Can We Still Trust Attribution? Brian Quinn on AppsFlyer, Privacy, and Measurement
- Jul 9
- 53 min read
Attribution was supposed to get easier. Instead, marketers today operate in a world of disappearing identifiers, privacy restrictions, walled gardens that increasingly measure themselves, and consumer journeys that unfold inside environments few advertisers can fully observe.
So how much confidence should we still have in the numbers? In this episode of Signal & Noise, Brett House and Rio Longacre sit down with Brian Quinn, President & GM of AppsFlyer, one of the most influential—but often least understood—companies in digital advertising.
For more than a decade, AppsFlyer has been at the center of mobile measurement, helping brands understand where users come from, which campaigns actually drive installs and engagement, and how to measure effectiveness in an ecosystem transformed by Apple’s ATT framework, signal loss, and evolving privacy regulations.
Together, we unpack one of the industry’s most important questions: Can attribution still be trusted?
We also discuss:
• Why AppsFlyer became much more than an attribution company
• Whether the walled gardens are effectively grading their own homework
• The lasting impact of Apple’s App Tracking Transparency changes
• Why fraud remains a massive hidden tax on digital advertising
• The growing role of clean rooms and independent measurement
• Whether privacy has made advertising measurement better—or simply more uncertain
• How AI may reshape attribution and optimization over the next five years
• What a modern measurement strategy should look like in a world where signal loss may be permanent
Apps account for nearly 90% of mobile time, consumers spend hours each day inside apps, and marketers collectively spend more than $150 billion annually on mobile advertising. Yet many advertisers still struggle to explain how mobile attribution actually works—or what replaces it when deterministic signals disappear.
Brian offers a candid and practical perspective on where measurement is heading next, what marketers misunderstand about attribution, and what parts of today’s measurement stack may disappear entirely.
If you’ve ever wondered whether the industry’s most trusted metrics deserve your trust—or what comes after attribution—this episode is for you.
Watch the full episode and join the conversation.
🔑 What We Cover💡 Key Takeaways🎯 Why This Episode Matters
Read the full transcript below.
Brett House (00:02.45)
Welcome back to Signal and Noise. I'm Brett House, and I'm joined by my co-host and one of my favorite people on the planet, Rio Longacre. today's guest is Brian Quinn, President and General Manager of Apps Flyer. Welcome, Brian. Thrilled to have you on the show. for those of you that don't know Apps Flyer, it's one of the most influential companies in mobile marketing, in mobile measurement, and general measurement. though many people outside of the mobile ecosystem may not fully appreciate the roles it plays.
Brian. (00:14.24)
Thank you guys.
Brett House (00:30.072)
And we're going to use this episode to really dig into that, to dig into Brian's knowledge. You've been around in the industry for quite a while. I know you guys are not positioning yourselves as the mobile marketing cloud and the measurement foundation for AI ready data, spending mobile, CTV, PC, console, basically data collaboration. There's even some fraud prevention solutions that you guys have have brought into the mix, which I find super interesting, having come from New Star and TransUnion.
That was certainly a place that we played in the identity ecosystem. And you spent a bunch of years both you're on the on the MMA Global board. We just introduced yeah, interviewed Greg Sirt. He's a friend of the pod. I've known him for a number of years. and you were also at Kinshoo, you're a global ki VP at Kinsho. You spent a bunch of time at Xperian and ATT in your earlier the earlier part of your career. And I've heard I've heard you advise a lot of startups, or at least you have in in recent in recent memory.
Rio (01:05.036)
Yeah, Greg Stewart was on recently. Yep.
Brian. (01:06.841)
nice. Great.
Brian. (01:19.542)
That's right.
Brett House (01:26.512)
So thrilled to have you on the pod. I think you're gonna add a lot of interesting perspective when it comes to to measurement in general and mobile and the app ecosystem specifically. and I've seen a lot of your posts recently about really helping marketing leaders fight that battle of being able to speak the language of the CFO, which we've talked about a t a ton. Would love to definitely unpack that and really how you think about it. Yeah. So so
Rio (01:49.422)
Yeah, it's a gap, right? And it's been a a trend to push towards that, yeah.
Brian. (01:52.564)
Yeah, huge gap. Yeah, I would love it.
Brett House (01:54.77)
Yeah, so Brian, welcome to the pod and and would love for you to just to give a brief intro to your to yourself, your trajectory, kind of what what got you excited about the space that you're now sort of leading. And did I miss anything?
Rio (02:03.052)
Yeah, and did House miss anything? What what in your background did he miss? Or
Brian. (02:07.416)
that was a great, great, great intro. Thank you. Thank you guys. Good to be here. yeah. So you know, I think just as way of background, you know, I've been on sort of the business side of tech companies in the media, mobile, performance marketing space for a bit now. So get a chance to sort of work with customers directly across a lot of different, you know, industries and verticals. you know, in the capacity here at Apps Fire, we have some of the more sort of
performance-oriented advertisers who've been in mobile and mobile is their business. The app is the product. But I also get to work with like large enterprises where they're huge household names and they, they, they are in the mobile game, but but it's helping them understand what is the business of an app, what does the app do for the customer journey and for the brand. So it's a pretty big spectrum of of customers that we work with. And I just get a kick out of you know helping to drive the business and helping to kind of move the industry
forward through a lot of the the changes that we've seen, whether it be privacy or regulatory changes or platform shifts and now AI. So it's a very evolving space. I'm having a lot of fun doing it. And yeah, happy to happy to jump into these topics because this is some of the stuff I like talking about the most.
Brett House (03:24.115)
Nice.
Rio (03:24.15)
So well, Brian, one of the reasons we really wanted to have this conversation with you is I think Brett and I both agree that there's kind of a reckoning coming for digital advertising writ large, right? You think about the last 20, 25 years, we've really operated all under this assumption that measurements getting better, attributions getting better. And I think underpinning that was we have more data, more identifiers, more platforms, and of course more sophisticated measurement and attribution tools, right? But I don't know, as you talk to different people and you see what where the rubber meets the road, I'm not really sure that's
So true, right? many ways, even though we're in the age of outcomes, like what's really driving those outcomes, right? Our consumers are spending a lot of their time inside environments that marketers really can't fully observe because of signal loss. And you mentioned privacy laws and some of those privacy changes. And many of the signals we relied on, they're kind of disappearing or maybe even disappeared. Then you look at the walled gardens, like they increasingly gr grade their own homework, right? I mean, can we even trust platform signals or platform
Attribution anymore. And yet at the center of this, this is I think is so wild, like mobile, huge part of the digital it digital economy, right? But kind of flies under the radar. I don't think people appreciate really how big mobile is, how much time people spend inside apps, shopping apps, banking apps, consuming apps, or even play games and apps, right? It's kind of amazing. Yet when people talk about the future of advertising, they tend to focus on AI, agentic, walled gardens, but overlooking the place, people frequently spend.
A lot of their time and I would argue pretty soon, probably most of their time. So we want to dig into these things, look about, you know, how has what's been solved for mobile, what hasn't. And Apps Flyers role in this 'cause do think you're a company with a really cool story that's like like mobile flown a little bit under the radar as Burt mentioned for a lot of people. So welcome.
Brian. (04:54.592)
Yeah.
Brian. (05:06.552)
Mm. Thank you. Yeah, I'm happy to get in. It's it's great to be here.
Brett House (05:10.012)
Yeah, so so considering that Apps Flyer, not everybody knows the the the minutiae of what you guys do. Could you kind of explain to the audience this is gonna be one opportunity to shill a little bit, Bri Brian? Yeah, this is it. This is the one shot. so so your key messages can come out right now in the next two minutes. No, no exactly what like what you guys do and and sort I think the evolution is what I'm sort of interested, like where you got your start and then how it's evolved. Because I mentioned a whole bunch of of services solutions that you guys are currently providing.
Brian. (05:21.418)
This is it. This is the one shot. This is it.
Brian. (05:28.672)
Perfect.
Brian. (05:39.978)
Yeah. Cool. Well, we've been in business for 14 years. And when you do that and you grow, you add a lot of products, capabilities, and features along the way. So now it's sort of a suite of of different services and and capabilities for marketers. But the company was born to solve a problem of mobile attribution. So when the app stores were created and people started downloading apps, the the brands and the developers who built those apps had no visibility into who was downloading them and why.
