← Next in Media
Loading summary
Transcript104 lines
- [00:00]
A
At the end of the day, it's not just a story, it's actually just something that drives better performance. When you have real high fidelity and durable data. The core ability to have an individual based graph with a medium like ctv, that's what leads itself to better performance based marketing and better outcomes. Because we don't think having agentic is a strategy, it's how you operate. It is a strategy.
- [00:22]
B
This one feels like it could be fewer people, 10 people doing the job of 100 kind of thing.
- [00:27]
A
Table stakes are huge. The quality of content has to be really good and accurate, written with good context. We've evolved the thinking of what used to be more of a tech writer type functionality. It's really an AI librarian type function becoming really critical. Let's just call it open Addressable Biddable Inventory the amount of live sports we've seen double on the platform platform over the last nine months alone. The amount of addressable biddable premium CTV inventory, audio inventory has gone up so much. You talk about an enterprise marketer's needs and the table stakes there, you have no margin for error.
- [01:09]
B
Digital advertising is facing growing signal loss across browsers, devices and connected TV that makes accurate identity more important than ever. Intent IQ is a privacy first identity resolution leader, helping advertisers, publishers and platforms recognize and reach real audiences across both cookie based and idealist environments. Powered by a patented identity graph and advanced signal enrichment, Intent IQ delivers the scale and accuracy needed to drive measurable performance and better monetization for both advertisers and publishers. To learn how Intent IQ helps turn fragmented signals into real results, visit intentiq.com this week on Next in Media, I spoke with Adam Ruderman, General Manager of Yahoo. Dsp. Adam and I talked about the ongoing DSP wars and why Yahoo is in it for the long haul, as well as the company's somewhat accidental positioning as a leader in identity in ctv. Adam and I also dove deep into the future of agentic advertising and the very weird industry debate over protocols. Plus a troubling story about a fight between Adam and his own agent. Let's get started. Hi Everybody. Welcome to Next 2 Media. I'm Mike Shields. My guest this week is Adam Ruben. He is the general manager of Yahoo. Dsp. Hey Adam, thanks for being here.
- [02:17]
A
Thanks for having me, Mike.
- [02:19]
B
Adam, you were in a rare club of returning guests, so you should be excited about that. But I'm pumped to talk to you because you guys have had a lot of you're in a very interesting space. There's a lot of interesting News the last time we talked, I remember having the conversation Yahoo had made news for shifting its strategy a little bit in ad tech. Take me through the last couple of years and how that, like how that has evolved. You, you went, you remind people you, you are kind of on both sides of the market. You're really focused on the SP business. What's a lot, what's the evolution been like since then?
- [02:55]
A
Yeah, you're right. The last time we chatted we were just coming off the heels of going from what was a multi pronged ad tech sort of strategy around not just dsp but we also operated an ssp. We also operated a native ad network. Yeah, simplifies that. Around the time that you and I chatted, I can't remember how far along we were in the process. But to your point, now it is solely demand side, platform centric approach. If I think about what's changed, I mean a lot has changed. I'd say, you know, not so much from our strategy perspective but like the ecosystem has moved quite a bit in the last year. And I promise you I will wait till you say you want to talk about AI and I will.
- [03:36]
B
Yeah, but I don't think we talked about it last time. But we will, I promise you.
- [03:39]
A
That's the funny thing, I don't think it was even. No on the tip of our tongues and if it was, it was probably generative, not definitely not agentic.
- [03:48]
B
So.
- [03:48]
A
Right, no talk about the nuances there. But I do think the one thing that's most notable between then and now would be the supply, availability and footprint for addressable and biddable back then versus today only just a short number of years ago, particularly around premium ctv.
- [04:07]
B
Okay, I definitely want to get into that. This, I have a macro question that maybe this is like I don't know if you can answer this or this would be like for your, your bosses but I think it's interesting to people from the outside maybe that are not that close to it that Yahoo is even you could have argued like why stay? Why stick with ad tech at all? Like yeah, it's a, it's a dsp. Wars are really competitive. You see what's happening to the trade desk going back and forth with Amazon. You got this monster consumer business like why do this? Why bother to exist?
