. We definitely look forward to seeing that. I would like to see a lot of this action myself just to get an idea because obviously it all sounds great, but it's something that I think we need to kind of see how it works over time, especially with it being as nebulous as learning and modeling are all the rage these days.

So hopefully it works out, but I definitely would love to check it out when you guys have that live, but I did want to thank you for coming on today and talking about it. I think it's going to be a very interesting space going forward. Obviously, a lot of problems needing solved and you guys are definitely taking a big crack at it here.

So hopefully that'll work out, but thanks for coming and explaining everything. And if you want to plug the uh, the website, I think it was, was it?

Geeshan: Yep. That's right.

Devin: Cool. So definitely make sure to check that out. And you guys are doing a lot of cool stuff. So who knows, maybe even more. In the near future, as you guys tend to do a lot of cool different things.

Geeshan: Yeah, fantastic. Devin really good to be on the show and looking forward to catching up again soon.

Devin: Awesome. And thanks everyone for listening. I hope you uh, had a good time, learned some new stuff and definitely make sure to check out What Nefta are up to at It's N E F T A dot I O.

So you don't confuse it with nefta, as you mentioned earlier. But thanks again, both Geeshan and you listeners and we'll catch you next time.

If you enjoyed today's episode, whether on YouTube or your favorite podcast app, make sure to like, subscribe, comment, or give a five-star review. And if you wanna reach out or provide feedback, shoot us a note at [email protected] or find us on Twitter and LinkedIn. Plus, if you wanna learn more about what Naavik has to offer, make sure to check out our website there. You can sign up for the number one games industry newsletter, Naavik Digest, or contact us to learn about our wide-ranging consulting and advisory services.

Again, that is Thanks for listening and we'll catch you in the next episode.