← Next in Media
Next in Media
Charles Manning on Why Measurement Is the Secret Weapon in the Age of Agentic AI
00:38:48
Loading summary
Transcript115 lines
- [00:00]
A
Facing us over the next 16 months is going to be so profound that it's like Elon taking Twitter private combined with a crapload of roll ups that aren't going to be purchasing tech because tech's recreatable really easily with AI. They're going to be purchasing contracts and then in two years a like bonanza of public offerings. There's going to be more money to be put to work in media, there's going to be less tax because a lot of the things that were made up of all the pieces that would otherwise be taxing is now happening. What's going to happen is there's going to be all this disruption and transformation in people's jobs because of AI. And then when people lift their head up and they realize where they're at, let's say in nine months or in 13 months or in 16 months, they're going to be like, oh, would you look at that. Media mix has totally changed. Right.
- [00:54]
B
Hello everybody. I'm going to do my, my official formal introduction. Hi everybody. Welcome to nexty Media. I'm Mike Shields. My guest this week is Charles Manning. He's the founder and CEO of coachava. We are live at the coachava Summit in Sanborn, Idaho.
- [01:07]
A
Hey.
- [01:07]
B
Welcome Charles.
- [01:08]
A
Thank you.
- [01:08]
B
Awesome.
- [01:09]
A
It's great to be here.
- [01:10]
B
Yes. I was going to say let's prove that we, we have an audience. Thank you for proving that we are in front of a live audience. It's exciting, it's funny. We did this a couple years ago.
- [01:19]
A
It was awesome.
- [01:20]
B
Did we talk about AI? Like I don't even know if we.
- [01:23]
A
We talked about ML, we talked about.
- [01:25]
B
Apple, we talked about cards on trade desk which feels like it was a hundred years ago. A lot of things have changed. Help me, I want to take go big picture a little bit. Help me understand. I'm, I'm what's called a normal person. So I struggle with some of the things you people are talking about here. Agentic advertising. It seems like it's awfully real. There's a ton of hype. I have a little bit of metaverse PTSD where I'm like, hold on. I never, I realized I was a consumer driven thing or supposed to be. Help. Help us. Help me wrap my head around how profound this change is going to be. Are we talking about making workflow better which is a big deal versus like fundamentally changing how this industry functions and performs and works. Like give us a set the stage for the next couple years.
- [02:14]
A
Yeah. I overheard something introduction that Trevor had just made and one of the comments he made was that the last decade was really about digitization of signal and media. The next decade is about the digitization of the workflows. And I think that is a really good representation. And it's probably not a decade, it's probably 16 months. And I think one of the really compelling things about what's happening with this notion of agentic advertising and then just generally with AI as a whole, is that everything is happening super fast and it is transformative to how people work. And the combination of something happening really quickly that also happens to impact how people work means there's going to be a lot of disruption.
- [03:10]
B
Do you think it's going to. Is a theory. The analogy a lot of folks use is when Programmatic kicked off about what, 15, 12 years ago, where that was, it moved fast, but it was a gradual thing and then really exploded. You think it's going to be a lot quicker?
- [03:29]
A
Tremendously, like, breathtakingly faster.
- [03:32]
B
That seems hard for big companies.
- [03:35]
A
It does. It seems hard for big companies. What's super interesting is that I think the advertising ecosystem already has the Lego blocks established through kind of traditional APIs, where a lot can happen with APIs that people infrequently use to date. They use them. No question they use them. But when you layer on, on top of those APIs, things like MCP, which is Model Context Protocol, which can wrap APIs. And MCP is like the, for anyone listening is like the API layer of AI and LLMs in that an LLM can trigger a tool call, which then in fact is like an MCP hook that queries something. And I'd say that Programmatic is to the auction as Agentic is to the workflow. Okay, if that makes sense.
- [04:37]
B
First of all, I like it. You have at least four acronyms in one sentence.
- [04:40]
A
Oh, yeah, that's right.
