Latest / Signal & Noise: Bringing Clarity to Media, Data, Tech and AI / The Godfather of MarTech on the Great Convergence: Scott Brinker on Al, First-Party Data, and the Future of MadTech
Transcript
- 0:03This is signal and noise, bringing clarity to data, tech
- 0:07and AI. Hey everybody, welcome back to
- 0:15signaling noise. I am Brett House.
- 0:18I'm joined by my Co host Rhea Longacre and we've got a very
- 0:21special guest today, Scott Brinker.
- 0:23Welcome to the show. Give you a little intro, Scott,
- 0:26coming out of Boston. I'm a Boston boy originally is
- 0:30widely regarded as the godfather of Martech and I wonder did you
- 0:34give yourself that or did somebody else give you that
- 0:36term? I think you've deserved it.
- 0:38I think you've definitely, you know, I actually.
- 0:40Resisted it for years and now I'm finally like, you know,
- 0:43that's great. Yeah.
- 0:44It's. Like you're like, you're pulling
- 0:46a James Brown, like the godfather of soul right over
- 0:50there. And, and you know, you're one of
- 0:52the most influential thinkers in the marketing technology
- 0:54industry. You're the chief editor of Chief
- 0:55Martech, which I read all the time.
- 0:57I've written about some of your, your recent reports, which we'll
- 1:00talk about. You're the creator of the iconic
- 1:02Martech landscape. Don't mistake it for the
- 1:04Loomiscape. The Martech 5000, right?
- 1:07Yeah, the Loomiscape's a content marketing play.
- 1:09The martech landscape is actually a industry analysis.
- 1:12It's like a true industry analysis.
- 1:13I would tell you the difference between my landscape and Terry
- 1:17Kwaja's is Terry Kwaja makes an awful lot of money off of what
- 1:22he does with his landscape. Yeah, selling that much better
- 1:26operating system for this than I.
- 1:27Yeah, exactly. And you're the author of the
- 1:30best selling book, Hacking Marketing, which I have not
- 1:33read. They need to read and I
- 1:34apologize for I have. To it's out of date 10 years
- 1:37ago, yeah, that's another like that.
- 1:38Things are changing, right? And your martic landscape has
- 1:41become the industry's definitive sort of visualization as we
- 1:43talked about, but with with and we've seen it go from a few.
- 1:46We all use it. We heard articles about it.
- 1:48We include it, right? We quote it.
- 1:49Yeah. A few 100 vendors just like it
- 1:51was, I think it was like 283 or 276, something like that, up to
- 1:5515 / 10 / 10,000. 15,000 now, yeah, it's yeah.
- 2:00Which is remarkable, but you were the VP of, of platform
- 2:03ecosystem of HubSpot, which is headquartered in Boston, you
- 2:07know, and you've helped, you helped build their ecosystem to,
- 2:10to more than 2000 eurations. You are what I actually thought
- 2:13was really interesting. Your background is that you were
- 2:15an MIT and a Harvard grad, but you did what I thought was the
- 2:18reverse of what I would normally assume is that you went for to
- 2:22Harvard for computer science and MIT for your MBA.
- 2:26I always thought it was the reverse of that, but the really
- 2:29interesting background. It's an MBA program, though I've
- 2:31not a few very smart people I know have graduated from there.
- 2:33So, yeah, and my brother was an HBS grad and, and we love to
- 2:36have, and I and I could see why you've had such success and sort
- 2:38of understanding and, and building in the ecosystem is
- 2:41because it brings together sort of the commercial Intel of an
- 2:44MBA with the sort of hard data engineering of a computer
- 2:49science degree, which is often the people we talk to product
- 2:52leaders that also have commercial, you know, CE OS
- 2:54founders that also have commercial acumen.
- 2:57But welcome to the show. We're thrilled to have you.
- 3:00We've read all your reports over the last few years with, you
- 3:03know, welcome. And and yeah, tell the audience
- 3:06a little bit about yourself. If I, if I, if I missed
- 3:09anything. Wow.
- 3:10I mean, like, yeah, that's a pretty thorough intro on that.
- 3:13Yeah. I started life as a software
- 3:15engineer, entrepreneur Once Upon a time, build a web development
- 3:19agency back when those were things that we did.
- 3:23I built a SaaS company after that that was an interactive
- 3:27content platform for marketers. And it was after I sold that
- 3:31company that I joined HubSpot for eight years to build out
- 3:33their tech ecosystem. And so that that arc was always
- 3:38how I made my living professionally.
- 3:40All the work I'd been doing around chief heart attack,
- 3:43frankly, it was just my labor of love and you know, for these.
- 3:46Your side gig, huh? Just fascinated by this
- 3:49intersection and that landscape, you know, it was always a bit of
- 3:51A, and it's like one of these things where like I, I, I, I'm
- 3:57very grateful that so many people like latched onto it and
- 3:59it's, you know, had a certain popularity.
- 4:02But to me that was never really the story about martech.
- 4:06I mean, the story about martech was like, this is changing how
- 4:10marketing organizations run. Who, who, who works in
- 4:14marketing, You know, what skill sets do they have?
- 4:16I mean, there's some of these things.
- 4:17I mean, the technology is sort of obviously a factor to that,
- 4:21but arguably it's the human and organizational dynamics of this
- 4:25what originally attracted me to the space.
- 4:27And still to this day, that is where, like really, the hard
- 4:31work remains. Yeah.
- 4:34Well, Scott, we're thrilled to have you on here.
- 4:36And like the more tech assessment as I mentioned, like
- 4:38we've all used it at some point in our career.
- 4:40Fact, I've been a management consulting off and on for close
- 4:43to two decades now. And I remember when I built a
- 4:45martech practice at a previous place, we'd actually taken that
- 4:48and put it on a whiteboard and dissected it and broke it down
- 4:50to a way where we could go in and do maturity and capability
- 4:52assessments for clients. Those kind of the starting
- 4:54point. So it's, it's really helped a
- 4:56lot of people in terms of just a structured discipline way of
- 4:59approaching things and, and given a methodology that you
- 5:01can, that you can do a lot with. So kudos to you.
- 5:04We've all followed has been very helpful.
- 5:06And but I mean, the reason why we wanted you on here, besides
- 5:08just wanting to have you on our pod generally and we both and
- 5:12knowing we both follow you is I think it's really timely.
- 5:15We're one of the few podcasts that I brought.
- 5:17You know, we kind of pride ourselves on going deep on
- 5:19martech and ad tech and data tech.
- 5:21But you know, I think one of the few that kind of really does
- 5:24touch those two worlds, which, let's face it, have been pretty
- 5:27separate over the last couple decades.
- 5:29Martech has run by marketing operations people, marketing
- 5:32technologist. IT AD tech by and large has been
- 5:35run by, by media agencies, right?
- 5:37It's a very different model. I'm we're going to dig into that
- 5:40a little bit. They've run a separate world and
- 5:42when you look at the data, they, they capture it ad text, more
- 5:45anonymous, ephemeral data points and interactions, right, related
- 5:49to paid media execution. While more tech is more like
- 5:51known customer relationships to support and building those
- 5:54longer like like CRM relationship management over
- 5:56time first party data and identity increasingly are
- 6:00becoming the connective tissue between the two.
- 6:03I'm doing so much work now with connecting CRM and CDP's over to
- 6:07DSP's demand side platforms to through clean rooms, not through
- 6:11clean rooms I and it and it is in different places.
- 6:14It's getting very complex and kind of the connective tissue is
- 6:17that first party data and then enter now into that the whole
- 6:20Legentic AI thing and and agents potentially do some of these
- 6:24things that we've been building pipes and reverse CTLS and using
- 6:27AP is to connect. I mean, is, is that in the
- 6:29future? So it's a really wild time.
- 6:31So we wanted to get you on here, unpack these issues, talk about
- 6:35these two worlds and really see, are they really colliding?
- 6:38Are they combining? Are they like, what does the
- 6:40future look like? I don't know if we're going to
- 6:41solve it all, but we at least wanted to talk about it with
- 6:44you. So welcome.
- 6:45Awesome. Thank you.
- 6:46I love this thesis. OK, that's fun conversation.
- 6:49Yeah, we might be talking for 3 1/2 four hours.
- 6:51We're going to go full Joe Rogan and Joe.
- 6:53Logan got nothing on us. Yeah, we, we, you know, there's
- 6:57a lot of that stuff that's that's interconnected and
- 6:59related, but certainly critical. I mean you've had, you've had
- 7:02this vantage, but it's sort of watching marketing technology
- 7:05evolve over the last few years, which a lot of us have.
- 7:07We were there in the very beginning.
- 7:08So I want, I want to hear you talk a little bit about some of
- 7:11the stuff that you've recently done.
- 7:12But what? When did you realize that
- 7:13marketing and software were sort of on a collision course?
- 7:18I mean, that goes back to the early days of these web agencies
- 7:21where, you know, the one we were running, we were always being
- 7:24hired by the marketing department because there's
- 7:27always their vision of like, OK, how do we start to get customers
- 7:30this way? And here's what we're doing with
- 7:31the brand. And you know, and then the
- 7:33marketing team would hire us, But since I was on the tech side
- 7:37of it would be my job to go and talk to their IT department
- 7:41because they couldn't talk to their IT department and not
- 7:44actually in my opinion, because of like some sort of hostility,
- 7:47like they just didn't like speak the same language.
- 7:50They like who serves as they were just in.
- 7:53Worlds you know yet if you set aside that you know I always
- 7:57think of the high school guidance counsellor you know
- 8:00having like OK on this end of this background you could go
- 8:02into IT and software and on this end of the spectrum you could go
- 8:05into marketing and you. Know they're communications of
- 8:09liberal. Arts, right, It could become a
- 8:11peace activist or you could become a general in the army.
- 8:14I mean, you make some choices, you know, but it's like, even
- 8:18though there was that, that that distance, you know, like
- 8:20culturally and just from a mental model perspective, when
- 8:24you looked at like where the world was actually heading and
- 8:27like what customers wanted, I mean, like behavior is so clear.
- 8:29Like, OK, this is this is going to be how business runs.
- 8:33Is it just going to be marketing sales go to mine.
- 8:36It's just going to be totally entangled in these digital
- 8:38channels. And anytime you have anything
- 8:40that's digital, you have software.
- 8:42That's that's what makes it all run.
- 8:44Yeah, it's funny. I wrote a blog.
- 8:45I used to run a blog for I think one of the one of the marketing
- 8:49magazine like 20/10/2011 at least Target marketing.
- 8:52And then I remember one of the one post I wrote was of how this
- 8:55new CMO was going to be more like making decisions based on
- 8:59data was going to be really deep in technology.
- 9:01I've never gotten so much hate mail.
- 9:03I mean, it was like, you know, you don't know any marketing.
- 9:05You're complete idiot. How many years ago was this?
- 9:07Yeah. 20 I think 201020. You're failing to separate art
- 9:10versus science. You don't understand that, you
- 9:12know, you don't understand the art of it.
- 9:14It's about persuasion. It's about creative, like not to
- 9:18say it's not, but like I've never seen so many people get so
- 9:21angry and irritated. It was, it was it was using a
- 9:23block. I would get like one or two
- 9:24comments. I think I got like 30 really
- 9:25nasty ones. But I think it said something
- 9:27about about kind of the transition back then.
- 9:29You were really early doing this.
- 9:31So back then I wonder like did you have any idea that you would
- 9:34be over 10,000 or not over 15,000 Martech tools?
- 9:38Oh, God, no, no. When we did that first
- 9:39landscape, and the whole reason I did that original 1 was
- 9:43because that exact phenomenon you were talking about, you had
- 9:46all these like senior marketing executives who didn't want
- 9:49anything to do with technology. And I was originally just
- 9:52assembling the chart to be like, hey, listen, you're relying on
- 9:56this technology for this thing, this technology for this thing,
- 9:59this. You might have no interest in
- 10:01technology whatsoever, but your actual ability to execute here
- 10:04is now becoming dependent on all these things.