Brett House (05:40.11)
but yeah, we'd love to hear that.
Brian. (06:07.828)
Right. Traditional web analytics and cookies weren't in the app stores. Apple and Google weren't giving information back to brands. So it was a very sort of no visibility. And so we and a number of other companies started and we created the ability for brands to understand, you know, who's downloading these and installing these apps. And that was the very first use case. And that pr that continues to be a very important problem that we solve. That was how the company got its start.
Brett House (06:08.115)
Yeah.
Brian. (06:38.386)
Once you can connect the media touch point to the download of the install, which is sort of the basis of attribution, then you could add a lot of other capabilities onto it. So then you could look at, well, what happens after I install an app? You know, what do I do in that? What are the what are the actions that I'm taking? And how do I sort of aggregate those actions and better understand the value of those actions? And now I can better understand the value of the media.
And the marketing, whether it's paid media or own media. And so you you can imagine the analytics and the understanding that comes with this. Now I I'm able to effectively understand all of my marketing and what's driving performance. In the mobile space, we hadn't used the word outcomes because it's all outcome. And I kind of find it humorous over that's become like a buzzword in the in the broader ad tech ecosystem.
Because it it hasn't been outcome driven for so long and it's now that's cool. But in mobile, it's it's I think we laugh because it's it's only outcomes, it's only performance that really gets any type of scale. And so this is the business we're in. Brands look at us as infrastructure that they write into their apps and their websites and their CTV apps, and so we
Have SDKs and APIs and server side integrations. And we are all of our revenue comes from advertisers. And so we sort of sit in their stack and we act as their advocate. And when they go out and they deploy capital and they deploy spend on advertising networks, walled gardens, non walled gardens, C T V companies, you name it on the open web, we're the ones that tell them what what's working in their marketing and how they can begin to optimize that. So that's the business that we're in.
Rio (08:29.922)
When looking so Brian when looking at like what's working, so you're tracking like downloads and installs, it sounds like, but also like you mentioned like what they're doing. So this sounds like almost like the the product analytics, what like usage understanding, like like are they actually doing anything with these apps once they're installed? Do you also venture into that?
Brian. (08:48.746)
Yeah, I think, you know, you look at product analytics, you you're you're you're taking a user view and it's it's built for a product manager understanding, okay, once you enter my environment, what's your behaviors and how do I understand you better? A lot of that data also helps understand, okay, what's the lifetime value of a user? And how do I sort of connect that back to my media touch point or my marketing spend? And so a lot of it's the same, maybe data points, but we use it in a different context and it's serving a different purpose. But there's a there's some overlap there in the in the space.
Brett House (09:18.354)
Yeah, and when you say a different kind sorry.
Rio (09:18.606)
Okay. And did you work do you work with sorry, do you work with like tag like tag man like like product analytics platforms and like and actually use the information from their tags, or do you have your own tagging system that you would then like like install on the apps themselves?
Brian. (09:31.37)
Both. We have our own sort of IDs that we set with customers so that they understand what's the apps flyer sort of way of looking at things. But the product analytics companies are great partners. We have 10,000 integrated partners. It's one of the largest B2B partner ecosystems out there. Part of it is just like we're so woven into the fabric of our ecosystem that it's not just media companies that we're measuring that we integrate with, but it's
Rio (09:39.907)
Yeah.
Brian. (09:58.514)
It's product analytics, it's CDPs, it's engagement platforms, it's agentic tools, it's anyone that needs to understand, you know, and and that where we can connect the data for the brand to give them sort of a more holistic picture of of of of what we're working on.
Rio (10:15.336)
So some wonder like w let's say I had Amplitude or Pendo installed, I could use that and then like like share the information with with Apps Light. Okay.
Brian. (10:21.19)
Great for sure. And great great great great partners of ours and and a lot of kind of use cases where at attribution data in enhances the value you get from those products and and product analytics vice versa helps the marketing team. Let me tell you, you know, a simple sort of situation that you'd see. Monday morning, teams come into the office and some type of you know retention metric fall off a cliff.
And so the product manager will look at, well, what did I do? What did my team deploy over the last 24, 48 hours? And let me revert that. Their view will only be about what changed in the product, what onboarding flow changed, you know, what code was pushed, and they'll quickly revert any changes to kind of like reverse that that KPI. The marketer will look at the same data and say, wow, what did we do? What launch did we go? You know, what what what did my team?
Do that, we have to re revert or or you know, and so it's the same problem looked at from two different perspectives with two different tool sets and two different data sets. And the companies aren't talking. They're not, there, there's not a is it a product issue, is it a marketing issue, is it something in between? So by connecting the data sets and integration of these partners, we attempt to provide a more holistic view. So, you know, the the the the business can understand.
actually what went wrong and how to address it and not only solve the problem from the very narrow, you know, view that they have of the of the customer or of the marketing.
Brett House (11:55.795)
Yeah, well let me ask you this, are it like so is it mostly B to B use cases or are you extending into B to C use cases as well from
Brian. (12:04.512)
So we our clients are all businesses. So we're a B2B company. we help attribute you know, consumer touch points. So it's sort of B2B to C in that sense. but you know, our businesses have consumer facing apps and that's the world that we live in.
Brett House (12:07.965)
Yeah.
Brett House (12:15.23)
Yeah.
Brett House (12:19.614)
Yeah. Yeah. And you and you mentioned that that yeah, I just wanted to confirm that. The the you mentioned that that you guys it's all been outcomes, right? So so how when you when you think of outcomes, are you thinking just any sort of KPI that has the ability to be detested t tested or the ability to have some sort of incremental contribution to the to the benef you know to the benefit of the advertiser, right? Like like how do you actually define outcomes? Because, you know, does that is that every aspect of marketing from brand awareness all the way down to
the acquisition play. where do you see that fits?
Brian. (12:52.706)
It this. We we we we sit in the performance marketing arena. And for performance marketer, you don't need to define outcomes. It's it's in inherent in what they're working on, right? So I have a return on advertising spend goal, I have a an LTV goal. I'm trying to drive new user acquisition. I'm trying to re-engage my users. I have very clear KPIs on what I'm looking to do. It could be revenue-based, it could be LTV-based. I might have predictive algorithms that help me.
Brett House (12:56.914)
Yeah.
Rio (13:01.964)
Pretty clear, yeah.
Brett House (13:02.334)
Yeah.
Brett House (13:14.557)
Yeah.
Brian. (13:22.504)
Inform that we can get into attribution and incrementality and modeling. But the reality is I'm trying to drive a a business result and outcomes is sort of like the like I said, it's like the new world, new word, but it's always existed in mobile. And performance marketers don't scale any spend without seeing comparable performance improvements, you know, across different channels. That's just the nature of the of the of the business. And and mobile has sort of grown up that way.
So it's a very performant channel. And if you look at, you know, different industries to kind of like put a fine point on that, you look at like the gaming business where you're almost actually generating revenue immediately or within days of that advertising spend. And so it's it's it's it's the the exactly. And so that the they're very confident with our tools of what kind of spend I'm willing to go to acquire a user.
Brett House (13:51.646)
Yeah. Yeah.
Brett House (14:06.856)
Yeah.
Place your first bet type of advertising, right? Where you're logging in and yeah.
Yeah.
Brian. (14:19.36)
How quickly that user is going to return revenue to me on what, you know, day three or day seven or whatever that might be, to whatever business metric I find important. You know, sports betting apps are going to wanna see, you first deposit made, ride sharing apps are gonna wanna see first ride booked, or the first three rides in a couple of weeks. All these businesses understand what's the key KPI that's gonna signal what's gonna be a long, you know, a high lifetime value user.
Brett House (14:33.609)
Yeah.
Brian. (14:46.762)
And they can optimize on that. And that that's really performance marketing.
Brett House (14:50.088)
Yeah, well, and I was trying to get at the the topic of sort of brand versus performance, because you have a halo effect at the upper funnel that that drives that may not be a you know a buy now type of advertisement, but it has an effect on that that down funnel activity that a consumer might eventually take, right? And so you think of of bit mass reach mediums, whether it's mass reach C T V or Linear Television, radio, to to to kind of place that
Brian. (15:04.721)
for sure.
Brett House (15:17.252)
brand in a person's brain. Now do you guys even talk on those topics or is it really more down funnel type of stuff? I mean do you look at the entire picture? Yeah.