- [04:43]
A
Yeah, it's a good question. Sometimes I ask myself, every morning I
- [04:47]
B
ask myself a lot, am I doing this? But
- [04:52]
A
the real reason that it lends itself well to a programmatic ad stack strategy would be the data footprint and the consumer footprint. Does give us a very strong advantage in terms of delivering incredible performance because of that owned and operated inventory footprint that we get to take advantage of and build off of. And you know it's that that core and I think the differentiation in market is, is what is well received but at the end of the day it's a, it's not just a story. It's actually just something that drives better performance when you have real, real high fidelity and durable data. That's, that's the, the answer.
- [05:33]
B
So that, that leads to the next question because I think again people might not have expected everybody in adtech all in the last few years before we started talking about Agentic was talking about getting into ctv. We got into C CTV more as that becomes open to to players. You again Yahoo doesn't is yah is not a inherently a TV company or by historically at least. So you might have thought well what do you, what are you guys doing there? But you have, you have talk about your identity layer and how that can be advantageous in television even though you're not, you're, you don't have your own, you know, streaming service with your own logged in user base look. But it's, it's, you have a lot of, you have a lot of users.
- [06:13]
A
That's right. So you're right in that like the original Yahoo DSP value prop going, you know, well before you and I spoke, you know, back in time, had a more of an inventory flair to it. The DSP synergy with the inventory. Yeah right now. But, but correlating back to what I just said around the data footprint, you know, when we built that identity graph and started building it, CTV was nascent at best.
- [06:42]
B
Right.
- [06:43]
A
It was obviously what a lot of the folks in our industry were doing which is planning ahead for a world without cookies. And so we looked at the data footprint that we had and we invested in you know, this people based graph with actual identity resolution to it at a time when we thought it was going to be used predominantly there. And really what it's turned into is this juggernaut for CTV buying where you do need to use you know, an email to log in and that email based identity graph has turned into something that we see activated on over 90% of the CTV we buy. So when you have that like, you know, deterministic ability to do so, your, your, your accuracy goes through the roof. And when you have higher accuracy you can bid with more conviction and you can deliver a better outcome. And so it's, it has turned into a I can't say that, you know, that was the original use case for building the graph that it's turned into an absolute, you know, pleasure to have it in our arsenal.
- [07:46]
B
Talk about that because I think the, it's like you, you, you, you spoke about that there. A lot of, a lot of people, a lot of companies made preparations for this, you know, cookie Armageddon. Yeah. And CTV Cookies was never a cookie, but based medium. But it doesn't have a universal identifier. I could argue that there is no universal identifier and maybe some would say, well do we need that in television or not? Or is that, is it going to work like that? But you're finding a huge amount of interest in using your data. Like give me. So that's, I guess what is the state of identity based targeting in CTV
- [08:19]
A
and where is it going? Well, it's, it's, you know, obviously there's the ability to do to and the desire to use household graph targeting as well. That's, that's, yeah, obviously.
- [08:32]
B
And there's some debate about whether that's the, we go one way or the other.
- [08:35]
A
I guess like that's a, that's a tactic for certain folks. Again, do you think we would have come up with household based targeting 30, 50 years ago if we had the ability to do individual? Probably not. You know what I mean? And so now that we have the option to. There may be circumstances where you want household and we support both. But the core ability to have an individual based graph with a medium like ctv, that's, that's, that's what leads itself to better performance based marketing and better outcomes.
- [09:07]
B
That's interesting because I think some would say, well, no tv. The reason TV is we should go the household direction because it's a communal thing. It's never going to be one to one. But the way that younger people watch or really all of us, sometimes you watch TV on the couch with your family and sometimes you're doing your own
- [09:25]
A
thing, but you get more and more rare.
- [09:27]
B
Yeah. So it's not like, it's not like one way should. We should just by nature, household is the way to go.
