The CDP Reckoning: Reality, Hype, and Life After the Magic Quadrant
- Feb 10
- 48 min read
Updated: Mar 6
The Customer Data Platform, or CDP, was supposed to be marketing’s system of record. Instead, it’s become one of the most fragmented—and misunderstood—categories in the stack.
In this episode of Signal & Noise, hosts Brett House and Rio Longacre sit down with Matthew Niederberger, Founder of MarTech Therapy, for a candid, practitioner-level conversation about where the CDP market actually stands today—and where it’s headed next.
Matthew brings 20+ years of hands-on MarTech experience and a global perspective from working directly with brands across Europe and beyond. Together, the group unpacks why so many CDP initiatives stall, why buyers remain confused despite a decade of “category maturity,” and why analyst frameworks like the Gartner Magic Quadrant are becoming increasingly misaligned with how marketing technology is really bought and used.
This isn’t a vendor ranking episode. It’s a reality check.
Topics include:
Why the CDP category feels both overcrowded and incomplete
Platform CDPs vs. composable stacks—and where agentic AI actually fits
The hidden costs of monolithic ecosystems and switching friction
Why “data quality” still matters more than architecture diagrams
How regional maturity (especially in Europe) changes buying behavior
What marketers should actually trust when evaluating CDP options
Whether “CDP” survives as a standalone category—or dissolves into something else entirely
If you’re a CMO, CIO, MarTech leader, or practitioner navigating CDP decisions in 2025, this episode will help you separate signal from noise—and avoid buying technology that looks great on a quadrant but fails in the real world.
👉 Subscribe to Signal & Noise on YouTube, Spotify, and Apple Podcasts for weekly, no-BS conversations on data, tech, and AI—and what actually matters next.
Read the full transcript bellow:
Brett House (00:03)
Hey everybody, welcome back to Signal & Noise. And for those of you that are new to our show, we are a No BS Media franchise decoding data, tech, and AI and the marketing transformations that actually matter. That's my sales pitch. I'm joined by my co-host, Rio Longacre, and our special guest today is Matthew Niederberger calling in from the great country of the Netherlands. We're thrilled to have you. Yeah, and so I wanna ⁓ give you a moment to introduce yourself, ⁓ but before I do, ⁓ I wanna say everybody subscribe. We're on Apple Podcasts, we're on YouTube as well as Spotify. We have dozens of new episodes slotted with some incredible guests over the next six months. And if you're into depth of perspective from founders, they're all available. Yeah.
Rio (00:53)
Plus all of our previous episodes, right, Brett? They're all live on our YouTube site now. They're all there.
Brett House (00:58)
Yeah, so if you're into deep insights from founders, product leaders, innovators across technology and increasingly AI, you've come to the right place. We're moving beyond ⁓ the standard ad tech ⁓ podcast content into some brave new world. So it's super interesting. We're super thrilled to have you, Matthew. Matthew is a marketing technology veteran, 20 plus years experience. I think we're all saying 20 plus years experience at this point in our lives.
Matthew Niederberger (01:27)
Hahaha
Brett House (01:28)
Like, sure,
Rio (01:28)
But who's Karen here,
Brett House (01:29)
I'm
Rio (01:29)
right?
Brett House (01:29)
like, when do I up into 25 or 30, right? Is that gonna age me out, right? And you've got digital analytics, server-side tagging, and then I loved your bio on LinkedIn because it said CDPs, CEPs, CMPs, and I was like, oh, these acronyms are terrific. And so for the sake of our audience, I actually defined some of those things. I think everybody knows what a CDP is, which is a customer data platform.
Rio (01:32)
Never.
Matthew Niederberger (01:56)
Yeah.
Rio (01:56)
Most people probably do, yeah, but not
Brett House (01:58)
CEP
Rio (01:58)
everyone.
Brett House (01:59)
and then am I right in saying that's complex event processing or is that something else? Huh? customer experience platform or engagement platform Okay. Yeah, my search was wrong and then and then CMP is just a content marketing platform. ⁓ Consent management see I I knew I was gonna get this wrong right cuz these things there are content marketing platforms which are similar to CMS's
Rio (02:02)
Experience platform.
Matthew Niederberger (02:04)
customer engagement platform.
Rio (02:07)
Engage and apply.
Matthew Niederberger (02:12)
Consent Management Plan. Ha
Brett House (02:22)
but slightly different for workflows. But anyways, we're glad to have you. You've built a career in MarTech. You're now running your own advisory, right? MarTech therapy, another therapist in the AdTech MarTech space. Yeah, we know a lot of therapists, right? So, hey, so tell the audience a little bit about yourself and what you're doing.
Matthew Niederberger (02:23)
Yeah, absolutely. That's correct, Not a therapist.
Rio (02:39)
or we know it had tech therapists.
Matthew Niederberger (02:44)
And I'm sending out the bills this time, that's the difference.
Yeah, no, absolutely. And about the CDPs, CEPs and CMPs, I think back in the day, if I was friends with the lads over at the Beastie Boys, they would have come up with a great song about that. But unfortunately, that wasn't the case. ⁓ My name is Matthew. I've been in MarTech, I think around 21 years now, kind of rolled into it. think 99 % of the people in this industry kind of rolled into it if you started back then.
Uh, my boss came up to me and said, Hey, uh, I want you to become an administrator for, uh, I'm going to your site core. I'm sorry. Uh, I'm not your site catalyst. That's the one. What's your site catalyst? was a digital analytics platform and it just opened my eyes. And from there, it's just been a continued discovery of solutions, but also the headaches that comes along with these kind of tools. And, uh, you know, there's no Holy grail in this. There's, there's no one perfect tool.
No one ring to rule them all. And it's a, it's a continuous, ⁓ like I said, headache. And that's why I came up with, ⁓ Martek therapy. wanted to call it CDP therapy because I'm really focused on CDPs, but I think I was narrowing myself to a quite a small and as we've experienced now so far, ⁓ a bit of a, roller coaster market. So I widen it to Martek and it's been great so far.
Brett House (04:03)
Are you implementing any of Jungian ⁓ psychoanalytical theory in your therapy or is this more of a ⁓ I understand your problems. I'll help you fix them ⁓
Matthew Niederberger (04:13)
The last, no, well, it's funny you say that. no real theories, but the arguments that I do bring to the table now is that you see a lot of these tools being commoditized. ⁓ It's like the Pepsi challenge. It's very difficult at some point to really tell the difference between all these tools. And ⁓ one of my most recent clients actually said, listen, you need to think about two things. One is, regardless of the stack, the technology, that was all,
We ticked those boxes, I said, what tool is going to make you excited to come in to work for every morning? Is it going to be some tool that you need to write SQL codes in and just kind of wonder if it takes 24 hours for your feedback to come back into your system to see what your campaigns are doing? Or are you looking for something that's so easy to use that's going to get you motivated, that's going to give you the insights, that's going to give you the tools to create the campaigns that you want?
And I see the eyes light up of these junior marketeers who don't know hardly anything about SQL. And that's not their fault. well, they're learning, but these systems are becoming so inherently smart, working with smart marketeers now, but without all the heavy baggage. And I think that's doing wonders. what makes them happy in the morning? And the second one, which I learned about 11 years ago.