Brian. (15:27.55)
More and more. So C T V is a great a great medium for that, right? It's it's got the promise of brand marketing. It's got a big screen, captive audience, rich media, thirty second kind of slots that we're used to. but it's it's addressable, it's measurable, it it you know, a lot of a lot of the people buying performance C T V buys come from the mobile ecosystem. They understand the analytics, they understand the tooling and
Brett House (15:38.204)
Yep. Yep.
Rio (15:40.878)
But it's addressable, right? Yeah.
Brian. (15:56.778)
Anyway, I see we're seeing performance on TC grow. It it should be growing faster. We can get into that, but it's it is growing. But you know, it's not a bottom of the funnel direct response channel, right? You know, and so it
Brett House (16:07.858)
Yeah.
Rio (16:08.59)
I mean some people like Amazon are trying to turn it into that, right, by putting QR codes or click here. But but you're right, it's generally not. It's generally more of a in above the line. Yep.
Brett House (16:14.441)
Yeah.
Brian. (16:14.73)
There are there are good examples of use cases like that, which are very kind of immediately understand like what was the action of what was the call to action and what was the purchase. There is that's happening, but it's not at scale. Right. And for most most advertisers, what they're seeing is it's it's somewhere in the middle of the funnel. It's a consideration type kind of play. And then our customers are saying, Well, how do we look at this channel? Right. Like we want to open up user acquisition. We need to understand the flow.
Brett House (16:35.656)
Yeah.
Brian. (16:42.207)
So if if I'm sitting in my living room and I see something on TV and I go to a website and I search for that brand and I ultimately end up on the website and make a purchase where I go to an app and download the app, Apps Fire tell me that journey. So we're in that business. We're helping them understand the journey that's originates on a TV and how it how it results in a excuse me, a purchase or an account registration or whatnot and the steps involved there.
Brett House (16:55.656)
Yeah.
Brian. (17:07.68)
But the CTD impression may not have been performant in the sense that it was, you know, higher in the funnel. So the fragmentation and the connecting of the user journey on different platforms, you know, you're you're you're you know, it's sort of moving upper funnel based more on the on the platform and the marketing strategy.
Rio (17:25.464)
So looking at mobile itself, as I brought up earlier, I think that when you look at the s the scope and size of mobile, it's probably like undervalued in some ways. And I remember it was I think it was the end of last year, but I don't remember a push out that article house. It was an article on another company in the mobile space that also starts with app. And I remember like it was in this article I'd brought up, I wasn't really saying there like there's anything sketchy, but I was like I had reported on some people who were talking who had maybe even disparaging them a little bit. And then I remember I got a lot of like people like
Clapping back at saying, hey, that's not actually some of these short sellers are full of it. And one of the comments, which I actually thought was very smart, is from Matt Barash, who's the the C CEO at Nova Studio. He said, he said, listen, say what you want about them. Mobile is so damn big. He said, you go in a bus, look around you, everyone's on their phone. He said, Half those people, their own games. They're gaming, it's huge. The amount of the amount of time people spend, it's undervalued. Any company that's doing well in the space and is big, they probably may maybe take a look at it, you know, maybe maybe they actually are, right? So
Love your thoughts on that. And why do you think for a lot of reasons it is maybe undervalued or underappreciated? And certainly you look at the amount of money spent in mobile relative to to other spaces, I think that it's certainly certainly not as big as it should be. Thoughts?
Brian. (18:39.806)
Yeah, a lot of thoughts. It's it's a it is it's significantly undervalued and underappreciated what's going on in mobile. I mean, since COVID, I think average adults in the US are somewhere between five and a half to six hours a day in apps.
Brett House (18:54.642)
Yeah. It's like s it's like eighty percent of time spent, right? Something like that.
Rio (18:57.794)
That's crazy.
Brian. (18:58.964)
Yeah, it's it's wild. And the and the disparity between you know, advertising in apps and time in apps still has a a lot of of room to grow on behalf of advertisers, right? So there's more dollars that could be deployed there for return than a lot of businesses. They just haven't figured it out yet. There's a lot of reasons for this. I think it's always sort of been put to the side, particularly in the US ad tech market, right? When you say ad tech in the US and
you know, we've we've met at conferences and we talk a lot of that is sort of browser-based thinking, TV and browser-based thinking. very different. Asia markets, China, Japan, Korea, India, Eastern Europe, most of Europe, frankly, it's it's it's mobile first. And you see a lack of innovation and adoption in the US market over the years with relative to mobile, because it was a web-first market and we had the infrastructure and
Rio (19:32.066)
Yep. Whereas India may be totally different, right? Or or someplace like that.
Brett House (19:32.424)
Yeah. Yeah.
Brian. (19:55.764)
That's where, you know, people knew and that's where the vendors were. And, you know, mobile grew up much quieter. Like you said, gaming, you know, probably the first industry to really like like bring mobile to scale. You know, the mobile is the game. These, these, these businesses, these publishers have a number of apps. They are the game. They live and they die by user acquisition. So they really became sort of experts in a very kind of like narrow and silo version, you know, space of of the broader ad tech.
And a lot of things, I think, are the reason here. You know, initially frameworks, Apple and Google's sort of frameworks were separate between mobile and web, and those have collapsed. So the privacy frameworks that leads to product development, right? Because there was mobile identifiers like IDFA and Google ID, they didn't really equate to a cookie, a web cookie, right? They didn't, they didn't connect one to one. So analytics
Brett House (20:35.571)
Yeah.
Brett House (20:46.61)
Yeah.
Brian. (20:50.442)
were different and separate. And the idea of an install on an on an app is very different than a website visit. and so the be the behavior for consumers, while we as a consumer are flying in and out of apps and websites and the the the way that the businesses looked at these was siloed. The vendors were different. Even from a talent perspective, you know, people in the mobile industry have stayed and remained in it and it's a very kind of tight knit organization. And they're they're they're different skill sets. They they know
Or I should say more that sort of they they they know that ecosystem differently than the than the broader ad tech. And it's it's beginning to change, but you know, it blows my mind that I go to conferences and you know it's there's a lot of kind of talk about broader ad tech topics. And I think of myself, all this stuff happened in mobile first. All we were talking about cookie deprecation for a long time in the broader ad tech market, and then Apple made a change and IDF, it it dropped from about 80% to 25% overnight.
Brett House (21:43.282)
Yeah, they
Rio (21:43.628)
Yeah, IDFA, yeah.
Brett House (21:49.16)
Yep. And that's
Brian. (21:49.726)
And the market got it wrong.
Rio (21:49.9)
And Safaris blocked third party cookies for years, right? I mean it's so peop people don't realize that, right? So
Brian. (21:53.757)
Exactly.
Brett House (21:54.033)
And and and that's Apple's ATT their their app tracking transparency framework, right? When they when they introduced that. Yep.
Brian. (21:59.55)
That's exactly right. Yeah. So it you know, it was in COVID. iOS 14 was their operating system. They launched it. It was it just simply switched the setting. It it was gonna it defaulted to the opposite of what was. So rather than brands automatically getting this identifier from the phone, they had to ask permission. Apple was very tight, controlled in, you know, how that that prompt looked. The brand, the app developer couldn't, you know, didn't have lot of room to to to control what was said there.
You could kind of play with like when you showed the prompt and if you showed the prompt, but there wasn't a lot of like freedom, if you will. And you know, all of a sudden for iOS, there was incredible signal loss almost overnight.
Brett House (22:30.547)
Yeah.
Brett House (22:39.57)
Yeah. What what that what percentage of the population I
Rio (22:41.314)
Well yeah, I well I remember I remember there's I remember there were th there were th rumors they were gonna like like eliminate UTM parameters completely, right? I mean I don't they never didn't end up doing that, but it certainly restricted what you can do.
Brett House (22:52.7)
And what percentage of the population actually opted in to the ATT framework, right? Or or to to tracking? Yeah, that's what I thought. Yeah. So you lost a huge you know, and so that fundamentally altered the measurement landscape a bit, and you guys being in that world, how did you guys adjust to that that new world with with the iOS changes?
Brian. (22:58.518)
I think it's landed in like the mid twenties, like twenty seven, twenty five, twenty seven percent.
Brett House (23:19.092)
Did he freeze?
Brett House (23:30.258)
I th I thought my question he was pondering my question for a second.
Rio (23:34.296)
It was it was a stumper.
