Latest / Signal & Noise: Bringing Clarity to Media, Data, Tech and AI / From Strategy to Scale: George Musi on What Actually Drives Growth Inside Agencies
Transcript
- 0:03This is signal and noise, bringing clarity to data, tech
- 0:07and AI. Hey everybody.
- 0:14Welcome back to Signal and Noise.
- 0:16I'm Brett House, joined by my Co host Real Longacre.
- 0:19We've got a great guest today, George Musey.
- 0:23We're talking about something every company says that they
- 0:26want, but few actually achieve it consistently growth, right?
- 0:30We've decided to broaden that topic to, to, to all companies,
- 0:35although agency hold codes is certainly a focus considering
- 0:37George's experience. George, you've spent decades in
- 0:41big agency hold codes from Publicis to WPPIPG and then most
- 0:46recently Horizon, right? As an EVP, one of the largest
- 0:49independent agencies, you know, you, you've seen it from every
- 0:53angle. You've seen it from indie
- 0:55challengers to global sort of complex, you know,
- 0:59organizations. And now you're an advisor
- 1:02working directly with executive teams that are trying to scale
- 1:06and trying to grow, whether it's agency holders or tech or vendor
- 1:09companies as well. One welcome to the show.
- 1:12Great to have you also welcome as an executive contributor.
- 1:16For those listening, we've just added George as an executive,
- 1:20executive contributor to Signal and Noise or he's going to be
- 1:23bringing up. Thank you.
- 1:25Writing articles, joining podcasts, and one thing I I
- 1:29thought so refreshing when we first spoke was George brings
- 1:33kind of a no bullshit mentality to his.
- 1:36Conversation, New York. The New York attitude, right?
- 1:38Love it. Whether you've seen about panels
- 1:41or speaking and you made it. Really clear when we first
- 1:43talked that. You're like, I just hate the
- 1:45surface level stuff and I like to dig down and really ask the
- 1:49hard questions, which often times the events especially are
- 1:52sort of ignored or, or or avoided.
- 1:55But yeah. So introduce yourself to the to
- 1:57the to the audience. We'd love to hear a little bit
- 1:59more about you and your story, George.
- 2:01I appreciate it guys. It's a pleasure both for the
- 2:05podcast as well as the the ability to kind of partner on on
- 2:09signal noise, right? It's a it's definitely a time
- 2:14event venture. Definitely need more straight,
- 2:19no nonsense insights in the marketplace.
- 2:22God knows our industry is notorious for bullshit.
- 2:27So you definitely need a lot less of it in this day and
- 2:29night, particularly with this advert of AI.
- 2:32So my background is unique, I would say in many ways, right?
- 2:36I started as a data scientist, events and a local person and
- 2:41then I've I've evolved as the need showed me the evolution was
- 2:47warranted right. So instead of having 75
- 2:54different conversations where an agency person goes with the CMO
- 2:57or CFO or even have to get in front of the Cfoi told myself is
- 3:02I'm going to hedge my bet that data analytics and technology
- 3:04are going to be key into the future.
- 3:07I could have been wrong, but I wasn't.
- 3:09But this notion of translating what that means to the speak of
- 3:14ACMO, to the speak of ACFO, to the speak of chief data officer,
- 3:18chief digital officer, chief information, whatever version of
- 3:21chief. But more importantly, is this
- 3:23notion of systemic? How do you get those people to
- 3:26speak the same language? Because they're all being asked
- 3:29to do the same exact thing from different vantage points.
- 3:31So they hide in that ambiguity versus bringing that
- 3:35transparency to the table and saying, gentlemen and ladies,
- 3:38right, growth is the agenda, right?
- 3:40How do we architect that? So what I, what I then said is
- 3:44how do I become the best solutions person that exists?
- 3:48It doesn't matter where agencies, consultancies,
- 3:50technology company doesn't really matter at the end of the
- 3:52day. It's how do you bring that
- 3:55repertoire of solutions, technologies and architect a
- 4:01growth and, and that could mean something different from from
- 4:05every angle of an organization. It typically does if you're
- 4:08talking about US and global, if you're talking about CPG,
- 4:11whatever the case would be. So that's that's kind of where I
- 4:13started from where I've evolved. Yeah, and George, we're.
- 4:17Thrilled. We're thrilled to have you on
- 4:18here. I remember when you reached out
- 4:20a couple weeks ago, we had that and we ended up speaking for
- 4:23over an hour to three of us. We thought, OK, we we got
- 4:25something here. We got it.
- 4:25We got to make this happen. So I'm glad we could.
- 4:28We could do it. It was a late Friday afternoon
- 4:30if I remember correctly. That's right.
- 4:32Yep. And then, yeah, in terms of
- 4:35thesis for this episode, because we, you know, signal and noise,
- 4:38we, we try to make every episode episode have a thesis, right,
- 4:41that will resonate out there. There's a good one.
- 4:43Growth is hard. It's very hard.
- 4:45Every company has a new AI product, a road map software.
- 4:49They're they're rolling out. But the reality is where the
- 4:54rubber meets the road, it's really tough and you really need
- 4:57a vision. And when you look under the
- 4:58hood, even below the vision, a lot of companies, they really
- 5:01struggle to execute, or at least consistently execute.
- 5:05Generally, growth doesn't break because of lack of ideas.
- 5:07We see a lot of those now, right?
- 5:09It breaks because the system itself, how it's structured and
- 5:13how the incentives are structured and how they impact
- 5:16really who makes decisions and why they make decisions.
- 5:20And that controls how things go to market.
- 5:22Although things need to interlock and they they need to
- 5:24work together and very often they don't.
- 5:26So you spent your career in the middle of this got had like as
- 5:29Brett mentioned, you've worked in a bunch of agencies hold
- 5:33codes, including most recently Horizon is a former client of
- 5:36mine. So we really wondered, Deegan,
- 5:37why is growth so structurally hard inside these organizations,
- 5:41inside agencies hold codes, but I think generally inside a
- 5:43professional services organizations everywhere,
- 5:46they're all going through some of the same challenges.
- 5:48Why are we seeing these shifts like the whole whole code
- 5:53operating system thesis we, we talked about in this web signal
- 5:56noise website a couple weeks ago?
- 5:57Is that real? What's actually happening?
- 5:59How does AI fit into all of this?
- 6:01Everyone's asking these same questions.
- 6:02So we're thrilled to get you on here, excited to talk about.
- 6:05I know there's tons we can dig into.
- 6:06Brett, first questions over to you.
- 6:08Yeah. What are these environments?
- 6:09I mean you in your intro, George, you kind of mentioned,
- 6:12you said systemic and I think you were referencing sort of
- 6:14systems thinking right across technology, data analytics and
- 6:18then the the language that different departmental heads,
- 6:21different leaders speaking, right.
- 6:23So that everybody can be thinking around the same kind of
- 6:25operating. Operating system is kind of a
- 6:28overused term nowadays. It seems like everyone has an
- 6:30operating system. And they last everything but
- 6:32nothing, right? Right.
- 6:34I would argue that nobody really has.
- 6:36One, yeah, the boat really has one.
- 6:37Let's let's deconstruct that. So what did you learn in that
- 6:39working at at these big agency hold Coast Horizon just
- 6:42recently? What it really teach you about
- 6:44growth and about what's real and what's not.
- 6:48Yeah, if you, if you look at the agency world generally and where
- 6:54it came from, where it is and ultimately where it's going, it
- 7:01evolved as as much as the agencies say they have.
- 7:05They set the tone of the pace of the industry.
- 7:08They're very, they're very reactive to it.
- 7:12They're not very prescriptive about it.
- 7:13They will say that, but that's that's.
- 7:15Not. Not first movers more, more
- 7:17laggards. And sort of.
- 7:18Adapting and reacting and, and what's changed is their partners
- 7:23have become competitors and the the consultancies have come in
- 7:30and some of them didn't realize that that was happening until it
- 7:33was too late. Some of them like Accenture,
- 7:35Accenture song. Yeah, TA, ta is going to come is
- 7:41coming Slalomed. There's a bunch of everybody
- 7:44sees the opportunity and they come from a position of strength
- 7:48because they hold I would argue higher order relationships in
- 7:53true business defining decisions.
- 7:56Yeah, and. Certainly with I, so.
- 7:58That's right. Well, I would even be CEO any,
- 8:01any person that is selling anybody else in marketing, what
- 8:04to do at the end of the day, right.
- 8:06So if they're, if they're, if, if you're going in doing
- 8:08massive, massive growth architecture within a Fortune
- 8:14100 company, right, marketing will be a component of that, a
- 8:18component of that. It's not the whole conversation.
- 8:20Yeah. And the media will.
- 8:21Media will be a component of that.
- 8:23And a component of that, right? So you have.
- 8:24Audience, the table with that Regency, that media play, some
- 8:28tests, a little bit of tech behind, but you're saying that
- 8:30the consulting firms come in with a much more systemic
- 8:33business growth sort of playbook.
- 8:35And strategy correct their job is growth and and now what they
- 8:38look at it is oh, like this component, this add on of media
- 8:42marketing that's easy for us. I mean, it's we're not we won't
- 8:46build it ourselves. We're just going to go by and
- 8:48that's like a censure song. Did they want more drug?
- 8:50And they said that we have this or, or let's go buy this because
- 8:53God knows they have the coffers to do it right.
- 8:55And they and they and they have the money sitting on the side to
- 8:58do so. So across these across these
- 9:00agencies, you know, it's always to the agencies has always been
- 9:04the pure precipice of it has always been the agents on behalf
- 9:07of. And that has changed
- 9:09dramatically over the years, right where you're seeing
- 9:12independent agencies try to position themselves as mutual
- 9:17right stewards of a brand. But the end of the stage,
- 9:22they're still they're still pitching services and solutions.
- 9:27Some of them they've built a lot of PM Geo tenuity, some of them
- 9:30they've partnered with. But the end of the day, they're
- 9:32still trying to do the same exact thing, which is how do I
- 9:35make the most money for the work that I'm providing a client?
- 9:40Hold coves are interesting because they've taken a
- 9:43position, not all of them, but most of them have taken a
- 9:45position on the thing that they said that they would never do,
- 9:48which is media, right? So now, now by definition, they
- 9:53really can't be agents anymore because they're buying and
- 9:56selling, right? And so they're their, their,
- 9:59their positioning should change. They lose.
- 10:02Their sense of independence Because.
- 10:04Because. It's hard.
- 10:06I mean, how do you, how do you, how do you tell somebody that I
- 10:08own the media? I'm going to sell you the media
- 10:10and then I'm going to, I'm going to judge how I'm doing against
- 10:13buying and selling. And it's, it's, it's, it's for,
- 10:16I'm not saying it's impossible. I'm just saying this is very,
- 10:18very, very hard for somebody to look at and say, how do I know,
- 10:22right? How do I know, behind the shadow
- 10:23of a doubt, that that you are actually doing what what you are
- 10:26telling me that you are doing? Question that.
- 10:28So I'm a former Yep systems integrator consultants employee
- 10:34and I've also worked agencies. So like remember when I was at
- 10:38the systems integrator and a consultancy like we, we had
- 10:41great relationships with IT, like great relations with the
- 10:44execs generally, but weakest with the CMO and agencies
- 10:47typically held those. Remember the Gardner had that
- 10:49thesis that the CMO would outspent the CIO, which never
- 10:52actually happened, right? And there was that thesis and
- 10:54then there was this big, this big.