Brett House (05:19)
or marketing.
Brett House (05:25)
Yeah.
Matthew Niederberger (05:39)
was when I was doing a comparison for a client and I said, they asked me, do we stick with kind of Adobe and what were they using at the time, Adobe and Tealium IQ or do we move to the Google stack? So I did a research for them kind of comparison and generate on and build a report. Couldn't generate at the time, there was no AI, but we built a report and it basically ⁓ came down. One of the main decisions for the Google one is that if you look at the experience that's in the market, what people are really asking for.
It was the Google stack, it the Google Analytics, the Google Tag Manager. That's where companies were seeking experience in. ⁓ yeah, it's sad to say that a lot of those more expensive tools like Adobe and Teeling really pushed companies towards agencies into lock-ins. And Google kind of opened that up. So I really made a point saying, hey, choose a tool that makes you happy and that offers you more career opportunities going forward.
Rio (06:33)
Cool. we're thrilled to have you on this. I I personally wanted to have you on this pod for a while. pretty frequently, don't know if shitposting is the right term, but we comment on each other's LinkedIn content pretty frequently. You're a very avid poster. Your LinkedIn game is strong, I will say that. I pride myself on mine as well.
Matthew Niederberger (06:36)
Yeah. Yeah.
Brett House (06:53)
High compliments for real Loghaker.
Matthew Niederberger (06:54)
Thank you, thank you, I know, I know, a big smile here.
Rio (06:58)
And I'm sure, well it's funny, I bet it's, and I'm sure you ran into Adam Greco, who by the way is gonna be on the pod in a couple weeks, we already recorded a session, ⁓ very smart guy. I see you nodding your head, so I'm thinking the answer is yes, but the reason why we really wanted to have you on Matthew is, as everyone, I think anyone who's a Martech nerd knows, the Gartner Magic Quadrant dropped.
Matthew Niederberger (07:08)
Absolutely.
Rio (07:22)
I think was about a week ago officially. I had a preview a couple days before that, I think as did you. It dropped and as is the case every year, tons of controversy, people complaining, some people saying they liked it. I really wanted to dig in that, but more generally, just wanted to talk about CDPs or customer data platforms. I wrote this post that got a surprising amount of traffic a couple weeks ago. ⁓ actually just more for holidays about the different areas of CDPs, right? From just being identity resolution, data unification to really becoming activation engines to the whole composable thing that started with high touch and action IQ and some others. And now to maybe a future that's a little...little less certain and I think that uncertainty was very obvious when you looked at the magic quadrant of okay where is this whole thing going? Is our CDPs going to be around in the future? Will they, if all the things CDPs did migrate into other solutions, what's left, right? So there's still a bunch of vendors, it's a pretty big market. So what did it dig into that? So starting with the magic quadrant, look at CDPs, understand...
Maybe we can assess, we think it was report any good? But beyond that, what does it really mean for the industry? And more important, what does it mean for marketers? If we're helping them make their decisions, decide which tools to get, which tools to not get, how to connect things together, these are pretty important questions because certainly you're not going to get the truth from a SaaS salesperson, let's face it. ⁓
Matthew Niederberger (08:47)
Yeah, slightly tainted advice.
Rio (08:51)
a little bit. So with that, let's dive in. Love to talk about it. We've got some great questions, but I'm expecting that you'll have a, the conversation will go where it goes, right? So maybe first question, Brett, I think you had one,
Brett House (09:02)
Yeah, well, maybe we start off by defining CDPs, right? And I'd love to hear what your definition is. And then I'll compare it with the magic quadrant definition from Gartner, which is interesting. And we can sort of deconstruct and interrogate that. But like, what is your definition? Let's start there. ⁓ Of a CDP and what key problems is it solving for?
Matthew Niederberger (09:26)
It's, um, I'm kind of old school in that sense. I, uh, I believe the CDP as it is, um, relatively defined by David Rabb and the CDP Institute and that, you know, there are four parts to it. And, um, um, and that, that's great. I you've got your ingestion, you've got your, um, you've got your, your processing, your cleaning up the data. You've got your, um, identity resolution and you've got your activation layer, the building of the audience and sending and making that data available downstream. And. yeah, I, no, no, no, no, I was sorry. Downstream.
Rio (09:57)
What about analytics? Is analytics part of it too or?
Brett House (10:01)
Yeah, activation is typically downstream in the sense that it's not necessarily built into the platform, which distinguishes it specifically. And this is the space that I came from, ⁓ the data management platform space. Remember, DMPs, right? And how those sort of went the way of the dinosaurs. But it was interesting when I first started. Yeah.
Matthew Niederberger (10:10)
No. Yeah.
Rio (10:22)
They're still around on the supply side.
Brett House (10:25)
When I first started to see, they were around the supply side, but it was much more of an activation play. There were cookie warehouses, right? They started to ingest mobile data to create more, you know, viable household graphs. But oftentimes when I first started seeing this was Exalate that became, that was acquired by Nielsen and it became the Nielsen Market Cloud. So that was the sort of ⁓ context of that. But what I did see is that when these, when a lot of these early CDPs came out, they sounded a lot like DMPs. It was like, is this just a different flavor?
Matthew Niederberger (10:30)
Yeah.
Brett House (10:55)
of exactly the same type of core positioning but but go on
Matthew Niederberger (10:58)
I think the goal of the CDP was comparable to that of the DMP. But DMP mostly dealt mostly, I mean, little asterisks there, with anonymous data. And ⁓ especially when you started to sell that data on. But I think the CDP was more focused on what can I, as a company, do with my data. And when you start looking at, back in the day, God, I say it all when I say that, but it's the gray hairs.
Brett House (11:27)
Yeah, it's not 20 years anymore, it's 25.
Matthew Niederberger (11:27)
Back in the day, no, no, exactly. mean, back in the day, we would just start with analytics. And the thing that frustrated me more than most about analytics is that it's just so passive. You collect all this data and you had quite a lot of freedom in the data that you could collect. You might not be able to collect email addresses, but you could collect user IDs, but you were not able to do anything with it. It would just sit there and you could build out these aggregated reports and see the drop-off rates in your checkout funnel.
But then what? And so, you the next steps in my journey, at least, was more towards qualitative data. So we would be looking into tools like user testing or usability, any kind of in-app research tools to ⁓ further kind of enrich that data. was this great book called Complete Web Monitoring by Alistair Crowell and Sean Power that really inspired me because they said, look, you need at least kind of two types of data to really make a story. If you look at analytics, It doesn't tell you the complete part. people drop off, you need to know why people dropped off. So you would put in a kind of an exit survey to ask them the questions. And then you could make a story about the, the, the drop off that you're seeing, you know, what you've witnessed. And that has always fascinated me. And I, I think I mentioned their book, their books outdated, it's 12 years old, but it's still the premise of it is fascinating and bring it to the CDP. Now that is like the amalgamation of everything.
It's you've got your data collection, you've got the qualitative data that you can put into it, mesh everything together based on a single identity. can clean it up, you can do the governance over it, and then you can send it out to other destinations that are willing to pick up that data. we started with advertising. That's the golden one from Google Analytics because Google Analytics is a great tool, but it's built to kind of see the results of your marketing campaigns.