Brett (23:37.4)
I guess there are gonna have to be some sort of edits after all.
Rio (23:40.781)
Yeah.
Brett House (23:41.372)
Yeah, they're we're gonna have to jump in at this portion of the of the show. Yeah, he's he's totally frozen.
Rio (23:48.834)
Yeah. They get lost connectivity.
Brett (23:51.066)
Messaging him, see what's going on. Maybe something in the office just tech wise.
Brett (24:06.638)
Yeah, I think he was right. I think that 27, 28% give or take, is where that opt-in ended up. then there's like different ways some other companies measure it, which is I'd have to like reconfirm, but it that's if you include the people who are like auto-opted out. You know, like if you basically like one time select, like
Brett House (24:14.526)
Yeah.
Brett (24:35.298)
I never wanna receive an opt in message.
Brian. (24:39.447)
Hey, you guys there?
Rio (24:41.23)
Yeah, we're here.
Brett House (24:41.522)
Yeah, we thought you we thought we really stumped you with one of our questions. Yeah.
Brian. (24:44.557)
Yeah, it's this I think my internet in my office dropped.
Rio (24:48.344)
Yeah, no yeah, yeah, and well it's back. Okay. So so Brett, you had asked about how you responded to A ATT and how and how it how it actually impacted
Brett House (24:55.356)
yeah, yeah, so yeah, I yeah, so maybe I'll re ask that question. So so a so so I so here we go. Start now. Apple's ATT changes them and obviously they fundamentally change the measurement landscape. How did you guys adapt to that?
Rio (25:17.314)
That question's just a killer, Brett.
Brett House (25:20.948)
Did you hear us?
Brian. (25:23.437)
Guys, hold on a second. If you can hear me, I'm still trying to get on.
Brett House (25:27.228)
Okay.
Brian. (25:35.873)
It's like not letting me into the video.
Brett House (25:39.656)
Or you can't see us?
Brian. (25:40.939)
No, I it's it's not even letting me into Riverside. It's like
Rio (25:44.77)
We can hear you and
Brian. (25:46.333)
okay, here I am. Sorry, I was in a different tab. Okay, so I'm back. I pa I my my we lost internet for a minute.
Brett (25:52.91)
Yeah, we can see and hear you.
Brett House (25:52.917)
All right, so I'm gonna hold on. Okay, I'm gonna close that because you run out of time. Okay, so yeah, so we were just talking about AT ATT's the Apple ATT change and how it fundamentally impacted the measurement landscape and how you guys adapted to that. Right? So yeah, so let me just re-ass that question. So so Brian, so clearly these ATT changes, which we've confirmed, you know, about twenty-seven percent, something around there, twenty to twenty-five, twenty to thirty percent.
Brian. (25:54.497)
Okay, cool. I'm back.
Rio (26:21.996)
Or opting in, yeah.
Brett House (26:22.618)
of of people opting in to be tracked basically for targeted advertising or cross app tracking. How did you guys which is a huge signal loss, right? You know, in in the Apple iOS ecosystem, how did you guys at AppSpot adapt to that from an analytics perspective?
Brian. (26:39.991)
Yeah. I think right away I remember our CEO, Orin, deploying about 80 engineers on everything privacy. You know, I think he saw that this was a a sea change in consumer privacy led by Apple and and we were gonna go all in on it. So we did a number of things. We sort of created privacy preserving APIs and probabilistic modeling in our measurement tooling.
We built the beginnings of what ended what would end would become our current kind of data clean room and data collaboration platforms. We became Apple's CS team. So all of our clients who had questions of what does this mean, we be we were the experts. and so what ended up happening is you know, we really leaned into sort of privacy safe measurement.
Brett House (27:20.404)
Yeah.
Brian. (27:31.277)
and became leaders and sort of advocating for it, ensuring that the data that we were, you know, working with our clients was in sort of privacy-safe frameworks. We built up a lot of our methodologies to regain that accuracy. and we worked really deeply with the market with partners, with Apple, with with with Walled Gardens, with all other kinds of companies, because everyone was sort of like grappling with the same thing, right? Brands lost signal, but a lot of ad tech companies and networks, that was their.
main signal for targeting, right? And you saw what happened with with Meta, you know, stock prices plummeted and then and then there was an enormous kind of investment in gathering kind of different signal, more privacy safe signal. So it's interesting. but
Brett House (28:02.835)
Yeah.
Brett House (28:15.432)
Yep. It was it was all the topic for like a dec for five, seven years there, right? When it was cookie deprecation plus ATT and and I mean I can't tell you how many cycles I've spent in product teams. Yeah, at at New Star, to your point, we were doing a lot of consultation to clients that thought, like, is this entire like building that we just built, this data building or house of cards, is it gonna just collapse, right? Because of because of the loss of both of both mobile IDs as well as cookie deprecation.
Rio (28:25.73)
Yeah, how many panels I sent on talking about signal loss and
Brett House (28:45.17)
So so
Rio (28:47.426)
Yeah. Well, but qu turning to attribution. So this is one of the main things that your organization does. And I think this is an interesting one to talk about because I mean I guess the the bigger question I'm going to ask is I really can attribution be trusted? And let me preface that by like giving like the audience a an example. So this is a famous case study in 2018. Uber turned off its Facebook and Instagram ads for three months to test if their actual value just could be I guess considered an incrementality test, right?
And it discovered they had no measurable business impact whatsoever. They had been paying for all of these app downloads, right? And they've been, you know, I guess that you know the the the Walled Garden had been saying, these are, you know, the actu, you know, we're we're we're we're claiming attribution for these these downloads, right? But they were it made no it had no impact. Like they were spending billions of dollars, right? And like it wasn't driving a single additional download because the whole thing was being games. So I guess just with that is the the framing, like can at
Can attribution be trusted anymore? Thoughts?
Brian. (29:50.517)
I have a few. so I listen, I think you bring up incrementality. It it's it's the the market's finally maturing to understand what that is and how it needs to be sort of an always on experiment based complement towards traditional attribution. and so you see the market maturing. I don't see this as a loss of confidence. our business is growing very well.
Brett House (29:51.771)
Ha ha ha.
Brian. (30:19.945)
Our partners, media partners, walled gardens, are investing significantly in attribution and the collaboration around attribution. In fact, some of the walled gardens, I can mention Snap, I was just on a panel with them and they launched they they released their unified or universal attribution. The these walled gardens are taking Apps Flyer's attribution to feed their optimization because they know that our marketers will look at
our kind of cross platform data to make budget decisions. Exactly.
Rio (30:52.622)
'Cause you're a third 'cause you're a third party, right? They're not relying on the walled garden itself to actually provide us a check its own homework, so to speak.
Brett House (30:58.388)
Yeah.
Brian. (30:58.659)
So think think about that. Think think about all the data and signal that a walled garden has of their users, yet they're gonna take Apps Flyer's attribution signal as a priority in their own optimization. And then ask me if attribution is is is you know still a thing. It's it's the signal that attribution produces is incredibly valuable. And here's what I don't think the world quite understands. You know, attribution in one level at an advertising level is sort of like, okay.
You know, what what was the result of the ad that was, you know, served? incrementing incrementality is going to tell me sort of like, did that matter? Did it drive incremental returns or incremental revenue to the business? And both are very important and they're different. But one thing about the mobile attribution ecosystem is how is the architecture and our position in it? So we are telling brands.
Exactly which media touch point drove a conversion. that's how the economics of mobile actually work. That's who gets paid. so it
Brett House (32:07.612)
Yeah. And and and and is that across your signal that's being pulled from the app ecosystem? But do you have out of app mobile web, be beyond beyond just that channel? So you're seeing you're getting a much broader view?
Brian. (32:16.161)
Yes, yes.
Exactly. Exactly. So so so let me first say this. So the the and and the way the industry developed is we are sending in real time postbacks. We're sending signal to all the the networks in walled gardens, and they're using that signal for their own optimization. And what we've done in the last several years is we've built web attribution and CTV attribution.
And PC and game consoles. We're showing brands this journey of their customers, but we're then computing that into a signal that goes back to the walled garden for optimization in a cross-platform manner. And that infrastructure is incredibly valuable, both for the networks, the advertising, the businesses who generate their revenue. They need that signal for optimization. It's very unique. You can't create that signal without.
Brett House (33:15.625)
Yeah.