- [09:34]
A
Yeah. For, for me, I, you know, every marketer has their own, you know, preference to how they want to operate. But the ability to have the individual and have it be, you know, accurate. Especially when to your point, the, the family is probably the original use case and if the family is streaming on different devices right then same house and you want to be able to, I mean, I know my family, you definitely don't want to target me with the stuff that you intended for my teenage son. That's probably not my consumer differentiant. So I, I would say that that is a new development that is well supported by having individual based identity graphs with people resolution.
- [10:12]
B
All right. Now it's still, it's still relatively early in trying to turn to you. You. Well, you. I guess you could argue, should TV be a performance vehicle? Most would say, I guess. I think so, but in a different way. Um, but how, how much of what you're doing is what like you said, targeting very. Either very specific audiences with messages or really trying to do some of the performance things that we think hope CTV can do.
- [10:36]
A
Well, I think those two go hand in hand. I mean targeting individuals lead to good performance. I do think this medium allows you to go full funnel and so you don't have to pick.
- [10:47]
B
Yeah.
- [10:48]
A
Or that's.
- [10:48]
B
And there's still lots of categories that just want to reach a lot of people and don't want to go crazy at Oreos or whatever.
- [10:53]
A
Yep. And, and I still too. You know, we don't do this stuff in a vacuum either. I think the, the power of the platforms today, you know, ours and others is being omnichannel and not just isolating CTV by itself and being able to have a full funnel that plays nicely with all the other mediums. There's obviously great growth in audio as well. So. And that's a very similar medium with you with the authentication opportunities. So I don't think it has to be one versus the other. I do think there's a weird interest around asking whether something's performance or not. Everything can be performance.
- [11:29]
B
Right.
- [11:30]
A
And maybe it makes for good panels. Maybe you just did a panel on this this week and it's not mine,
- [11:35]
B
but there, there were, there were a few of these topics that. This was a topic at Market Extra Live for sure a couple times. Performance tv. Yeah. Can't get on my mind.
- [11:43]
A
You know, everyone at the end of the day is looking at some level of performance and how granular they want to get. Whether that's, you know, performance can be measured in networking media towards a target household or it could be an in store purchase. There's just such a wide gamut. But it all relatively comes down to some form of performance. Right.
- [12:06]
B
All right, let's. Let's table CT viewers for a moment here. Let's talk about you. You're gonna be in my, my agentic translator a little bit because. And you guys had a bunch of announcements at CES you're. It seems like again I have market extra live in my mind, but there's like a segment of the community that is like going crazy. I can't wait. They can't wait. So much enthusiasm around this. They want to talk about protocols. There are others, I think that are not sure where it's headed or even a little. Maybe skeptical is the wrong word but like wary of how all in we're going to be with agents. So all right, that's a big question. But give me like give us what you guys are announced, announced recently. What are you doing there? And then I want to talk about like where do you think the industry
- [12:51]
A
is going to go? Sure. So what we've been external with today going back to ces like you mentioned would be our, you know, approach to Agentix specifically because we don't think having Agentic is a strategy. It's how you operate it is a strategy. And so one thing that we made really, really clear with folks is that it's not all going to come within development inside of our native, inside of our DSP or native to our platform itself. We see an opportunity to be a little bit more flexible than that. We look at, you know, going back to what makes a DSP so potent for performance would be the core assets of the identity graph, the data that we use for targeting, for measurement, our machine learning that hooks into our bidding, all that kind of stuff. Right. So that's the core of what makes the DSP special, is the ability to get access to those tools is what we want to enable. And if we can do a better job of that through Agentic, then we need to make sure that we're being flexible to the buyer. So what we announced is yours might and our strategy round, meaning if you're a customer and you want to use our platform and want to create your own agents for simplifying your workflows with DSP and other tools, you can develop those agents on your domain and then access through our mcps.
- [14:19]
B
Okay.