- [04:42]
B
But so again, is that going like. I think the vision I've heard is that, okay, that is going to just. It's going to be so much easier for companies to pull reports and not have to have mega spreadsheets and, you know, like, get. Get their work done faster and spend more time on strategy, which is a pretty profound change. But then, then you, you. There's a vision of it's much more revolutionary where advertising will be so much more effective and machines will optimize well beyond human capacity. And which, which Programmatics sort of did. But like, how, how. I guess, how profound can you, can you envision things coming, the change becoming.
- [05:21]
A
I think what we've seen in Advertising because it is such a technical space and the signal is already digital in large part is that machine learning has been doing an awfully good job at optimization. In the early days that optimization was really around reach and frequency. Increasingly it's now about business outcomes because CFOs want to establish what's the net impact of these dollars spent. And so the ML is slightly changing to be really around business impact and outcomes, which is largely what we're doing with this Atlas performance product, enabling premium publishers to, to handle that and to have that be turnkey. What's interesting, what's new, what's on the horizon and on the frontier is that what's now going to be automated are the workflow steps that previously you would gather a number of different spreadsheets and gather data points in order to present it to someone who's a decision maker. And that is probably a six step process workflow. If you really unpack the pieces, that process can now become automated with natural language and LLMs and even SLMs. And what that means is that instead of spending, let's say a week preparing a Post Campaign Brief 45 days after the campaign was run and then making subsequent decisions, that workflow can happen daily. And we've been in the business of real time for you know, since we. Right, since we started. And that's what you know, programmatic is really driven around. Now all of a sudden the processes are going to be close to real time along with the signal, which is kind of remarkable.
- [07:08]
B
I want to come back to some of the processes and the way you're talking about workflow changing. But a big theme in the last couple years as, as AI has really taken hold in the industry. I think people got excited about generative AI and they should be. But it, there was much quicker traction in media buying optimization mostly from the big companies.
- [07:28]
A
Right.
- [07:28]
B
Like you have Performance Max and Facebook's.
- [07:32]
A
Meta's products kind of his internal deliveries.
- [07:36]
B
Is that what you mean? Like, and it's, it seems like it was shaping up to be maybe not a fair fight. Like the big guys got even stronger. They have so much data, they can optimize on things we can't even, you know, the buyers can't even understand. And the worry is the rest of the industry can't, the agencies can't compete there, the brands can't compete there. What do we do about that? Is that a worry? And then how do we level this playing field?
- [07:58]
A
I think there are two major worries that every brand should have and they in large part have Demonstrated that they have these worries. So I think we're on the right path. The first worry is that their data is their proprietary differentiator and they need to use it and leverage it to their benefit, not to another company's benefit.
- [08:22]
B
Right. So you don't want to make bluntly like you don't want to make a meta smarter on your hard earned data and then you don't get as a brand.
- [08:31]
A
That's right. That's right. And what it will enhance, what it will favor that on that first point is those companies that know how to use their data as a kernel or a seed of the reinforcement learning of what's working and what's not working has to happen within their workflows. They not conform to someone else's workflows and by virtue of that leak their data in the process of it. So that's kind of point number one, what they should be worried about. The second one is the how is as important as the what. And I would argue that they need that organizations need to think about certainly things like regulatory compliance like in finance, legal, and you know, remaining kind of above board in terms of all the things that they do. But then barring those elements how they do, growth is going to be a even more compelling competitive advantage because it's going to be very hard when everyone has access to AI to differentiate. Unless the how is unique.
- [09:43]
B
Well, that, that sort of leads me my next question because I think, I think if anyone's worried at a big picture supposed to be bad for the agencies holding companies, oh my God, we, we don't, we either can't compete with the big eyes on technology or we're all going to use the same tech LLM and build products on top of it and it's all going to be the same. How do, how do they avoid that fate?