- 10:07You might you might just want to consider the possibility of
- 10:10getting some tech. You don't have to become
- 10:12technical, but you might want to get some technical talent inside
- 10:15your organization. And that resist that resistance
- 10:18from the marketing teams to your point Rio is is kind of is
- 10:22problematic, right and and we keep.
- 10:24Talking to you to. This day that CM OS have to be
- 10:27more technical. They need to understand exactly
- 10:29how these things are architected, built, and how they
- 10:31interact with other platforms. And even if they don't, they
- 10:35better hire people to do. I mean, I think that's the
- 10:36thing, right? Yeah, it's a critical, it's a
- 10:39critical place. So I think I think would be
- 10:40helpful for the audience. And I know you're going to love
- 10:42this. Is that is that I wrote an
- 10:44article that you commented on Scott around with data bricks.
- 10:47I believe it was right called the new Martech stack for the
- 10:50AIH. Yeah.
- 10:53And I and I, I was, I like read that report.
- 10:55One of my, my partners of high signals, Mark Sabatini sent it
- 10:59over to me and it, it just published and I read it and I
- 11:02thought, God, I got to write about this.
- 11:04So that that shows that you kind of inspired me.
- 11:06I'm like, thank you. You know, it might have been
- 11:07imitative, but I was like, this is worth it, worth the time to,
- 11:11to analyze and understand its implications, right.
- 11:14And you've outlined basically three ages of Martech.
- 11:18And so the reason why I wanted to kind of get you to chat about
- 11:20what those ages are, the first age of Martech, which was the
- 11:2390s to 2000 tens, I think of like the sales forces of the
- 11:27world. Tell me if I'm interpreting this
- 11:29right, the second age of martech, which like mid 2000
- 11:32tens to 2000 twenties, I think you start to get the marketing
- 11:34cloud world, right? You get to a bunch of platforms
- 11:37that are that are becoming platforms, sort of ecosystems,
- 11:40right? And then this third age of
- 11:42Martech, which I think is super interesting.
- 11:44And we've seen this term a lot in the ecosystem composable
- 11:48canvas. You kind of divide it slightly
- 11:50different from kind of a classic composable architecture.
- 11:52But can you get a walk the audience through that?
- 11:55Because I think that'll give us a really good grounding sort of
- 11:58martech and how you think about it.
- 11:59Yeah, I mean, again, like the the original days of martech, we
- 12:04were so much of this is tied to not just martech, it's tied to
- 12:08basically the the evolution of software as an industry, as a
- 12:12business, as a discipline. And we came from a world of
- 12:16traditional client server apps where, OK, there's going to be 1
- 12:21there, there'll be two or three companies, they'll battle it
- 12:23out. You're going to have them
- 12:24finally or or whatnot is going to win and this is going to be
- 12:27the thing you want. And so that was the mental model
- 12:30that so many of these companies had when they were coming into
- 12:33Mardek as a space was like, OK, well, we will either build or
- 12:36acquire and just assemble everything and we will be the
- 12:38solution. And they were very resistant to
- 12:42like. And this is precess this before
- 12:45there was really a cloud of any sort, right this.
- 12:47Right. But then even when they moved to
- 12:48the cloud, right, it was still this thing as they brought that
- 12:51mental model with them of like no one vendor to rule them all.
- 12:56And so for years, when I started doing that early martech
- 12:58landscape, you know, you want to talk about like hate, boy,
- 13:04people really had a lot of hate associated with that thing.
- 13:08Ironically, we'll come back to this, The second thing I've
- 13:10gotten the most hate for in my life and I'm on.
- 13:13Thank you for inviting me for this particular podcast is
- 13:17anytime I'd said, you know, I really think Martech and Antech
- 13:20should be, you know, viewed as like 1 cohesive hall.
- 13:23Boy, does that unleash hate online.
- 13:25It's very tribal. It's very tribal, yeah.
- 13:29Well, you know, it totally does. I mean, if you if you find
- 13:31people who are like really consider themselves deep in
- 13:33Antech, they will scoff at martech.
- 13:35Like they'll say like we're not interested, totally different.
- 13:38Like and with Martech and the most martech people, they don't
- 13:41understand the model behind ad tech about what the role of the
- 13:45agencies, what, you know, working media versus tech tax
- 13:48versus, you know, like any of this like so I think so it's
- 13:51very different world, different languages.
- 13:53I agree. The technologies have become
- 13:55like, especially as we've moved to the cloud, there's been a lot
- 13:57more and we'll talk about this, there's been a lot more
- 13:58convergence. But, but I'll come.
- 14:01Back to that he. Hasn't he hasn't gone through
- 14:03his evolution for for the I think we're moving to the second
- 14:06age of Martech. Yes, You know, I think the
- 14:09second age was basically when those companies finally, when
- 14:11they came to the realization they couldn't build it all.
- 14:14There just wasn't enough things to build and I couldn't buy it
- 14:17all, you know. And so somewhat reluctantly they
- 14:21started to say like, OK, we'll open up, we'll open ecosystem in
- 14:25some cases. I mean, for Salesforce, they
- 14:26build actually a pretty profitable business around, you
- 14:30know, an ecosystem on that too. So it worked out.
- 14:32Yeah. Remember Salesforce used to have
- 14:33the no software. That was their whole whole
- 14:35thing, right? Yeah, the new software, the
- 14:37begin T and then that's where it's sort of API started to come
- 14:39into play and and direct integration for capabilities
- 14:42that you didn't have Hubspots, another company that built a
- 14:45huge ecosystem of partners for capabilities outside of their
- 14:48kind of core competency. So instead of it being like this
- 14:50choice of you're either sweet or best of breed, it was this
- 14:54recognition like, well, you can kind of have a little bit of the
- 14:56best of both, you know, like have a major platform at the
- 14:59center of your stack. Some cases people have multiple
- 15:02big platform that but you know the the nice diagram would be
- 15:06yeah have you know a major system that's the center of
- 15:09gravity, the cohesion, you know in the sack, but then augment
- 15:13that with any other specialist apps that you need and so you.
- 15:17And the high watermark scout for this would be like 2018-2019, I
- 15:20think 2020, like like the peak of that kind of like monolithic.
- 15:24Like, let's say we're going to go, you know, going on sweets,
- 15:26but get a suite maybe like. Whether you're doing even a
- 15:29little bit earlier than that, I want to say in Martech like
- 15:332014, 2015, like I remember very clearly Marketo was one of the
- 15:41first ones to like really lean. I think the original name might
- 15:44have been might have been launched point and whatnot.
- 15:47They've gone through changes over the years, but like
- 15:49originally, you know, John Miller, you know, Phil Fernand,
- 15:52like they're like, Nope, we're now opening up this ecosystem.
- 15:55Here's his marketplace. And then immediately after that,
- 15:58Elliqua, you know, was like, OK, yeah, we've got a marketplace so
- 16:01we'll do that. You know they.
- 16:03Both get acquired. Yeah.
- 16:06And hopefully that's part of it, although both of those ended.
- 16:09Up do you think the marketing animation platforms that you
- 16:11just Eloqua, HubSpot, Marketo, do you think those are places
- 16:14where that really started to show itself earlier maybe than
- 16:17some of these other places or that?
- 16:19Mean certainly have been marketing the.
- 16:20Current evolution. Yeah, certainly with Martech
- 16:23they became the prime examples of, OK, this is the platform
- 16:28ecosystem play. You are still selling a
- 16:32something, you're still selling the sweet you had before, but
- 16:35it's out open and you now actually have a management
- 16:38control over other integrations or partners in that ecosystem.
- 16:43So that you could. Again, it didn't have to be
- 16:45entirely sweet or best of breed, it could be the best of both.
- 16:49And it's kind of like an App Store or a partner place where
- 16:51you could go like you could that had built in connectors to
- 16:55different partners that added complementary services, that
- 16:58being that the premise, right? And then people started to
- 17:00compete on that. They're like, oh, yeah, do you
- 17:03have more connectors than I do, you know?
- 17:05But anyway, yeah. And now, now the third What
- 17:09about the Third Age of Martech? And all right, so here's the
- 17:11canvas we're talking about. What we've mostly been talking
- 17:14about is that sort of convergence around, you know,
- 17:18major suites versus, you know, like these these app ecosystem.
- 17:21Yeah. The thing is, there were a lot
- 17:23of other things that were treated as dichotomies before.
- 17:26Like there was this dichotomy of you're either a services company
- 17:28or a software company. If you're a software company,
- 17:31you're not allowed to do services because we're going to
- 17:33penalize you for. You're not beyond a certain
- 17:35amount, not beyond me, like 10% or something.
- 17:36Yeah. And if you're a services
- 17:38company, like we're not sure we trust you building software.
- 17:41There's also like the dichotomy of build versus buy.
- 17:44It was pretty binary. I'm either going to build this
- 17:47or I'm going to buy it, you know, and there were.
- 17:49Organizations that would say like we like philosophically, we
- 17:51like to build and or philosophically with and they're
- 17:55I would run into that a lot. That's funny.
- 17:56Yeah, Yeah. You know, in a lot of cases they
- 17:58were building things that they didn't need to build, you know,
- 18:01but so the second age was like, OK, those dichotomies started to
- 18:06blend. So it was, you know, sweet
- 18:08versus best breed platform ecosystem.
- 18:10On the services side, you started to see more and more
- 18:12blending of software companies offering more and more services,
- 18:15services companies actually starting to bundle up apps often
- 18:19in these ecosystems, you know, and turning that into a real
- 18:22business lever. And then even the build versus
- 18:24buy because because of the AP is that got opened up as part of
- 18:27the platform ecosystem play. You started to see more and more
- 18:30companies be like, OK, well, we're going to buy the
- 18:33foundational elements. Like I don't want to write my
- 18:35own e-mail server, you know, but on top of Walmart, you can start
- 18:40to yeah, customize and build, you know, specialist thing.
- 18:44So that was the second age to me.
- 18:45What makes the third age like this pinnacle of convergence is
- 18:50even the blend between like platforms and apps and services
- 18:59companies and what am I building and what are you building?
- 19:03It's kind of all converging, you know, And then this is, you
- 19:06know, there's several different reasons for that.
- 19:09But the most obvious example you look at is what's happening
- 19:11right now with AI where people are using like, you know, Clod
- 19:15and ChatGPT and these MCP servers and they're connecting
- 19:19across all these different systems, commercial systems,
- 19:22they own custom things they built themselves.
- 19:25And it's almost like at that level, you know, inside POD, it
- 19:29doesn't really care what was a commercial product, a custom
- 19:32built product, a service or whatnot.
- 19:34And so is this latest age got a continuation of the composable
- 19:37age to just put an AI like veneer on top of it or is it or
- 19:41is this, do you consider this a distinct age?
- 19:45I think it is composability like the the, the art towards greater
- 19:51and greater composability where it's not just composability from
- 19:54a technical perspective. It's like am I buying a software
- 19:59or am I buying a service or am I building something myself?
- 20:02Even at a strategic level, the interchangeability and
- 20:06composability of these things is becoming much more fluid than
- 20:11like I mean. It ever was before.
- 20:14And and I think we're very early in this.
- 20:16We're in the middle of this. So it's like it's it's very
- 20:18chaotic and confusing. Yeah, part of mine's chaotic and
- 20:21confusing is because you can intermix these things.
- 20:23Yeah. And I think it was your quote,
- 20:25you said Martech no longer sits on top of the data.
- 20:28It is the data, which I mean if you think about the profundity
- 20:32of that statement, simple, but it's like it where the it's not
- 20:35data is that's what. AI is doing to it right?
- 20:39I mean. If yeah, like can we throw that
- 20:42comment if data is the new oil just into the garbage can?
- 20:44I, I can't, you know anybody that says that again, data is
- 20:48actually the the operating layer and the way that you can you
- 20:50break down how you actually how you actually talk about data.
- 20:53It's not literally just data, it's content.
- 20:56It's the semantic capabilities of controlling and in the sort
- 21:00of the the dictionary to how to manage and control the the the
- 21:04data underneath. Can you talk about how you think
- 21:06about that from an operator perspective?