And now we're just seeing this more and more of these integrations to allow you to do more with that data. And I think that's golden. I we can trash talk CDPs. It's funny that you mentioned DMPs and dinosaurs. And the metaphor that jumped to my head is like this meteor being the CDP that hit the earth. But then at the same time, I'm seeing this image in my head of a meteor breaking up in the atmosphere into dozen little pieces. I'm like, yeah, that's what the CDP category looks like right now.
Rio (13:49)
That sounds about right, Yeah, it's funny, so I've been working, I've been working with CDPs around 10 years, that's probably similar to, yeah, I guess same timeframe as you. And it's funny too, I remember we used to do these MARTEC assessments, you being in consulting side, right? And then we would look at all the different.
Matthew Niederberger (13:56)
Yeah, yeah.
Rio (14:05)
map all the different capabilities and tools the clients had and then look at what do they do, what do they not do. You had a maturity model. It became really obvious around 10 years ago that there was a big missing piece. There was a missing piece that the marketer had no place that was the system of record for them where they could go, where they could activate data. without going to IT necessarily for marketing use cases, whether it was segmentation, whether it was activation, whether it was just some basic analytics on the customer, there was no place they could go. So it became really obvious, I was not surprised when we began to see CDPs emerge. And the first one I worked with was Impairty way back in the day, this is about 10 years ago, and have a bunch of patents on identity resolution, right? So like, I like the way you described it. So there's ingestion and let's say normalization and identity resolution, right? And then there's segmentation, then there's activation, and then...It's interesting you didn't include analytics, but I do understand why with the way things are going.
Brett House (14:59)
Well, and what we learned when we were...partnering with a bunch, know, the treasuredata and the action IQs when we were at Newstar. And one thing we actually realized that they didn't have was identity resolution. So we were bringing the machine learning and AI based, which was one idea at Newstar, which is basically what TransUnion bought to them to give them that architecture for how do you map in hygiene cleanse and align this data across multiple formats and IDs, et cetera, because they didn't have the built in capability, but that might vary a little bit. just to bring this full circle back to, because I think what you just said, Rio, aligns.
And I think we're going to interrogate that, right? Like on how CDPs have been defined, right? And the Magic Quadrant says it at a very high level, but let me ⁓ read this out to you so we can interrogate this. A software application that supports marketing and customer experience use cases by unifying customer data from multiple sources into a persistent, unified customer database accessible to other systems. That's pretty straightforward, right? And then they expand that, right? So just they expand that to almost all the things that you said, Matthew, right? Ingest data, resolve identities, store the data over time, make profiles available for other use cases like analytics, segmentation, schemas, etc. activation, and then support downstream activation across various channels. ⁓ But to me, if you go by this definition, they're sort of saying that CDP is a software that unifies customer data from multiple sources into persistent profiles for not just storage but orchestration. It seems like it's too broad and I'm not sure ⁓ CDPs effectively cover ⁓ all of that. Analytics activation, all of these things. mean, do you think it's a, is it all in one or do you think that we're starting to see a of a fragmentation on how CDPs are focusing on certain use cases?
Matthew Niederberger (16:55)
It's a name game. mean, CDPs is a label for a category for a specific set of features. look, I mean, even these days with composable solutions, does the label even matter anymore? We've got tools like Zing AI for identity resolution that you can just dump on your data warehouse. Is that what we call that a partial CDP? or ⁓ a quarter CDP, there's, I think that's, know, when we can touch, dive into that a little bit later about what the core premise is, core solution that the CDPs needs to deliver, and that's supporting the business and its use cases and any other kind of requirements that it has. but you know, if we, like I said, I am kind of pure in heart when it comes to the term CDP.
And I think the CDP Institute and I'll give gardener that the definition is, is solid analytics doesn't necessarily fit in. You don't need analytics to run a successful CDP ⁓ solution.
Brett House (17:56)
Yeah. And when we say analytics, we're defining, you're defining it as, as marketing attribution is some sort of connecting the dots between, ⁓ customer journey analytics. Okay. Yeah. Okay.
Rio (18:03)
Well, I think more customer journey analytics, right? Yeah.
Matthew Niederberger (18:05)
Yeah, yeah, exactly. Yeah, yeah. Spot on.
Rio (18:11)
So looking at like the CDPs, like it started out, they started out of these, I don't know if that makes, sweet's not the right word, they started out doing many things and then the composable revolution happened where it became pretty obvious that, okay, well, if...maybe you can have the system record be the Cloud Data Warehouse. Maybe it could be a Snowflake or Databricks, right? And maybe segmentation could sit somewhere else, or maybe if analytics is something somewhere else, it could be activation. So I think you started to see these core capabilities of what was the initial CDP start to migrate out, right? And it became obvious, okay, maybe we can just have reverse CTI. Maybe we don't even have to store data, right? Maybe we can just, you know, the way that some of the composable systems do have zero copy and just be facilitating this, right? So you started to see these different models.
And I guess that takes us to the magic quadrant. Like, are they even evaluating one thing? That's why I think it's very challenging what they're trying to do. And I think in your blog post, Matthew, you called out the split. I called it a bifurcation. It's really the same thing. There's two very different markets Gartner was looking at. Can you maybe explain that to the audience, like, what you meant by the split, what they meant by that, and what the implication is?
Matthew Niederberger (19:17)
Yeah, no, absolutely. So when you look at platform platformization, ⁓ and agentification, so that was kind of the split along the, along the middle there. I, I still feel that it's incomplete because, ⁓ agentification infers that people are using AI on top of their data. ⁓ it doesn't necessarily mean that, but let me start with platformization. I mean, these are the huge ecosystems. These are the, ⁓ self, ⁓ what's
Rio (19:46)
Adobe Salesforce, the monolithic solutions, right? Yeah.
Matthew Niederberger (19:47)
Adobe Salesforce, exactly.
Brett House (19:48)
Yeah, Oracle, yep.
Matthew Niederberger (19:50)
I mean, yeah, lot of a lot's been said on LinkedIn, very colorful language towards those companies. And let's
Rio (19:57)
Yes.
Brett House (19:58)
And based on where they were ranked in the Magic Quadrant, right? That's where the Colorful Language is coming. And for those of you that don't know the Magic Quadrant, ⁓ it does rank across a bunch of kind of core categories. you don't have to, you can either pay and become a member of Gartner. ⁓ So there's arguably a pay to play aspect. Yeah.
Matthew Niederberger (20:02)
Oof. Yeah, yeah.
Rio (20:19)
Well, that's the accusation, right? Is that like people are like the rankings are based on.
Matthew Niederberger (20:22)
Pay for play.
Brett House (20:22)
⁓ But they but they do reserve the right to cover you whether you pay or not And you have the visionaries which your top right quadrant and you sort of have the laggards which your bottom left quadrant And then there's some strong very similar to the the Forester did plot the Forester wave reports ⁓ In Salesforce is like almost off the chart right same with Oracle same with Adobe to a degree ⁓ and I think I Yeah
Matthew Niederberger (20:48)
And then the newcomer high touch on there as well.