- [10:06]
A
Yeah. So a few grounding comments, your question makes me think of what are other patterns we can follow? Now that's only a few months in advance, but it's a pattern we can follow and that is that of software development. So LLMs AI has impacted how software is being built. It is going to continue to impact it massively. And for those who are capable and willing to listen and see what's happening, there is a transformed process on how software is being developed and will continue to be developed. The same thing is going to happen in the media space because AI is impacting sectors by vertical in different ways. There was a lot of feedback a couple of weeks ago about how AI was interpreting legal contracts and Legal is largely language, so it makes loads of sense. Programming is largely language, it makes loads of sense. Is going to be disruptive on a vertical by vertical basis. That's our belief. And if we, our position has been really for the last couple of years, if we are going to remain relevant to our customers, we have to be the vertical leader in media in that process of workflow. So to your question, what is going to change and how is it going to change? I think the modification of these workflows and facilitating that process makes Kochava look like a workflow enablement company when in fact we're a measurement company. But it's our position that measurement is the kernel of all of that feedback loop for learning and so that's what makes us competitively advantage.
- [11:58]
B
So talk about this idea that's been a theme this week at your event. The measurement is less of a scorekeeper, how did I do? But something that's more integral and more pushing the business forward, like how does that change? How does that work?
- [12:13]
A
That's right. If you were to come in from the outside and you would say let me understand how this advertising ecosystem works, you would think of measurement much like an odometer. There's a distance, it's measured, it's consistent, it's a black box. You went 1/10 of a mile, it shows 1/10 of a mile. The reality is measurement is really about multiple different methodologies based on approach. It could be certainly directionally driven on Last Touch. It could be self attribution or outcome based. It could be MMM based or media mix modeling based and that fundamentally the winning organization is one who has a agile and easy approach to looking at how they measure with what their goal is. Right, right. And so with that in mind, I would say, you know, the, the objective for, for us is to be a company that can be, that can enable competitive advantage to. Because they can use those things like a, you know, like a really well designed clutch in a transmission for, you know we added this analogy of racing this week and that's tough for those from the outside who don't know our space. Which is why I think vertical AI is so specific. Everything matters when you're in the space. You can't just apply the problem from the outside.
- [13:48]
B
Who, given the way that software development has changed the way you talked about who is going to the like. I actually started my career in the agency world and it was a lot of marketing majors or liberal arts students and then in the last 10 years we got to get data scientists in here and Figure that out. What's it going to look like now? Who do you need to come in there?
- [14:05]
A
I think it starts to flip. So we're seeing this in software, like that's our precursor. Right. You don't need the skill that is most useful when engaging with a model is the ability to articulate your desire. Okay. It is not data science and programming.
- [14:26]
B
So it's literally writing and reading well and being able to communicate well.
- [14:32]
A
Right. All right.
- [14:33]
B
I'm okay.
- [14:34]
A
Yeah.
- [14:36]
B
But that is a really different set of skill. I mean, I'm already old. It's over for me. But no, but that, that is a different kind of recruiting mentality because I think that people wonder who is going to, who are going to fill the seats for the next generation. What's going to, what's the next wave of talent look like in this industry? With what. With the needs, where the needs are going.
- [14:55]
A
Yeah, well, I think, you know, if you extrapolate and you run the tape fast forward, there's a whole bunch of tasks that are getting done by people that don't need to be done by people and they can get reallocated into other more meaningfully interesting things to do. Strategy is going to become really important. Uniqueness of inventory is going to be really important. Access to data is going to be really important because it's the driven that's the, it's the mechanism or the scaffolding upon which decisions are made. If you and I both can do the exact same thing, but I'm informed in a less effective way than you are, you're going to do it better. And so it's going to be, you know, the things that have, that we have taken for granted of free access to data, I think start to change and those, those knobs start to turn.
- [15:52]
B
Coming back to this, the, the agentic world we're envisioning. Right. You've seen a lot. There's a lot of experimentation going on and, and test cases. How far away are we from or will we get to a place where the agents are not only doing a lot of the grunt work and the research and helping letting the humans freed up to think, but where they're going out and executing buys for clients and optimizing in real time without someone driving? Because I think there's some worry that that's going to be dangerous with clients dollars. We never really let them go crazy like that. How far away is that?