- 21:08Well, in some ways this is a theme that again ties back to,
- 21:12you know, our main topic around like, you know, the silo ization
- 21:15between antech and Martech, you know, is we have treated all of
- 21:19these different kinds of data in very hard silos.
- 21:23All right, here's where I'm keeping my customer data.
- 21:25In fact, in multiple silos of different customer data.
- 21:28Here's my campaign data that's going to be in some separate
- 21:30thing. My finance data is in something
- 21:32else. My content, OK, I'm going to
- 21:34have that in a dam. That's a separate system.
- 21:36I'm building some code or software, OK, I'm going to
- 21:39manage that in some other thing. And the thing that I mean, the
- 21:43reason why that came up in the report that I did that was in
- 21:47partnership with Databrex was because the sort of thought
- 21:50exercise there was well, like, OK, well wait a second.
- 21:53You know, if all of these different kinds of data at some
- 21:57level can like be all deployed is wrong.
- 22:03More important, if they can basically all live inside like a
- 22:06common data substrate, then like this distinction between what
- 22:10was customer data versus company data versus content versus code.
- 22:15You might. Not want like certain
- 22:17governments distinctions between them, but the the ability for us
- 22:20to treat that as a much more fluid relationship.
- 22:25That's actually a really mind blowing concept.
- 22:28Yeah. And then?
- 22:31But I was saying but that but that distinction like let's say
- 22:33breaking down those walls as it becomes just a substrate you can
- 22:36access with they can a gentic layer using AI tools that hop
- 22:39across it, you can really get rid of the front end.
- 22:41I mean, I think philosophically you even think, OK, does
- 22:44remember Brett, our first episode of Signal Noise we
- 22:47interviewed been wild. Well, then, well, the second one
- 22:50been wild. We talked about the feature.
- 22:52This is even before the SAS apocalypse.
- 22:54This is early. We were even talking about
- 22:55what's going to be impacting the SAS business model.
- 22:57So, but you think philosophically like it
- 23:00challenges it, right. But then financially when you
- 23:03when, when the SAS business model is building for seats, for
- 23:05licenses, for number of people using things, I mean it it, it,
- 23:09it does call, but it does call it into question or at least put
- 23:12pressure on it. Yeah, 100%.
- 23:14Yeah. Whereas this kind of new model
- 23:16it's it's you're basically building sort of a compute
- 23:19layer. You're paying for compute and
- 23:21you're campaigning for, you're paying for decisioning and all
- 23:24of that. If you've got agents that sort
- 23:26of sit, sit on top of this content and this data which the
- 23:29data you're you're defining is very different than I think most
- 23:32people define data content most people wouldn't put into the
- 23:36data category. People think ones and zeros,
- 23:39decimal points, etcetera. With a big data and IDs you know
- 23:43ID graphs things. There is content as metadata,
- 23:46content as a taxonomy. You know, there's a lot of data
- 23:48that goes to the contents. It's it's all the raw fodder,
- 23:50right that that powers this learning system to power some
- 23:54sort of decisioning layer? Is that the right way to think
- 23:57about this? Well, and also like, I mean
- 23:58content used to be expensive, you know, I mean it just keeps
- 24:03getting cheaper by the minute. A whole whole debate we could
- 24:07have about quality of content. But again, at the end of the
- 24:10day, just this ability to have a fluidity with content means we
- 24:15actually can start to think about how we manage and
- 24:18operationalize that in a way that looks a little bit more
- 24:23like, you know, a software than it did just some sort of like
- 24:27static asset. Yeah, it looks a little bit more
- 24:30like data, right? And is.
- 24:32This the other kind of data exactly.
- 24:34Yeah, the other kind of data. So is this a conversation
- 24:36because like all the AI companies that I'm talking to
- 24:39and that we're consulting with, you know, talk about sort of
- 24:42context, you talk about context rafts and and you know, and how
- 24:47you build these sort of context engines that are kind of these
- 24:52these learning systems. Can you talk a little bit about
- 24:54how you think about that? Yeah, I I have nominated Context
- 25:00as the word of the year here in Martech for 2026.
- 25:04That officially. We, we are all going to be sick
- 25:07of it by the end of this year, partly because the, the word
- 25:10keeps getting overloaded. Like we use context to mean a
- 25:15lot of different things. You know, there's the context
- 25:18window of the, you know, the particular AI is like just kind
- 25:22of almost, it's like working memory.
- 25:24Yeah. Talking about context
- 25:25engineering of like, oh, well then how do I optimize what I'm
- 25:28putting into that working memory?
- 25:30The context graph concept, which is something the folks at
- 25:33Foundation Capital, you know, sort of pulled out of thin air
- 25:36in December of being like, oh, you know, what's kind of missing
- 25:39and all this stuff is nobody actually has good records of
- 25:42like how decisions get made and, you know, like how the, the
- 25:46actual operating institutional knowledge, you know, of a
- 25:49business or if you could have that.
- 25:52And since they wanted to come up with a name that wouldn't 'cause
- 25:54any confusion whatsoever, they're like, oh, well, let's
- 25:56call that the context graph, you know?
- 25:58So now people talk about context and it means so many different
- 26:02things. But if you try and synthesize it
- 26:05to what the English language, simple, you know, colloquial
- 26:08definition of, it's like, yeah, just understanding like the
- 26:11thing we are doing right now, what is that in the context of
- 26:15that is a really powerful opportunity for us overall.
- 26:20I I think the thing that I just flag is every time I scratch at
- 26:28that the depth of the rabbit hole like the oh.
- 26:31My God. Work that is going to be
- 26:34required to actually turn that fluffy vision into reality.
- 26:39Into something, yeah. And and there are companies,
- 26:40there are companies that are doing that on the Rev op side.
- 26:43So we talked to a company all the time and and are partnering
- 26:47with them at high signal. It's called full cast that's
- 26:48doing that where they're looking at at multiple like in the
- 26:51amalgamation of e-mail, Slack, CRM and using that for basically
- 26:57signal predictive analytics. So they're looking for signals
- 27:00and, and signs and, and to help predict Chung to help predict
- 27:06likelihood of, of upsell cross sell, right.
- 27:09So there's all these insights that can be derived from stuff
- 27:11that people are are storing and communicating across multiple
- 27:16platforms within their organizations and, and it or, or
- 27:19within their own head. And none of it can be decided
- 27:22upon because it's so fragmented and siloed.
- 27:25And so there's a lot of. People analyzing it would be so.
- 27:27Different, yeah. So they're trying.
- 27:28To play. With AI right?
- 27:30But how accurate is that those outputs is the question right?
- 27:33No, yeah. I mean, again, this is super,
- 27:37super early on it which is this goes back though to actually why
- 27:40I'm relatively bullish about these nature martech platforms.
- 27:46Not to say that I don't think they have their challenges, but
- 27:50I think the work involved in managing context at scale is
- 27:55massive. And I think it's different than
- 27:57being, I mean, all these systems have had contacts, but that
- 28:00hasn't really been what they've been selling.
- 28:01They've been selling AUI and a workflow and all that.
- 28:04And that's stuff that's getting disrupted to hell, you know, but
- 28:08the context piece of it, that's actually becoming more and more
- 28:11valuable. And so I think the degree to
- 28:13which you see whether it's the incumbents or new players who
- 28:16come into this space are like, OK, we can actually provide
- 28:20context as an infrastructure capability in domains like
- 28:23marketing or customer service or watching.
- 28:25Scott, it's interesting what you just said though, because like
- 28:28one thing we wanted to talk about was convergence of Martech
- 28:30and AD tech, but I think this is like, this is like this is an
- 28:33interesting point. So Brett, in order to do some of
- 28:35the things you're talking about, like taking, let's say financial
- 28:38data, which may come from your accounting system, which, you
- 28:41know, maybe from your ERP. That's that's definitely not
- 28:44martech. That's more like a different
- 28:45types of tech that are used by different parts of the
- 28:47organization on a finance tech or sales tech, whatever, right?
- 28:50I mean, like or accounting software, right?
- 28:52But like I'm seeing some interesting stuff go on kind of
- 28:56like we just described, Brett, we're like AI is enabling us to
- 28:58maybe begin to look across the organization into different data
- 29:01silos and use that for insights, combining it with it's got like
- 29:06different data and signals. And you know, I think we tend to
- 29:09call stuff and media more signals than than data these
- 29:12days. But looking at that across the
- 29:14enterprise, it's more enterprise play more than it's then it's a
- 29:17martech play. Even that makes sense.
- 29:19Oh yeah. I mean, 100%, yeah.
- 29:22I mean, this is, this is, I think this is actually why the
- 29:29theme of this whole conversation, like all ties
- 29:31together. We have come from a world where
- 29:33things are remarkably siloed and fragmented, and we are finally
- 29:38now starting to see pathways where those silos are being
- 29:42breached. And boy, it's starting to open
- 29:45up some really interesting ideas.
- 29:47Look like, well, wait a second, we can now actually start to
- 29:50connect these dots. You know, the ad tech martech
- 29:52thing in some ways is like a subset example, you know, of
- 29:55things that for a variety of reasons had their silos.
- 29:59There's a lot we can do when we can rise at a level above.
- 30:02That in fact, I think some of the other enterprise stuff is
- 30:04probably easier to do because at least everything's in the same
- 30:06enterprise, right? It might be they, they have the,
- 30:09they have the SAS accounts they'll have, I mean, to a
- 30:11certain extent, right. But I think with ad tech is, and
- 30:13we'll get into this and start, you start to see the data
- 30:14sitting other places entirely right.
- 30:17But but yeah, you're right. I think it does unlock some the
- 30:19ability to do things. And you know what, like so much
- 30:21you're combining all these data sets together that never work
- 30:23combined together. Like like Brett, your your
- 30:25point, Maybe you could unlock some incredible insights that
- 30:28would allow you to identify propensity to turn or identify
- 30:31higher value customers or identify maybe white spaces for
- 30:35organization, combining the stuff together that you never
- 30:37would have seen before because there's just so much data and
- 30:40just the ability to process, to gather and process it would have
- 30:42taken forever. Yeah.
- 30:43And then I've, I've already seen a ton of tools like this that
- 30:46are, that are kind of BI tools in addition to the one that I
- 30:48talked about that look across a stack and pull a bunch of
- 30:51insights and give you actionable reports in, in, you know, could
- 30:55be in preparation for QBR and, and to be able to do that
- 31:00agnostically. I think that the whole notion
- 31:02here is that you're moving away from sort of a world that was
- 31:05relying on integrations and there was sort of an integration
- 31:08tax that was not just money and time, it was money and it was
- 31:12also time, huge amounts of time we've spent.
- 31:14And then think about all the CRM integrations.
- 31:16I mean, I've been through multiple and.
- 31:18The integration spread, like even connecting the systems, you
- 31:21have to then normalize the data and then you have to make sense
- 31:24out of it. It's not easy.
- 31:25Yeah. So, so I mean, do you think this
- 31:27is freeing up people to innovate, you know, when you
- 31:31remove some of that time that's normally spent on, on data
- 31:34integration, on tech integration on etcetera, etcetera?
- 31:39I think it's opening up the integration space because for a
- 31:42lot of people, yeah, it's, you know, the just the cost, you
- 31:47know, you know, time, expertise, availability, opportunity costs
- 31:51and, you know, doing one things like people weren't just only
- 31:54willing to do so much. And so as a reality, most
- 31:56systems were not integrated. Even systems that were
- 31:59integrated, most of them are pretty shallow integrations.
- 32:04There's usually a pretty narrow subset, you know, shared across
- 32:08them. And so I had to.
- 32:11You couldn't just dump all the data in there, You wouldn't
- 32:13really do anything with it. Right.
- 32:14Well, again, like to to to actually then make it work, you
- 32:17had to do a lot of work of like, let's understand the con context
- 32:20of that, you know, I mean the semantics of it, I think.