- [14:20]
A
And. And maybe never even see our ui if that's the way you want to transact with us. If you don't have those resources and you're working with the many, many innovative companies in the space that are popping up, that we have the ability for those third party agents to port into our MCPS as well and work with our agents and go agent to agent just as flexible. And then lastly, obviously the HOURS concept, of course we're going to have agents built into our dsp. We've got a handful of them live already. And if you want to log in and use the agentic workflow within our dsp. So it's that flexibility of being able to either develop on your own, develop with a partner, or just leverage a new version of our DSP that we wanted to make very, very clear to the market that we're not picking one way to use agent. So.
- [15:13]
B
So it sounds like what you're describing is at least in the short term you probably know that some of the big holdcos are going to build their own tools. So you want to be ready for that. There are going to be some big time third parties that emerge that'll that are offering these kind of things. You want to, you want to be able to work with them. But then there's what small agencies, smaller brands that are not going to have this stuff and they, you want to have stuff for them to do that.
- [15:37]
A
You nailed it.
- [15:38]
B
That's. Do you think the is there analogy, like again, we're in the early, early stages of this. If is is it going to evolve like the DSP world where there was a time when agencies bought their own DSPs and there were a million of them and eventually it settled into, you know, they're mostly, it's mostly a third party thing that is like are, are agents going to be easy to maintain that so many companies are going to ultimately use their own or is it likely to consolidate in terms of DSPs or like IS. Are the, are agents going to be. Is the yours, mine and our ours thing eventually going to settle into one?
- [16:22]
A
I'm not sure. I think, I think if I understand the question. You want to know if, if folks are going to pick one agent, like
- [16:31]
B
yeah, or I guess or one way of working. Yeah.
- [16:37]
A
I do think the workflow element of a UI doesn't need to go away immediately. But I do think, I think if I follow the question. I don't know. I don't know.
- [16:52]
B
Okay, thank you. I love that answer.
- [16:55]
A
I would love to tell you I know how this is going to play out and I've got lots of theories and I definitely think a lot of folks have some theories as well. I just know one thing. I know that what makes the DSP platform go and work and be the most powerful has the least to do with the familiarity with the buttons and the navigation through the workflow. It has most to do with the graph and the data and the proprietary bidding technology. And if using that is a big switching cost, you know, to, to leave a platform that you're comfortable with to move to, to ours or any other and this eradicates that then I think the ability to use multiple platforms and see what works best, it's going to get easier and easier and that's a win for marketers. How the trend will follow, I. We'll see.
- [17:50]
B
Okay. Help, help people understand how again there are so many unknowns and this is all going to evolve rather quickly. Yeah, you know like I saw there was a guy at Market to bring it up again that spoked about this is the most profound change in marketing since the.com boom. And then. But if I take a step back as a non, you know, as a, as a third party, I'm like Programmatic is, was already billed as high speed, automated, powerful. What's the big deal here? Is it just saving, you know, spreadsheet headache time or is this, is this potentially far more dramatic than that the agent revolution we're talking about?
- [18:29]
A
So what Programmatic sped up was the act of bidding on multiple on 2000 sites instead of having to transact and all that kind of stuff. The support and use case of that for analytics and trading and management and reporting and trafficking. That that's still a lot of work that, that it didn't solve for something. It created some new workflows that we didn't have pre Programmatic. So Agentic does come in at a time to address that stuff in the workflow. I mean some of the most boring use cases for MCPs around trafficking are going to be the biggest time savers. So it's, it's Programmatic was never really sold as like this, you know, time saving mechanism.
- [19:15]
B
Well if you go back though, it was like we're not going to need salespeople, we're not going to need people at all. And then the opposite happened. This one feels like it could be fewer people. 10, you know, 10, 10 people doing the job of 100 kind of thing.
- [19:28]
A
It's changing the output of a power worker today, that's for sure. And the out if the output goes up then it changes what these power workers are capable of. Whether that impacts the ratios of, you know, like a good example of that would be for us as a dsp, the most common, you know, ticket that we probably solve for his account log. I forgot my password. It's the most commonly, you know, submitted ticket for support. And you, you look at someone like that and you go man, that, that seems like a very easy one to be solved with an agent. It's not a strategic, you know, troubleshooting event. Now what do you do with that bandwidth, Right, Right. And so certain sort of support workers can really uplevel what they can work on and now the bandwidth just changes completely. But it doesn't mean that you change the number of people just because of a use case changing overnight.