Yeah. But like I said, I've had a few demos with, the Salesforce CDP. Pardon me if I don't know the term of what it's called today, cause it's, it changes very frequent and, I've never had a sense seeing it like, ⁓ this is going to work. This is, this is what a company needs.
Rio (21:00)
or changes frequently.
Matthew Niederberger (21:12)
But I can understand and the same goes for Adobe and I don't want to be Adobe bashing or any kind of brand, but a lot of these solutions are built around incorporating it within a single ecosystem. A Adobe real time CDP. Look, it's good, but it flourishes when you combine it with other Adobe products. might have it. A Adobe target. Yeah. ⁓ mean, ⁓ I think you can, I think I've seen them. Don't quote me at all.
Rio (21:33)
If you have like AEM as your, let's say for your .com, yep.
Matthew Niederberger (21:41)
being recorded, but don't quote me on this. think even within Google or sorry, Adobe analytics, you can build out segments and sync them to ⁓ ACM and Adobe target for campaigns and for targeting based on the analysis that you do in Adobe analytics. I mean, and that's fine. There are companies who do have the resources to be able to deal with that kind of monolithic presence within your architecture. But and I mean, and that is what platformization is. I kind of label it together with enterprise. But if you look at the, yeah.
Brett House (22:15)
Yeah, that's exactly what I was thinking. Yeah if you're a big, super complex organization, especially one that needs centralized governance, ⁓ cross-functional orchestration where there's so many silos within these big, complex multinational organizations, that... Yeah.
Matthew Niederberger (22:24)
Yes. Yeah, but it's not a requirement.
It's not a requirement because there's so many solutions out there that can do ⁓ multi-brand implementations without these monolithic tools in there. based on my own experience, and there have been a few, I think it's also a lot of political clout that these brands have. The Adobe's, they do great deals with these larger agencies⁓ same with Salesforce. have, they have great partner plans. So when you, when you look at it that way, you see, and I'm, I'm not blaming anyone. This is just based on my personal experiences. They push these tools onto clients. ⁓ there was, there was one client, ⁓ that I worked with and at the end, ⁓ they needed to do a screening for a solution to see if, ⁓ just for, just for.security. So their security team could get an idea of what a CDP was, but, they, they did it on Adobe real time CDP only afterwards. said, look, this process was too immense to, ⁓ it dragged on for too long. They didn't want to do anything else. So they just stuck with Adobe real time CDP. And I mean, that's fine. And then even Salesforce, you'll get, you'll, you'll, you'll have your, you've RFP finished. You've, you've actually, ⁓ shortlisted it down to three. And then a ball, a call comes, from somebody from the board saying, Hey, you need to include Salesforce in this process. Look, it's above my pay.
Rio (24:01)
or straight from the CEO who had dinner with someone, right?
Matthew Niederberger (24:04)
Yeah, played golf or had got seats to some concert that the daughter wanted to go to or whatever. I have no idea. That's way above my pay grade and I don't get involved with it, but these situations happen. ⁓
Brett House (24:09)
Mm. You never got fired for hiring Salesforce, right? Or you never get fired for hire... You never got fired for... Yeah. Yeah.
Rio (24:22)
Well, you might, but for not including them, you might get in trouble,
Matthew Niederberger (24:23)
Seriously? that's true. That's your probably, yeah, probably that, yeah. There are countries where that happens and, ⁓ no, but it's, it's a really crazy, crazy system, but that's the platformization. then gentrification is more towards the it's I like to split that one in two. So you have the composable and then you have the composable plus AI, which if you look especially around AI decisioning that tools like, braise now have, but also high touch that
Rio (24:27)
right?
Matthew Niederberger (24:51)
I think is really where the market is slowly headed. ⁓ Paul Meinshausen from Amp said perfectly, it's the billion decision problem. There are so many, once you start working with AI and optimizing a lot of your campaigns, humans can't keep up with the pace of the decisions that need to be made to optimize those campaigns. So you leave it to that gentrification. And I ⁓ think we're slowly heading there and optimizing and performing the campaigns at least and we'll see what happens, what follows next.
Rio (25:24)
You know, it's interesting. I would almost, I would agree with what you said and I think I'd break it into two steps. think number one, there's the monolithic versus composable. I mean, that is still a thing, right? I know everyone's saying they're more composable, but you know, the reality is an Adobe and a Salesforce because they have a platform. have, you know, it is kind of monolithic. You can maybe break certain things off. They just are not going with a composable approach, architected differently. So how would that be? That would be the first, let's say, split or bifurcation, then I think the second one, which is interesting, is the agentification, right? And this one's interesting because when I first, let's say three years ago, post-GI GPT, I remember thinking, wow, the winners of this are going to be the platforms. And that was my initial thought, right? I said, well, because they're going to be able to embed automate things in their platforms using AI tools very easily. they like to think about how quick did Adobe have Firefly in there where they could just, want content generation, boom, it's done. You already have, you already host your .com, we're already pumping out your emails, we already have all your data, it's easy for us to do. So I think that was the initial.
I think that was correct, but what I think very few of us saw back then that has happened now is as the tools have gotten better, as they've gotten more sophisticated, as they've become much easier to build out standalone tools or integrate AI into existing tools, even if you're not a big Adobe or Salesforce, even if you're a smaller company, you can do it really well now. I think kind of leveled the playing field. I think that the gentrification of that approach is starting to happen as like the front end of software gets rewritten and how these systems interact. I think that is the...I agree that's the more important one, but I think you have to look at the two together.
Brett House (27:01)
Yeah, and would you assume that agentification's strongest when the warehouse is already sort of the source of truth for the organization? They've got a centralized data warehouse, like a snowflake or a Databricks. And teams, need modularity. That's where Composable came in. They need speed to market.
Rio (27:19)
Well, they also need consent.
You mentioned CMPs before. There's a connection there.
Brett House (27:22)
Yeah, and do these monolithic platforms. And I've seen more Salesforce instance integrations than I could care to think about, where half the time that your rev ops team spends is on integration and data data duplication and hygiene. And they're not getting any outputs from the platform because they're trying to ⁓ combine these modular.
I'm sorry, these monolithic systems that are super hard to do and it just doesn't make you nimble.
Matthew Niederberger (27:51)
Yeah. No, no, but that plays right into the hand of these composable solutions. And to kind of get back to your question about agentification, does it depend on the data warehouse? Look, in a sense, it does. We need to store the data somewhere, but I think the real key is going to be data quality. And I think so many people are overlooking that. ⁓ Shit in is shit out. It's very simple. ⁓ So as we can spend time, we can spend money.
Put resources on implementing the most fanciest tool, the most composable solution, Mac approved ⁓ stack that you can have. But if your data quality is crap, that's the only thing you're being served, or at least your clients are.
Brett House (28:33)
Yeah, yeah. And do we think that the kind of smart hub, this notion that, go back to where we were defining CDPs, this idea that this enterprise system, right, which I think points to the oracles and the adobes and the sales force, it, you know, becomes like your end to end platform. Do we believe that there's a future there? You know, when we talk about this bifurcation with the magic quadrant?