- 10:57This is around the time Accenture song started.
- 10:59Like, you know, those acquisitions, they they even
- 11:00bought a meat in a programmatic company, even tried to get into
- 11:03media. I know the guy who who built
- 11:05that. It didn't go well for them.
- 11:06It's I was a solemn. We even tried buying media for a
- 11:09while, but we like we wouldn't like they refused to they'd be
- 11:13solemn, refused to set up a line of credit.
- 11:16So the problematic playing the platforms net 30, we actually
- 11:20had campaigns. So we decided not to do it.
- 11:22So where do you see that like that playing out?
- 11:25Right, Because like they've tried like you see this, the
- 11:27consultancies are trying to get in like they they own the other
- 11:30relationships they've been trying to get in.
- 11:32And then the hold codes are kind of trying to go the other way
- 11:34and AD systems integration and engineering like like, do you
- 11:38think like Accenture song comes a media company or do you what's
- 11:42your? Opinion on that?
- 11:43I, I, I looked at where all the agencies, the hold code
- 11:48specifically and maybe some of the largest in the include
- 11:51horizon and you look at them and you're you, you start saying,
- 11:55well, they don't look like media.
- 11:57They don't look like agencies anymore, right?
- 12:00They don't look at the. Horizon matter just what?
- 12:02Not 9 billion a media if I remember right.
- 12:05Probably a little less than that.
- 12:06Now they've lost, they've lost a couple of counts, but they made
- 12:08it up with with Spectrum. So let's call it yeah, probably
- 12:11about 7:00. But if you look at them, they're
- 12:14not really, or at least they're positioning themselves.
- 12:18Horizon OS, right Omni. Interact.
- 12:22People, people cloud, right, interact right, you have
- 12:27converge AI you have. WPP Open.
- 12:30They all have their own, yeah. So you look at this and you're
- 12:33like, wait, that that's, that's you're no longer an agency.
- 12:36You're you're not creative, you're not experienced, you're
- 12:39not media, right? What you are selling is
- 12:42technology, data and system integration, right?
- 12:44That's, that's what they want. They want the solutions
- 12:47architect and then they want the implementation.
- 12:51That's interesting because that's exactly what the
- 12:53consultancies do, right? Exactly.
- 12:58So I will argue is you're going to go have a fight against the
- 13:01guys that have been doing this for a very long time
- 13:04successfully that have deeper pockets and deeper relationships
- 13:07with the with the most important stakeholders in the
- 13:10organization. Not, not to, not to marginalized
- 13:14what the CMO does, but you're talking about the people that
- 13:16control this, the purse strings at the end of the day, right?
- 13:19So then, but you're also then competing against the platform
- 13:25companies, right? The trade desks of the world,
- 13:27the Microsoft's of the world, the Googles of the world, which
- 13:30by themselves also start looking at saying, hold up a second, in
- 13:33this world of AI, do brands really need an agency?
- 13:37Yeah, I mean you. You you still run into the
- 13:40grading your own homework problem with the platform.
- 13:42You you you do but but but then you make that same argument for
- 13:45the hold coves, right? Creating your own homework.
- 13:47Yeah. I mean, do you think, do you
- 13:49think these operating systems are, are like a service Raptor
- 13:52wrapper or do you think that depending on the agency hold Co
- 13:55and the operating system that they're claiming to have at the
- 13:57platform 'cause they're all moving in that sort of?
- 13:59They're all trying to build their own, yeah.
- 14:01Yeah. Do you think it's a?
- 14:01Services wrap, do you think it's changed enough to where they can
- 14:04change their sort of FTE heavy business models?
- 14:11It's it's no. In order for that to happen,
- 14:14truly happen, that means it is the operating system of the
- 14:18whole code. And I don't see that yet.
- 14:22Yeah, I don't. So you need a complete.
- 14:24You need a complete. A complete break from the FTE
- 14:26model going to something different, correct?
- 14:28I look, you still need dedicated people to to a brand.
- 14:34I think you you have to in order to to immerse yourself into that
- 14:37brand DNA, right? You have to.
- 14:39You can't. You can't have.
- 14:40People walking the halls. Yeah, yeah.
- 14:42But you need, you need somebody to live and breathe, right?
- 14:45And that's why I think consultants did a really good
- 14:46job. They just immersed themselves
- 14:48into an organization and they'd learned everything.
- 14:50And that's what made them even better.
- 14:51Even if they were mediocre, then you had so much institutional
- 14:54knowledge that it made them, it made it really impactful you.
- 14:59Even look at you even look at the most successful AI companies
- 15:02that they're, they're, they're offering forward deployed
- 15:04engineers, which is really consultants.
- 15:05You're you're, you're having work embedded in client teams.
- 15:08Yeah. It's easy to say that you've got
- 15:09this capability, this tech, but it's much harder to get the
- 15:13organization that you're supporting to change into, to
- 15:16Orient themselves around the tech right effectively.
- 15:19Yeah, I would argue that historically the agencies never
- 15:24really had company people that were designed trained to act
- 15:31like that. So what they did is they try to
- 15:33go by and some of them bought, some of them bought
- 15:36successfully, some of them bought without really realizing
- 15:39what what they were supposed to do right and how to do it.
- 15:42And that that that created a lot of friction and disorientation.
- 15:46At the same time, they try to go to client brands, prospects are
- 15:49current and say we have a new way to service you, right.
- 15:52So don't judges by don't judges by this, judges by this.
- 15:55And the brands look at them and saying it looks great on the
- 15:59PowerPoint slide. Oh yeah, I don't see that.
- 16:01Yeah. And really your article, your
- 16:03agency Holdco article on on signal noise kind of referenced
- 16:06this is you you're still seeing it sounds to me like you're
- 16:08still seeing a collection of tools that came through
- 16:11acquisition that came through either it was organically built,
- 16:14but it's not really a unified system in in the true sort of.
- 16:19Way of thinking. I have yet to see true systems
- 16:24thinking, true connection of purposely built created entities
- 16:33that can individually contribute, but collectively
- 16:37they really guide the organization through a growth
- 16:41agenda, right, right. It's not, it's not that it's,
- 16:45it's, I don't think it's anywhere close to that in my
- 16:47opinion, no matter what anybody says on slides or pitches or
- 16:50anybody else. And I think if you look at put
- 16:56aside the mandated 3 year reviews or whatever the case may
- 17:00be with brands, I will argue if you're, if you're actually
- 17:04delivering what you're saying growth, true, true growth,
- 17:07right? There is no brand in their right
- 17:10mind that's going to make a switch, right?
- 17:12If they truly see you as a systems integrator, as a systems
- 17:15thinking that that understands their business, understands the
- 17:19mechanisms, the levers that if you do this in marketing, you do
- 17:22this in product and you do this in pricing, you do this here and
- 17:25you do do that the collective things because it's it's
- 17:28interdependent ecosystem. At the end of the day, things
- 17:31will move. Yeah, and the switching process
- 17:35would be huge. Well, systems integrators, I
- 17:38mean, one thing I always admired, like working like you,
- 17:40you see an Accenture, Deloitte, right?
- 17:41You know, suddenly they sell a project with four or five
- 17:44people. You turn around, you close your
- 17:45eyes, the clients and complain about this like, you know,
- 17:47there's 150 people all billing there and then they'll be there
- 17:50for years. They'll be locked in and they'll
- 17:52they'll have a seat at the table.
- 17:53Like with the CEO, which is typically in fact, I remember
- 17:56when I was working at big pharma, we we had to change all
- 17:59our sows to say we didn't. They didn't have strategy in
- 18:02them because McKenzie science may exclusive deal with the CEO
- 18:04because he was former McKenzie. So like having that level of
- 18:07access and be able to like land and expand system adjudicators
- 18:10are really, really good at that. I would argue that something
- 18:12agencies definitely, if they want to get into this business,
- 18:15they need to learn it. But I guess the other thing too
- 18:17is like both businesses because of AI are starting to go and run
- 18:20into some pressure because the base of the pyramid is being
- 18:24clients are starting to push back against, you know, do you
- 18:26really need 3 BAS to take notes? AI is probably better than that.
- 18:29So I think systems integration model and, and the agency model
- 18:32both go into similar things and it's kind of been a push towards
- 18:35more value based or performance based contracts.
- 18:38What are your thoughts on that, George?
- 18:39Love to hear what you're seeing. Yeah.
- 18:44And everybody is going out and saying AI is changing, changing
- 18:51the paradigm. And that is true.
- 18:54But you mentioned something real earlier, which is also very much
- 18:58true, which I think is the current reality.
- 19:00No matter how much money has been spent in AI, the vast
- 19:03majority of it has failed, right?
- 19:06So you're, you're trying, you're trying to, you're trying to put
- 19:11a square peg in around holding an agency world and it's not
- 19:14working. And they're like, no, no, no,
- 19:16no, we have AI, right? We're doing AI.
- 19:18And what they're really, really doing is basically wrapping AI.
- 19:23So you're wrapping, you know, you wrapping your.
- 19:25Stream around workflows or you're.
- 19:27Just you're just saying. You're automating things.
- 19:30Yeah, which is conversational AI chat chatbot or workflow
- 19:35integrate. Yeah.
- 19:37It's not true applied AI, right? What I mean by that is you've
- 19:42built something from the from the foundational core of your
- 19:46organization, right, moving up right.
- 19:49That is the only way through the app, the true application of AI.
- 19:53Now what they're saying is they have that true or to whatever.
- 19:56Most of the time it's not it's not being built internally,
- 20:00right. I don't, I don't, I don't see
- 20:03that. So they're partnering with with
- 20:06what exists out in the world from the world, right, which
- 20:09then they start realizing that those things cannot really
- 20:14deliver what they're asking them to deliver because.
- 20:17Those world models or those horizontal models or those off
- 20:22the shelf models, whatever you want to call them, they're not,
- 20:25they're not built to understand the entities of any
- 20:28organization. That's not their job, right?
- 20:30That's what they called world models.
- 20:32So they're promising we're going to, we're going to be AI native
- 20:37organizations, which it's impossible because that's not
- 20:41where they came from. And #2 is because of that, we
- 20:44are going to recalibrate the relationships that we have and
- 20:49the things that we do. So relationships that we have is
- 20:52this is how we, we, we will service you moving forward.
- 20:55This is how you would should value us moving forward, our
- 20:58compensation. And these are the types of the
- 21:00services solutions that we cannot now also provide.
- 21:03So what they're saying is, yeah, we want to shift the value based
- 21:08pricing. But by the way, we also want to
- 21:10want you to give us more services and solutions, right,
- 21:14that historically you have not given to us.
- 21:16So what they're doing? Is more risk on your behalf
- 21:19right? Which?
- 21:19Is well, give us more, give us, give us the ability to because
- 21:23now we are an AI agency. Give us the.
- 21:25Buildings. The things you saying you have
- 21:26right? Yeah, so, so, so, so George.
- 21:30But the so you said they're not really truly it's not a core
- 21:34competency, let's say being an age AI 1st.