- 32:25Yeah, AI is particularly, I mean, not just AI, it's like
- 32:29this perfect storm of MCP is a standard that's essentially
- 32:35like, OK, we can actually sort of simplify the connection
- 32:38portion of this. The fact that the demand for a
- 32:41for people do stuff in AI is really forcing more vendors to
- 32:44make more things, more API access available and then to
- 32:49have AI be something that's actually pretty good at them
- 32:52being able to synthesize things when you connect them across.
- 32:56Yeah, I think it's like a for, for an industry that is spent as
- 33:02long as I've been around 20 plus years rest.
- 33:05Oh, you stop. You stop counting at 20.
- 33:08At that point in time, nobody. I stopped at 25.
- 33:13Oh, you're 25? Yeah, I'm 29.
- 33:15Wasn't that obvious, you know, But I think martech in its
- 33:20heyday is probably like a 20 year.
- 33:23Yeah. That's true.
- 33:24So you've kind of painted this picture of and you've tracked
- 33:28this, which must have been a kind of a mind bending
- 33:30experience of going from 200 and whatever the number was just
- 33:34just out of 300 to 15,000. And actually you probably
- 33:38leveraged a little AI to do some of the counting because it's
- 33:40sure that's pretty take some rigor.
- 33:42Right now we have an AI system that like, yeah, it is like the
- 33:45first pass of like evaluating like, OK, are folks still in
- 33:48business or they still operating like, yeah, things that get
- 33:50submitted. So you can validate the count
- 33:52rate. Yeah.
- 33:53I mean, we still have like a human layer, like the human in
- 33:56the loop to oversee that and double verify.
- 33:58But yeah, the first pass of all of that.
- 34:01Well, have firms been acquired? Have they, you know, gone out of
- 34:03business? Yeah, there's a lot of things
- 34:04you can sort of pick up programmatically.
- 34:06So you describe this shift, you know, so with that in mind of
- 34:09tracking that and kind of getting to this large number
- 34:11that we're seeing and now it's starting to decrease a bit.
- 34:14You know, you talked about this commercial long tail as of a
- 34:19custom build Martec, right, which is what AI, you know,
- 34:22anybody with a with a cloud account and in a replit account
- 34:26or a lovable account can build some sort of prototype prototype
- 34:29that eventually needs some engineering document behind it.
- 34:32But you call this the sort of hyper tail, which I thought was
- 34:35a really interesting idea of of sort of, you know, it's kind of
- 34:38similar to your your first, first, second and third
- 34:42generation of martech, sort of this notion of like a head,
- 34:45which was your major platforms, a torso, which was sort of
- 34:48horizontal software, which probably integrates into some of
- 34:51those things a tail, which. Can you walk people through that
- 34:55that sort of thinking? Because I thought that was
- 34:57really interesting and. Then by the way, that's a really
- 34:59interesting connection. I had never made that
- 35:01association between, yeah, sort of the dynamic of those three
- 35:04ages and how that. Yeah, I think conceptually,
- 35:09yeah. All right.
- 35:12Whether whether it's inaccurate, but the hyper tail is this
- 35:14notion that that, that you're going to have a lot of custom
- 35:16built democratize build in terms of both startups and people that
- 35:21are able to do this internally to build core capabilities.
- 35:25Yeah. And then you get this massive
- 35:26multiplication of agent build software of, of people
- 35:30developed, citizen developed software, as you call it.
- 35:34Walk us through how you guys are thinking, how you're thinking
- 35:36about that because I found that super fascinating.
- 35:38What we're going to see, that 15,000 number is going to grow
- 35:41dramatically. So this becomes the the very
- 35:45interesting questions of where you want to draw the line on
- 35:47that, which just there's one example of that.
- 35:49For instance, there's a case I could make that there's maybe
- 35:571000 martech products out there just by saying like, you know,
- 36:01I'm gonna go to the WordPress. Through the App Store, right?
- 36:04And I'm gonna say every single like plugin that exists for
- 36:07WordPress. Technically speaking, it could
- 36:09be a martech, so you have to you choose your judgement line on
- 36:14that anyway. So then not go down that rabbit
- 36:17hole too far. Yeah.
- 36:18I mean, the long tail was always the commercial apps.
- 36:21And to me, there's a line that gets drawn and says like, OK,
- 36:24when a company starts building their own things.
- 36:27And they've always done that to a certain extent.
- 36:30But for a while it was like, OK, professional engineers, ITA
- 36:34small number of stuff. For like the past 10 years,
- 36:37we've seen a general rise in what was called no code.
- 36:41So when you start doing things with like Air Table or Zapier or
- 36:45things like that, you know, we could debate.
- 36:47Is that making software or not making software?
- 36:50I think if you just let go of any sort of philosophical debate
- 36:53there, the reality is, yeah, there are people, like using
- 36:55this stuff to like, OK, this is kind of building my app or my
- 36:58workflow or my automation the way I want it.
- 37:00Yeah. AI has now obviously like
- 37:03accelerated that phenomenally for professional developers, for
- 37:08amateur developers, vibe coders. But the thing that's actually
- 37:12really wild is these AI's are like building and executing
- 37:17software at a phenomenal rate that nobody even realizes.
- 37:21And there's an example I give where if I go to Claude and I'm
- 37:25like, hey, can you get me a graph of I want to see top 20
- 37:29most valuable companies in the world over the past 20 years,
- 37:32divide them up by tech versus non tech.
- 37:34And, you know, put this in an area graph for me, you know, and
- 37:38turns away, it comes back, I get my graph.
- 37:40I'm like, wow, wasn't that nice, you know, behind the scenes
- 37:43because you click the icon and see what did it wrote a friggin
- 37:46Python program. You know, I went off and got the
- 37:48data, crunched this and then executed the program and then
- 37:51threw the program away. Now maybe it's a philosophical
- 37:55debate. Was that actually a piece of
- 37:56custom software or not actually proposed to you?
- 38:00It is. It is a little piece.
- 38:02Of it might be used case narrow. These cases specifically.
- 38:05It might only be used for a short time, right?
- 38:07But it's. Still might be used for 10
- 38:08seconds exactly. But I think it is this reality
- 38:13that the reason I do think of it software is because it is
- 38:17actually running these programs as we increasingly figure out
- 38:20how we're going to manage governance and what tools it can
- 38:22use, what data can access one of the store stuff.
- 38:25It is very much a part of the overall tech stack operations
- 38:29for the company. And I don't think we've even
- 38:31begun to get our arms around. I mean, there's a lot of this is
- 38:36happening, but. It's almost like we need it
- 38:38well. Understood.
- 38:39But it's almost like we need a new definition of software.
- 38:41Software becomes something almost ephemeral that could be
- 38:44created quickly, dissolved rapidly after a couple uses,
- 38:48reconstituted in different ways. I mean, that's, that's
- 38:52definitely not like what when we think of software today still in
- 38:54the enterprise software, that's not how we think of it.
- 38:57Yeah, yeah. Although it's ironic, it's
- 38:58enterprises where most of this like hyper tailed custom stuff
- 39:02is like flourishing. Yeah, it's it's sort of
- 39:04composability, next generation composability.
- 39:07I mean it's like you're, you're going down to kind of this micro
- 39:10use case level and developing something.
- 39:11To your point, I've seen Python more in my life as just a code
- 39:15base that I have in the last, I don't know, 10-15 years just
- 39:19because I was not a a developer, right.
- 39:21But you're seeing and it and so, yeah, you're seeing that that
- 39:24that's a it's, it's like there's a multiplication of that within
- 39:26organizations, which it does. There need to be a governance
- 39:30around this. I mean you.
- 39:31Know this is what I mean by like it.
- 39:33Right now it's actually pretty chaotic and it's not very well.
- 39:39It's not not just about governance, it's about even
- 39:41understanding what's going on and how do we choose to like
- 39:45orchestrate this? Is there a strategy of like how
- 39:47we get these things to like align and work together?
- 39:51One of the big open questions right now, in my opinion, is
- 39:55people use different language for it.
- 39:57We use the term most popular called orchestration.
- 40:00It's like I've now got all these moving parts inside my
- 40:04organization and they're multiplying at a ridiculous
- 40:06rate. Now every other AI company that
- 40:08comes to market is talking about orchestration.
- 40:10They've they're beating the hell out of everyone.
- 40:11Talking about it, but when you actually like I have, I have yet
- 40:15to come into an an actual organization where they're like,
- 40:18yes, and this is what we have that like orchestrates all of
- 40:21this shit for us. Well, they have one orchestrator
- 40:23agent right to rule the other agents.
- 40:25That's. That's the aspiration, of
- 40:28course. Yes, You know, the orchestration
- 40:30right now is done through a bunch of like humans who are
- 40:33like trying to hold it together with bailing wire.
- 40:35And the trade to build agents to do to help with that, but.
- 40:38I've heard the subscription of a CMO as like the ORC, like it's
- 40:41becoming more of an orchestrator in this new environment.
- 40:43Anything, Bret, I start to cut you off.
- 40:45Yeah, the so, so you had a really interesting quote, I
- 40:48believe it was in your in that same paper from from Megan
- 40:50Eisenberg, the CMO of Samsara or Samsara Samsara.
- 40:55And she said, and this was an interesting point about kind of
- 40:58SAS evolution, right, and defensibility from a product, a
- 41:02tech product perspective. She said if a SAS product is
- 41:04just making data easy to access, when I've been with SAS
- 41:08companies that do that, companies are increasingly going
- 41:10to do that themselves. So vendors need something more
- 41:12defensible at the experience layer.
- 41:16Can you talk about like what that means to you?
- 41:18Because that experience layer is, is critical as opposed to
- 41:20just a wrapper around that's, that's sort of protecting A
- 41:25valuable asset and you're just getting access to that asset
- 41:28like data. What do you, what do you think
- 41:30about that quote? How do you interpret that?
- 41:31Yeah, I mean, I would use a slightly different term than
- 41:36experience, because I think experience can be 1 subset.
- 41:38Like, there's a lot of debate right now about the death of UI.
- 41:42I actually don't think it's the death of UII think there is a
- 41:45new UI paradigm that is subsuming a bunch of things.
- 41:48But I do not think the text chat box UI is the universal cheat
- 41:53code to everything we want to interact with in the world.
- 41:56I think this is the final Test. I think this Star Trek, Hey
- 41:59computer, I think this is where we're going.
- 42:01Yeah, well, even they have visualizations I.
- 42:03Don't think you're just going to die, I.
- 42:05Understand I'm thinking about stuff computer.
- 42:09Yeah. So you're saying your paradigm
- 42:10shift in UI, which means we have to redefine what UI means in
- 42:15conversation. We might be part of that, but it
- 42:17but it's brought much broader than that.
- 42:18It's not just what we do with LLMS. Yeah, OK.
- 42:21But anyway, so there's the experience piece of that.
- 42:24That's one subset. I would argue what we were
- 42:26talking about earlier, this idea of like a context management
- 42:29layer is actually where there's far more value and opportunity
- 42:34because again, the like the raw data and you were making this
- 42:36point earlier, Rio, I mean, why integrations were never easy.
- 42:39It's just throwing the raw data around and actually now that
- 42:42we've got even more data than ever, it's more of a problem
- 42:46when it's just the raw data on its own.
- 42:48You need this sense making system that provides some sort
- 42:52of, you know, taxonomy semantics, you know, like
- 42:56organizational structure for it. And I think that's there's so
- 43:01much opportunity for platforms to, in their domains of
- 43:04expertise, be that authoritative arbiter of context for all of
- 43:10that. Data.
- 43:10So they're building their own semantic layers, they're
- 43:12building their own sort of data dictionaries or taxonomization
- 43:15of their data, right? Is that?
- 43:17So that's one variation. Like if you talk to the data
- 43:20layer companies, they talk about this very much, you know, from
- 43:23like, oh, well, this is the semantic, you know, layer that
- 43:26you put on it. But I think it's more than just
- 43:29semantics, because I think it's a lot of relational process.