Brian. (33:17.237)
Apps flyer or without a company that's in the attribution space. It's very critical for the brands themselves because they're getting better returns on their advertising spend. They're getting better performance. And so the more cross-platform and the more trusted that signal is, the key here is the independence of it. Right. So we are, you know, the one of the biggest values we bring the market is that we're trusted and we're independent. We're friends with everybody.
But we're we're we're calling the shots like we see them, and those shots effectively get packaged up as signal and sent in real time billions, if not trillions, of signals in in a in a you know, you know, daily, which is kind of powering this entire ecosystem.
Brett House (34:02.012)
Yeah. And that's and that yeah, and that's everything. I mean that's behavioral, contextual, app download, which would be sort of more acquisition based. And when we're talk when we're talking about signal, it's exposure, I'm assuming, engagement, right? Like can you can you kind of deconstruct what the signal means in terms of what you guys see?
Brian. (34:23.331)
So different apps, different websites, different brands send us different information. some don't want that to go anywhere and they're they're very sort of locked down and they want sort of minimal information passed about what happens after a download, let's just say, and they're only looking at at a a very kind of narrow slice, and that's their preference. Others want to send us revenue and post-install events and
Brett House (34:35.454)
Yeah.
Brian. (34:50.989)
connect us to their websites and their web traffic. And they want to get a really holistic picture of what happens in my business. And we want that to go to walled gardens because we want better performance. And it comes down to the business that you're in, the sophistication of your marketing strategy, your own kind of privacy frameworks and philosophy. It it it ranges by, you know, is it more of an acquisition play? Is it a retention play? We see it ranged by enterprises versus startups and games versus
you know, banks. but at the end of the day, you know, we are infrastructure that allows, you know, these these businesses to figure out whatever signal is important. It's it's a word just, you know, for sort of data, whatever data is important for them that they want to see their campaigns optimized for.
Brett House (35:34.45)
Yeah, yeah. Yeah.
Brett House (35:40.297)
Yeah, you're increasing sort of person's level fidelity, you know, like which which is being passed back to the Walt Gardens, and they're using that for enhanced audience segmentation, enhanced targeting, personalization. I'm assuming improved measurement, just because you have higher fidelity signal across at a person's level, which is which ties back to a device in a household. So so that that's how you should be thinking about it, I guess, is is right? Is is that you're
Brian. (36:05.259)
Yeah, and I would just say they're in a very sort of privacy safe way, right? I think that's the difference between, you know, what the what what was being attempted with sort of multi touch attribution and these models of the past to sort of where we are now is it's all sort of very privacy first. It's got privacy preserving a APIs, it's got clean rooms, so it's got a whole series probabilistic modeling that allows that to happen at scale that sort of protect user privacy.
Brett House (36:08.84)
Yeah.
Brett House (36:17.234)
Yeah.
Brett House (36:23.571)
Yeah.
Rio (36:32.8)
Speaking of MTA, I remember when I first got into this business, I mean, you know, a while ago. I think we all probably started around the same time. Yeah, at least really there was like last click, and then there was this realization that this was not definitely not the right way to look at attribution. And people started experimenting with different models, and then MTA became a big thing around 10 years ago. There were all these companies that had their own models, and there was debate about which model was better, and it became this really big data exercise of trying trying to untangle it.
And then you had signal laws, which we talked about before, which made it much more challenging. So like, is MTA still realistic in today's environment? Have we moved past it? And just wanna wanna get your thoughts on that, Brian.
Brian. (37:11.245)
I think MTA was built for a world where you could track a single user across every touch point and that doesn't exist. That no longer exists. So it it's it's sort of an academically ideal state, which isn't feasible. I hear a lot less of it. Some companies are still have a and most of the vendors in the space are like s you know, service providers are coming in and doing very deep, hands-on consulting type environments. And I I don't want to say there's no application for it. I think a lot of big brands are still
Rio (37:17.059)
Yep.
Brett House (37:18.099)
Yeah.
Brett House (37:36.84)
Yeah.
Brian. (37:41.025)
Looking at their version of it, maybe they've been doing it a long time and that consistency matters. We offer elements of an MTA, even though the mobile ecosystem is sort of the economy is driven over last click because companies do want to understand a more peer view of what's happening. But what I think is really changed is the industry has moved towards sort of the triangulation of attribution, incrementality, and modeling. And I think this has replaced sort of the ideal.
Of what MTA could be. Right. So we talked about attribution. This is in real time. It's used for optimization. While it's not perfect, it's a utility that everyone has adopted and understands, and there's a lot of scale there. And now the game in attribution is keeping it fraud free, bringing in more platforms beyond mobile, doing it in a privacy-safe way, and sort of making sure that the signal that attribution data provides is.
Brett House (38:12.382)
Yeah.
Brian. (38:39.819)
is leveraged for the benefit of of of advertisers. But we talked
Brett House (38:43.336)
Yeah, yeah.
Rio (38:43.438)
It's funny. That's a good it's a good point you made. It's funny. We talked to Greg Stewart a couple of weeks ago he he had said MTAU's calling it multi-touch analysis, which if it aligns with your your your hypothesis they're about it really being a services. Yeah, yeah, for sure.
Brett House (38:52.508)
Yeah. W more of an estimation. Yeah. Well, and I think I think it's critical to also realize that when you're pulling in because we did a ton of this at New Star and in in partnership also with with MMA Global and Greg Stewart, we did a ton of research with them. that when you're br you're when you talk to about modeling, it you're starting to bring in data signal that is not platform and not media specific, right? So because those are the things that could potentially influence, you know, it's its propensity.
Brian. (38:56.761)
Yeah.
Brett House (39:20.25)
natural propensity, it's seasonality. So we would pull in all these third-party data sources non-tied to the actual advertising of the media that might have some sort of impact on a person's actions, right? And and you might be and so what we realize is you're is is we're often overcrediting a media channel for some sort of performance when in fact you know they should know really how how much to credit it so they can make better strategic decisions on budget allocation.
Brian. (39:46.743)
Listen, I I believe when I talk to CMOs and marketing leaders, I believe the combination of modeling, which in my opinion, it it really needs to be done sort of in house because it's so customized to the specific business. It's got so much we've looked for years about building like an MMA MMM product. And, you know, as opposed to sort of incrementality or attribution, I I believe it needs to be so highly customed to really be trusted and believable. It
Brett House (39:57.118)
Yeah.
Brett House (40:05.639)
Yeah.
Brian. (40:15.945)
It it at least in our view, we stick to where we are are good and we work with MMM vendors. But the the the businesses, I feel from a measurement perspective, that are most advanced is they're triangulating these different methodologies. They're they're working with them together, they're laying them together, they all begin to feed one another. I don't know that there's a perfect recipe yet, but it's it this is the world that we're operating in now. These are where the thought leaders are thinking.
Brett House (40:32.724)
Yep.
Brett House (40:37.715)
Yep.
Brian. (40:45.345)
As opposed to sort of MTA in in in my view.
Brett House (40:48.424)
Yeah, it's certainly a hybridization, right? Because you j you're just getting as close as you can to an an analytics results that sort of gives you high a high degree of confidence of how you could shift your budgets or how you can adjust or or or whatever. So yeah, for sure.
Rio (41:03.256)
So turning to clean rooms, this is something I know that AppsLive does offer as one of your one of your capabilities. I remember learning about this a couple of years ago. Like curious, why did you decide to go into the space? Was it some of the privacy laws you mentioned earlier? Was it customer demand? Was it something else? Was it a combination of factors?
Brian. (41:21.315)
So going back to IDFA deprecation and iOS 14, a lot of our advertisers and customers got hit with signal loss overnight. And it had a material impact on their business. And both our publisher, the ad networks, and our advertisers were saying, You're you're sitting in the middle here. You know, how can you help us? So we started to develop a data clean room. initial use case was to sort of return signal.
Brett House (41:44.179)
Yeah.
Brian. (41:50.977)
that you had lost on on meta. and we could sort of do some privacy safe way to allow meta and and advertisers to understand sort of each other's views without at that without them exchanging data. and it became an implementation for sort of core mobile marketing. We then saw sort of a a big opportunity to point that to retail and and commerce media networks, you know,
This everything is an ad network sort of thesis where we saw all kinds of businesses in grocery and in finance and in travel creating advertising businesses. They didn't have any tech stack. They were scaling up, hiring advertising, you know, sales executives, but they were building a ad product, you know, stack from scratch. And they found our value proposition interesting. You know, here is a neutral, independent measurement forward company.