Matthew Niederberger (28:58)
I've not necessarily well on CDPs. I'm you know, I think the jury's a little bit out on that I think it's headed that way. I've seen cases where Gentic is actually ⁓ doing awesome work with SEO You know connecting it to Google search Asking it for plans to update your content and actually generating that content for you putting it back online and you're you're you're seeing effects I mean, CDP is going to be headed in that way. And Duarte Garrido, the guy's over at Dojo AI, I think his vision around the platform that he's building, Dojo, especially around marketeers, is that, look, you need to get these agents to do the work for you. You still need to have these guardrails, you still need to do due diligence, but as a marketeer, your strongest...
The strongest thing that you can deliver as a marketeer is your creativity. But like you said, we're wasting so much time on integrations, trying to get tool A to talk to tool B, understanding the data model between them, make sure that we're only sending what we are allowed to send based on consent management, et cetera, and et cetera. That just wastes, it's important, but it takes away so much time from the marketeers to be creative. And I think AI will definitely...change a lot for that because one thing AI is not good in is coming up with new and unique ideas.
Rio (30:30)
So look, turn back to the magic quadrant. As a marketer, should a marketer find it valuable? Is this something they should look at? mean, does it have information there that should be helpful when they're going through their vendor assessments? Or do you think it's kind of, is it past the sell-by date?
Matthew Niederberger (30:44)
Yeah, I'm on the latter part sell by date. Listen, if they, I don't want to ⁓ bump heads with anyone at Gartner. They're doing their work. They're doing their research, but in all honesty, if you look at the entire quadrant that I think there were this year only 10 vendors on it. Yeah, only 10 out of 200.
Rio (31:03)
Only 10, only 10 made the cut out of 200.
Brett House (31:04)
Yep. Yeah, and let's name those. Salesforce, Oracle, ⁓ Unifon, Hightouch, Adobe, Treasuredata, Imperity, Telium, Twilio, and Blueconic.
Matthew Niederberger (31:16)
Exactly. And, ⁓ there's one client right now that I'm working with and we're replacing blue conic at the end of the day, you can be on that magic quadrant. You can be at the, the top right hand corner alongside holding hands with Salesforce if you want. But if your solution does not match the requirements, technical or business or marketing that the, that, know, that the end user needs, it doesn't matter if you're in the magic quadrant or not. And I think.
That is the problem that the magic quadrant ⁓ highlights, is that it's only focusing on those 10. And I'm going to, I'm writing an article right now that talks about the shadow market, because as Rio just said out, or just said, there's 200, there's 200 CDPs alone. We're not even counting the CDPs or other related tools in the market, but there's more than 200.
Brett House (32:09)
Yeah, We'll call this the shadow gap, is the market. Procurement teams might perceive the CDP market as those 10 companies we just named. But in reality, there's a there's a gap there to your point because there's this other market. But does does this act as an effective, especially for larger enterprises that have procurement teams that need to get to RFP quickly and they need to streamline that selection process, which is often internal, right?
Matthew Niederberger (32:12)
Yeah.
Rio (32:36)
Right, I think this addresses one way of selling. I think this addresses the procurement route. If procurement is going to make you a short list, you almost kind of have to be a gardener magic quadrant, forster wave, because they don't know the market, right? They're just going to go to this thing. But I think informed buyers, and I think you're just getting to this, I'll let you jump back in, Matthew, are going to LinkedIn, they're talking to their colleagues, they're going to conferences, and they're talking to people who have actually been in there, practitioners who really know their stuff, and who aren't...
Brett House (32:43)
Yep.
Matthew Niederberger (32:45)
Yeah, exactly.
Brett House (32:47)
Yep. Yep.
Rio (33:03)
I'm not making any accusations necessarily, aren't like... Don't have a cozy relationship with an analyst for them in order to get their logo included. Thoughts?
Matthew Niederberger (33:14)
They don't drink Kool-Aid.
Brett House (33:18)
Yeah, well, so it's an imperative for the CMO, right? To be educated and informed, or the people on the CMO's team, to know, or the CTO, whoever it is, to know who these other players are.
Matthew Niederberger (33:27)
Well, let me tell you what I'm seeing. this is like I said, like you mentioned at the beginning, I'm in the Netherlands, Western Europe, and there is a, the digital literacy or at least the, the maturity around marketing technology in Western Europe is quite high. So the magic quadrant doesn't really fly. Wow. That rhymed. Okay. We can, we, we can turn that into a rep. Exactly. So the magic quadrant doesn't really fly.
Brett House (33:48)
Back to the Beastie Boys.
Matthew Niederberger (33:56)
because the intelligence of the people working in marketing technology in ⁓ customer side or at least client side is incredible. They are putting together their composable stacks from day one. ⁓ I think that is where they've lost kind of touch with reality is that in these more mature markets, the magic quadrant
Look, it might speak to a CEO or a CTO, but even then, if you look at what a CEO in the Netherlands is compared to a CEO in the US, the market sizes are different. So the CEO in the Netherlands, for instance, is probably a little bit more knowledgeable about what's happening on the ground. And the hierarchy in the US is even higher. So he's even more distance from reality, what's happening within his marketing teams. Now, if you go over to the Middle East, yes.
More magic quadrant. If you go to the far East, yes, magic quadrant, South America, magic quadrant. Maybe it's a lack of knowledge of what's available within the industry. Maybe like Rio said, it's a lack of ⁓ consultants who have real world experience or people who are switching between jobs from one company. Hey, you we used to with that tool over there. that interesting for it? doesn't, yeah. The influence depends. I think it's more regional.
Brett House (35:22)
Yeah, and more based on the maturity of that market in terms of the...
Matthew Niederberger (35:24)
Yes,
Rio (35:25)
Yeah. Yeah, you know, it's interesting to imagine question. was reading through it. mean, like it. I suspect that I don't think that companies pay money and like the money determines the ranking. But what I suspect, and if you look at the questions and in fact, I talked to a couple of vendors who did participate in the process, who remain unnamed because they did share this anonymously. Apparently, if you looked at the questions themselves, one question was related to do you have to use Intergroup with Agent Force?
Matthew Niederberger (35:52)
home.
Rio (35:52)
Another question related to CRM capabilities, case management, other things. So it's like the questions were baked in a way to give a preordained result that Salesforce was gonna be near the top. Now I'm not saying Salesforce doesn't check a lot of other boxes, but apparently that is the case. Again, that's inside information I was given. I would question, I think whether consciously or unconsciously it influenced their, certainly their...the way they want about this, methodology, right, more than anything else. So I'm not accusing, again, that's not an accusation of pay to play, it's more I think, you're working with people, you're getting information from, they're helping you actually, as a consultant, look, the best thing is you help the brand build the RFP for the work you're gonna bid on. So, that's prob-
Matthew Niederberger (36:33)
Yeah, no true.
Brett House (36:34)
Yeah.