- 43:38Know how, you know, I was just having this debate with, you
- 43:41know, someone else earlier today that, you know, this idea that
- 43:45the AI companies are going to come in and like, OK, yeah, you
- 43:48don't need your sales force, your marketo, any of those
- 43:50platforms. We just automatically can do
- 43:53this for you. Yeah, we'll create a CRM for
- 43:55you, we'll send out emails for you, we'll manage case, you know
- 43:57all that stuff. Yeah.
- 43:58I just. To me, it just shows a, a, a
- 44:00lack of just how much and a lack of understanding of just how
- 44:05much sophistication in the actual process of how marketing
- 44:08runs and how things get done. It's it's massive.
- 44:12And even if a lot of that has also been in the human layer of
- 44:17how we run marketing, if we truly want to bring more a
- 44:22gentrification to those things, well, it has to translate from
- 44:26the human layer into systems. And I at least the way things
- 44:30are structured today, yeah. I just don't see the Frontier
- 44:33Labs, Yeah, being being that source of that system.
- 44:38Well, even looking at UX Scott to like I I, that's a good point
- 44:41you made that it's not decided how things are going to look.
- 44:45It's not as though like the the chat, the prompt is the final
- 44:49destination or anything we've seen now is the final
- 44:51destination. Even like SAS UX took like a
- 44:54good decade to even mature React and other like technologies
- 44:59needed to be created in order to make it more dynamic.
- 45:02And I mean, there was a lot of technology went into that and
- 45:05thinking. And so I don't know, it's people
- 45:08who are proclaiming the end of the end of like the front end.
- 45:13I think it's a little premature. I think it'll look different
- 45:15than we expected, probably different than it has been, but
- 45:18I don't think it devalues necessarily the UX profession.
- 45:21I just think we're in the unknown right now and it's so
- 45:24and, and, and people are guessing at what the answer
- 45:27might be and many of these guesses will be wrong.
- 45:28I think that's where we are now. I'm sorry your camera is just in
- 45:35this like I know. And again I go like fuzzy OK
- 45:40autofocus. You are here.
- 45:43Isn't here yet. I'm like computer autofocus
- 45:46well. The sound is great, so we can
- 45:47hear you fine. So like most people listen to
- 45:49stuff. OK.
- 45:52Well, I guess, but going back to one of the topics we wanted to
- 45:56talk about was kind of ad tech martech conversion, Scott.
- 45:59So like one way to look at it, you know, looking at the data
- 46:02would be martech is more deterministic and durable
- 46:05identities versus ad tech that's more kind of ephemeral and
- 46:10probabilistic. Again, one way to look at it.
- 46:13Love your thoughts on that. And maybe we can use that as a
- 46:15starting point for kind of talking about the 2 worlds.
- 46:19The short answer is yes, my I just know I'm going to get a ton
- 46:25of hate mail. From.
- 46:27This. Speak your mind.
- 46:30My point about architecting adtech and martech as part of
- 46:35the same system, I don't mean the same system, like oh, the
- 46:39same piece of software or even necessarily the same vendor or
- 46:43necessarily even the same team. I'm just saying there is some
- 46:47level you go up, we don't have a chief marketing officer separate
- 46:51from a chief advertising officer separate from a chief.
- 46:54You know, there's marketing and this whole spectrum of how we
- 46:58engage with people across owned channels, paid channels, earned
- 47:02channels. Like at some level marketing has
- 47:05to be thinking about how are we orchestrating across these
- 47:08things and really like getting the synergy and the alignment
- 47:12and the consistency for them. And that's just, this is, yeah,
- 47:16again, this is the thing that just gets me in trouble every
- 47:18single time. I'm just like these debates that
- 47:22happen at the level down of like, oh, well, this is a
- 47:25different system than it works on different algorithms.
- 47:28And this one we had the agency do and this one we have, I'm
- 47:31like, yes, but there's got to be a level above this, my friends
- 47:36like. This is a long.
- 47:37Level to be having the debate. Those are the people, yeah,
- 47:41those are the people within the organizations that have moats to
- 47:43protect. They're fearing
- 47:44disintermediation. So I think it's fundamentally
- 47:47driven by fear to say we've got a boat and we've, by the way,
- 47:51boat. Ad tech people, if you say Mark
- 47:54advertising is a subset of marketing and AD Texas subset of
- 47:57Mart tech, that's when you get hate mail.
- 47:59Yeah, yeah, that, that's. Interesting.
- 48:01That'll drive people crazy, right?
- 48:02Right. Yeah.
- 48:02I don't think I agree with it either, by the way, but let's,
- 48:04but let's keep going on this. So, so right you were.
- 48:06Saying it's a world they're. Protecting their moats.
- 48:08Yeah, they're protecting their moats and they I think it's,
- 48:10it's, it's, it's self-interest and it's, it's short termism.
- 48:14But I mean, if you think about marketing as a whole and that we
- 48:16talked about this in in a few episodes ago, is this notion
- 48:19that, that the entire media ecosystem, the advertising
- 48:23ecosystem is competing for basically $0.25 of every dollar
- 48:26spent by big brands, right? The remaining 75% is being done
- 48:31in shopper marketing environments.
- 48:32It's being done by the CPG companies within, within stores,
- 48:37right? The end caps you know the.
- 48:39That's a good point you just made.
- 48:40So like if you look at how marketing spend divvies up,
- 48:42basically it's 25% goes to martech, typically 25% goes to
- 48:46media, right? Obviously this will change
- 48:48depending on the industry, but 25% goes to media, then you have
- 48:50operations and salaries, right? So, so martech is roughly the
- 48:54same as the same as is paid for media like which I think is kind
- 48:57of fascinating. And but in but that's when you
- 49:00follow the money. I think that's where you start
- 49:01to get this difference, right? Because martech is paid for,
- 49:05usually paid for out of IT or marketing budget is procured by
- 49:08marketing in collaboration with IT.
- 49:11Lot of time it's managed by IT or at least supervised, right?
- 49:14And then, I mean, that's a debate, right?
- 49:16A lot of times. But then I think media is
- 49:18different because media comes out of a media budget, which
- 49:21pays for AD tech and it pays for media.
- 49:23And out of working media, you have your tech tax that pays for
- 49:26the, that pays for the ad tech, as well as other things that go
- 49:29into advertising, including brand safety, brand suitability,
- 49:33ad verification. All these other little taxes
- 49:36that come out of there come out of working media And it's
- 49:38usually invisible to the head of media and the CMO.
- 49:42They just know they're paying this much money for media and,
- 49:45you know, get performance reports back.
- 49:46So it's a very different way of actually buying technology.
- 49:49You're buying through intermediary in one case, you're
- 49:52buying directly from the vendors for the most part in the other.
- 49:55Thoughts on that? I mean, again, this is where I
- 49:59feel like I'm speaking a different language than everyone
- 50:04else on this. Like I will grant you all of
- 50:06that. Where, where do they converge?
- 50:10Where is the level at which someone actually connects the
- 50:14dots here? Because I and I know we're going
- 50:18to get into a little bit of like we are now seeing more
- 50:22intersections between these worlds.
- 50:24And they're actually proving to be really exciting because this
- 50:26ability to like align insights we have on the 1st party side,
- 50:31you know, to, you know, the opportunities we have with
- 50:34audiences, you know, you know, in these media.
- 50:36But there's like some actually really creative innovation
- 50:39that's happening there. And it kind of pisses me off
- 50:43that I don't see any reason why we couldn't have potentially
- 50:46been like pursuing that innovation a hell of a lot
- 50:49earlier in this process, you know, But all right, I'm, I'm
- 50:52willing to let that go. Here's the analogy I would give
- 50:55it. You were saying earlier, Brad,
- 50:57about like, OK, this thing like say with Claude, you're like,
- 51:00OK, you know, I can connect. I'm not putting words in your
- 51:04mouth. But this idea of like, Oh well,
- 51:05with Claude, I can say like, yeah, connect to my Salesforce
- 51:08or my CDP and you can connect to my accounting system and you can
- 51:11connect to my customer service thing.
- 51:13And now I'm able to like see insights across this.
- 51:16Now it is 100% the case that every single one of those
- 51:19applications that you're connecting to the run by
- 51:21different companies, they're run by different teams, the
- 51:23economics totally different. It doesn't change the fact that
- 51:26holy shit, you know, now that I can actually start to connect
- 51:29the dots across this like the opportunity for me have like
- 51:32like greater insight and like rethink my strategy.
- 51:35It's a huge unlock. And that's the analogy I keep
- 51:38winding the feel for like ad tech and martech.
- 51:40Just I get all the distinctions between the systems and The Who
- 51:44owns this and. How it operates and and those
- 51:46are all. We need to operate at a level
- 51:48above this. Please for the level of.
- 51:50God, it's a unified operating environment that you're talking
- 51:52about it And Rio, what you just described is basically the
- 51:56economics, the incentive structures and how deals are
- 52:00made in technologies bought, which has nothing to do with
- 52:04where we should be going. To your point, nothing.
- 52:06To do with what the customer like.
- 52:08If you think, if you really think about CX as the customer,
- 52:10put in a customer first working backwards.
- 52:11There's no way in a million years we would design it this
- 52:13way if we could purposefully design this entire system.
- 52:16Yeah, if we could say we could help you better predict churn
- 52:19rates if we pull in all. And this is your context engine
- 52:22argument, right, Scott? You could pull in context, not
- 52:24just from this narrow subset of the stack that you use in the
- 52:28marketing department or you use in the in the, you know, Rev OPS
- 52:32department, but from a broader set of places that are part of
- 52:37your broader set. Maybe it'll give you just more
- 52:39predictive capacity on who's likely to churn, who's likely to
- 52:42upsell, whatever the the use case might be.
- 52:45But you need you need to give more context because you don't
- 52:47know where that data or those insects are going to lie.
- 52:50And the fact that we separate it out doesn't, to your point, make
- 52:53doesn't make a lot of sense in terms.
- 52:55Is the insights I get from what I'm running with Antech and
- 52:59media, can the insights from that help inform the decisions
- 53:02of what I want to do in my first party Mart and vice versa?
- 53:05You know, like, can it inform that?
- 53:07And that's all I'm saying. I'm just like I, I just like us
- 53:11kind of like to work together in the mode of like, OK, how do we
- 53:14think about the marketing experience as a whole and ways
- 53:19in which we can optimize across this fragmented side we.
- 53:24Talk about journey orchestration, right?
- 53:26If you really want to orchestrate incredible
- 53:27experiences and journeys right across the entire like every
- 53:31single touch point, every different channel, I mean, you
- 53:33would, you would certainly need to look at it holistically and
- 53:35you certainly need to be look at tied together.
- 53:38So, so no disagreement with there.
- 53:40But I think that you would need a complete reordering and
- 53:42restructuring. Brett, to your point of the way,
- 53:45the way things are bought and sold, right?
- 53:47I mean, you would need like even now, if you said to a brand,
- 53:49hey, go get, go get data from, they wouldn't even know how many
- 53:53DSPS they're using for activation typically.
- 53:55But let's say they're using 5 DSPS, right?
- 53:57Their agencies are, they might not even have the ability to
- 54:00access those DSPS because an agency would say, we have other
- 54:03brands in there. We can't give you access to
- 54:05that, right? You don't like it's not even
- 54:07your instance. It's ours.
- 54:08We're we're paying for the license coming out of the
- 54:10working media. So even getting access to that
- 54:12would be incredibly problematic. Some brands are doing that.
- 54:14They're actually going paper direct with ad tech platform
- 54:16Scott. She started because they went
- 54:18log level file access. They want the ability to look at
- 54:20like what actually look at the bids themselves and they want
- 54:23the ability to again, take what the learnings from there and
- 54:27pipe them back into like a more holistic CRM program.
- 54:31We're starting to see some of that, but like it's it.
- 54:33That's just like that's a small subset of the world.
- 54:37Most of it does not operate. You would need a complete
- 54:38restructuring. I think that's very fascinating.