- 21:36And you, you've got a long history and sort of machine
- 21:38learning back when it was, you know, before all of the AI hype
- 21:41that's affected us for the last probably 20 years, but
- 21:44especially in the last five years.
- 21:46Is it, does it require an agency to be, do you have to have that
- 21:49centralized unified data source, all of the semantic kind of
- 21:53middleware logic built on top of that to interpret the data in
- 21:59order to get to a point where you're sort of a as a core
- 22:02competency, you're an AI first agency.
- 22:04Is that what we're looking at? So I think all models are
- 22:08commoditized. That's always been the case,
- 22:10right? Everybody, everybody assumes
- 22:13who's going to win is going to be who has the the best model.
- 22:15So it's all good commodity, right.
- 22:17So at the end of the day, so is the data.
- 22:19I would argue that that you have a big enough check, you can act,
- 22:23you can acquire any data source that you want.
- 22:25Yeah, I don't. I don't see that as a because
- 22:28even in. One in one brands audience data
- 22:31starts looking a lot like another.
- 22:32You go, you go to publicist and say and say, Hey, look, I want
- 22:36to buy epsilon. They're not going to sell you
- 22:37epsilon. Of course they're going to sell
- 22:38you epsilon. You write another big enough
- 22:40check, they're going to give it to you, right?
- 22:41Because that's that's the structure of how they put those
- 22:43deals. Because they have to be
- 22:45compliant and and they can't, they can create non competes in
- 22:48in the marketplace like that because it's important data
- 22:50source. What I would argue is the most
- 22:52important thing which every organization does a really,
- 22:56really bad job is, is institutional knowledge
- 22:58retention. Yeah.
- 23:01That is the most important thing in any business.
- 23:04Agencies no different. And I will tell you agencies are
- 23:07are probably the worst at it, right, because they have a hard
- 23:09time keeping people. So where is the knowledge coming
- 23:13from? It's not data sources.
- 23:14Data without context is meaningless.
- 23:16Yeah. Where's?
- 23:17The context, The play for context is critical, and that's
- 23:20the the people and and. Heads of subject matter experts,
- 23:23implicit, explicit tactical information that most employees
- 23:30cannot figure out a way how to institutionalized that into an
- 23:33organization because there's never really been the mechanism.
- 23:35So when people say, well, I'm going to dial, I have AI, so I'm
- 23:39just going to upload stuff into AI.
- 23:42OK, So but, but, and we're going to, we're going to put some
- 23:46props against it and that's going to give us, that's going
- 23:49to give us all this magic that's going to all of a sudden pop
- 23:52out, right. And the reality is, well, OK,
- 23:56but you're giving it data, you're not giving it context.
- 23:58Although you might think you're giving your prompts A context,
- 24:01that's not the way it works. You still need the right people
- 24:04writing the prompts. That's that's a big limitation.
- 24:06But here you needed some other experts.
- 24:08That that's right. So you need to be able to encode
- 24:13the true expertise of an organization, the true DNA of an
- 24:17organization, into a machine. Not just once, but you need to
- 24:21do that on a continual basis. Now go back to the.
- 24:25Optimizer and almost make it specific.
- 24:26It's, it's based to like sort of tasks that you're setting up for
- 24:30specific agents and agentic teams is they all have very
- 24:32specific guidelines and, and and approaches and then A and then a
- 24:36memory system built on. That so.
- 24:38I'm an expert and creative, right?
- 24:39How do you keep that thing going and optimizing over time?
- 24:43But but here's also. Here's also the other component
- 24:46parts. It's doing that, that constant
- 24:50encoding, right? Every single time you're working
- 24:55every, every single time you're working, right?
- 24:57That's how you build a system of intelligence.
- 24:59You don't build a point solution, right?
- 25:01And right now most agencies, most agencies I still
- 25:05fundamentally have just a bunch of point solutions.
- 25:09Even with AI as underpinning this top point solutions, nobody
- 25:12really truly has a system of of of intelligence.
- 25:16Solutions, George, and a nice presentation, a nice Yeah, I.
- 25:19Mean, that's right. You got to you got to give it to
- 25:21them. They did they if agencies are
- 25:23good at one thing, they are really good at pitching.
- 25:26The reality is quite different of delivery, right.
- 25:28So I would argue the other, the other aspect of it all, which I
- 25:32think is critical is this notion of creating a neural network,
- 25:38right? System in a different way, but a
- 25:40neural network, right? This notion of connecting
- 25:43individuals in the collective, right?
- 25:46That's really the power of what AI True knowledge, right?
- 25:51Knowledge graphs. Yeah, capturing and development
- 25:55and growth. Yeah, knowledge graph for the
- 25:57organization. Well, I would argue it's even
- 25:58more than there's knowledge grabs I would take.
- 26:00I would argue it's this thing called temporal intelligence
- 26:03graphs, which is knowledge, context and time.
- 26:06Yep, right over the history of an organization, right?
- 26:10You got to be able to look at all that stuff and then tell
- 26:14yourself, if I ever wanted to understand how that came to
- 26:18being over a course of time, what, why and the what
- 26:22conditions, who did it, what they do, how do they trade it
- 26:25off right? Because this notion of
- 26:26knowledge, if I don't, if I can't access the way we know
- 26:30real things, but not just what he said and when he said under
- 26:34what conditions, but I can trace it back, I can audit it, I can
- 26:38explain it. And by the way, I can override
- 26:41it should I want to decide to do that.
- 26:43But then we all could do the same exact thing to me and to
- 26:45you. That's a system.
- 26:46Because then the AI then truly learns your DNA.
- 26:51Yeah, that's a closer approximation to an actual human
- 26:53being in the way we operate and think.
- 26:56Correct. Which is which is what AI is
- 27:00supposed to be doing, right? Which is augmenting the human
- 27:03being. It's not replacing the human
- 27:05being. It's it's supposed to augment
- 27:07us. But George, looking at the egg,
- 27:08let's see the AI native or AI enabled agency, which yeah,
- 27:12we're very compelling. No one's there yet.
- 27:16I don't think any one of us would argue.
- 27:17But even unless they're turning the clock back a couple years,
- 27:19remember, Publicist made the acquisition of Epsilon that you
- 27:22called out before they had a sapient, which is like the kind
- 27:25of their services org. There was a perception as
- 27:28recently as a couple years ago that they had kind of not only
- 27:32created a model but also had a pitch to go along with it.
- 27:36That was there was winning them a lot of deals, right?
- 27:40That was that was leading the way.
- 27:41And then you know, we've seen Omnicom, you know, we, we've
- 27:43made a lot of moves here that are I think are very similar,
- 27:45some some somewhat different, but some very similar.
- 27:48And then there was a perception of that that was the model.
- 27:51And then, but I mean, what I've heard is a lot of the, the
- 27:55deals, even the, the publicist was winning there.
- 27:57They did at, you know, for very low cost and that they're even
- 27:59struggling to support them, right?
- 28:01I, I don't know if I don't know how true that is, but you know
- 28:02that those are whispers we hear, right?
- 28:04And then the question is, OK, So what is the model going to be?
- 28:07I mean, is AI, does AI shake it all out?
- 28:10Is, is that the right model? But then you, you put an AI wrap
- 28:13around it, Brett, to your point, or does the whole thing really
- 28:15like it? Does the FT model crumble
- 28:18quicker than people are thinking?
- 28:20And the whole thing has to be rethought.
- 28:21And I think that's kind of where we are as an industry.
- 28:23Love your thoughts on that George.
- 28:25Yeah, I think, I think the this notion of selling people is
- 28:29going to definitely go away, right.
- 28:32I've always argued that you should be judged based on the
- 28:37ultimate impact to the business, both top line and bottom line.
- 28:40I don't really care what version of it because marketing has,
- 28:44will and should always be about incrementality.
- 28:48Why the hell would you be spending any money if if you can
- 28:51get that doing without spending money?
- 28:53So if that's the if that is the pure fundamental goal of
- 28:56marketing, right, you are judged.
- 28:59You should be judged by that so the relationship that you have
- 29:03with a brand the brand has to understand here's my growth plan
- 29:07to you architect. Here's what I'm going to give
- 29:09you right, forget about the people.
- 29:11This is what I'm going to give you right These are these are
- 29:13the things that will give you that right you agree to that
- 29:16yes, and that's what you're being held against That should
- 29:19not come to with an FTE that should not that should be here's
- 29:22how I'm packaging this cell to answer your question like
- 29:25there's. AP and L behind it.
- 29:28Right. That.
- 29:28Well, yeah, because you, you, you, the branch should not the
- 29:32branch should understand who a core group of people that they
- 29:37can pick up the phone, whoever the case may be.
- 29:41They should never care about how you're ultimately enabling them
- 29:46to do whatever they're doing right now.
- 29:48They might want to understand what, what technology are you
- 29:52using? What, what AI are you using?
- 29:54What data are you using? It's their right because if
- 29:56you're using those component parts and particularly if you're
- 29:59charging them right, however you want disclosed or undisclosed,
- 30:03right, they should understand right, the architecture of how
- 30:08you're going to drive that growth.
- 30:09I I believe that a. 100% what does that mean?
- 30:11What it does that mean value based contracts outcome be like?
- 30:14What does that mean in in. In practice?
- 30:18Media buying. Transparency.
- 30:19There's all these. Things yeah, I, I, I'm not a big
- 30:21I, I am not a big fan of of principle based buying.
- 30:25I understand it. I'm never been a big fan of it.
- 30:28I I. And we and we just saw the WPP
- 30:31Richard Foster news. I around just just, there's all
- 30:35this undisclosed sort of stuff that if you're getting rebates
- 30:38from principalities buying and then you don't pass on those
- 30:41benefits to the client. I just think I I I think there
- 30:46are many different ways for agencies to ensure that they're
- 30:51making ample enough margin where they can make sure that they're
- 30:58investing in the most important asset that they have, which is
- 31:00their people. Yeah, right.
- 31:02No agency and I would argue no services business even with the
- 31:06advert of AI and this doesn't really matter how much AI5 you'd
- 31:09do anything. The end of the day, any services
- 31:12businesses is what people based business.
- 31:13Yeah. So I.
- 31:14Think people versus tech, right, because you said tech is
- 31:17probably in the core competency of the consultancies that.
- 31:22They think eventually I I think the agencies are starting to
- 31:26again through acquisition, mainly through hiring and into
- 31:30partnerships. They're starting to become more
- 31:36tech competent, not tech enable because you can you can buy an
- 31:38organization doesn't make doesn't make you that an expert
- 31:41in at the end of the day, right. God knows agencies have had a
- 31:45long history of buying stuff and failing, right?
- 31:48And they're writing it off, right?
- 31:50You can name a bunch of them, the last 10 years that have done
- 31:53that. Well, tech companies.
- 31:53Too. Yeah, not only tech companies,
- 31:55right. A lot of failed integration,
- 31:57yes, and failed. But, but, but I would, I would,
- 32:00yeah, the, the FTE model will, will go away.
- 32:03I'm saying it's going to go away to more, but it will and it
- 32:05should ultimately turn into true outcome based or true outcome
- 32:09based. And, and so whatever, whatever
- 32:12you sign up to now, I will say that in order for that to be
- 32:17successful, the brands have to be cognizant that they can't
- 32:22change the goal posts every single day, right?