- [20:31]
B
Reaching the right audience on streaming TV takes more than guesswork. It takes intelligence and SA delivers. Savio is redefining what it means to be a modern media company where data technology and creator driven content work together to deliver real results at scale. At Its core is AppScience, the only company building audience segments for mobile and CTV behavioral data, reaching over 70% of the US streaming households. How do you activate SAO? SAO is platform agnostic, meaning you can run campaigns through your preferred DSP or tap into sabio's proprietary platform with its owned and operated Creator television network. SAO unites media, data and culture, giving brands a direct line to engaged, authentic audiences. Sabio's full stack approach, from ad serving to audience insights to creator content, brings every piece of the media puzzle together. So campaigns don't just perform better, they connect deeper reach modern audiences with sabio, the media company built for how people stream, watch and connect. Learn more@sabioctv.com that's S A B ioctv.com okay, this is hard for me not to think. As a former journalist who's covered a lot of crazy ad tech things gone bad stories, how does this not become a ripe with, you know, fraud or mistakes? How do you avoid the hallucinations or just the crazy stuff that happens when, when the bad guys figure out how to, how to work these, work these systems?
- [21:52]
A
You got. Oh man. And if you, I mean, look like I, I got into a fight with one of our internal agents literally this week.
- [22:02]
B
You got in a fight with the agent?
- [22:03]
A
I did, I did. I got in a little bit of a verbal spat.
- [22:07]
B
You know, is that healthy or is that maybe I'm not.
- [22:09]
A
I'm frustrated because. And it turned out it was because I didn't realize how much setting was to Greenwich meantime and that I was in Pacific Standard Time and I was saying something about yesterday and I had the dates wrong. And so, you know, there's a lot of opportunity for humans to get it wrong as well. But to your point, what we found to mitigate, you know, classic hallucination, we've taken approach of making sure that everything we do is isolated into specific knowledge bas information. These knowledge bases, which historically been written for things like documentation, are now being written for agentic consumption of, of of information. And the table stakes are huge. The quality of content has to be really, really good and accurate, written with good context. We've evolved the thinking of what used to be more of a tech writer type functionality. It's really like an AI librarian type function becoming really critical. So it's, it's, it's who's doing the work, it's how you structure it in these individual knowledge bases and then what we build on top of these knowledge bases. So you can have a knowledge base on frequency capping, a knowledge base on traffic and knowledge based on all the different aspects of the demand side platform. But then you need a graph and not the identity graph stuff that we've been talking about, but you need to graph all those different knowledge bases together so, so that when a orchestrator type prompt comes in, it knows how to ping different knowledge bases. And I might be losing half of your audience right now with a little boring level.
- [23:45]
B
I am following along perfectly.
- [23:46]
A
I am obsessed with it because it's all you. To answer your question, you have to be really good at documentation and you have to be good at structuring your data and your information and your content so that things don't trip up. Because to your point, it's one thing to come up with an agent that's autonomously managing, you know, $500 a quarter of, you know, self serve marketplace type activity. When you talk about an enterprise marketer's needs and the table stakes there, you have no margin for error. So you have to be very, very careful, make sure that there's human interventions, SSO authentication, all that kind of stuff so that nothing comes off the guardrails.
- [24:26]
B
All right, what about this is. This is probably a very naive question, but you were talking about this excitement around agentic advertising and buying at the, at a time when everyone is sort of like, well, the web is shrinking and it's not going to matter anymore. So that opportunity seems smaller and then I, I'm, I'm not, it's not clear to me whether the CTV opportunity is going to welcome agents because they are so the TV guys like to control things and they're still doing a lot of direct deals and maybe I'm wrong. Maybe. But like how does, is there going to be enough of an opportunity for this to really amass what we hope it does?
- [25:07]
A
Yeah, I'm seeing. So our inventory footprint of what we consider the Open Web is. It's going up, not down. Yeah, there are pockets of, you know, web inventory that maybe are not trending up, but there is so much new addressable Inventory in audio and CTV that I'm not.