- 54:42I and I don't want to downplay the the hurdles and the
- 54:45challenges to it. I think the don't let the the
- 54:54improvement, the perfect be the enemy of the incremental of
- 54:58anything, right? It's just like, listen, yeah, I
- 55:01I fully get why we are not going to turn around tomorrow and
- 55:04there's going to be this perfect alignment, you know, across the
- 55:07entire ad tech mark tech universe.
- 55:10I just like the resistance to even wanting to consider, are
- 55:14there incremental ways in which we can make this better and
- 55:17smarter? You know, And when there's just
- 55:20been such cultural resistance, like vicious cultural resistance
- 55:25to it, and I'm like, OK, I mean, come on.
- 55:27And I think what's fascinating here is now that the hand is
- 55:30being forced, you know, variety of reasons of these changes.
- 55:34Yeah. People are like, oh, actually
- 55:35there are. Turns out there are some
- 55:37interesting things we might be able to do here.
- 55:39And the fact that it's imperfect and it may never be perfect.
- 55:43Yeah. Welcome to technology.
- 55:46But one place you mentioned incremental change, one place I
- 55:49am starting to see differences with like let's say within
- 55:51advertising, the use of first party data from owned in CRM
- 55:55channels for activation and media, whether it's actually
- 55:58activation of a direct audience or creating a see or look alike
- 56:01in order to extend it. I'm seeing some incredible stuff
- 56:04going on there where there's a realization.
- 56:08I think it was initially pushed by the vendors, but I think many
- 56:10brands and CMOS are getting on board like hey, we have this
- 56:13incredibly valuable first party data, we should be using it.
- 56:16We're spending a couple of money in media if we can use to get
- 56:19better return and ad spend, better targeting, better
- 56:22measurement or better, you know, better customer experience, God
- 56:24forbid we should be using this stuff.
- 56:27And I've seen some pretty outstanding results.
- 56:28I think that's one place I'm starting to see a little bit of
- 56:32change. Yeah.
- 56:33Well, and that's on the buy side.
- 56:33And then you have on the sell side, we just start to Joe Root,
- 56:35the CEO of a primitive or primitive.
- 56:37This is super cool. Yeah, yeah.
- 56:39And on the publisher side, they have a huge store of first party
- 56:44data, especially for authenticated audiences.
- 56:46That are and beyond bitstream. We're talking just the first
- 56:48part data they have. Yeah, and and this data is not
- 56:52being pushed into the the audience segmentation and
- 56:54targeting ecosystem because because it just would cause, you
- 56:58know, kind of bandwidth, you know, they're, they're they're
- 57:00still narrowing the pipes to maintain cost control over
- 57:04sucking in all. Of you neither buy or sell side
- 57:06could absorb that congest all this data, the processing would
- 57:08just be astronomical. Also, be too slow and you have
- 57:11two milliseconds to serve up serve up ads.
- 57:13Yeah. And so they're, they're finding
- 57:14a way where it's basically more of a direct connection between
- 57:17publishers, premium publishers that have authenticated traffic,
- 57:20subscribers to brands and, and removing the middle lens of it
- 57:24gives a direct source of incredibly rich contextual,
- 57:27behavioral, historical, so temporal data for their
- 57:31audiences to the brands, which brands right now and the digital
- 57:35ecosystem have an access to a a super small sub segment of that
- 57:39data simply because it would cost too much and it would cost
- 57:43too much. The old it might have is the ID,
- 57:45right? And the DSP is prioritized like
- 57:47IDs over everything else, right? So all you're getting is IDs and
- 57:49basic and maybe a couple of a couple texts, contextual
- 57:52signals. You're really not getting a lot
- 57:53right. And so the bid, so you know the,
- 57:55the argument then is if we had access to this, you'd be more
- 57:59intelligent bids, better targeting.
- 58:02It'd be better for publishers too, because they would keep a
- 58:04bigger, bigger percentage of the share.
- 58:07Yeah, yeah, for sure. So, so data unification, so, So
- 58:11what are your, what are your kind of top takeaways from some
- 58:15of the research that you've done most recently?
- 58:17I mean, we talked about the report.
- 58:18Do you have another report coming out?
- 58:19I think I've missed one of your most recent.
- 58:22I seem to be writing a lot of reports these days.
- 58:25You'll be forgiven for like losing track of them.
- 58:27Well, we did publish this State of Martech report just a couple
- 58:32weeks ago that has the data on the landscape.
- 58:34But also, yeah, some data on like trying to answer actually
- 58:38this build versus buy question where people were using existing
- 58:41SAS, new AI native products. And we did that over like 70 use
- 58:46cases across marketing. I guess the the punchline
- 58:51version of it. Is it it, it speaks of the
- 58:54convergence story. It wasn't build versus buy.
- 58:57It was on almost every single one of these use cases.
- 58:59People are like, yeah, we're buying these things here and
- 59:02we're building these things here.
- 59:03And we have a mix of that and, you know, some use cases more
- 59:07than others, but the, the overall arc is unmistakable.
- 59:10It's like, yeah, we're we're building and buying.
- 59:16Yeah. So, so how do you, how are you,
- 59:18you know, and do you consult with companies in terms of how
- 59:20like as we move towards this, this notion of like a third era
- 59:25of martech, right in this composable canvas you talked
- 59:28about not to not to belabor the point, but how would you suggest
- 59:32brands prepare for that? I mean, is it, you know, like
- 59:35how do they set themselves up for success in kind of a
- 59:38changing world or when it comes to SAS and martech?
- 59:43Let's see, can we sum that up in a few short things?
- 59:46I guess there's three things I would suggest.
- 59:49First of all, getting that data layer in order.
- 59:53I mean, again, for a lot of companies this is still a mess.
- 59:57It isn't just a martech thing, it is very much an enterprise
- 1:00:00wide thing, but. Almost all of the really fun and
- 1:00:03sexy stuff we want to do moving forward, it's predicated on
- 1:00:06getting like the data layer in a much better shape.
- 1:00:09So there's that. The second is when you start
- 1:00:11looking at applications and platforms that you're doing.
- 1:00:14On top of that, openness matters more than it ever has, you know,
- 1:00:18So it's certainly, you know, a dimension that you know, as
- 1:00:21you're evaluating both your vendors and renewals and what
- 1:00:23you do next. Those that are embracing this
- 1:00:27world of yeah, open AP is and connectivity and composability.
- 1:00:32I I think those are the wide choices they're going to give
- 1:00:34you the most optionality and the most flexibility in what is
- 1:00:37still a very foggy future. Anyone who tells you they know
- 1:00:43how your marketing organization is going to need to run like 2-3
- 1:00:46five years from now, it's yeah, I mean, they're.
- 1:00:49Probably not. Smoking something or selling
- 1:00:50something, you know? But we do know.
- 1:00:54Yeah. Maybe both, but we do know it's
- 1:00:57going to change. And so a lot of this comes down
- 1:00:59to like being able to build optionality into your
- 1:01:04infrastructure and your system so that you can adapt.
- 1:01:08And then the third thing is, yeah, the people dimension of
- 1:01:11this, this I am very bullish in marketing employment overall
- 1:01:19because I think the amount of work to be done in marketing in
- 1:01:23this next era far exceeds, you know, the hours in the day.
- 1:01:29But it is very different skills that are having to be developed
- 1:01:33and a lot of learning. And I think both for individuals
- 1:01:36but also for companies who really are investing in that
- 1:01:40change management and that learning and that upskilling,
- 1:01:45those are the three things I would.
- 1:01:47Suggest by the way, that last point I really like I kind of
- 1:01:49liken it to what we're seeing with coding right like there is
- 1:01:51this hypothesis like we've heard a lot of people say, oh, it's
- 1:01:55between Claude code and codecs and cursor.
- 1:01:59All the developers are going to be unemployed.
- 1:02:00Well, it turns out like very few of them are if any, right.
- 1:02:04And it turns out that there is a huge dip.
- 1:02:06There is like unmet demand for coding and now that coding the
- 1:02:10cost has dropped down go guess what, like there's that there's
- 1:02:14way there's there's way more demand for it.
- 1:02:16I actually. Think.
- 1:02:18Paradox exactly. I think I, I think we're going
- 1:02:20to see the same thing with marketing.
- 1:02:21I think we're going to realize marketing has always been really
- 1:02:23hard. It's been very labor intensive.
- 1:02:25It's been very operational, like 7080% of it like the iceberg has
- 1:02:29always been operational, right. And I think we're going to find
- 1:02:31as less of it becomes operational in terms of
- 1:02:33percentage of over marketing, we're going to find, wow, we can
- 1:02:35do a lot more marketing. Well, guess what, like we'll
- 1:02:37need more. We're going to need more people
- 1:02:38anyway that what they do may change.
- 1:02:40It may be less operational, but we're going to need more people.
- 1:02:43So I totally agree with you. And they definitely less Mad Men
- 1:02:46oriented. But like some of my key
- 1:02:47takeaways from from, you know, reading last couple reports that
- 1:02:50you've written, Scott was 1. You know, the obvious one, get
- 1:02:53the data foundation right, right.
- 1:02:55And, and think of the data foundation as beyond how you
- 1:02:58normally or traditionally define data, right?
- 1:03:01It's content. It's, it's the whole notion of
- 1:03:04building a context engine, right?
- 1:03:06Give, give your decisioning systems the ability, as much
- 1:03:10context as they can possibly have, because that's going to
- 1:03:12make their predictive capacity better and you'll be able to
- 1:03:15just do things faster and more accurately, right?
- 1:03:17The other one is sort of it you kind of hit on this before it's
- 1:03:20sort of without using the word semantic, but I'm going to use
- 1:03:23it just for lack of a better word.
- 1:03:24It's sort of build the semantic layer before you.
- 1:03:27Semantic layer, by the way, every uses every podcast.
- 1:03:30Yeah. But before you over build the
- 1:03:32layer, meaning build because you talked about it as sort of core
- 1:03:35IP, meaning, hey, you, you've got this IP on how you manage
- 1:03:41the all of the data storage that you have and and you know what
- 1:03:44the governance layer is what how you organize and structure this,
- 1:03:48right? You want to get that right.
- 1:03:49It sounds to me like that's where a lot of the IP lives
- 1:03:51before you actually plug in a bunch of agents.
- 1:03:53Is that? Otherwise, like, you're just
- 1:03:55plugging into chaos. Yeah, yeah.
- 1:03:58Yeah. And then, and then I think it's,
- 1:04:00it's like moving, moving the attention from plumbing.
- 1:04:02We spent a lot of time with integration plumbing.
- 1:04:05I see it, I saw it in my last company from a CRM integration
- 1:04:08perspective, the amount of time it's, it took to combine
- 1:04:12instances of Salesforce in this particular case, which really
- 1:04:16slowed down our our ability to act on very important statistics
- 1:04:21within the organization like churn rate, who was churning
- 1:04:24based on what products they bought, based on their history
- 1:04:27with the organization. And we had none of those
- 1:04:29insights for months and months and months.
- 1:04:31So it's sort of this notion of how do you get from plumbing to
- 1:04:34decisioning and and and is that a you think those are three good
- 1:04:39areas and and would you add anything to that next?
- 1:04:42No, I think, yeah, I mean, start there.
- 1:04:45Like, let's face it, just those things we name like that's going
- 1:04:48to keep you busy for the next year.
- 1:04:49When you get that homework assignment done, come and see
- 1:04:52me. I can give you the next step.
- 1:04:54Yeah, yeah, yeah. It's like we're still wrangling
- 1:04:57the same problems. Scott looking at the Martech
- 1:05:00framework and I, I sent you over an ad tech one as well, which
- 1:05:04yeah, we, we'd, it was purposely simplified because it's
- 1:05:07something we work with in clients and we're helping
- 1:05:09identify. OK, what tools do you have?
- 1:05:10What tools do you not have? Let's map the what capabilities
- 1:05:12do you have start and then not having.
- 1:05:14And we eventually map those to tools they need us to.
- 1:05:17It's funny, I showed those to a couple of martech guys here and
- 1:05:20and they said they showed me their Martech framework, which
- 1:05:22had like a little box or ad tech, right?
- 1:05:24Like all all your stuff fits right in there.