- 32:25They have to say, this is what this is what we agreed to,
- 32:29right? Understanding there's some
- 32:30nuances in there. A war happens, supply chain,
- 32:33logistics goal, what whatever the case may be, I get it.
- 32:36Black Swan events, I get that bit.
- 32:38But then you recalibrate, right? But if you're if nothing happens
- 32:41and then the brand keeps shift for whatever reason, shifts
- 32:44things, then the agency should not be held accountable, right?
- 32:47It should be compensated because at the end of the day, if that
- 32:51if a brand which they always say no, no, I will always, always be
- 32:55credit of a brand as well. They always want growth.
- 32:57They want the best people, they want, they want the best tech,
- 33:00they want the best everything, right?
- 33:02And then and then they say, well, would this is we don't
- 33:06think it's worth. We don't want to pay for it.
- 33:07Yeah, but George, it's funny you say that about outcome based
- 33:11contracts to say I've had that experience multiple times over
- 33:14the past 10 years, clients have approached me and said we want
- 33:17to negotiate some kind of outcome or value based contract.
- 33:20And then let's say you approach to be okay, we'll do what it
- 33:23cost normal consulting services, your margins or projects or
- 33:27anywhere from like 48 to like 50 to 55%, right.
- 33:30So let's say we'll do it a cost which is a fraction of that.
- 33:33And then if we, if we, if we hit, hit these milestones, the
- 33:37up, you know, the downside for us is we don't make any money.
- 33:39The upside for us is if we hit these milestones, it's got to be
- 33:42upside, otherwise not worth it for the consultant services org,
- 33:45right? So you, you, if you hit these,
- 33:47whether it's accelerators based on time or quality or whatever,
- 33:50right, you're going to pay us extra.
- 33:52Every single time I, I've shown that to a client, they've balked
- 33:54and said we, we don't want to sign up for something if we
- 33:57don't know how much we're actually going to pay someone in
- 33:59the org probably usually procurement says no to that.
- 34:02I found that really fascinating, actually, even.
- 34:04If it's tied to grow, the number should justify the costs.
- 34:09Spread I even know I, I, I George, I mentioned this
- 34:12earlier. I know one, one guy.
- 34:13He he he sold a project. They crushed all their all of
- 34:17their targets. The client owed them a lot of
- 34:19money backed out of the contract.
- 34:20When you're in before this payment, this balloon payments
- 34:23came. So I mean the outcome based
- 34:25contracts sound great, but like I think it's not only
- 34:28consultancies and services need to get better and I think brands
- 34:31need to understand how to buy them.
- 34:34Yeah, I think we talked a lot about agencies, right and their
- 34:42need to kind of rethink or re evolve.
- 34:44But you're right, real brands because you would argue
- 34:49increasingly so probably over the last five to six or seven
- 34:51years, most of the people that work in brands are ex agency
- 34:55people, right or wrong, right. So what you what the branch
- 35:01procurement is a different beast.
- 35:02But but but you know, the people that have that have kind of in
- 35:05procurement that have worked in the media world, work in the
- 35:09marketing world versus not just a fine widget world.
- 35:12I would argue that they have a much more innate understanding
- 35:17of kind of the the the commercial mechanisms that
- 35:21affect brand needs to also change, right.
- 35:24They can't say we want performance based and then the
- 35:28agencies delivers and exceeds the performance as well.
- 35:30No, we want to be in a go. It can't work that right, right.
- 35:33So what I what I and and yeah, the brand also has a logical
- 35:36point is if it becomes so economically infeasible for a
- 35:42brand to pay something, then what you have to go in there is
- 35:46do something like look, if if our performance exceeds X,
- 35:52right, Danny, it trigger, it triggers a, a, a conversation.
- 35:57Yeah, right. But if it's yeah, some of this
- 36:00is dependent on on your ability to attribute performance, which
- 36:03in and of itself MMMMTA and all of the related solutions,
- 36:06incrementality measurement, it's it's tough to do, especially as
- 36:10you start to move across channels into offline worlds.
- 36:13So yeah, your, your data analytics team, you're a data
- 36:16science guy, right? Come back and just say I'm not
- 36:18100% confident that this is an actual.
- 36:21Attribution and I think that's I think that's the I think that's
- 36:26where a lot more a lot more thoughtful conversation, which
- 36:33which said a different way, a lot more educational
- 36:36conversations need to happen on both sides of the fence where
- 36:40you can sit down and saying, OK, it's not perfect.
- 36:43We agreed to a commercial construct.
- 36:46Here's how we're going to judge that.
- 36:47Everybody, Everybody and. Here are the assumptions that go
- 36:50into this map and everybody's getting stepped that these
- 36:53assumptions. Ironically, it may actually be
- 36:55easier to do this in technology than with media because
- 36:58technology you can deliver, you can deliver and ship something
- 37:01before a time frame, right. And with media, I mean, you
- 37:03know, there's so many dependencies in terms of
- 37:05creative campaign, a lot of things.
- 37:06So sorry George get you off the going.
- 37:08Yeah. No, no.
- 37:09So that's why I think that's where you're seeing some of the
- 37:16foundational models for them becomes easy because they're not
- 37:19really being judged by by outcome.
- 37:22They're just being judged by utility in and out, right.
- 37:26The number of prompts and the number of responses, right is
- 37:29very mathematical to them. Easy, right, scale right, Ask
- 37:32questions, we'll give you answers.
- 37:33Great you're talking about. AI, specifically from.
- 37:37AI specific but but I, I say that just because that is now
- 37:41being introduced into the agency's world right and so now
- 37:46they're saying well you have questions we'll give you answers
- 37:51right and that's a mathematical it's very, very simple right you
- 37:54you ask questions we'll give you answers right now the value of
- 37:58those answers right as long as. Ultimately the same every time
- 38:02they come out. Well, that's that's a different
- 38:04that's a different story. So I think it's going to get, I
- 38:08think it's, it's you're definitely seeing a lot of
- 38:12tension and friction and even even with with quote UN quote,
- 38:17the most happy on the surface level relationships between the
- 38:23agency and the brand. And I've been on both sides of
- 38:26that table, right where I've been inside immersed in brands
- 38:28as a consultant and I've been conversations agencies and
- 38:32everything in between. And I will tell you what brands
- 38:34say when there's nobody listening versus what they say,
- 38:40what they say when it is completely different, right?
- 38:43They are very. Negative.
- 38:46I mean, is it? Does it tend to?
- 38:47Be like, I think I, I, I think, I think you got to always, so
- 38:50strip it down to kind of, you know what that, what that
- 38:53negative in that context is. But I will tell you that it's
- 38:57always, it's always skeptical. Yeah, that's the one thing I've
- 39:02always come out to. It's always very skeptical no
- 39:04matter what. Like I, I really like this
- 39:06person. Do I really trust this person?
- 39:09Not really. But I do like the person.
- 39:11But I don't trust the person. I don't believe that that person
- 39:13has my best interest. Yeah.
- 39:14Trust. Just you have no like and trust
- 39:16you have to have all three of those three of those boxes
- 39:18checked. Yeah, I agree.
- 39:19So, and they're also, and they also say, do I believe that the
- 39:25that whoever I'm partnering with truly has my best interests at
- 39:28heart, truly has my best interests at heart, right?
- 39:31Truly is, is unbiasedly bringing to me the best combination of
- 39:37people product, right? Are they bringing that
- 39:41irrespective of where it comes from, irrespective of where it
- 39:44sits? Do I believe that?
- 39:45And, and I would argue that the that answer is most of the time
- 39:52is no. They don't believe that because
- 39:53they believe everybody is trying to put their interests above and
- 39:59beyond the interests of the brand.
- 40:01Right. Well, this is consulting side,
- 40:03George. I think this is one thing that's
- 40:05happening that's interesting too is that like consulted, like
- 40:07with a large language model, you can get really smart answers.
- 40:10So I think the value of just bringing smart people has been
- 40:13diminished and clients know this, right.
- 40:15So they're expecting deep, deep expertise.
- 40:17You're like really specific subject matter knowledge that
- 40:21you can bring to things. They're and they're expecting
- 40:24proactive thought leadership, right?
- 40:26I mean, I think that's where the consulting model is changing
- 40:28quite a bit. I think the big guys are they're
- 40:30starting to feel this and they're all they're all bringing
- 40:32in McKenzie doing this far. They bring in like they'll hire
- 40:35like MDPHDS even if it's just part time entertainer just to
- 40:38bring them to calls and to support on projects.
- 40:40So I think we're seeing this as a general trend.
- 40:42Yeah, I would say I would. You know, I think a lot of the
- 40:45things that ultimately once you can charge for premium has been
- 40:53commoditized, right? That's what AI, I mean, AI is
- 40:56both an enabler and a detractor, right?
- 40:59I think everybody's feeling the blunt of both of those cups on
- 41:04our blade, right? So it can definitely reduce the
- 41:06latency between our steps. It can definitely make you much
- 41:11more efficient across steps. For real.
- 41:14What you, what you mentioned has always been the case.
- 41:17I will argue it doesn't really matter if AI came or didn't come
- 41:20at the end of the day, right? You can have all the data in the
- 41:23world. You could, you can put all the
- 41:25machine learning behind it. You can put all the models
- 41:27behind it, right? It's going to present something
- 41:30to you. Doesn't mean that that
- 41:32presentation is right or wrong. It just means it's presenting
- 41:34something to you. What ultimately creates impact
- 41:38is somebody that has understands the world of that business,
- 41:43understands the entity of that business, understands how that
- 41:46business makes money, understands how that business
- 41:48competes because. Then they can interrogate that
- 41:51output. That's it.
- 41:53Or or, because everybody assumes that's the end point.
- 41:57To me, it just reduced the latency between data to what
- 42:02potential impossibility exists. Somebody still has to take that
- 42:06and say, here's the reality. Somebody has to refract that
- 42:10through the reality of what a strategist truly does.
- 42:14I will argue the the, the, the world or the value of a
- 42:20strategist is going to ultimately become much more
- 42:24impactful. Well, a high level strategist
- 42:28because that's also an overused term, but it's you're basically
- 42:31saying the iteration, that sort of drafting process for lack of
- 42:34a better way of saying it, right, which could take a ton of
- 42:36market research, it could take a ton of client Intel and research
- 42:40and analytics. That time has been compressed
- 42:43exponentially. Very much now.
- 42:45You need a strategist, a high level strategist, to be able to
- 42:47say I'm going to run this and if it's wrong, they might go
- 42:50through 10 iterations of that output before they come up with
- 42:54something that they. Would I'm already seeing that
- 42:56happen, I'm already seeing a premium and more value like
- 42:59shift towards that type of role. Yeah, yeah.
- 43:01And so I would argue even in the agency world, I would argue it's
- 43:07not your typical strategist, right?
- 43:09I would argue it is this notion of combining the component parts
- 43:16of what makes the true strategist impactful.
- 43:20So yeah, marketing strategist, but I'll also argue is a
- 43:22business strategist. Oh yeah, you can't argue as a
- 43:25product, right? So, yeah, so you have, you have
- 43:27and this and This is why I think building a core AI system,
- 43:33right? Because I mean.
- 43:34Back to that core competency of like what?