Matthew Niederberger (36:36)
Are you guys slowly starting to understand where the word therapy comes from now? ⁓
Brett House (36:39)
Yeah, we need some therapy man. Lay me down on a couch. Right? You know, and it's interesting because having actually run, and I mentioned this at the beginning of the hour, having actually run a number of both magic quadrants for the marketing cloud space when those existed with Andrew Frank and some other analysts, as well as ⁓ Tina Moffat at Forrester, who's now at Analytic Partners. She ran the DMP wave or there are other folks that run the DMP, and MMOS, which is the ⁓ measurement and attribution ⁓ wave. And it was interesting because they have their criteria. We actually figured out how to sort of reverse engineer their weighting, meaning weight questions and which topics were weighted more heavily than others ⁓ and then you had this opportunity when you part of it was that there was a little influence game gamesmanship right where you were developing a relationship and the people the companies that had the tightest relationships with the analysts directly ⁓ especially at Forrester ⁓ could sort of sway the way that they thought right sway their their inherent biases and one of the biases that we saw in the MMOs wave was a kind of a
Matthew Niederberger (37:23)
Okay.
Brett House (37:50)
lean towards a perspective that having a managed service consulting led measurement practice for the MMOS sort of vendors was critical, even though it's not as good of a economic model as more of a SaaS tech driven model. And so we always felt we were at Newstar at a disadvantage because we were being perceived as tech first managed services, even though we had a very large managed services team. Second, there were other players in the space that were positioning themselves as who didn't have the tech, but they were able to
Matthew Niederberger (38:00)
Yeah.
Brett House (38:25)
influence perspective enough to say this is critical to the success of these huge measurement implementations within organizations. ⁓ And so you could see how that influence sort of colored the results of the surveys. And it happened again and again and again. were being, companies sort of got typecast. And no matter what you did, ⁓ trying to play the game, it was tough to get out of how you were being perceived by the analyst.
Rio (38:51)
Okay, so that being said, I do think what Gartner had in front of it was a very difficult job. And so I'm not tired to take anything away from that. So it's a very big market, there's 200 players. We mentioned the other, and are they even this, like could you, is, you know, let's say Adobe's...
Matthew Niederberger (38:56)
yeah.
Rio (39:09)
AEP or Salesforce Data Cloud, is it even the same as a high touch? Are they even the same product? mean, you can maybe even argue they're not, right? And I think when you look at the monolithic versus composable, very different solutions. I think it was a very tough job. And I do understand why they would want to reach out to vendors to say, listen, this is confusing. Help us navigate this. I'm not surprised. I would probably do the same, right? If I would reach out to people I know who know, and so I don't know. So comments on the difficulty of the task they took on.
Is it even reasonable to expect anyone to do a good job at it, Matthew?
Matthew Niederberger (39:41)
No, no, no, absolutely not. mean, like you said, the definitely big shoes to fill for the analysts over at garden and all these other analysts firms. but you're right in thinking that, ⁓ look, it's not comparing apples with apples anymore. even, even if you look at, ⁓ the C customer engagement platform space, you've got customer engagement platforms with and without analytic solutions built in. It's there is no, there is no one-on-one comparison possible.
Brett House (39:56)
Yep.
Matthew Niederberger (40:11)
And I think that is why I feel that, you know, as much I appreciate the Gardner view on splitting up the category to platformization and gentrification, I think that was the best thing to take home from that report. But if you start looking at what really works, don't look at the reports. Maybe use it as a guideline to kind of, you know, use it in your chat GPT and say, hey, what is an alternative for these Magic Quadrant listed vendors for a company my size, you'll probably get a better answer. You can hire people like myself or like you guys and do that work with you. Because I'll tell you, mean, it's great seeing these junior teams kind of realizing that the MarTech that they've been working with is actually the thing that's been holding them back. And once you start opening up their eyes to the industry, the tools and the capabilities,
Look, they don't care if it's a Salesforce, they don't care if it's a small time startup in that space like Zepic from Australia. They don't care. They want the problems solved. That's the only thing they care about.
Brett House (41:18)
Yeah, get your strategic decisioning. That's what I do a ton of is get that correct. And you can gentify that whole process. You can compound your decision-making ability in terms of like, how are we bringing a specific product to market? How are we speaking to different ⁓ ICPs? How are we adapting to competitive threats and market changes? And doing this in kind of a way that's compounded from just people doing research online.
Matthew Niederberger (41:25)
Exactly. Absolutely.
Brett House (41:48)
And to me that that's where marketers need to embrace the decisioning power that some of these ⁓ even if it's not using LLM ⁓ Data, it's still using the decisioning compute power of LLMs to help scenario plan and to me that's where the magic happens because everything else can be filled in with with is tactics and you can plug in your platforms to solve Hey, we've got a demand generation ⁓ problem. We've got a customer success, you know problem that's tied to product
Matthew Niederberger (41:48)
Yeah.
Brett House (42:17)
analytics, you can plug in the point solutions into those.
Matthew Niederberger (42:21)
Exactly. And, and, and, and that comes back and that comes back to the foundation. If you have the time, if you have the resources and the patients, cause it does, does take a lot more money, go with composable own that data, make sure, make sure that the data is clean, make sure that you can use it and then start from there. And, ⁓
Brett House (42:33)
Yep. And does that make more economic sense too because you're investing in a small piece?
Matthew Niederberger (42:42)
I'm not, ⁓that's a whole different discussion because like I said, you need different resources if you go down the composable route, but what it does do ⁓ is that it gives you that freedom. Look, you might have a Databricks or Snowflake that has all of your business data coming in. You've got your customer service data. You've got your analytics data going in. You've got your customer data going in there as well. Your product feeds, everything. And you can plug a customer engagement platform on top of that.
Brett House (42:49)
Yeah.
Matthew Niederberger (43:12)
to do the emails, the audience building, whatever you want. But if you grow as a company, if you grow more mature and you say, hey, we've outgrown this solution, it's so much easier to say, we're gonna trial another solution in parallel on top of that data and to kind of see, is this what we're looking for? So instead of kind of rip and replacing, you ⁓ can kind of rip off that bandage and put a new one. Well, that's a bad metaphor, but.
Brett House (43:37)
Yeah. Well, and you're testing a of a discrete use case with ComposableLogoutCase, as opposed to trying to sort of re-implement.
Matthew Niederberger (43:40)
Exactly. Yeah. You own the data. So it doesn't matter, you know, what kind of a plug you put into it to send it downstream or send data into it. As long as you own the data, especially in this composable market, there's so many different ways you can go. unfortunately with, well, unfortunately, but on the other end, you have the Salesforce is these ecosystems again, that have their own benefits, but it's going to be very difficult at a certain point to exit.
Rio (44:07)
Yeah, well, mean, use case approach makes sense, right? I mean, if you're number one use case, you know, if you're on AEM and your number one use case is web personalization, you kind of have to go Adobe, right? Because, you know, they have real-time CDP, it's already integrated. So I think some of it depends on what you've invested in already, where your money is, where your data is, and, you know, what your main use cases are. If your main use case, if you're not on AEM and you want to just personalize emails, well, I mean, it's...It's very different, right? If you're a heavy CRM user, you're B2B, and you want to send it a crap load of personalized emails, maybe Data Cloud is the right solution. So I think it really depends on that, right? But looking at the composable approach, and this is where I love to get your insight, Matthew. I think if you really take the logical conclusion of composable, right? I mean, we started taking core components of CDP, like again, like segmentation, identity resolution, data ingestion, and you start breaking them out and saying, like...