- 1:05:26I said, come on, like, like you just like you're itching for a
- 1:05:28fight, aren't you? Right.
- 1:05:29So, so it's funny. So I don't, I don't think that'd
- 1:05:32be the right way to do it. I do think there's probably some
- 1:05:34better way to display, let's say the world's of Martech and AD
- 1:05:37tech together, probably based on Brett, to your point earlier on
- 1:05:40the customer, right, like based on the customer jury.
- 1:05:43I don't know, there's got to be a better way to do this.
- 1:05:45I would, I don't know if you've any thoughts on that, but I, I
- 1:05:47think there's got to be a way that can keep everyone happy and
- 1:05:50can be up because they should be if not combined, they're at
- 1:05:54least converging, right? And there's a ton of like
- 1:05:56overlap now. And let's say if the goal is to
- 1:06:00service the customer, whether it's the CMO or the end
- 1:06:02customer, whether it's B to B or B to CI, think these things
- 1:06:05probably should be be looked at in one continuum.
- 1:06:08Yeah, I, I guess, and this is where I understand why it
- 1:06:11becomes fighting words is with an Amtech and Martech.
- 1:06:17I mean, even within Martech, there's just a tremendous amount
- 1:06:20of specialization that is needed on the actual ground of where
- 1:06:24things happen. I'm sorry, but the way the
- 1:06:26social media system works and the way you manage your e-mail
- 1:06:29for deliverability and you know, the way in which you manage
- 1:06:32personalization on your website, these are very, very different
- 1:06:35things. And you know, you generally
- 1:06:37don't have people who are absolute experts in every single
- 1:06:40one of them. And so I have a lot of respect
- 1:06:42for again, everything from the Amtech universe as well too.
- 1:06:45That's just, there's a lot happening there to me getting
- 1:06:50this alignment at a higher level of like, and I think Brett, it's
- 1:06:56the perfect way to look at it's like, like, can we just look at
- 1:06:58this through the customer's journey, the customers lens, you
- 1:07:00know, like, how are they going to do it?
- 1:07:03And like, given that, how do we start to leverage what we have
- 1:07:07across all of these different systems and, you know, to make
- 1:07:11that better and more effective and more efficient for us?
- 1:07:14And that's it doesn't mean that it's poo pooping or somehow
- 1:07:16downplaying the expertise or the specialization in the particular
- 1:07:20pieces. It's just recognizing there's a
- 1:07:23there's another level at which this game is being played and
- 1:07:26we've ignored it for just. Yeah, yeah.
- 1:07:27Well, it, it reminds me of like it like some of the leadership
- 1:07:30lessons I've learned when you're, when you're sitting back
- 1:07:32and you're bringing in various executives to kind of present
- 1:07:34their side of the story. You know, the head of product,
- 1:07:37the head of sales, the head of customer success, They all seem,
- 1:07:40you know, the, the good leaders, the good executives will think
- 1:07:43much more holistically about how this applies to the, the overall
- 1:07:46organization. The more myopic leaders or more
- 1:07:50immaturity in their bloodthirsty ones.
- 1:07:52Yeah, well, yeah, bloodthirsty ones will think, will think very
- 1:07:55narrowly about their lane, what their, their, their Moat, their
- 1:08:00team. And they won't be able to see
- 1:08:02outside that lens, which which is fundamental for for when
- 1:08:05you're making large enterprise tech in data decisions, you need
- 1:08:10to be able to think outside of the lenses of where you're
- 1:08:13coming from. I appreciate the real politic of
- 1:08:16it. I'm just saying, like the way
- 1:08:17this ultimately plays out in the market is there will be some
- 1:08:20companies that bridge that. And I my bet is they're going to
- 1:08:25be more effective and more efficient and they're going to
- 1:08:27win. And they're going to be
- 1:08:28companies who are going to resist that for all sorts of
- 1:08:30variety of reasons. Real politic probably being at
- 1:08:32the end of the day, the real 1 and they're going to be
- 1:08:35disadvantaged. And that's ultimately, I mean,
- 1:08:38maybe I'm wrong. This is my thesis.
- 1:08:40It'll either prove out in the market there is, or it proves
- 1:08:43out that I'm just blue. Yeah, well, in real, we had a
- 1:08:46conversation with with another guest Jennifer that talked about
- 1:08:50sort of the death of specialization, right, How a
- 1:08:52genetic AI is sort of enabling us to think more holistically
- 1:08:57because you don't need to be specialized in each, you know,
- 1:08:59like think from a. You can get specialized in for
- 1:09:01like a detailed information or anything, right?
- 1:09:03It's. Yeah, because you don't need to
- 1:09:05be an e-mail specialist because you can actually have something
- 1:09:08automated that's running and has got the detail in terms of how
- 1:09:11to manage the, the, the, the, the nut, you know, the nuts and
- 1:09:14bolts of that system. Right.
- 1:09:16And so she talked about it as sort of the rise of the what,
- 1:09:19the hyphenated job description or the hyphenated job title.
- 1:09:22Like I am a, you know, it's a cool concept.
- 1:09:24Yeah. Yeah.
- 1:09:24It's like data tech, you know, dash, dash, dash, dash, which is
- 1:09:29interesting because it, it sort of shows that that as executives
- 1:09:32in the industry, you start to think of yourself as more of an
- 1:09:35orchestrator and a conductor. Right as if I, I, I read, I read
- 1:09:40something crazy I tweeted about this morning, Brett, about how
- 1:09:42the and I thought this was the same.
- 1:09:43They're saying like 89% according to some Gallup poll
- 1:09:46of, of leaders. I don't know what that what they
- 1:09:48mean by leaders said that like they found AI did not help
- 1:09:51productivity. It's like who, who do they talk
- 1:09:54to? Like this is the most incredible
- 1:09:56hack for productivity. I like I, I, we use it like for
- 1:10:00almost everything. I used to write emails, right,
- 1:10:02because it's dictate. Boom, the emails done in 2
- 1:10:05seconds is very good. Maybe added a couple of things.
- 1:10:07I mean, it's such an incredible productivity.
- 1:10:09I just find it astounding. And I think you, these people
- 1:10:12don't want to use it and are avoiding it.
- 1:10:14Or I mean, I don't know how you could not know how to.
- 1:10:17I mean, I see some more junior people struggling.
- 1:10:19Maybe they don't have the right context to feed it for inputs to
- 1:10:22get the right outputs. But like, if you've been in, if
- 1:10:24you've been in the business world a long time and you can't
- 1:10:26leverage this tool to make your life easier and do things
- 1:10:28quicker. Like, I don't even know where to
- 1:10:30tell. I don't even know what to tell
- 1:10:31you. Yeah.
- 1:10:32Deep research, Think of the countless, you've done a lot of
- 1:10:34deep research. I think about the countless
- 1:10:36hours that it took before you had the deep research models
- 1:10:39that could go and just scour the world in humanities information
- 1:10:44and come back. Yeah, you have to curate it and
- 1:10:46manage it and validate it. But pointing that type of work
- 1:10:49could take dozens if not hundreds of hours.
- 1:10:52And you can do it in 15 minutes and go, Oh my God, I didn't even
- 1:10:55realize. What did they get this
- 1:10:56information? It's good, yeah.
- 1:10:57Then you check it, but it's really good.
- 1:11:00Well, this is yeah. The irony is a lot of those
- 1:11:03executives who haven't really leaned into figuring out how to
- 1:11:07use his hands on are also some of the same very people saying
- 1:11:10like, all right, and all of you people need to use AI.
- 1:11:13I want to see how AI agree more. And yeah, and you wonder why
- 1:11:17they're sort of like chaos. Yeah, yeah.
- 1:11:20You gotta be able to show versus just tell.
- 1:11:22It's a basic instruction. They're the same as hiring
- 1:11:25Accenture to do like ChatGPT training.
- 1:11:26Yeah, it's. It's a good time to be a
- 1:11:29services. Yeah, it can't amplify the
- 1:11:31noise. I mean, that may be part of
- 1:11:33that. That what that that research
- 1:11:34from the Gallup was, was hinting at though.
- 1:11:36Is that it real? Is that it that it amplifies
- 1:11:39noise? It amplifies because everybody's
- 1:11:42got access to this kind of machine, and if it's not
- 1:11:44governed and managed and structured and strategic, you
- 1:11:48can just have every room, every voice in the room is suddenly
- 1:11:52amplified the larger and is backed by the world's knowledge
- 1:11:55base. You know, Well, yeah, Slop
- 1:11:59cannon I love. That word like.
- 1:12:01If you don't have judgement, but you have access to AI, you can
- 1:12:03therefore become a slob cannon pretty quickly.
- 1:12:05Or if you don't have judgement or taste, but OK, look at AI SAS
- 1:12:09apocalypse. I I think we have to touch this
- 1:12:11a little bit. Got loved like like do is that a
- 1:12:13real thing? I mean obsessed valuation have
- 1:12:16gotten pounded. I think I think I have a spot
- 1:12:18where you were for a long time. They're like their stocks way
- 1:12:20down. I think it's like 2 times
- 1:12:22earnings now, right? I mean the sales force has
- 1:12:24gotten pummeled. I unfortunately had some Adobe
- 1:12:27stock and has not done very well and Figma got absolutely crushed
- 1:12:31right? So like thoughts on SAS
- 1:12:33apocalypse? Is it legit?
- 1:12:35Will these stocks? I mean, obviously no one can
- 1:12:36predict the market, but long term prospects for these
- 1:12:40companies or at least midterm Thoughts.
- 1:12:42Yeah, this is where I disclaim like this is not financial
- 1:12:45advice. Yeah, do not make your financial
- 1:12:48advice based on this podcast. Just yeah, that's a disclaimer.
- 1:12:50Right. I have to assume completely
- 1:12:52insane here. So it's interesting.
- 1:12:55I actually think, I do think for a number of companies the SaaS
- 1:12:59pocalypse valuations have gone too far in the other direction.
- 1:13:02But for the most part, I actually think it's a very
- 1:13:05rational movement. And the fact is that, yes, those
- 1:13:09companies by operating standards are doing phenomenal today.
- 1:13:13Like they're growing, their margins are growing, they're
- 1:13:16providing real value. In fact, actually this whole
- 1:13:18headless thing of now people figuring out how they use AI on
- 1:13:21top of that, it's like growing the value and usage.
- 1:13:24That's all. That's all true and legit.
- 1:13:26The thing is, SAS had gotten to be a model where the financial
- 1:13:34industry, I mean the industry overall, it's sort of felt like
- 1:13:37there's a predictability of this.
- 1:13:38You hit a certain amount, you do this, we can extrapolate this
- 1:13:41out over the next 10 or 15 years.
- 1:13:44And this is how we come to the valuations that gave us all
- 1:13:46these like, you know, crazy multiples on revenue.
- 1:13:48It wasn't insane. It was like, no, we think this
- 1:13:51is pretty stable thing. You can you get to this, you
- 1:13:54keep doing this, we're going to be able to take that to the bank
- 1:13:57for 10 or 15 years. I think the recognition is like
- 1:14:00we do not know what 10 years is going to look like from.
- 1:14:03Now, yeah, these kind of creates a lot of investor kind of
- 1:14:06concern, right And. Well, it's just uncertainty, you
- 1:14:09know, and it's not saying that like the sales force and the
- 1:14:12hubspots and the Adobe's, I mean they very well might become even
- 1:14:15larger giants 10 years from now, but it's not an extrapolation
- 1:14:20from where they are today. And there's this uncertainty and
- 1:14:23the market is priced that in again, I think possibly is
- 1:14:26overcorrected, you know, in a few cases.
- 1:14:28But the overall thing of like, hey, we're in an uncertain world
- 1:14:32for the future of software, I think is actually a very
- 1:14:35rational assessment. Yeah, that's.
- 1:14:37A really good, that's a good assessment because you know,
- 1:14:39people forget stock prices, you know, current performance like
- 1:14:44combined with future projections and future prospects, you got to
- 1:14:47factor both in. Yeah, yeah, no.