- 43:36You're how many times, how many times you're going to find the
- 43:38Unicorn that has done market strategy, has done economic
- 43:43theory strategy, has done behavioral strategy, has the
- 43:46marketing strategy. Maybe there's one person in the
- 43:49entire world that actually fits that bill, right?
- 43:51Versus if you have an AI system that can capture a behavioral
- 43:55strategist that can capture a marketing strategy, you can.
- 43:58Synthesize that and you're acting as the.
- 44:00Curator and can connect that and saying hey, by the way, these
- 44:04are how these strategies ultimately come together to to
- 44:08and now on the builders. Oh my God, I didn't realize that
- 44:12that that because no human being ever, I don't care how smart we
- 44:16ever become, is going to be able to work under that multi
- 44:19dimensional frame. Without FTE teams of hundreds,
- 44:24even thousands of people, right? That's what that's what's
- 44:26actually being sort of compressed and eliminated.
- 44:28And it's and it's kind of all of our, the three people on this,
- 44:31on this podcast, it's all of our futures that's being able to do
- 44:35that sort of conducting and orchestrating.
- 44:37But you don't have to hire, you know, the oboe section of the
- 44:40violence section of the drummers.
- 44:41You can actually have a lot of that stuff done, but you have to
- 44:44have the the the Intel in smarts to be able to conduct connect
- 44:48conductance in in separate. Good.
- 44:50Well, George, one of the reason why I think this is more
- 44:52important than ever is that like you look at like, I think the
- 44:56agency model, right? The media business I is under, I
- 44:59think increased pressure due to demands for transparency, right?
- 45:05Make sure you have things like principal based media where
- 45:07you're probably never going to get the transparency.
- 45:08But you know, I would argue maybe principal, principal based
- 45:12media. I, I don't know if in 10 years
- 45:14it's around, right? Maybe we'll be, maybe it won't.
- 45:16I mean, I think agencies need it now because it's, it's, it's,
- 45:18it's, it's, it's a way to keep the keep the gears turning
- 45:22right. But generally there's need for
- 45:24more transparency. AI is making it easier to bring
- 45:26certain tools and processes in house.
- 45:28Brands are doing that. Even if agencies can be buying
- 45:31media, which I think for large like global organizations, they
- 45:33most, they probably will for the foreseeable future.
- 45:37I think the margins on that are falling.
- 45:38And I think this creates a problem for agencies, right,
- 45:40Because brands are demanding lowered like, like, I mean, if
- 45:43you're, if you're, if you're picking your agency, when you
- 45:46put any counter for review just based on your CPMS with the
- 45:48major pubs, it's a race to the bottom.
- 45:50And it's, it's hard to your point earlier, it's hard for
- 45:52agencies to make money without principal based media and other
- 45:55ways to do that. So if, if the media sides under
- 45:58a lot of pressure, you almost have to invest these other
- 46:01things that you can that add value, you can charge transfer.
- 46:04Can you get your thoughts on that?
- 46:06Yeah, I think I, I fundamentally still believe that media is
- 46:13probably the most important part of marketing because it is the
- 46:18closest thing to ultimately creating the necessary impact to
- 46:23the consumers, right? I, I I I speaking about like I
- 46:27it becomes increasingly important in this world of
- 46:31massive fragmentation, massive attention, nutrition, right?
- 46:35Massive over saturation, effective media.
- 46:40True. Effective.
- 46:41Media to come through and find people where they are in 1001,
- 46:45Bulk and I. So to answer your question, it's
- 46:50the wrong, it's the wrong vantage point on the media,
- 46:53right? If you believe that that is the
- 46:55most important aspect, right, the closest thing to where a
- 46:59consumer is susceptible receptor, right, then it doesn't
- 47:03really matter if you pay a dollar for this thing, you're
- 47:05selling it for $0.50. That shouldn't matter at the end
- 47:08of the day. It should matter is how are you
- 47:09getting that specific media in front of them, the most
- 47:13receptive consumers, right? And so don't don't judge
- 47:17somebody's component part on ACPM.
- 47:20Yeah, we'll do exactly because that, that race to the bottom
- 47:22ends up that's it. That's, you know, you're paying,
- 47:24you're paying cheap CPMS and guess what, You may not even be
- 47:27reaching a human. That's right, that's right.
- 47:29That's a whole system on the platform.
- 47:30That's a whole that's that's and that's a choice, right.
- 47:33People gaming the system is a choice, right?
- 47:35You have a choice not to game the system, but there was a
- 47:37choice to game the system, right?
- 47:39And most unfortunately decide to game the system because there's
- 47:41money to be made. I would I would say is if an
- 47:46agency fundamentally believes that their investment in data
- 47:51and technology in AI right can ultimately increase the
- 47:57contribution value of a media activation, then they should
- 48:02have no issues being transparent.
- 48:06Yeah, no, exactly. That, yeah.
- 48:08And I mean by that is all transparency, right?
- 48:11Yeah, this is the data, yeah. How much it cost you?
- 48:15Yeah, maybe that needs the brands even go platform direct
- 48:18with contracts, right. And the agency is just, you
- 48:20know, I mean, you could even say that if you really want full
- 48:22transparency. Yeah.
- 48:24But no. But do you deal with
- 48:26overwhelming complexity? I mean, isn't that why principle
- 48:28based buying is attractive to big brands?
- 48:29Because you have hundreds? Well granted a lot of the media
- 48:32spin is in Google and in Meta and a handful of Amazon, but
- 48:36still there's massive fragmentation across CTV and all
- 48:39these other walled garden type channels.
- 48:41How do you manage not? Just duopolating, Yeah, there's
- 48:43a lot of channels. It's not so, but here's the
- 48:47argument of that. And I think you're starting to
- 48:49see some of that play out, particularly with the publicist
- 48:52and The Trade Desk. Healthy conversation, yeah.
- 48:55Healthy conversation. We have Fool coming on next week
- 48:57to talk about that. Oh yeah, you can argue AI, the
- 49:06introduction of AI, right? You can make it compelling
- 49:10argument that is going to disintermediate an agency,
- 49:15right, because if AI can do what it everybody says it can do, it
- 49:21can handle almost any not yet, but it can handle almost
- 49:24anything inside a a campaign life cycle everything right.
- 49:30You can assume that and if the Trade Desk for Google or Credo,
- 49:34whoever the hell decides to say like we can handle that, right?
- 49:40If this notion of is creating a system that creates the best
- 49:45match between what a brand wants and the consumers that want that
- 49:49brand, right, then you are you are collapsing all the
- 49:54complexity into a single environment now.
- 49:58Yeah, you might, you might have to give up a little
- 50:03transparency, right. But are you getting
- 50:07significantly more in return because you don't have to deal
- 50:11with all the intermediaries anymore.
- 50:14You're just dealing with a central platform, right?
- 50:17And you have more control. You would argue having more
- 50:20transparency into what's going in, how it's being done that
- 50:23with the cost structures, right, Assuming that the Google and
- 50:27Trade desk of the world, yeah, will open, open up their
- 50:30environment. Visibility into where those
- 50:32impressions? Are going and then if like
- 50:36agentic trading ever really becomes a thing, right?
- 50:38And then, then, then you could argue maybe the DSPS get, you
- 50:41know, they really get turned into dump pipes, right?
- 50:43Because if all the decisioning and all and all the allocation
- 50:45happens upstream of them, what are they really doing then at at
- 50:48that point, right? Here's the.
- 50:51Here's the argument right, I think.
- 50:53I think agentic will become will become a thing.
- 50:57I mean, we had this argument 15 years ago and programmatic came
- 51:00out and nobody's going to do it, but nobody's going to allow
- 51:02algorithms to make a decision. And here we are.
- 51:04Algorithms make a decision, right or wrong, right?
- 51:06So agentic, I think it's a question of not if, it's just
- 51:09when I go back to the fundamental thing, right, Who is
- 51:12telling the agents what to do, right?
- 51:15If you don't have a core DNA of your organization built into
- 51:22this right AI infrastructure, right AI system that is telling
- 51:27an agent this is what you should be doing, or a series of agents
- 51:30or a system of agents right, that is doing that, guess what's
- 51:34going to happen you're. Going to amplify a lot of.
- 51:37Crap, right? So, so the current environments
- 51:40in the trade vests of the world, the Googles of the world, they
- 51:44had built their own version of their internal because that's,
- 51:49that's why, that's why they're never going to unleash it
- 51:51because they don't want anybody understands how they built it
- 51:53and what goes into it, right? And so that's why they can say,
- 51:57well, algorithms in our in our environment are better than
- 52:00anybody else's because we have built this, this collective
- 52:03memory of context over the course of 25 years.
- 52:07Yep. Unfortunately, they're not going
- 52:09to. They don't want to share with
- 52:10anybody, right? Which is the whole, yeah.
- 52:12Collective memory of context is critical.
- 52:14That's that's correct. That's the core DNA.
- 52:17Eventually, I do believe that that that will happen, right?
- 52:21And then if that happens truly in camp, in marketing, that you
- 52:25have a series of agents that understand how to work in a
- 52:30system that is being empowered by a core, right, Let's call it
- 52:37hive mind. Well, they would be negotiating
- 52:39back and forth by site agents and sales agents.
- 52:41They would be talking together and negotiating, structuring
- 52:43deals in different ways, right? Yeah, yeah.
- 52:46I like to. I like to.
- 52:47You just said hive mind. That reminds me of the Borg.
- 52:49This notion. We're moving more and more
- 52:51towards the Borg. Right here to be assimilated.
- 52:55Of collective memory temporal over time right which gives some
- 52:59of these these core tech companies that have been around
- 53:01like Google that have you know huge lot volumes of data gives
- 53:06them just a natural advantage over any independent startup
- 53:09that's tech forward. I mean, do you?
- 53:11Think those, I think so. I think the the only difference
- 53:14and this is where brands I think will have to take ownership of
- 53:20that and the agencies ultimately want to compete, truly compete,
- 53:23we'll have to take ownership of that.
- 53:26Data is, is a commodity, but when you have enough of it,
- 53:32right, it really defines who you are.
- 53:34And I mean data as in as in just ones and zeros, but also data as
- 53:42as expertise that sits in people's heads once you put it
- 53:45inside a machine, right? That's true IP.
- 53:49The moment you get the moment you put that into Google, a
- 53:52trade desk, open AI perplexity, whoever it is, you are giving up
- 53:57your differentiation. That's it, right?
- 54:00So once brands realize that, once agencies realize that
- 54:04they're what they're going to do is never do that right?
- 54:07They're going to say, we're going to keep it inside our
- 54:09walls and we're just going to build this hive mind inside our
- 54:14walls. So any organization, you're
- 54:16seeing some of them and one of the companies that I consult
- 54:19with this company called Madsense is exactly that, right?
- 54:22That's why I consult with them because when we started this
- 54:26conversation, I've said, you know, I don't, I don't, I don't,
- 54:29you know, my bullshit meter is extremely high.
- 54:30So when I look at companies and say, OK, that sounds great on a
- 54:33piece of paper, can you actually do what you say to do?
- 54:36Right. Once I dig into it either
- 54:38quickly, yes or no, these guys are actually legit and, and
- 54:42they've built this notion I want to describe to you, which is a
- 54:45true AI system by taking what is the most important thing, which
- 54:50is knowledge inside an organization and encoding that
- 54:55into an AI system that compounds over time.