Matthew Niederberger (44:51)
for a CTL. Yeah.
Rio (44:56)
That's going to sit in the data warehouse. That's going to sit in Stoflicker Databricks. that's going to sit in Braze. So that's going to sit in what's left. Is there a CDP? And then you start to gentrifying it. Is there a CDP in a few years?
Matthew Niederberger (45:08)
That comes back to the beginning of the discussion. It's just the label. You're not solving for a label. You're solving for your use cases, for your business to do better. And if that takes a ⁓ mashup of all these solutions, then that's the way it's going to go. And you'll get some alphabetic spaghetti soup of acronyms in front of you. But listen, at the end of the day, if it's delivering and you're making a revenue off of it and you're not sinking because of the high costs,
Brett House (45:16)
Yep. Yep.
Matthew Niederberger (45:37)
or other constraints that it might have, then you're doing pretty good.
Brett House (45:42)
Yeah, and it sounds like...
Rio (45:42)
Does this survive as a standalone category? Like in let's say three, four years. Is there a CDP category?
Matthew Niederberger (45:48)
⁓ I know some people would really want to keep the CDP category, but I think we are going into a more kind of fractured system where people are looking for point solutions to use cases within their system.
Brett House (46:05)
Yeah, endpoint solutions can be orchestrated by the marketing team with the proper strategy in place, right? The proper decisioning mechanics.
Matthew Niederberger (46:10)
Yep. Well, not a, and, and if you look at the 2025 state of MarTech report from Scott Brinker and Franz Riemann, micro-SAS, do not underestimate the power of micro-SAS.
Brett House (46:21)
Yeah, but we're dealing with one big hairy audacious thing that companies that are invested in the big monolithic players switching costs.
Matthew Niederberger (46:30)
Yeah. Oof.
Brett House (46:32)
switching costs, You reference it. Yeah, never underestimate that, like, hey, we're gonna, you know, but if you've got the ability to test use cases and sort of build, there seems to be a path where you could start small, you're not trying to boil the ocean and replace the entire thing, but you can replace pieces of it. I mean, how do you think a company navigates that? And are we gonna see a world where enterprise sort of leans towards...
Rio (46:34)
Never underestimate that.
Matthew Niederberger (46:54)
Ooh.
Brett House (46:58)
considering governance and compliance, especially certain industries like financial services? Yeah, yeah, and multi net, you know, are they gonna lean towards the big monolithic players to be able to manage all this in one place versus the emerging threats in the category contenders, the smaller brands that don't have the size or complexity or budgets?
Rio (47:01)
Guys, those things are important. You can't underestimate them.
Matthew Niederberger (47:20)
That's a lot of questions there, Brett. But ⁓ I think it's a hippo situation for some companies. If you look at these enterprise solutions that are in place as a marketer working with it on a daily basis, you might get frustrated about certain features not being available or data points not being available. But at the end of the day, you could have a strong case to replace it. But it's the highest paid person's opinion if you're really going to replace it.
Brett House (47:22)
Hehehehe
Matthew Niederberger (47:49)
And again, a lot of factors come into play there. A lot of factors that I will never see, like I said, it's in a lot of cases is above my pay grade on what happens in those boardrooms and outside them. But I think if your use case is strong enough, I know I remember in the past, and true story, ⁓ I was working for ⁓ TUI, the travel agency, and I had set up Google Analytics. were tracking Uh, the checkout flow we were seeing where people abandoned, put in a tool called campile. It was, uh, uh, a kind of an exit survey tool. was really cheap. Fifth, uh, it was like $40 a month. I was paying for this with my own credit card and I snuck the tool in, um, for exit surveys to ask questions. Why are you abandoning the booking of your holiday? And I got great amount of data. I sorted it all out. I created a presentation and I said, listen, I know exactly this is.
These are the things that we need to do to improve the checkout. And the, the feedback that I got from my manager at a time says, great, this is awesome. This is awesome work, Matthew, but we can't do anything with it because at the end of the day, as strong as your use case might be, it's always somebody else's agenda and their opinion about if they're going to do anything with it.
Brett House (49:08)
Or if there's like implementation life cycles, which they know are there's like a cost benefit analysis. Do you think it was that that was going on in the head?
Matthew Niederberger (49:15)
No, it was, it was just other priorities. I think this was at that time, I think it was political. Yeah. And it's a, was a Dutch saying that ⁓ it cuts wood, but we can't do anything with it. And that's kind of a Dutch expression. It's night out, meaning it's, it's very logical. It's very clear what you're trying, the point that you're trying to make, but we can't put it into action. And ⁓
Rio (49:17)
Just purely political, and priorities.
Brett House (49:40)
Yeah, and your boss might have been the person that championed the implementation of the software that this was... Might have been a... Yeah.
Matthew Niederberger (49:45)
It might. was gone two months later because I think this is an attitude we'll see with a lot of marketeers moving forward because a very young generation of marketeers is coming up and they're getting used to all these cool SaaS tools. And if they're confronted with something like Salesforce Marketing Cloud where they need to string everything together, put in a SQL query to kind of get the data on the audience, look, they'll find another job they'll find a tool that they want to work with. And I think that's a serious consideration these vendors need to make.
Brett House (50:14)
Yeah. Now we might have a point solution bloat ⁓ for a little bit like we did back in the early ad tech days, right? Where you've got 101 point solutions. Eventually there's a consolidation play and then brands start to just cut budgets and say we can't have all these tools. They're redundant.
Rio (50:18)
Yeah.
Matthew Niederberger (50:23)
Yeah.
Rio (50:30)
Well, even, well, even Martech, I remember like we used to do Martech assessments around, let's say circa 2015, 2016, like before kind of this, that there was a big, I think there was a trend right before and especially accelerated during COVID like where Adobe and Salesforce were eating everyone's lunch, right? You know, there's big kind of platformization of everything, right? And I think before that we would do Martech assessments. Remember I did one with the Beauty brand. I think there was like 92 tools and only four were integrated into the, the actual common data core because a lot of a shadow IT kind of what you just described, like the loyalty program, the loyalty team wanted a better solution. So they ended up getting like MailChimp or something, right. And then this other team, they already had some other solutions. So you would see a lot of these things. And then I think there was a consolidation and then that reverse, I think probably because of AI we're in another.
Matthew Niederberger (50:58)
my God.
Rio (51:20)
period of proliferation where people are getting and they're getting exposed to like better tools because it's easier to make better tools and then, but it's also the question, yeah.
Matthew Niederberger (51:26)
Yeah. And the real savvy people will build their own tools.
Brett House (51:30)
Yeah, you just need Claude and a lovable accountant. Boom, MVP.
Rio (51:34)
Yeah, well
Matthew Niederberger (51:34)
No,
Rio (51:34)
this.
Matthew Niederberger (51:35)
but seriously, on these data warehouses, if you take into consideration all the security and the safety precautions with working with data, you can do this. And I'm not saying people should go out there and start doing it, but I think we are reaching a point where the usability of these tools and the effectiveness, wow, four out of 92, that's crazy. It's just buzzing in my head here, Rio.