- 1:14:48And I think it was in February of of this year that that like
- 1:14:51285, almost 300 billion in SAS market value was just erased in
- 1:14:5648 hours, right. And it's some of that concern
- 1:14:58like is this predictive capacity that we had to say that, hey, if
- 1:15:01you keep adding this number of seats and you keep growing your
- 1:15:04service value by this, you can predict that 10 years is
- 1:15:06suddenly kind of under threat, right.
- 1:15:08That's why real, that's why I don't invest in individual
- 1:15:11company stocks like you talk about a I'm like, that's a fools
- 1:15:15errand. You're never gonna win that
- 1:15:17game. You gotta have a.
- 1:15:19Portfolio, if you follow closely, like if you follow
- 1:15:22closely, you can do pretty well. You're about as capable of
- 1:15:25predicting whether a company can grow as a chimpanzee is, is.
- 1:15:30Yeah, I think no, The reason, the reason why I say that no,
- 1:15:34no, they say that your ability to predict the stock market is
- 1:15:38a, is a. Chimpanzees can write, could
- 1:15:40write, could write a Great American novel if.
- 1:15:41You let them they could is is literally a complete a coin
- 1:15:44flip. But I think they've done some
- 1:15:46statistical analysis to show that anybody who thinks they're
- 1:15:48an expert in a particular category trying to predict stock
- 1:15:51market movements actually has less chance than complete non
- 1:15:56experts, including chimpanzees. That I'll pull that data, we'll
- 1:15:59put in the notes. But that's that.
- 1:16:01I learned that in my stats class MBA and I'm like I am never
- 1:16:05investing individually in company stock.
- 1:16:07And everybody, well, everyone's, when the stock market is going
- 1:16:08up, everyone thinks they're a genius stock trader, right?
- 1:16:11And then? And the more you know, sometimes
- 1:16:12the less likely you'll succeed. And so you gotta have a
- 1:16:15diversified portfolio anyways, Scott's like, Oh my gosh, what
- 1:16:19have I got myself into this? Podcast again I I've got no
- 1:16:22financial advice to offer anyone.
- 1:16:25So, but Scott, I do think they're to respond to you said
- 1:16:28that there's a lot of uncertainty, totally agreed.
- 1:16:30And I do think too that a lot of these SAS companies during the
- 1:16:33boom years, right? Like for them, which was, let's
- 1:16:36say 20/19 until, you know, obviously it's the blip for
- 1:16:39COVID, then everything skyrocketed during COVID when
- 1:16:41everything went digital. And these companies, they, they
- 1:16:44doubled, tripled during COVID. They double, triple their
- 1:16:47headcount, right? They, they, they, they got
- 1:16:50their, their revenues skyrocketed, right?
- 1:16:52And some of this might just be returned or turn to earth to a
- 1:16:56certain extent, right, because, you know, maybe they never
- 1:16:58needed 100,000 people, You know, maybe they only needed maybe
- 1:17:01they over hired, right. So I think that's probably still
- 1:17:03weighing on them as as they as they.
- 1:17:04I think some right sizing is still going on, to be honest,
- 1:17:08you know, met, you know, meta, for example, the 8000 more
- 1:17:10today. It's sad, right, But I think
- 1:17:12there was a lot of over hiring. I think we're still seeing some
- 1:17:14of that and that's and then as they right size their
- 1:17:17workforces. But I think also a lot of these
- 1:17:20the software got very expensive, right?
- 1:17:22As they got so big they you know, they did less cost of
- 1:17:25licenses went up, cost of cloud went up.
- 1:17:27I mean it like the the cost for some of these platforms got
- 1:17:31ridiculously expensive. So maybe this is just like the
- 1:17:34normal reaction, right? Yeah, it's probably a
- 1:17:37convergence of a lot of things, right?
- 1:17:38I mean, those what's happening with AI, those, you know, what
- 1:17:42you just had mentioned about the economics of it.
- 1:17:44I mean, it could be also just that like software has been
- 1:17:47pretty unusual as an industry, you know, that it had that sort
- 1:17:52of evaluation and economic structure to it.
- 1:17:54And this kind of goes back to that convergence thing where you
- 1:17:58know, like even the lines between what is a software
- 1:18:00product, what is a service, you know, how do you do it?
- 1:18:04It could just be like what they call that perversion to the
- 1:18:07mean. Like at some point in time there
- 1:18:09is not some guaranteed cheat code that like, Oh well, if I
- 1:18:11start one of these businesses, I'm going to get like A50X
- 1:18:14multiple on things like no, it's just really hard to build
- 1:18:17businesses like at, you know, those kind of multiples.
- 1:18:20But actually turns out building businesses that 2X3X5X webinar.
- 1:18:26I mean, the world is largely made-up of businesses like that.
- 1:18:30People make a lot of money, you know, building these things.
- 1:18:33People build with modest. Groceries.
- 1:18:34I mean, like, right, you know, I think the software industry,
- 1:18:37yeah, it needs to kind of get over itself a little bit here.
- 1:18:40You know, you aren't a special snowflake.
- 1:18:42I mean, I guess if you're snowflake stuff like, but you
- 1:18:44know, yeah, otherwise you're not a special snowflake.
- 1:18:48Yeah. So what do you think we were?
- 1:18:50We've definitely taken this one as far as this has been a nice
- 1:18:54long podcast. Yeah, I think we're quick hit,
- 1:18:58Scott, if you're good. OK, let's do it.
- 1:19:00Yeah. So one martech category that
- 1:19:03looks more vulnerable than others to AI disruption.
- 1:19:11Because they're all like so incredibly vulnerable to AI
- 1:19:15disruption. All right, this is an easy 1.
- 1:19:18The analytics category already disrupted essentially.
- 1:19:23We had Adam Greco talk about that a few a few weeks ago,
- 1:19:26yeah. I love Adam, that's the best.
- 1:19:28Yeah, yeah. Vibe analytics, OK.
- 1:19:33What's 1 capability or category that you think becomes more
- 1:19:37valuable as I'm not going to say your prediction, but your desire
- 1:19:40for AD tech and martech to to converge, let's say that does
- 1:19:43happen. Like what?
- 1:19:44What becomes more important or valuable?
- 1:19:47I think it was what we're talking about here of context,
- 1:19:51because to me, that's the thing that's so exciting, is the
- 1:19:54customer has a context in which they're operating before they
- 1:19:58get to our first party environment, you know, and then
- 1:20:00at the first party environment and the degree to which we can
- 1:20:03start to see that context and carry it along across that, Oh
- 1:20:08my goodness, I think there's huge value to be unlocked there.
- 1:20:11Yeah. And it doesn't live in one place
- 1:20:13and it doesn't live in one category of technical
- 1:20:16technology, right? It lives in multiple places
- 1:20:18where everything that you can pull together within an
- 1:20:20organization should be leveraged for that.
- 1:20:22Kind. And then again it comes to the
- 1:20:23thing, you can have this specialization in MARTEC and
- 1:20:25ADTEC. As long as we start looking at
- 1:20:28more of the opportunities for alignment and synchronization
- 1:20:31across things, I get excited about the possibilities.
- 1:20:35Yeah. So one thing do you, what is one
- 1:20:37thing that every CMO should should ask their team about the
- 1:20:40the kind of future of the stack? I mean, how should CMOS be
- 1:20:43thinking about? It start with the data layer.
- 1:20:46Do we actually have that in any semi like coherent form at this
- 1:20:51point? Yeah.
- 1:20:53And it seems like there's plenty of solutions out there that
- 1:20:55facilitate speed to value in terms of that type of work,
- 1:21:00right? It used to be a much slower
- 1:21:02integration and normalization process.
- 1:21:04Is that arguable? I don't know.
- 1:21:06I still feel like the reason why data is so hard is it's
- 1:21:08ultimately very little of a technical problem.
- 1:21:12And very much, you know, like, like you talk about the semantic
- 1:21:15layer. The hard part is not, do I have
- 1:21:17a technology that can manage, you know, the semantic
- 1:21:19definitions, right? Actually, that's pretty easy.
- 1:21:22It's like, can I get people in my company to actually agree
- 1:21:25what the semantic definition for this thing is?
- 1:21:27That's really frigging hard and so, but it's even.
- 1:21:30Within a company, yeah, that's wild.
- 1:21:36What's like 5 years from now, I think I know what you're going
- 1:21:38to say. But five years from now, like
- 1:21:42looking at that, let's say the Martech stack like in in the
- 1:21:46framework you put out more solutions, fewer solutions.
- 1:21:54I believe two things are I have high confidence in two things
- 1:21:59and a question mark about 1/3 the things I have high
- 1:22:02confidence in is I do think this like commercial infrastructure
- 1:22:05that powers this stuff is absolutely still going to be
- 1:22:08thriving. I think it'll be pretty
- 1:22:10consolidated. It's not going to be 15,000
- 1:22:13infrastructure solutions. You know, the second thing I'm
- 1:22:16very confident is there's going to be an absolute, like it's
- 1:22:19going to be dwarfed by the amount of custom things that
- 1:22:22people are just running inside there or at just the hypertail.
- 1:22:25Yeah. You know, where I have a
- 1:22:27question mark is in between those two extremes, will there
- 1:22:31be a market for smaller packages of, I mean, whatever they are,
- 1:22:37agents, apps, skills, you know, and will there be marketplaces
- 1:22:41that facilitate those things? I mean, none of those are going
- 1:22:43to be like multibillion dollar businesses, but is there an
- 1:22:47aggregate something really interesting that could happen
- 1:22:50there that I don't know. I'm I'm that that's the one I'm
- 1:22:54most curious about. Like maybe, maybe.
- 1:22:56Yeah, Yeah. Well, hey, good answer.
- 1:22:58I like. That's a great answer and and
- 1:23:00thanks Scott for for joining us and entertaining this.
- 1:23:02Is cool. Yeah.
- 1:23:02We're looking forward to this. Did not disappoint.
- 1:23:05Our our crazy meandered conversation.
- 1:23:07We kept it thematically tight. I think we kept coming back.
- 1:23:10I. Just want you to know, as I
- 1:23:11start getting all the incredible hate mail that's going to be
- 1:23:14coming my way and hate LinkedIn both, I'm tagging all of you on
- 1:23:18this. Yes, a bad press is good press
- 1:23:22any way you look at it, right? And it was.
- 1:23:24No such thing as bad press, right?
- 1:23:25Yeah, One thing I love about you, Scott, is that your, your,
- 1:23:28your influence sort of goes beyond the world that Rio and I
- 1:23:32often, most often see them and even beyond Martech.
- 1:23:35And you know, my brother works for a company called Elastic
- 1:23:38Path, which is a composable commerce company, subscribes to
- 1:23:41your feeds like read your, read your reports.
- 1:23:44And so it's a much broader view of the tech.
- 1:23:47Oh, you were. Almost shiven Miles like a
- 1:23:50couple weeks ago, Right. I saw that too.
- 1:23:52They're they're great good friends of ours.
- 1:23:53Yeah. It's all interconnected.
- 1:23:55All of this tech and data is all interconnected, and we sort of
- 1:23:58have to break down the silos of the Bartek, the AB tech, the
- 1:24:01data tech, you know, the AI, the BDB, SAS of varying types,
- 1:24:04right? To me, it's all decisioning on
- 1:24:07underlying data. That's pretty common, all of our
- 1:24:10organizations that we service, right?
- 1:24:12Amen. Amen.
- 1:24:14Well, for everybody that made it this far, this is one of our Joe
- 1:24:17Rogan episodes. We told you in advance that this
- 1:24:19is going to be a good long one. And visit us at Dev dev, dev dot
- 1:24:23signal, noise dot AI, YouTube, Spot, Spotify, as well as Apple
- 1:24:28Podcasts. We've got a ton of content, but
- 1:24:31editorial and other editorial deaths from our executive
- 1:24:33contributors, as well as video and audio content available, and
- 1:24:38we're publishing a ton more. Thanks Scott, and we will see
- 1:24:41you all next time. Thank you.