- 54:59So true memory, true context, true understanding, right, true
- 55:03traceability, true explainability, true
- 55:06orderability, right? And somebody can say, I get it,
- 55:09I understand. And which is to me was probably,
- 55:14if not one of the most important things is the most important
- 55:17thing is transferability, which basically says if a brand says
- 55:23I. Want it.
- 55:24Yeah, that that was my next. Question.
- 55:26All of it. Can I take a?
- 55:27Brand a brands like I want to outsource that or have any sort
- 55:30of limitations to access that for going forward.
- 55:33So is it, is it in brands better interest to bring that in, in
- 55:36house or do you think they don't have the core capabilities to
- 55:38do? That no, no, I think eventually
- 55:40I think what? What what you're starting to see
- 55:44and and this is the you guys have seen, you know WFA, you
- 55:49have seen McKenzie, you have seen all the consultancies, you
- 55:52have seen everybody says the same exact thing, right?
- 55:55So massive amounts of money invested in AI, very, very,
- 55:59very, very few under 5%, right. Our arguments probably even
- 56:03under 4% have seen true production ready AI that has
- 56:07given any value into the margin business.
- 56:09Why? Because everybody assumes you're
- 56:12bringing AI and you do nothing, right?
- 56:15It's magic. It's going to move your business
- 56:17from A-Z. Yeah, right.
- 56:20It's exactly right. Right.
- 56:23And then what they, when they start putting the level of work
- 56:25into it, they start dealing with the realities of wait, wait,
- 56:29wait, wait. It's telling me stuff that's not
- 56:31real because they're using foundational world models,
- 56:35They're using these these open source models, right, that are
- 56:39predicated on the agamation of the world data, not your data,
- 56:43the world data. Well, this is part of the reason
- 56:45why I don't see the all these predictions of AI displacing all
- 56:48these jobs. I mean, I don't see that
- 56:50happening at all. I mean, it's creating jobs.
- 56:52I just don't see it eliminating jobs at any.
- 56:53It it's it's it's it is re. It is recoding what people will
- 57:03do and how they do it that's ultimate.
- 57:05And I think that's probably the other big reason why you see
- 57:08massive failure. Organizations don't understand
- 57:11that they need to change management component.
- 57:14Skills shift and they're not prepared for that and so they
- 57:16get this capability. And here's the AI make it work.
- 57:22The way I work. It's going to fail.
- 57:24I can tell you that from day one.
- 57:27Have it have it automate this existing process, which maybe
- 57:30you don't need with, you know, like say, I think we're that's
- 57:32we're going to where we are. Yeah.
- 57:33So companies, so companies like companies like like Madsense
- 57:37that have basically come out and says, look, we will build an AI
- 57:42system inside your walls, right? It's yours.
- 57:46You can do whatever you want with it, right?
- 57:48It's our proprietary architecture, but it's yours to
- 57:52begin with. At any given time, you can pull
- 57:54it out, you can push it somewhere else.
- 57:56It doesn't really matter. It's never your right.
- 57:58It's yours from the from the get go.
- 58:00Those types of companies that understand true application of
- 58:03AI, right, irrespective of or said differently, not bound by
- 58:08time, place and space. Those are the companies that
- 58:12will ultimately enable brands, agencies, or anybody else to
- 58:16truly have native AI capabilities that will be force
- 58:22factors or force multipliers for people.
- 58:24Yeah, And what is Madsense? What are the outcomes?
- 58:26Not to shill Madsense, but what what are what are they pushing
- 58:29from an outcomes perspective in terms?
- 58:30Of SO. It's.
- 58:32It's exactly what I told you before in the marketing world is
- 58:36true incremental output, right? So definitely it's true marginal
- 58:40gains through the business, right?
- 58:42You can say it's top line revenue, yes, but it's also down
- 58:47to understanding the component parts of a business because once
- 58:50you help the organization build that level of understanding,
- 58:54right, Madsense types companies have an understanding of the
- 58:58business component parts of what a business works, right?
- 59:01Because what Madsense is not just marketing intelligence,
- 59:05it's true business intelligence, right?
- 59:08What? Because I again, I said this in
- 59:10the beginning, you cannot have marketing without an
- 59:13understanding of a business. There's no way you can have
- 59:15marketing be a demand driver of a business if you have no
- 59:18context of a business, it's impossible, right?
- 59:20That's why a lot of these companies get it wrong, right?
- 59:23Because they're saying what why it's not working, because
- 59:28marketing was not the problem. That's why it's not working,
- 59:30right. It was never about marketing was
- 59:32always about this notion of a bit.
- 59:34So once they do that, then they have this natural understanding
- 59:37is the interrelationship between those component parts.
- 59:40So they're they're in a much better position, much better
- 59:43position based on what I've seen so far.
- 59:47Recommendations and. Well, it's not even strategic.
- 59:51It's yes, but it's very tactical, executional,
- 59:54operational that you can always tie it back to a net incremental
- 1:00:01outcome, which is sales and margin.
- 1:00:05But you can reverse engineer that back to exactly where it
- 1:00:07came from. Like how many times can any
- 1:00:10organization trace decision making chain of thought?
- 1:00:14Through the entire organization back to a specific aspect,
- 1:00:17right? It never happens most agencies.
- 1:00:19Never. It's because because that's not
- 1:00:23how people work, right? You campaign specific point
- 1:00:26solution specific and you're like, well, does anyone what we
- 1:00:28did a year ago? I don't nobody even remembers
- 1:00:31what we did a week ago, right. So you've lost all that
- 1:00:34intelligence, and then you're like the.
- 1:00:36Source of this bad decision or this bad outcome?
- 1:00:39Like nobody can trace it back and so yeah.
- 1:00:41Companies like that, I think, I think are going to be the true
- 1:00:45catalyst of, of anybody that wants to own, I would argue the
- 1:00:52most important thing in any business, which is knowledge.
- 1:00:57Yeah, that is the most compounding intelligence.
- 1:00:59Compounding. The ability to build their IP
- 1:01:02into into compounded over time, right?
- 1:01:05Regardless of who leaves the organization or who comes.
- 1:01:08And that, and that's, and that's the thing.
- 1:01:10It doesn't really matter how many times you switch an agency,
- 1:01:13how many times people come and go, what they right.
- 1:01:16You have never broken that system, that system right, The
- 1:01:21more you give it now. It gets better.
- 1:01:24Well, but here's the thing, right?
- 1:01:25And this is for brands as well, right?
- 1:01:27Just just so I know. This this is for brands and
- 1:01:29agencies. It doesn't really matter at the
- 1:01:30end of the day because they mean they're all trying to do the
- 1:01:32same exact thing, right? They're all trying to buy, build
- 1:01:35IP. They're all trying to win in the
- 1:01:36competitive marketplace. They're all trying to grow.
- 1:01:38Yeah. So the the objectives is the
- 1:01:40same. The way they do it is different.
- 1:01:42But all AI is stupid in the beginning, right?
- 1:01:48I mean, it's it's called machine learning for a reason.
- 1:01:50It's kind of learn off of something.
- 1:01:52So it's all stupid. What people don't realize is
- 1:01:56this notion of you have to you, you have to give it something
- 1:02:02and constantly give it something, right?
- 1:02:05The AI is not going to know. The more you you interact with
- 1:02:08the AI, the more you give it context, the more you give it
- 1:02:11memory, the more you give it logic, the more you give it
- 1:02:13decision traces, the more you give it right.
- 1:02:15It is going to learn and eventually it will learn enough
- 1:02:20about the way your organization thinks and works.
- 1:02:23And then AI truly becomes, I don't want to say because I
- 1:02:28don't think it's anything about artificial intelligence, right?
- 1:02:32I would argue it becomes this notion of augmented
- 1:02:36intelligence. It now knows how to move the
- 1:02:39things. Well, I think it becomes true
- 1:02:42intelligence you can really harness and use in your
- 1:02:45day-to-day. I I think that's the thing.
- 1:02:47And I guess since Speaking of knowledge, I mean like George, I
- 1:02:49mean, you know, we're see, we're over an hour here.
- 1:02:50Like looking at like, like, like, say, moving forward signal
- 1:02:54and noise. What types of topics are you
- 1:02:56excited to share? Are are you, do you want to
- 1:02:59write or talk about like love to give the the viewers a little
- 1:03:03preview of that? So I would, I would say, given
- 1:03:07how much noise that they exist in the marketplace today about
- 1:03:12AI, nobody, because I don't think they know how to speak
- 1:03:17about it. Nobody really talks about the
- 1:03:20true application of AI. How do you really apply AI,
- 1:03:23Right? Everybody, every and all you're
- 1:03:25seeing right now is most companies, we have an agent for
- 1:03:28this, We have an agent for that. We have this for agent and OK,
- 1:03:32who's telling the agent what to do?
- 1:03:34Yeah. Yeah.
- 1:03:35And that's such a critical theme.
- 1:03:37I mean, I think your point is that this is, let's dispel the
- 1:03:40the myth that this is easy. This is not easy.
- 1:03:44And Mike Visenzino, who's one of our contributors as well wrote
- 1:03:46an article about that on signal and noise that it is extremely
- 1:03:49like a builder that's in there building and testing and, and,
- 1:03:53and eliminating bugs and doing bug testing and just the the
- 1:03:56trial and error behind building this correctly, especially for
- 1:04:00IP purposes within a company, it's extremely difficult.
- 1:04:04And I mean, otherwise everybody would be doing it so.
- 1:04:07Everybody, everybody would be doing it correctly at scale.
- 1:04:11Yeah. And so I say, which is not what
- 1:04:14we should, quite the opposite of that actually.
- 1:04:18Like all the agencies and even all the tech they call operating
- 1:04:21system, right. We have an OS for this.
- 1:04:23We have an older 1. So you say, OK, so what's your
- 1:04:27definition of an operating system then?
- 1:04:30Because what you have is not that they're like, no, I have
- 1:04:31not. How it how is that possible when
- 1:04:3498% of the organization have failed to apply AI into
- 1:04:38organization? How do you, how do you claim,
- 1:04:41how do you make this bold claim that you will have an operating
- 1:04:44system that that is ultimately at the core of an AI?
- 1:04:48And they're like, well, you know, it's well and they try to
- 1:04:51and you're like, well, then you're not really applying AI,
- 1:04:53are you? Right, right.
- 1:04:54If your definition of AI is you built an agent, well, that's
- 1:04:58commodity. I can go to Claude right now and
- 1:05:00give. It a build it Asia and you have
- 1:05:01a download LEM wrapper right? Yeah.
- 1:05:04And that and, and, and thank you.
- 1:05:06That is exactly what when you pull back the layers, that is
- 1:05:09exactly what probably 98% of the organizations have.
- 1:05:13It's. Product marketing, it's not
- 1:05:16operating. System so so to me and they try
- 1:05:18to hide it into all these these words and phrases that we have
- 1:05:22this and architects and at the end of the day, if they truly
- 1:05:25had that truly had that right they they would be it would be
- 1:05:30like such a no brainer of an organization right now.
- 1:05:34The other aspect of it all, which is the true application of
- 1:05:37AI. The true application of AI has
- 1:05:39to have component parts of data, has to have the component parts
- 1:05:44of true understanding of what what AI is or isn't, and then
- 1:05:48has the applications of what it means to truly operationalize it
- 1:05:52and and have production ready. Yeah.