- [16:30]
A
When you hire a trader in media, you give budget allocation so that they can't make too big of mistakes. When you hire a trader in equities, you give them allocation of security budget so that they don't make too big a mistake. I think the same thing is going to happen with agentic execution that there are going to be guardrails. And what's really important is that those guardrails are productized. So we're, you know, with Station One, which we've talked a lot about this week, and for, for your audience, you know, it's a. It's an integrative AI hub. It enables you to integrate with any model, connect any set of skills that you can then kind of codify for your team. You can syndicate those skills across your team. You can integrate things like knowledge bases which are vectorized rag interfaces and it's effectively like a slack for AI. Okay. What's slick about it is that it connects to your AI models. That is the point we made a moment ago that like this needs to be owned and controlled by you. Not. And then what's also slick about it is that you can containerize these things into workspaces and then syndicate them. So when we talk about team dynamics and how do you enable transformation of your team to do these things, you are going to. I think there's going to be layers of an onion where you're going to start by having your existing team using a common set of tools. Those tools are then going to observe what you're doing. It's then going to have an opportunity to say, can I have agents that have been doing the things that Charles and Mike have been doing on the floor, model out their daily activities from Monday to Friday? And I want to engage with this orchestrator thing and start to identify what things can Charles and Mike no longer do that. We can just mechanize agentically. Then it'll be, well, that was pretty slick. Let's amplify and let's create tiering and maybe we'll have some traders that have budget constraints and then we'll have agent managers that are observing the behaviors of those trader agents that then have guardrails. Right.
- [18:51]
B
What if it's realizing that Mike has barely done anything for the past couple of days? That seems like it.
- [18:55]
A
I mean, there's probably some reallocation opportunities. Yeah.
- [18:59]
B
Okay, what about, you know, when we last spoke, you got your coachaba has roots in mobile and you're very much pushing into ctv. Catch me up on like your journey there and how have you seen things evolve? Because television has historically been measured sort of one way, broad awareness, for better or worse. Everybody had the Same kind of measuring stick. Now it's really evolving. What has your journey been like and where are things headed?
- [19:25]
A
Yeah, you know, our origin story @Coachava started by deciding that we think apps are the future in a world where everyone was talking about mobile web and our position was mobile web is cute, but apps are going to be the thing because you have an embedded wallet potential. There's access to all these services, like a camera.
- [19:47]
B
Direct to consumer relationship.
- [19:48]
A
Direct to consumer relationship. It's like a tether between a brand and a consumer in the pocke of the consumer. I mean, phenomenal. Never been. You know, it's amazing, right? And we were like, we want to build DoubleClick for mobile, even though obviously DoubleClick already existed at the time. And that was our conviction. And so we did that. When we started to see what was going on with ctv, what we were observing was the OEM saying we're building an operating system. And we immediately had flashbacks to our origin where it was like, this is an 80 inch mobile device in the living room on the wall, which is.
- [20:27]
B
It'S going to be app centric really quickly.
- [20:30]
A
I mean, if they're building operating systems that aren't going to be some hideous thing, or they're licensing Google's operating system, this is an app economy all over again. And there's going to be a lot of interesting interplay because everyone's been fighting for years to have a living room device. And here it is. I mean, you could put a gpu, this isn't really happening at scale yet, but you could put a GPU and a tv. I think this is going to happen in the next year or two. And all of a sudden you don't need an Xbox that does compute and rasterization vectors.
- [21:01]
B
It's just powerful, it's strong, it's in the TV as whatever you got in your favorite gaming device.
- [21:05]
A
And if you have a game experience, you're going to have a really immersive commerce experience on the tv. And if you're going to do that, then you're going to have some really interesting streaming use cases that are going to be fascina. So we were like, we want to build an SDK for every operating system. Like, we need to be the perpetuated pushed solution set. And we thought it was as simple as building the SDKs. And as we engaged with all the CTV OEMs, they had come to the problem with great wisdom, having seen what happened on mobile. And they said, well, let me get this straight. If advertisers Buy my media and it's trafficked with your tag. Your coachaba tool has all this awesome ability to reverse engineer audiences and find them on other channels. That's not good for me.
- [21:57]
B
Yeah, we've seen how that plays out. It's not good for the media owner.