- [25:28]
B
Okay, so. So we're broad. That's partially broadening what we call the open. What we think of as the open web, I guess.
- [25:33]
A
Right. Yeah, I, I do think maybe there. I'm happy to come up with a better term because you're right. Open web sounds like websites.
- [25:40]
B
Yeah.
- [25:41]
A
But, you know, let's just call it open addressable, biddable inventory. The amount of live sports we've seen double on the platform over the last nine months alone. The amount of addressable biddable, premium CTV inventory, audio inventory, has gone up so much so that liquidity of inventory is.
- [26:03]
B
That's not a problem.
- [26:05]
A
Not concerned. It's not keeping me up at night.
- [26:07]
B
And so the big TV players, because they're all talking about, we could buy the Olympics this way. There's only so many more ways to get into television and. But then you hear from buyers. Well, mostly it's private marketplaces and direct deals. Those two things can both be true, I guess.
- [26:25]
A
That's right. Yeah. It's that simple. They can both be true. Yeah, a lot of stuff. You know, you mentioned the Olympics. That was a very programmatically available, you know, okay. Event. I mean, you don't even have to go that far back. You can go to the wbc. It's going on right now. We've got, we're seeing that come into the programmatic marketplace. That's just, you know, in the last five days. Right. And whether or not you need a PMP or it can be open market biddable, you know, both are supported. So for me, they're both considered opportunities.
- [26:56]
B
All right, let's. Let's really lose the audience here with this one. Why. Why might I have a very rudimentary understanding of. There is a debate going on about what language we use or protocol we use to build this future. And like the IAB is saying, don't mess with what we've got because it's crazy to try and rebuild all this. And then there are other parties that want to almost start over. Do I have that right? And why does this matter and where
- [27:24]
A
do you think it's going to happen? It does seem like that's what the narrative has turned into, but I don't think it needs to be that narrative. I don't. I mean, I know for us, for, for, for rdsp, you know, we have stuff that's live today using model context protocol.
- [27:42]
B
Okay.
- [27:43]
A
And it works. And it's operation. It's G8. We also, as a company are leaned into ADCP we're a founding member of ADCP.
- [27:53]
B
Okay.
- [27:53]
A
It might be more opportunity, candidly, in the near term for us as a publisher on that one for selling direct deals, but I'm still leaned into that one as well. And I'm not picking any sort of like, protocol per se.
- [28:07]
B
Yeah, but you're. You guys are kind of in the industry as much as you're a player.
- [28:12]
A
Yeah, I don't, you know, it does seem there's a lot of coverage for, you know, protocol versus protocol. But, like, that's not something that we're debating. We've got the capability to interact with multiple. And we're. We're developing where things can be shipped and.
- [28:29]
B
Okay.
- [28:30]
A
But today I do have to say, like, what we have put out in market that's live today is predominantly on mcp.
- [28:38]
B
Okay. As I would have predicted. I would not have predicted. I have no idea. Have you patched things up with your agent, by the way, that you're fighting with. I'm concerned about this a little bit.
- [28:48]
A
You know, they're so patronizing because, you know, usually.
- [28:51]
B
Well, AI is known for being, you know, overly complimentary.
- [28:55]
A
I gotta tell you, I gotta let them have it every once in a while. Or she.
- [28:57]
B
I'm not sure.
- [28:59]
A
They, they do a really good job. And mine is turned on to show its. It's inner thinking, which is kind of dangerous. So.
- [29:09]
B
Oh, wow.
- [29:10]
A
Call it out and say, hey, you did something wrong. It shows me, okay, I did something wrong, which is really bad. It shows all that sort of dialogue which is on the inside, which is both cringy and helpful for me to better wrap my head around how to, how to interact with it. It's. It's a process. We're getting to know each other.
- [29:31]
B
All right, good. Well, yeah, there's always this. Feels like an Asian therapist might be in the offing.