- 1:05:56Part of that, that's a critical point.
- 1:05:58That's a. Critical part of that part of
- 1:06:00that is change management and people and processes.
- 1:06:03Yeah. I'm not saying firing people.
- 1:06:05I'm just saying change management and the way you think
- 1:06:08and the way you work and the way your workflows happen, right.
- 1:06:12Everybody assumes you take the most sophisticated AI doesn't
- 1:06:16really matter, and you just say, oh, oh, I like this.
- 1:06:20I'm just going to take this and put it inside my agency.
- 1:06:23And you're like, what hasn't been working right?
- 1:06:25My people are frustrated. Yeah, because you've set the
- 1:06:28wrong conditions of why you've done this.
- 1:06:30So. AI education and organizational
- 1:06:32alignment. Although I think, I look, I
- 1:06:34think, I think our industry generally marketing
- 1:06:39communications, which is ironic to me because the whole
- 1:06:41fundamental precipice of why communications ever was
- 1:06:45developed was to communicate. So people make informed
- 1:06:47decisions. Our industry is notoriously bad
- 1:06:51by doing that. So I think like an A true and
- 1:06:56maybe maybe this is something for you guys, but true entity
- 1:07:00that can give true education for the masses.
- 1:07:07Like strip all the nonsense out like you really want, you really
- 1:07:10want to do AI correctly. This is the way you should be
- 1:07:12doing it, right? Everybody else is trying to say,
- 1:07:14and when somebody is trying to sell you something and they're
- 1:07:16telling you that their version of AI is better than anybody
- 1:07:20else's, but it really doesn't say why it's better than anybody
- 1:07:22else's, right? They use all these.
- 1:07:24It's the, there's probably some truth to it, but most of it is
- 1:07:28not they try to hide. So I think there's massive,
- 1:07:32massive, massive need for true transparency about how to true
- 1:07:37how to apply AI in any organization.
- 1:07:39What it really takes, right? It's not it's not a single
- 1:07:43thing. It is going to be like all
- 1:07:46things is going to be a continually investment and
- 1:07:49recalibration. What you said, Brent, a
- 1:07:51recalibration of thinking talent, understanding
- 1:07:58adaptability, OK. And and knowing that not
- 1:08:04everything right now is clear. Most of it is still very opaque
- 1:08:07right now, right? You don't know.
- 1:08:08You don't know what you don't know yet, but you do know that
- 1:08:11there's there. There is some power and truth to
- 1:08:13what AI can bring to organizations.
- 1:08:15Some, I would argue some core companies like a Madsense,
- 1:08:22right, that have come from a different way of thinking, not
- 1:08:25necessarily from the marketing world, but understand how to
- 1:08:29impact behavior and how to really aggregate true
- 1:08:34institutional knowledge and intelligence and apply that to
- 1:08:40organizations. I think those are the ones that
- 1:08:43are going to really help, right? Any business understand, OK, so
- 1:08:50this is not about a widget. This is not about an agent.
- 1:08:54This is not about pulling down a workflow because you go to many
- 1:08:58sites right now and says, oh, like I have a workflow for that.
- 1:09:00And I have this, I have an agent for this.
- 1:09:02And in reality is you've kind of missed, you kind of missed the
- 1:09:06bowl on what it's going to take to actually do this, right?
- 1:09:09I don't blame them from doing that because what they're doing
- 1:09:12is being reactive to what the agencies are asking them for or
- 1:09:15any organization is asking for. Like, I just need something to
- 1:09:17work in my workflow. I just need to show progress
- 1:09:19that I'm AAA, find something right versus saying wait, wait,
- 1:09:24wait. That's that you you're thinking
- 1:09:26about this room. You really want to be an AI
- 1:09:27agency or you won't be AI business.
- 1:09:29This is the way you should be doing, Yeah.
- 1:09:31Well, question about about the adoption of AI and how it
- 1:09:33impacts, let's say, like looking at the media space.
- 1:09:36So I mean, you see that, you know, the platforms like the
- 1:09:38Google Meta, they're all building their own kind of AI
- 1:09:42layer on top or they're re architecting, right?
- 1:09:44They already, they already have a lot.
- 1:09:46And you know, like Google's probably the furthest along, but
- 1:09:48everyone's doing it. Meta, you know, Meta is pretty
- 1:09:49far along too. The, the, the agencies are doing
- 1:09:53something similar, especially the hold coast are trying to
- 1:09:55build, let's say an interface, you know, across all the
- 1:09:57different DSPS. So a lot of the what's now done
- 1:10:00by 25 year old humans could then be done by some, some, some
- 1:10:04different AI tools, whether the agents stitched together or some
- 1:10:07kind of layer, right. And then, well, then you then
- 1:10:10you have the argument or brands should be doing this right.
- 1:10:12You know, they should be, they should be doing this for
- 1:10:14themselves because maybe the tools make it easier.
- 1:10:17Everyone seems to be trying to build the same thing, right?
- 1:10:21Like what? Like who do?
- 1:10:23Who do you think wins this or does Does it not really matter?
- 1:10:25Does everyone or no one? I, I look, I think, I think it's
- 1:10:32funny because it's, I think it's like Back to the future.
- 1:10:35Why the brand choose an agency? If you go back forward, why do
- 1:10:41the brand truly value an agency, right.
- 1:10:43It had unique thinking. That's what it was had your knee
- 1:10:47thinking, didn't we? Oh, which data source do you
- 1:10:50use? Right?
- 1:10:51What technology? It's just because they believed
- 1:10:54that they had the best thinking in the industry.
- 1:10:58So as much as we talked about AI, eventually that will become
- 1:11:02a commodity. And you, you would look at it
- 1:11:04right now, it's probably starting to look like that right
- 1:11:06now, right? It's commoditized.
- 1:11:08Data is commoditized. AI will be commoditized.
- 1:11:11The only thing that truly differentiates one over the
- 1:11:14other will be the true institutional knowledge.
- 1:11:17I am better or I am the best. What it is because I have true
- 1:11:24IP in thinking. Nobody can.
- 1:11:27Nobody can steal that away from me, right?
- 1:11:29Nobody can take what I have over the 15 years and take that and
- 1:11:34transpose them into something else.
- 1:11:36It does not work. So, George, it sounds like
- 1:11:38you're saying like the decision to split out media from like,
- 1:11:41let's say brand strategy and creating all those other things
- 1:11:44that happened, you know, decades ago.
- 1:11:46It almost sounds like that that reverses.
- 1:11:48It has to right? If like if, if it's just, if
- 1:11:51it's just buying media for the sake of buying media, then it's
- 1:11:53a race to the bottom. And we talked about and AI and
- 1:11:56AI becomes a commodity. So you you, but you still need,
- 1:11:59you're right, like. If you're, why?
- 1:12:00Would you work with an agency? A good one?
- 1:12:02You want for the sharpest minds who are, who know, who know The
- 1:12:06Who could present and propose the most interesting things,
- 1:12:09have the and have great experiences, right?
- 1:12:11Yeah, it's, it's the business knowledge, it's the creative
- 1:12:13knowledge that's gonna, you know in the audience, knowledge of
- 1:12:16who you're, of what, who, who brands are.
- 1:12:18Trying to reach. That is the IP and that that is
- 1:12:21often times in creative agencies at this point.
- 1:12:23Create and I think I think you will start seeing the morph of
- 1:12:34creative agency or branch strategy.
- 1:12:36You will start seeing the new version of what I would think an
- 1:12:39agency would look like, which is just which agency do I go by to
- 1:12:44be be my growth architect? Not execution, not, not not
- 1:12:49activation, but true growth architect of my organization
- 1:12:54because they have the smartest thinkers that have been enabled
- 1:12:59and empowered to truly think right, think right because they
- 1:13:05have this core on depending on them.
- 1:13:07It has not dilute. It has not diluted.
- 1:13:11Systems integration, same thing, George is happening, right?
- 1:13:13Because like you're not going to go with the systems integrator
- 1:13:15because they have like 100,000 people who can code for cheap in
- 1:13:19India. Like AI can code for much
- 1:13:20cheaper and much faster, right? What you're going to do is
- 1:13:23you're going to go with the one that brings the best thinking,
- 1:13:25the best ideas, that brings the 10X engineers to bring us that,
- 1:13:29the 100 X strategist, the ones who can do that.
- 1:13:31That's a good point. So it's funny, I look at, I look
- 1:13:34at, I don't know why it's not happening, but I think
- 1:13:36eventually it needs to because every time you see a pitch, I
- 1:13:40don't care which version of this pitch, oh, we decided to go with
- 1:13:44XYZ because they came out with the best data strategy
- 1:13:47technology. And you sit there, go wait,
- 1:13:49wait, wait, wait. That can't be right, right,
- 1:13:53Unless what they showed you, nobody else can do.
- 1:13:56Nobody else has, which is very, very hard for me to believe.
- 1:14:00Very hard for me to believe, right, that that like publicist
- 1:14:03wins because they got they have epsilon.
- 1:14:05You're like, OK, but but what makes that fun?
- 1:14:08Different than what axiom is at the end of the day, it's the
- 1:14:10same goddamn company. They do the same exact thing,
- 1:14:12literally have same access to the same exact data, right?
- 1:14:15And then now people will say that's not true, but we're it's,
- 1:14:19it's if you go back far enough, if you've been in this business
- 1:14:22long enough, you will know what those companies came from.
- 1:14:24They're all database marketing companies at the end of the day.
- 1:14:26They all, you know, this broadcast where you came from as
- 1:14:28well, right, right. So why not say, look, the reason
- 1:14:34we chose this company really, really chose this because they,
- 1:14:37they give us access to the smartest people that they have
- 1:14:39in the organization that will be working on our business.
- 1:14:42That's why you, that's what you should tell somebody, right?
- 1:14:44I chose and I would, and I would do it even better.
- 1:14:47I would call people out and say the reason we went with this
- 1:14:49agency is because Brett Rio and George, they wowed us with how
- 1:14:54much knowledge that they brought to our company into these
- 1:14:58meetings, right? Assuming, assuming everything
- 1:15:00else the same like I would argue, oh, we can buy, we yeah,
- 1:15:04well, no, we'll give you preferential rates.
- 1:15:06That's transactional because everybody you talk to, I want a
- 1:15:10true growth partner. And then everything that comes
- 1:15:13after that has everything to do with tactical transactional
- 1:15:19vendor based right conversation. And you're like, well, how about
- 1:15:24that? Those two things makes no sense
- 1:15:26to me. You said you wanted this and yet
- 1:15:28this is the things that keep coming out of your mouth versus
- 1:15:30going to anybody and say, look, yes, they came to us with a
- 1:15:35bunch of things. But what really, really, really
- 1:15:38made our decisions clear are these people that were really
- 1:15:42smart calling about by name. I think once you do that, once
- 1:15:46you do that, that will then recalibrate what it always
- 1:15:51should have been about. And I couldn't.
- 1:15:53And I will argue which which will always be about which is
- 1:15:57does this agency or this service provider have access to really,
- 1:16:01really, really smart people? Yeah, yeah.
- 1:16:04I'm going to give you, I'm going to give you the best.