- [22:00]
A
That's not fun. And we said, okay, we get it. Good point. What if we built so that in service of the advertiser, what if we built a separate piece of tech which we now call Atlas Performance, we call it Kochava for publishers where you can control your destiny. But we can also, in a clean and safe way, coalesce data so that outcomes can be observed. You know, lg, Samsung, Vizio, Roku, all of them became customers very quickly for that reason. So here we are standing today in 2026 and you ask, you know, what's your journey? So this was all built during 2019. We shipped it the end of 19 Covid hits February of 2020. And there is like a, you know, known desire to just binge stream from.
- [22:54]
B
Every household using this device a lot more than they were.
- [22:56]
A
Yeah. So now all of a sudden magically we're, you know, in CTV measurement. But it all happened because of that origin story where we saw patterns happening and we were following the patterns.
- [23:06]
B
Now so that you're right in that it mirrors a lot of the mobile app economy, but the behavior is, and it's getting to a place where you said there's probably more commerce, but behavior is different on television 100 we've been talking a lot about this week about the. How should we be measuring outcomes? Do they need, do we need a longer period of time? Should we, should all TV be performance based or not? Do you think we're going to, are we going to rally around one thing or one way of doing things or do you have a way of thinking about where this performance TV was headed?
- [23:37]
A
Outcomes absolutely matter. The view and do combo is you have an immersive experience on the TV where you're viewing and then where you're doing is on your mobile device. And a measurement solution has to account for that dynamism and ours does. And it translates it into a business outcome that's ultimately awesome for endemic media buyers. So folks that have like fast channels and they're promoting tunein and various different things, but it also really helps QSR or it helps financial services. So you know, to what extent do you get promoted an opportunity to have your Next buy on McDonald's have a free item if you sign up for the loyalty program and Then there's an ability to tie off the connection that that ad was shown from that household and then acted on from that mobile device. And that's something that is hugely valuable to brands.
- [24:39]
B
Ever wonder what the future of ad tech sounds like? Hint. It's happening right now in front of a live studio audience at the Kochava Summit in Sandpoint, Idaho. This episode is powered by coachava, where data meets innovation with little skiing on the side. Learn more about what's next for marketers, publishers and platforms@coachaba.com seriously, your future self will. Thank you. Now, it's still awfully early, but it's been interesting with CTV advertising. It's still, you know, it's a pretty small universe of like a dozen big companies and. But it's, you know, it's fairly fragmented. AI hasn't really taken hold in the way, at least the way that it has on with, on meta, Google, where the automatic optimization thing can happen because everyone's kind of doing their own thing. Do you think that can and should happen?
- [25:24]
A
Yeah. So there's two or three elements to your question. So why is it happening in the big companies like a meta and Google, they have control over the end to end experience, so they can deal with that. The CTV OEMs, they have this, they certainly have control over their own areas, but the reality is, is that they're one of a composite set of media plans that are being bought against and they have to conform to the standard workflows that are currently being done. So they're a little bit at the mercy of the media buyers at the agency or at the brand. And that's fundamentally what we're so excited about with Station 1 is that you're really facilitating this to happen in a way that's additive to everyone in the ecosystem. This is not additive to Kochava or one particular publisher. It's actually additive to make all of this more efficient. So that's kind of one point on why is it different between the metas and the Googles versus the the other OEMs or, or even programmatic against CTV. I think the second piece is that on the publisher side of, of CTV that there's still, there's still folks that are just announcing and we had it happen here today, like CTV is in fact tv. That was a great quote from today, you know, little reminder to everyone. And then second, it's absolutely performative and it's a function of the how and the approach. And if both of those things are right, this group understands that. But now it's about how do you get the trillion dollars in ad spend to change in their behavior in the context of those, those facts?
- [27:10]
B
And do you think that there's been a great hope in the last couple years that this will bring way more advertisers to television over time? And I think there's been some evidence of that, but it hasn't maybe hasn't been the wave that we were hoping for. As fast do you think that's coming?
- [27:26]
A
It's 100% coming and it's coming faster than we think. And I think what's going to happen is there's going to be all this disruption and transformation in people's jobs because of AI. And then when people lift their head up and they realize where they're at, let's say in nine months or in 13 months or in 16 months, they're going to be like, oh, would you look at that? Media mix has totally changed. Right? And it's going to be like, oh, there's actually a relationship. There were folks who wanted to keep it the same way it's been.