Brett House (42:23.047)
Yeah.
Brian. (42:48.739)
That's got a clean room where they can do audience activation and bring in, you know, identity resolution and a number of different enrichment data points. and they could work with anyone they wanted to. We weren't owned by an agency, we weren't one particular, you know, advertising publisher. And they like that interoperability. And that has become a business of ours, which which many folks in our core, you know, the Apps Fire brand is so tied to sort of mobile and attribution that they're not familiar with it.
I believe you'll see clean rooms. I look at the analogy of of cloud computing. You know, there was a time where, you know, moving to the cloud was a strategic decision. and you know, it was an it was a big initiative. and it's today, it's how all software is is developed, right? You know, l I'm sure we'll get to AI, but you know, for the last 10 years. Exactly.
Brett House (43:31.892)
Yeah.
Rio (43:37.634)
Yeah, most most software is cloud native now. Yeah. That didn't used to be the case 15 years ago. It's interesting clean rooms, though. I mean, like, so I remember I I started working with them around six, seven years ago. And then, you know, but now you look at it, and I don't think the clean room businesses really took off the way some many people expected it to be, right? Where, you know, there's universal adoption is certainly not that. But at the same time, you know, live ramp acquired, was acquired by Publicists recently that took Habu off the market. InfoSum got acquired by WPP.
Brett House (43:39.764)
Yeah.
Brett House (44:05.743)
The WPV, yeah.
Rio (44:06.722)
It's been a bunch of acquisitions, right? So and and look and I and I yeah, I work for a systems integrator. Like we are seeing or I'm working on a number of projects that include clean rooms right now. So I think despite maybe a slow beginning, like up up like we are seeing some uptick and adoption right now. What does that say about? I mean, are you seeing the same thing and what where does that say about where the industry is headed? is privacy becoming more important? Is this gonna become a standard like
like required piece of of architecture that that's gonna be in a modern stack and will people deploy their own or they rely on let's say platforms or measurement companies in order to provide them?
Brian. (44:42.615)
I do see more and more RFP questions talking about all our audience activation is going to be done in clean rooms. I see that continuing to be increasingly, if they're kind of designing infrastructure for the foreseeable future, they want to make sure that they're hardened from a privacy perspective. I don't know if, and I agree with you, like I don't know as a category if it took off the way that people had thought years back, but then I would ask myself, is it a category or is it a layer of technology?
Brett House (44:49.737)
Yeah.
Brett House (45:11.994)
Yeah, or or a or a a fundamental use case of like how do we prevent date data movement. Yeah. Yeah.
Brian. (45:14.283)
Yeah, does it does it serve a purpose? Exactly. Right. So I think we're seeing that more and more of while there's less maybe put aside the lab ramp acquisition, we could talk about that in a minute. There's less sort of buzz in the industry on clean rooms. I see it more and more as a requirement for implementation. So it's it's maybe quieter, but it's real. And it's in and I think I think companies are still very eager to do things in a privacy preserving manner. And they and so I I I see it like this.
Rio (45:31.267)
Yeah.
Brian. (45:43.651)
You know, I look at this live ramp news and you know, I think you know, whether a clean room is owned by a holding company or a platform, I I I question the incentives and and if they're structurally aligned with advertisers. You know, we again we have a business that's been operating for 14 years and it's been built on trust and independence. And as we look with our clients who who use our clean rooms and our data collaboration platform, they want independence.
And they want interoperability. And, you know, they want the ability to work with other clean rooms and and whatnot. And I I s this this is our this is our point of view on on clean rooms. I don't think they're going away. I think they're just gonna be a critical piece of how brands implement and integrate with one another to do partnership and and and audience activation. we bring a measurement slant to that, but that's how that's how I look at clean rooms today.
Brett House (46:37.937)
And
Rio (46:38.114)
Well, clean rooms have been good for measurement. That's one of the use cases that definitely has has stood the test of time. And it's inter with with clean rooms too. I mean, I like the point you just made. So I mean, personally, I don't see any commercial advantage for a Hold Co having their own clean room. I d I really don't. I do agree with you in what you said about it being a critical capability or piece of infrastructure, or Brett to your point, even like piece of tool that actually
Brian. (46:40.857)
For sure.
Rio (47:03.734)
addresses specific use cases that are probably gonna be in an RFP because they're gonna be critical for brand to if they want to activate or measure across dat channels or or or collaborate their data with second and third party. So for so no disagreement there. It makes sense.
Brett House (47:16.658)
Yeah, and should and should those use cases live where the data lives? And so if if the majority of big brands are managing their data in cloud, you know, native data warehouses, right? Whether it's a Snowflake, a a a a Databricks or an Azure, shouldn't the data shouldn't data collaboration and sort of non data movement use cases, right, which is the clean room use cases, all live where the data is versus versus having to go out to a to a separate platform?
Brian. (47:45.657)
That's a good question. I think, you know, our maybe it maybe I'm biased in in sort of who our persona is that we talk to in a in a in a customer, but th you know, there the clean rooms that most of our customers have used have failed them from a measurement perspective. And so you start talking well, I think a lot of clean room providers don't know measurement. They don't do measurement, right? It's not it's not their core competency. and so they're they might be provided a utility of a of a safe environment, but that's actually
Brett House (47:58.663)
Yeah. And what do you think that is?
Brett House (48:05.523)
Yeah.
Yeah.
Brian. (48:14.623)
Not there's not a lot of intellectual property around that, right? All the cryptography Exactly. Exactly.
Brett House (48:17.192)
Yeah, it was it was a it Exactly. It was yeah, it was it was like a solution looking for use cases, right? When it was when it
Rio (48:17.326)
Well just a place you ma you just a place you match data. I mean that's really all it's doing, right? I mean you join data.
Brian. (48:24.621)
Right. So when you when you when you bring measurement capabilities in a clean room and you have a significant market share of advertisers like App Store is now it's like, okay, now there's a real application there. Now I can make sure that, you know, if I'm an ad, if I'm a if I'm a burgeoning ad business and my customers are wanting more trusted closed loop measurement so I can sell more ads. Okay, now I have a now this is a more viable use case than I just have a kind of trusted environment, which is
Brett House (48:52.764)
Yeah.
Brian. (48:53.389)
You th that that that's sort of that that's kind of the commonality I think of of of any of these these these pieces of technology.
Brett House (49:00.69)
Yeah. And and where do you guys I mean, are you guys partnering with a lot of the the the kind of more modern MMM providers like the Mutant Nexes of the World, that are sort of trying to take the managed services heavy nature of MMM and even just attribution in general out of it by by powering these things with agents, right? Taking the you know, the the people out of the mix, so to speak. I mean like are you seeing that that's there's a future for these independent providers to provide this type of
and MTA type measurement or or and how do you guys partner with them in general?
Brian. (49:34.499)
We do. Again, we we partner with with most sort of businesses in this space. I think our clients might say, Hey, we brought on an MMM vendor. We want to feed our model, your attribution incrementality data, other data points. So we're helping our customers sort of feed some models, whether it's third party or they're building it in-house or or or they're kind of blend. So and you know, their service model, whether it's kind of like, you know, all tech or there's a kind of service layer.
it it it it doesn't really matter to us, right? I think they're looking for the quality signal that we provide them. You mentioned agents. I mean, I think this is where our world can get dramatically changed in the coming years. I think there is a case to be made that says performance marketing becomes sort of agentic marketing in a very fast time period, right? A performance marketing team that may have had.
Brett House (50:07.57)
Yeah. Yep.
Brian. (50:31.865)
six, seven, eight distinct roles a couple of years ago might now have a team of three running growth. And, you know, hundreds, if not thousands, of agents doing different things, a collection of your own agent, App Slier agent, a third party agent. and we see ourselves providing that underlying data to feed these agents for the guardrails so that they don't go kind of off script. and I think this is where
Brett House (50:37.97)
Yeah. Yep.
Brian. (50:58.413)
You know, I think that the vendor landscape gets gets really exciting because yeah, there's th this world or or it could take a lot longer, right? I think some of these things it's hard to put a time frame on. Is it a year away? Is it five years away? But we ask ourselves a lot like who is our customer of the future? and you know, what are we are we, you know, from an analytics perspective, from the products that we develop, we are prepared and I think well positioned for a world where most marketing is is.
is run by agents and autonomous workflows. and I think that that's that's a pretty exciting, you know, future to think about the innovation in.