- [29:36]
A
I know a lot of us, he. She may be listening right now. You might. Or you know what? I say.
- [29:44]
B
Okay, let me ask you another just like macro question because there is. I'll. I'll go to one event and someone will say, we're not moving fast enough as an industry. There's so much urgency. Look at what happened with some of the. Look at what happened with, you know, square or block when they, they have fired after the company because they're. It's. The businesses are going to be run by agents and then others are like, get real. It's so hard to change processes at these giants. The agencies are gonna have to, you know, you have to burn it all down to really get everything to place where we want to go. Where, where do do you think we, we are moving fast enough. Is it going to be sort of painful to get to the place of maximum agent power or the dream industry? We hope it will be.
- [30:29]
A
I think it's moving pretty quick. I think it's going to go pretty quick compared to. And I don't know what you want to draw a comparison to. Honestly. It's like if I think about the first iteration of AI that really came into the programmatic ad ecosystem, it was definitely ML and that saw ad networks evolve in anti platforms is probably the best way to qualify what happened and demand side platforms and sell side platforms for that matter. And that, that was probably like a five year sort of like metamorphosis from network being prominent to not being around. Yeah, this one's going much, much, much quicker. I think we got a false false positive when Generative sort of like came into the picture. It's been great honestly for certain things like UX and creative, but this, I mean, I mean we went from like learning what an MCP was to standing one up within I think six months. It was really quick.
- [31:23]
B
You're right. We're like, yeah, we've made people thought maybe Generative, we replace all the creative agencies in two days. And that's not happening necessarily. But this is, this stuff is getting traction quick.
- [31:32]
A
This is different, this feels different. And I can tell you that like behind the scenes what every company is dealing with is not unique to Yahoo, but I know it's happening everywhere is the ability for developers to use these tools and have through, you know, a cycle that looks anything unlike the software development cycle that we've had for the past 20 years. That's fascinating. And that's just gonna be, that's not limited to ad tech. That is just everywhere and it is very, very cool. It's gonna change the way we operate as product people, the way we operate as sellers and service people. So there's a lot cooking there and the early traction that we see, I, I think, I think it's gonna be quicker than most people realize. I don't think anyone should complain about speed.
- [32:16]
B
What about. Okay, last one for me is any advice. Let's assume that most agencies and brands are not thinking about this stuff yet, but advice for the ones that have not maybe moved as quickly as they need to. What, what should they be doing?
- [32:30]
A
Oh, I'd say that good. The good news is like the minute that anyone decides that they want to lean into an agentic investment, they're going to realize that the trick and the, the barrier to entry with say, standing up an agent or an MCP or a connection, compared to like the legacy of working with APX or whatnot, it's. The barrier is not that high.
- [32:55]
B
You're not three years behind. It can't and can't possibly catch up.
- [32:59]
A
No, I mean these things are to say someone could stand up an MCP is. That's like an overnight job. That's not the part. The part is structuring an organization to take advantage of it, understanding the capabilities and understanding that they don't need to think about platform decisions in the way that they used to because the ease of use and interoperability and everything goes up.
- [33:24]
B
Right. Whose job is it to do what and how do we do that work every day? That's the hard part.
- [33:29]
A
That's the transformation part. You know, I think, I think, you know, that can be, that can just be a level of conviction, challenge for people, not a, like, I need to start investing. And I think that's, that's the fun part for a lot of us right now. And I would say come on in if you haven't by now. I'm sure they will any minute now because it's pretty convict, it's, it's pretty convincing once you start operating in it.
- [33:58]
B
All right, well, I'll send people to you to get, to get help. Adam, you're the awesome AI whisperer here. Thanks so much for your insights and let's do this again sometime.
- [34:07]
A
Happy, thanks for having me. Good to see you again, Mike.
- [34:10]
B
Thank you. Thanks again to my guest this week, Yahoo's Adam Rudman and my partners at sabio and Intan iq. If you liked this week's episode, please take a moment to rate and leave a review. We have lots more to bring you, so please hit that subscribe button. We'll see you next time for more on what's next in media. Thanks for listening.