- [28:00]
B
We thought that they thought they had the right media mix. They felt confident in that or they'd.
- [28:04]
A
Been told that from endless number of people who no longer have a job at company X or Y or Z.
- [28:14]
B
The rethinking to happen, the retooling you've talked about and then all of a sudden you've got a lot more brands of television. Hopefully time's going fast. We didn't really talk to this. Does anybody want to jump in with audience questions? We could do that if we want. If not, I have more things to talk about. No one wants to do that.
- [28:33]
A
Here we go.
- [28:35]
B
I could barely see.
- [28:37]
A
How do I prove to my CFO that CTV is incremental to what we're doing on Facebook? And how do I get them to commit to spend across multiple months? That is an awesome question.
- [28:51]
B
Because it's hard to isolate connected television advertising versus other things.
- [28:55]
A
Right?
- [28:55]
B
Is that, is that at the heart of that question?
- [28:57]
A
Yeah. So I remember in the early days of Coachaba, there was this company called Machine Zone who was a customer of ours. And Machine Zone was made up of Gabe Layden, who was a very aggressive and awesome operator. He just plowed forward, had this guy named Deepak Gupta who was his CRO, so bought media and generated revenue and he was this kind of right hand man and helped drive things. What I learned from the two of them is media relationships and they did an amazing job of understanding and then keeping it a secret. From everyone else that when they do X and they do Y, irrespective of what they these signals are, this other composite signal says success. Okay. They were really the first gaming company that bought super bowl ads and at scale linear tv, let alone CTV typically.
- [30:10]
B
Is a very data driven world. Right. That it doesn't work that way.
- [30:14]
A
And part of the operating stance that they took, I mean this was 2000, help me out, 2014, something like that. So it was like before CTV was happening. And what I learned from that exercise was what I think we've perfected in mmm, which is it's all about the mix. And it's not just about one last click or one self attributing signal. It's coming from one source. You need to look at the mix together. The problem with media mix modeling or incremental measurement as a whole is it's like exercising really hard for four hours a day and not having it have an impact at all to your figure.
- [31:00]
B
That's why I gave up.
- [31:01]
A
It's like all the stuff that is not fun and none of the reward. And it's because people, once they go on this voyage of media mix modeling, they realize their data's not that clean. They realize they have to go exercise that and fix this and fix that. And so what I think we've done a really good job at and what we've learned in this process as well is that media mixed modeling is a function of making it easy to bring together data, leveraging AI tools so that you can have a lot of forgiveness of sins that are otherwise like really pointed if you have to manually deal with it. So I would answer that question with I don't think you can do what that person was asking without media mix modeling. And I don't think many organizations were really well outfitted to do media mix modeling until really just this last year with some of the access to these AI tools that vendors like us can bring to the table.
- [32:03]
B
I want to ask you, we've talked a lot about how agentic and the AI era is going to change things for agencies, for brands, for the media companies. What happens to ad tech? That's a broad question, but there is some thinking that once agents are out doing the things that we're asking to do, they'll take on some of the work that DSPs do or SSPs do and that could thin the herd or bring spending closer to the media companies, which could be good, but that's not good for everybody.
- [32:31]
A
What do you think happens when. What specifically what do I think happens.
- [32:36]
B
When what specifically are we going to have the daisy chain of, of folks in between brand and execution that we have today?
- [32:45]
A
Yeah. I think there's going to be an awful lot of collapsing.
- [32:50]
B
Okay.
- [32:51]
A
Yeah. And there's a re. Rationalization of what is a role. What is the role of something. So like SSPs who always said their job was just to aggregate supply magically now have ways in which you can buy media from their supply.
- [33:12]
B
Right, right.
- [33:14]
A
DSPs who have always said that their role is to aggregate demand and work with all the available SSPs have their, you know, in their own forms and their own names and their own approaches, unique ways to have unique inventory available.
- [33:26]
B
They're both kind of going around each other. Right.