Rio (51:35.342)
Yeah, it's f it's interesting too. Like the the agentification of marketing, hear that term thrown out quite a bit. And for what I'm seeing, a lot of what the implication, at least in the near term, is more on the org design, who's doing marketing, the size of teams, the collapse of multiple roles of specialists into one one role of generalist or a couple roles of generalists supported by.
agent an agent or swarms of agents. So I that is a very imp interesting discussion. So but like looking at fraud, I I thought that that was an it but at first I wasn't sure like I was a little surprised when I saw that you had rolled out some fraud prevention tools. But I actually started thinking about it more and thought it actually makes a lot of sense to have a measurement company also offering fraud detection services. Wonder if you can comment on that a little bit like what was the the thinking there.
how is it taken off? Like what what types of what types of clients or use cases are you seeing success with for the service?
Brian. (52:30.773)
we've had a fraud product we call P360 in the market for probably eight or nine years. And it's it's one of our best adopted products outside of kind of core attribution. Fraud is rampant in in all advertising, specifically, you know, in mobile. Yeah. it's it's the innovation of fraud with AI on fraudsters is you it's a it's a it's a forever count and mouse game. fraud has gotten very sophisticated, you know.
Brett House (52:46.174)
Yeah.
Rio (52:47.094)
Especially mobile.
Brian. (52:58.679)
These AIs can emulate a user that downloads an app, behaves in the app, even can make a fraudulent transaction. And so the type of fraud is is growing and and harder to detect. Their ability to fight it and prevent it and understand it ahead of time is also increasing. But I think, listen, you ask about like, you know, maybe you were surprised that a measurement company was in the fraud space. We sit on a pretty significant amount of data. And really, what in order to battle fraud.
You need a lot of data and you need strong algorithms and AI to look for anomalies. The larger your data set and the stronger the AI, the better the fraud you're going to be able to find. You're going to be able to find the more sophisticated fraud faster. So we're in a natural position because of the data we're set on. Nobody sees more data in mobile than we do. And so we look at it as if we can prevent.
Brett House (53:39.379)
Yeah.
Brett House (53:44.798)
Total. Yeah.
Brian. (53:53.123)
Bad data from coming into your environment as an advertiser. If we can help you block it up front, then the quality of that measurement signal is better. It's it's very s it it's it's a different mechanism, but ultimately driving towards you have a clean, trustworthy set of data to make business decisions on. And that's becoming even more important if we connect to the last point of now this data is feeding agentic workflows because there's
Less and less humans in the loop to kind of provide their judgment. And whatever data you're going to feed an agent, that agent's going to run with it and optimize whatever it needs to to hit that goal. And so if you're not protecting your environment from fraud, which is getting harder and harder to detect, you're feeding your agents bad data. So they're there, it's very germane to our thesis of you know measurement to be able to provide fraud.
Brett House (54:37.927)
Yeah.
Brett House (54:42.196)
No, exactly.
Brian. (54:49.411)
Prevention and detection services.
Brett House (54:51.612)
Yeah, well no, and Rio, th just with at New Star, that was a massive part of our the one ID solution that they sold to TransUnion is that it was it was there were three business units that sold. One was marketing solutions, which was heavy heavy in MMM and MTA, which was about sixty percent of our revenue. One was fraud and anti fraud behavior, and then another was caller ID. And I'm assuming which is yeah, j you know, obviously used in the financial services industry industry to prevent sort of fraud
Brian. (55:10.489)
Mm-hmm.
Brett House (55:17.93)
fraudulent activity by by you know people trying to get you know access bank accounts or otherwise. Do you guys actually get into that space or is that an area that is completely separate from from your fundamental data set?
Brian. (55:30.733)
Yeah, we're not actively in that space. but I will say that we we are very much expanding our thinking of what fraud is sort of in our court and what do we want to go after. There's a lot of there's a lot of fraud and there's a lot of there's a lot of interest our customers have in us helping them fight it differently. But we ask ourselves like what are we uniquely positioned to win at? Right. We don't we don't want to just get into the business that are from kind of like, you know.
Brett House (55:41.907)
Yeah.
Brett House (55:54.899)
Yeah.
Brian. (55:59.171)
There's more expertise in these other categories. But things are changing pretty quickly. And like I said, now now a lot of the fraud is sort of emulating these user journeys, which maybe a couple of years ago we said, Well, okay, that's not exactly our space, but things were blending fast. So you you you don't, you know, we'll we'll we'll we'll keep you updated on what we're thinking.
Brett House (56:01.204)
Yeah.
Rio (56:16.098)
Well well, you think about it like like these these these LLMs passed a touring test for Chat GPT too, right? Which is several years ago, right? So they can actually seem very much like humans, so much like humans that humans can't even discern if they're if their machines are not anymore. I'm so I'm imagining we've heard this quite a bit, how AI is really supercharging or or turbocharging fraudulent activity. What's some of the craziest stuff that you're able to share that you've seen recently in your space?
Brian. (56:41.623)
I don't have juicy stories for you on that one. Sorry. I know.
Brett House (56:42.964)
Yeah.
Rio (56:43.842)
Like I know the audience should love it, but if you don't, it's okay.
Brett House (56:48.83)
So so so a couple of years from now, what do you think I mean we kind of hinted at this, but what do you think measurement's gonna look like, right? Knowing what you know from your angle and the signal that you guys are providing. I mean, how do you think it's gonna change in the ecosystem?
Brian. (57:04.761)
So measurement and optimization and execution are collapsing into one thing. So you I think that this is first and foremost, right? So you need good measurement as a trusted foundation, but that data is already feeding your agents, which are performing autonomously for you. So the entire kind of workflow just collapsed. And I think measurement and measurement companies are in a very good space and position.
Brett House (57:12.69)
Yeah. That's a good point.
Brett House (57:27.614)
Yep.
Brian. (57:31.651)
To benefit from that collapsing of sort of workflow. That's one. I think you're gonna see measurement become more omni-channel, right? Kind of a marketing bud, but it's gonna map to the user behavior a lot better than it is today. You're not gonna have mobile measurement companies and web measurement companies and TV measurement companies and brands trying to sort of resolve what this all means. You're gonna have a measurement business and it's gonna understand.
Brett House (57:36.061)
Yeah.
Brian. (57:59.841)
Where in the funnel, what are you trying to do? What's the what's the anchor point I'm I'm trying to understand? And it's gonna provide you a single view of your marketing and your your your customers and and tell you what happened. I
Brett House (58:11.998)
Hall hallelujah. They've been say we've been saying that for like a twenty years. you know, Nielsen's been talking about this.
Brian. (58:15.639)
I I know we have, but now we actually have like I have dashboards I can show you, which talks you can see all these different platforms and it calculates an LTV. And I think a lot of things had to get right and and change for us to get here. And like with everything, AI is just accelerating this trend as well. So you look at every trend you see, and how is it going to be accelerated? Because of AI. So I think
Brett House (58:22.674)
Yeah, and they map it across yep.
Brett House (58:37.128)
Yep. It's the data mapping and the data learning and the ability to output something that's unified across all of these kind of siloed data work streams in a sense. Yeah.
Brian. (58:44.599)
Yeah. Yeah. So I think and then I think you're gonna see like I we saw we talked about earlier. I think you're gonna see companies think about measurement in layers or different methodologies, and not one is right or more right than another. There's attribution, there's causal inference, there's incrementality, there's testing, there's different kind of modeling, and they're all gonna be important. And I think brands are gonna understand how to resolve those different views together for a single view of of their of their business.
Brett House (58:56.553)
Yeah.
Brett House (59:07.187)
Yeah.
Brett House (59:12.072)
Yeah, and and and that's a critical point, is like how do you synthesize all those different approaches and in in in considering the volume of data in the channels? That's always been the the work of data science and data analytics teams, right? And AI is gonna really compound our ability to synthesize this information and give us actionable outputs that that would have taken months previously.
Rio (59:24.086)
Yeah, accelerates that big time. Yeah.