- 1:16:06And I tell this all the time to people, the best comparable that
- 1:16:10I have when you go to investment banking, right.
- 1:16:16And you look at it at all of them, and you say, well, Apollo
- 1:16:21is better because they have all these.
- 1:16:24I argue that doesn't really matter.
- 1:16:26Somebody can build those algorithms.
- 1:16:28Again, why do they give somebody a $40 million package?
- 1:16:33Why do they go after people? They don't go after people
- 1:16:37because that person can code better than any other person.
- 1:16:41They went after that person because that person has so much
- 1:16:43institutional knowledge and so many relationships.
- 1:16:46And why does that person have a leverage?
- 1:16:48Because those people follow that person.
- 1:16:51They don't. They don't care what trading
- 1:16:54platform, technology, AI, quantum physics, whatever the
- 1:16:57hell they have. Which I think they're at right
- 1:16:59if they follow that person. So, and you're like, oh, so you
- 1:17:04went after because that person is the one that is telling you,
- 1:17:08thinking, taking all these component parts and saying,
- 1:17:12here's what you should do it and here's what you should do what
- 1:17:14right. If our industry, which is
- 1:17:17increasingly, and this is where all trading and algorithmic and
- 1:17:21all came from, it came from that side of the world.
- 1:17:22People don't want to admit that, but we, we, we steal these words
- 1:17:25coming from that side of the business, right?
- 1:17:27If you just take that same construct and say, OK, what
- 1:17:31differentiates one over the other?
- 1:17:33Why should somebody follow one over the other?
- 1:17:35It has to always go back to these people.
- 1:17:38So once you have that, but here's the thing, if anybody,
- 1:17:42anybody, this is a brand too, because brand shifts through
- 1:17:46CMOS as agencies said, it shifts through leaders these days,
- 1:17:51right? But if any one of those two
- 1:17:53entities can systemize that internal knowledge, right, they
- 1:18:01will have created something that transcends anything else.
- 1:18:05So if an agency can do that, if a brand can do that and capture
- 1:18:10the last 10 CMOS that they've had right in an organization
- 1:18:16access to that truly access to that right or WPP, which they've
- 1:18:21tried to do. And I grant them at least trying
- 1:18:23to do it right, which is aggregating at least what they
- 1:18:26said, aggregating 80 years worth of knowledge into a system.
- 1:18:32It sounds great connecting all that data sources eventually, I
- 1:18:37don't know how they come, how they encode the context of
- 1:18:40people inside the organization. I think that will happen over
- 1:18:42time. But at least they understood
- 1:18:45this notion of systemic power, right?
- 1:18:46So I grant, I applaud them for that.
- 1:18:49Who figures that out first, brand or agency?
- 1:18:54I will tell you that that will be the inflection point within
- 1:18:57where where the entire relationship conversation shifts
- 1:19:02completely because you don't. If a brand has done and has that
- 1:19:09true understanding of what's worked, what's not worked, why
- 1:19:13it's worked, under what conditions has it not worked?
- 1:19:16What, why make? Why did this?
- 1:19:19Competitive advantage. That's true.
- 1:19:21I mean, I think, I think that's why that's, I mean there's more
- 1:19:23of a demand for like technology advisory and specialized
- 1:19:26services. I see.
- 1:19:27I see this stuff is so hard to your point earlier, George, that
- 1:19:30I think there's going to be more demand for, for for good
- 1:19:33partners. Yes.
- 1:19:35But I think at some point brands, some of them, are
- 1:19:39already starting to do this, start realizing that they need
- 1:19:44to control more of that. We need to collect the mass and
- 1:19:49manage this over the history of our business and it makes it
- 1:19:52less less dependent on vendors, on external parties, on tech, on
- 1:19:57media and and less dependent on even the people within our
- 1:20:01organization which are transient might come and go.
- 1:20:05Right. You have that institutional
- 1:20:06knowledge that that continues to to.
- 1:20:08Grow we are I agree with you. It does not mean that there's no
- 1:20:11there's no need for outside. It just means that they will
- 1:20:14have a much better to go back to what we started.
- 1:20:17They will have a much better understanding of why they need
- 1:20:20them and how they compensate that need, because I think
- 1:20:24that's what will be the ultimate recalibration of what value
- 1:20:29looks like once I once I know, because here's here's why I
- 1:20:32would argue brands just don't because they just don't have
- 1:20:36everything that they need, right, which I will go back to
- 1:20:39is, is they've lost right? And it probably makes them more
- 1:20:44susceptible to anything. The most important thing is, is
- 1:20:48that institutional knowledge. They they don't have it and they
- 1:20:51keep losing it constantly time over time, right?
- 1:20:54CMO leaves that. That's massive, massive, massive
- 1:20:59intelligence that goes out with that person.
- 1:21:01Massive it is, right. So that's what once they have
- 1:21:05that and they can, they can really understand that they they
- 1:21:08can be very purposeful when they engage any services and
- 1:21:12solutions provided. OK, here's what I need from you
- 1:21:15and This is why I need it from you.
- 1:21:17And this is this is how I'm going to value that need from
- 1:21:19you, right? And then I think you can clearly
- 1:21:22come out with an outcome based OK, got it right.
- 1:21:25So this, this and this and this. Great Awesome, right?
- 1:21:29Both sides of the fence have to change.
- 1:21:31I would argue that brands are in a better position to probably be
- 1:21:38more of the catalyst and eternal change than somebody from the
- 1:21:43outside forcing them to change. Yeah, that's true.
- 1:21:47They have access to all of the the access, the assets, the
- 1:21:51artifacts, the things they've used to train it that, you know,
- 1:21:53I'd say they probably do a better, better job keeping those
- 1:21:56things than their partners do. You'd hope anyway.
- 1:21:59Yeah. So.
- 1:22:00So this has been a tremendous conversation.
- 1:22:03It should be great. Discussions or or are we good at
- 1:22:06concluding where we are? I think.
- 1:22:07I think. We are.
- 1:22:09We are good. Yeah.
- 1:22:12Do you want to do quick hits? Are we, are we good?
- 1:22:13Are we good with this? Yeah, sure.
- 1:22:15Yeah, let's do some quick hits. So really, you start.
- 1:22:18So biggest myth about let's say growth in general or agency
- 1:22:22growth if you want to get more specific?
- 1:22:28Agency growth has been wait, wait, wait to get right.
- 1:22:34So growth in the agency world historically has meant new
- 1:22:37businesses waiting for RFPs to come in what the world that we
- 1:22:40live in today. It's it's about a hunter
- 1:22:43mentality, right, which is completely different than new
- 1:22:47business, right. So I think a lot of agencies are
- 1:22:48painfully realizing that they they're no longer in a position
- 1:22:53to wait for somebody to send them an RFP.
- 1:22:57Yep, Yep. So strategy or execution?
- 1:22:59What kills more companies? Which and what's harder?
- 1:23:04And what's harder? Strategy definitely right.
- 1:23:08I think that's, that's been marginalized quite a bit based
- 1:23:12on and and and for the reasons that we've talked about think
- 1:23:15that's going to be the biggest catalyst differentiator for
- 1:23:17agencies exit. I'm not, I'm not marginalizing
- 1:23:20execution. I think execution is critical.
- 1:23:23You can, you can mess up a great strategy by not doing it
- 1:23:26correctly. But I would argue that impact as
- 1:23:31in what can have more positive or negative impact into a
- 1:23:34business margin strategy will always and I've seen that play
- 1:23:40out many, many times in my career.
- 1:23:42AI will create more winners than losers or more jobs than job
- 1:23:46losses. True or false?
- 1:23:50I think net AI will create more jobs.
- 1:23:54Yep, I think you're going to need you're going to recalibrate
- 1:23:57a bunch of people. That's why you haven't seen
- 1:24:01people don't realize this. And a normal given healthy
- 1:24:04economy, about 1.4 million people get laid off for whatever
- 1:24:10reason. Awesome.
- 1:24:12Great. So that's normal, right.
- 1:24:13So if you look at where we are, anytime like this, we have we
- 1:24:17have massive is going to be a recalibration.
- 1:24:19Some people will make some, some I'll, I'll, I'll short time
- 1:24:24decisions. But what you start realizing and
- 1:24:26we had this one technology revolution team come out with
- 1:24:29the Internet, right? Oh my God, right.
- 1:24:31So people let go of a bunch of people and then what they
- 1:24:33realized was, Oh, I did that in earth, right?
- 1:24:37And I realized I need more action, need more people to
- 1:24:40actually service the demand, right?
- 1:24:41Because here's the thing, it's going to create a whole bunch of
- 1:24:44demand. And as far as I know right now,
- 1:24:48it's not going to be able to fulfill demand, right?
- 1:24:52Yeah, I think a lot of things it's creating.
- 1:24:53It's creating OK. It's easier to code now.
- 1:24:55You can use a codex and cloud code to code.
- 1:24:57But The thing is, I think it's showing there was a huge demand
- 1:24:59for code that was always unmet. I think that's true.
- 1:25:02That's right. More people, more, but it's
- 1:25:04democratizing the building process for sure.
- 1:25:06So fill in the blank and this will be our last one.
- 1:25:08And then we'll let this has been, this has been a marathon.
- 1:25:10This is like our Joe Rogan episode.
- 1:25:11We're going to go for 2 1/2 three hours, right?
- 1:25:14Like George, we can talk whatever.
- 1:25:15We should have a Part 2, but so fill in the blank.
- 1:25:18The modern agency is less blank and more blank.
- 1:25:23The modern agency is less opaque and more transparent.
- 1:25:28There we go, we can help George. So where can people get in touch
- 1:25:32with you? So on on LinkedIn, it's always a
- 1:25:36great place to try to keep that the central hub.
- 1:25:39You can reach out at george.museum@outlook.com.
- 1:25:44There's many I mean, if you want to get in touch or you can come
- 1:25:47on LinkedIn if people want to see in person that goes through
- 1:25:50you guys's too. We're actually I'm actually
- 1:25:52going to be at possible representing Madsense fortunate,
- 1:25:58fortunate enough they were one of six that was selected it for
- 1:26:01the innovation stage. That's awesome.
- 1:26:04So you guys will see, hopefully you'll see some of why I think
- 1:26:07they're an interesting well positioned company out of 100
- 1:26:11companies to be 1 of 6. Let's get let's get those guys
- 1:26:15on on where we're going to be broadcasting signals I.
- 1:26:18Will reach. I will reach out to you to to
- 1:26:22get that on the books. I think you'll be impressed with
- 1:26:25with with what they have. Well, hey, well, thanks, George.
- 1:26:27And you know, this has been a terrific call.
- 1:26:30Terrific. No, thank you guys.
- 1:26:32And for everybody that's listening, you know, subscribe,
- 1:26:35share it with your colleagues and friends on YouTube, Apple
- 1:26:37Podcasts, Spotify, et cetera, you can visit visit us at Deb
- 1:26:41Deb Deb dot signal and Noise dot AI as well.
- 1:26:45New content all the time. New content all the time and
- 1:26:47we're going to start pushing out newsletters with all of the
- 1:26:51content in case you've missed some or you're not going to the
- 1:26:53website. So thanks everybody and we'll
- 1:26:55see you. Next.
- 1:26:56Thank you. Thank you guys.
- 1:26:58Take care. Cheers.