- [33:29]
A
So there's a collapsing naturally happening at those two big goalposts. The second thing that's happening is we did this summit, I think it was in 15 or 16, and we had visual for every ad dollar that gets spent, there was all these different ways in which you could spend money that was actually not going to the media. And we reduced down, you know, how much of the quote unquote ad tax was being applied and why every dollar needs to get put to work. And it's one of the questions that's being put up here. Right.
- [34:07]
B
Is there a best, is there less money in the marketplace or more? Because there's the Zuckerberg idea that if there's more, I think there' but maybe less tax.
- [34:15]
A
Yeah. And so there's going to be more money to be put to work in media. There's going to be less tax because a lot of the things that were made up of all the pieces that would otherwise be taxing is now happening agentically. So you may spend more on, let's say a fleet of agents than you've ever spent in the past, but it's less than all the other middleware and.
- [34:43]
B
Components that go to your outcome based campaigns. What you care about.
- [34:48]
A
What's really going to be interesting about all this is are you going to start to see publishers building exclusive inventory access at ultra premium rates if they start to really understand how they drive the needle.
- [35:02]
B
That's interesting.
- [35:03]
A
So we have a customer who is in the sports betting space and they really know how to identify the whales and boy, do they have a very good approach to how they monetize that because they know what they have. And that's pretty interesting.
- [35:22]
B
You could see that playing out outside of the gambling world, but the rest of the rest of the industry.
- [35:27]
A
Part of the dynamic that happens between a Buyer, a savvy buyer. And we've always tried to position, historically we always tried to position that we want to make our buyers more savvy than their suppliers so that they can negotiate better. Right.
- [35:44]
B
Yeah.
- [35:45]
A
Right. And increasingly I think suppliers are going to be a lot more savvy. And so there needs to be an equal footing of capacity that of what is the value and what's being delivered. And there's a huge opportunity to flatten.
- [36:05]
B
So that changes marketplace dynamics, pricing, all kinds of things. That's interesting.
- [36:09]
A
Yeah. And I think by the way, you didn't ask this, but I think it's an awesome extrapolation. So whereas Programmatic was really around auctions on the per impression or the placement, I think we're going to start to see auctions handled at the IO level. Interesting. So it's like, it's almost like what, you know, effectively like a. It's almost like a at scale. Programmatic guaranteed.
- [36:37]
B
Okay.
- [36:38]
A
At an I O level. And I think that's going to be possible because that's administratively impossible to do with people. Agentically very possible. You get the workflow, you get the.
- [36:49]
B
Control that you like. But the scale that you've been missing.
- [36:51]
A
Theoretically, I want to put that as a byline. All right. I like that. The control that you like and the scale you've been missing.
- [36:59]
B
Last one, Charles. So what's in our way? What is in your way? What's your biggest obstacle or the future that we're describing here? What's going to hold us up, Brandon? What worries you? And I ended a negative like I always do.
- [37:12]
A
Yeah. Like. Like, right. Like I do. Right. So I, I think, I think what is facing us over the next 16 months is going to be so profound that it's like. It's like Elon taking Twitter private. It. You couldn't possibly and there's a lot of opinions about did he do the right thing with Twitter? Did he not like. And I'm not even, not even assessing that. But there are certain things that are really hard to do as a public company and I think being private over the next 16 months is going to be considered gold to not have to.
- [37:59]
B
Go through all this change in a naked feeling.
- [38:02]
A
Yeah. And so what's going to happen I think is a crap load of take privates.
- [38:08]
B
Interesting.
- [38:09]
A
Combined with a crapload of roll ups that aren't going to be purchasing tech because tech's recreatable really easily with AI they're going to be purchasing contracts.
- [38:23]
B
It's a time to regroup without the scrutiny.
- [38:25]
A
And then in two years, a like, bonanza public offerings.
- [38:31]
B
Interesting.
- [38:31]
A
Wow. That's what I think the next 24 months looks like.
- [38:34]
B
That is actually a nice, hopeful way to end this, but awesome conversation. Charles, thanks so much for everybod part of this.
- [38:39]
A
Awesome. Thank you.