Well well well Brett, even streamlining the integrations, right? Instead of having to manage all these all these like dozens or hundreds of or thousands even of APIs and they change all the time. AI's making that easier, the ability to ingest the data, normalize it, synthesize it, make sense out of it. I mean it's incredible, right? That it's just such an accelerant. So yeah, I I I think and we're only kind of at the start of this too, I would.
Brian. (59:52.163)
Yeah, yeah, very exciting time.
Brett House (59:52.979)
Yeah, this is exciting. So yeah, so you're so are you is this different than anything that you've seen in your career in terms of what we're seeing with with a lot of what we've been talking about with with the gentic AI capabilities, right? The the role that you guys play in is this is this different than anything you've seen? Having been in the space?
Brian. (01:00:09.561)
Yeah, I mean certainly the the the the pace and the acceleration certainly is the the sort of emotional FOMO and not knowing where one company every everyone's asking me, like, are we behind? Right? All my customers asked me, are we are we behind from your point of view? and so the FOMO.
Brett House (01:00:22.536)
Yeah. And you're like and your your answer ninety percent of the time is yes. You are you are behind.
Brian. (01:00:26.837)
I you know, it's it's it it's a the ones that think they are are are aren't. And and but you know it's like it's so there's an emotional sort of FOMO. There's a there's a the uncertainty in this space is something I've never seen. People don't know if they're AI proof. You know, they're trying to tell a good story, but they don't know if they're they got a tailwind or a headwind. I think there's gonna be a lot of dead bodies on the side of the road next couple years in these in these ad tech ecosystems.
Brett House (01:00:31.966)
Yeah.
Brett House (01:00:37.161)
Yeah.
Brett House (01:00:43.602)
Yeah.
Brian. (01:00:55.831)
you know, because of what cowork can do or or codex can do. And and, you know, analytics is is is sort of what is that anymore, right? Like, like, you know, I think so many things that were critical components of a of a of it are are are are kind of like dead on arrival already. And it's just like d d do do we see it yet? but I also think things take a long time. People take a long time. Organizations, I I wonder how quickly organizations are going to actually change, right? We work with
Brett House (01:01:12.723)
Yeah.
Brian. (01:01:24.619)
AI native companies who, you know, one person is doing the job of ten and they're very busy. It's hard to get them on the phone. They want to work by Slack and they work, they integrate very, very quickly with a couple of API docs.
Brett House (01:01:31.849)
Yeah.
Rio (01:01:35.266)
Well well there's that new well is well there's that new term AI vampires, right? People who are so enamored with AI they stay up on let you never sleep because they've you i they've got they've got fifteen agents doing things. AI I think that is that that is wild. And I think in ad tech too, like any company that's not that's just toll taking, that's not adding true like increment of value is gonna find themselves in a lot of trouble over the next couple of years as a lot of these things shake out, right? So I think that is going to be definitely an outcome. And then but I do think like to your point about
Brian. (01:01:41.144)
Yeah.
Brett House (01:01:56.04)
Yeah. Yeah.
Rio (01:02:02.968)
people being more more or less mature. So I as as a consultant, I would always get called in by these companies and they'd always ask, okay, get assessed my digital maturity or my mobile maturity or whatever it was, our analytics maturity. Right. And every it's so funny you said that. Like almost everyone we would work with would tell me in advance, especially when we do Martech and ad tech assessments over their stack. They'd always say, we know we're so far behind. We know we're we're terrible. But then you would do these and you realize, okay, well, yeah, there's things you could improve upon.
relative to you when you you benchmark them against competitors, you realize that everyone, everyone no one's as far ahead as they think all of their no competitors are as far ahead as anyone thinks they are. Most companies really struggle. To your point, it is hard to move quickly. Most people are moving at a at best at at like a moderate speed. It's it's difficult.
Brett House (01:02:47.452)
Yeah, well and the read and Brian, the read that I'm getting and Rio that I'm getting from the industry is so I'm working with a very large agency, global agency, independent, that is completely replatforming, AI native, building their stack and trying to introduce this sort of new way of delivering their solutions, you know, managed services solutions, but highly agentic driven, both data creative and sort of media activation. And they they're basically saying that most of the brands they deal with, marketing teams,
generally, you know, either mid or or upper level marketers are woefully under prepared, right? And and seem to be a few steps behind what the agency world is talking about, what the ad tech, market, data world, data tech world are talking about, right? Where it still seems like they're taking the baby steps versus versus you know being
kind of on a dead sprint with the rest of us. I mean, do you are you seeing that, Brian? Are you seeing that that all of our talk in this industry is a little bit ahead of where brands themselves actually are?
Brian. (01:03:48.477)
I, you know, maybe I I think a lot of, you know, larger enterprises are much more worried about governance and things going wrong than things going more right. And I think you know, some of you know the newer businesses that are around or AI companies, it's sort of like, you know, foot's on the gas and they don't have as much maybe to lose. so I think there's a lot of businesses that are trying to like lean in faster, but
Brett House (01:03:57.736)
Yeah.
Brian. (01:04:18.435)
There's a lot of security concerns. There's a lot of governance concerns. and so, you know, I think I they got they gotta be careful. So I I do see sort of a a disparity there.
Brett House (01:04:22.046)
Yeah.
Brett House (01:04:27.558)
Yep. Cool. So w we've pr we've taken you over the hour. Should we jump into quick hits and wrap up?
Brian. (01:04:34.231)
Yeah, sure. I g yeah, I g I gotta I could say for a little bit more.
Brett House (01:04:37.532)
Okay. so Rhea why don't you show us off with the quick hits?
Rio (01:04:41.134)
Okay, number one, are the walled gardens grading their own homework?
Brian. (01:04:46.681)
they are and they're also working with us in ways where it shows that they're not. So I think they're they they they're doing both. They're creating their own homework and they're also recognizing that the market wants and uses independent companies like App Slier.
Rio (01:05:01.816)
They don't have to, in other words. Okay.
Brian. (01:05:04.249)
They don't have to.
Brett House (01:05:04.498)
Yeah. Yeah. So is so we've kind of talked about this, but is attribution dead and how would you define sort of the the the the the next generation?
Brian. (01:05:14.957)
it's it's getting a lot of investment from large dollars. Apps Flyer is growing well. I don't think it's a dead category at all. I think it's maturing and I think you're seeing a recognition that it's a piece of a puzzle. and I think it's got a I I think it's got a good future.
Rio (01:05:33.494)
I've heard the argument that what happened to Meta with with the ID FA getting you know, getting bit really becoming opt in, right? Like actually long term helped to Meta. So I guess this question's based on that. Has privacy made advertising measurement better or worse in the long term?
Brian. (01:05:50.233)
It's a great question. I I think it's it's certainly better for consumers. And I think companies that have data, proprietary data, and have leaned in to be able to sort of monetize on that, it's made that better. It also hurt a lot of companies that sort of are over reliant on signal that they lost. So I think you maybe you saw sort of the better companies succeed from that and the better brands benefit benefit from that.
Brett House (01:06:14.612)
So what's the most misleading metric in marketing that you guys see?
Brian. (01:06:23.289)
Mm.
Brian. (01:06:26.857)
Good question. Misleading market.
Brett House (01:06:27.602)
I was g I was gonna I was gonna have a I was gonna have a leading question. Do you think ROAS is still an accurate mechanism for for measuring performance?
Brian. (01:06:36.729)
Yeah, I do. I think there's a lot of businesses that understand their customer in that in that way. And they're looking at how much dollars can I derive from my media investment and the ROI is pretty straightforward. So I don't think ROAS is misleading if it's the right, you know, the right way to look at it for your business. But I d I don't I the most I don't know. I don't know, I have to think about that. That's a good one.
Rio (01:07:01.91)
And we can wrap up with this one. Will data clean rooms still matter or be relevant in five years?
Brian. (01:07:08.471)
Yes, but I think they're because it's so common that they won't be a category. They'll just be a implementation you know, method.
Brett House (01:07:17.18)
Yeah. Yeah. A use case in a sense. well hey, thank you so much, Brian. This has been great. I know we I know we lost the internet in the middle, but we'll fix that in post production. And for everybody that made it through the end of this episode, visit us at deb dev dev dot signal and noise.ai. Find us on YouTube, Apple Podcasts, and Spotify. And we will see you next time. Thanks everybody, and thanks, Brian.
Rio (01:07:21.752)
Yeah, appreciate it.
Brian. (01:07:39.779)
Thank you guys.
Rio (01:07:40.918)
Thank you.