And you're right, people will find a way to do their work. And I think, again, coming back to the Martech therapy and where I like to help these companies is those silos. had one company last time, she said, yeah, I don't really like talking to this data engineering person. And I was like, let's go into this meeting. And a few weeks later, I got the feedback saying, hey, listen, that was great. Now I know how to communicate with these people because it's all these barriers in between.
But you need to understand what does the data engineer, what's going to make his life easier and happier?
Brett House (52:34)
Sometimes it might just be a scotch or two, right? ⁓
Matthew Niederberger (52:39)
It's not that easy. In this case, was acknowledging the composable architecture that they were working in. And I think as long as you can ⁓ change your mindset to talk to someone in another team ⁓ at their level, be the Uhura from Star Trek, be able to communicate any language.
Brett House (52:40)
You, Yeah.
Matthew Niederberger (53:04)
It's going to make real changes. And I think that is something that's underestimated well, and there's no tool on earth that's going to solve for that. That's something human, that AI won't replace it. That's where we as humans need to come in.
Brett House (53:13)
Yeah.
Rio (53:16)
Yeah, it's interesting you mentioned AI replacing tools. So remember I was watching, we were working on something recently and like this agent was able to spin up for its own needs, like a very rudimentary CRM. Now obviously this CRM, you know, it had case management, contact management, very basic things. So I mean, this thing would never pass muster for like enterprise procurement in a million years, right? Like, you know, like we have big platforms that are great at doing that and solve all the security and privacy needs.
Matthew Niederberger (53:30)
Yeah. And Salesforce was born. No, I'm just kidding.
Rio (53:42)
Right, I mean, it's got to hit, right?
Brett House (53:43)
Hahaha
Rio (53:45)
But you think about it. So I thought it was interesting. But if most of SDR activity and BDR activity is eventually done by agents, who cares? Maybe they can use the basic thing. They're spinning up on their own. So short term, think the big platforms are, SaaS platforms are going to be fine. But this is why I think if you look at Martech versus AdTech, I think Martech is much more...vulnerable to being disrupted by AI because of what I just described versus ad tech. mean good luck, good, yeah, good luck messing with real-time bidding. You have the whole freaking internet, whole open exchange. mean like you have to, you you're talking world scale versus enterprise scale, maybe not even that as you go to really agents doing things on behalf of marketers. Thoughts?
Matthew Niederberger (54:31)
It's, well, that's a, I find that a little bit, I still find it dangerous to put all my trust into AI. Like you said, if you do with ad tech, ⁓ real time bidding, you're not going to keep up as a human. But if you look at MarTech, you've got so many things to contend with. You've got your brand messaging, your brand, how should I say? Just your brand in general. Yeah.
Brett House (54:56)
narrative position. Yeah.
Matthew Niederberger (54:59)
I think if you put too much trust into AI right now, and sure, you could put up these guardrails, can put, ⁓ what's it called again? Human in loop, that's the one. ⁓ It'll work, but it's gonna take a lot of time to be able to train these models. And then still, are you willing to put your job, your career, your reputation on the line to hand it over to AI to do your entire marketing campaign from
Rio (55:08)
in a loop.
Matthew Niederberger (55:28)
Ideation, creation and execution and then performance reporting and then loop back again. think that's a, yeah, that's that's a huge, I don't think that will show up in Polymarket.
Brett House (55:40)
Yeah, no, I tend to agree with you. think the blocking and tackling and some of the hard execution work is going to be handled by the agentic workflows, right? Some of the decisioning and scenario planning, something that I'm in the process of building now where you supercharge that process and you have an always on learning mechanism that learns from the market realities, Which a product marketing team or a CMO or a product can't adapt to the constantly shifting sort of feedback that you're getting
Rio (55:42)
Not yet.
Matthew Niederberger (55:47)
Yeah, it's. Yeah. All
Brett House (56:10)
from clients, all that stuff.
Matthew Niederberger (56:10)
right, here's a scenario then, Brett. Airline pilots. Would you fly in a plane that's flown purely by AI?
Brett House (56:21)
Rio drives in a car that can be driven purely by AI, would I, at this point, no, because even if it gets 98 % of the process correct, you know, from the tools management to the whole, if they, you know, there's no room for error. There's no margin for error.
Rio (56:25)
You know, it's interesting. Well, right now I wouldn't, but at some point it will be very soon. Flying a plane with autopilot will be safer than human. Okay, it's not there. I wouldn't do it. I wouldn't get in a plane if they said there's no human on this thing. I wouldn't do it, right? Because I don't, because it...
Matthew Niederberger (56:56)
wait, no, no, I'm not talking autopilot. mean, have, yeah, yeah.
Rio (57:00)
no not a full self-fly like like a kid because you look at you look at you look at a test the right side i've got a test the free for five years i guess and i like when it first started the full self-driving it was it was okay but it wasn't great right but now it's it's really i mean it'll you know seventy eight miles an hour it'll switch lanes i mean it is really good i mean it's impressive it'll navigate the parking lots
Brett House (57:02)
Yeah.
Matthew Niederberger (57:21)
And if it's foggy, if it's rainy.
Rio (57:24)
well doesn't have light on it so like the fogginess like once the visibility degrades it does tell you not to do it right but it is really like I mean I don't use it I don't use it when my kids in the car because I don't you know my wife won't let me use it when she's in the car either
Matthew Niederberger (57:27)
That was a cheap shot. Yeah, sorry. ⁓
Brett House (57:28)
Yeah.
Matthew Niederberger (57:35)
Exactly. But let's use the business metaphor. You would drive it in a car. Would you drive it with your business in the car?
Rio (57:47)
I actually trust it a lot. I might. I don't think I'm quite there yet.
Matthew Niederberger (57:48)
Yeah, that's the, no, no, exactly. mean,
Brett House (57:52)
Yeah.
Matthew Niederberger (57:53)
humans, humans, we have checklists, the pilots, you know, before they, they, have take a pre takeoff after post takeoff, pre pre landing, et cetera. So they have all these checklists that they need to go by. And as many times as I've been working with AI, with AI, I have, especially with cloud code, you can create this cloud file where you put in kind of the rules.
And I always say, if I release a version, need to tell me, give me four options for the version history that we want to focus on. And it forgets it about 80 % of the time. And we might get there in the future, but AI is anything but flawless. And I think that's the risk a lot of companies shouldn't be willing to take at the moment. Yeah.
Rio (58:37)
Yeah, no, look, would, would, would people risk their, should the people risk their jobs over? Probably not. Should they risk their life safety? mean, I think there's going to be a long time before we have, you know, let's say pilotless planes, but, but you get in a Waymo or you get, it is, it is really amazing. And the accidents are actually lower than with people. So I think, yeah, I mean, how many people, 40,000 people die of traffic accidents a year in the U.S. Right? So like, if you had, if you had all Waymos, you'd probably have fewer, right? That's, that's the thing. You probably would.
Matthew Niederberger (58:41)
Yeah. Yeah, at that level. Waybo hit a kid last week.
Brett House (59:00)
Yeah.
Matthew Niederberger (59:04)
That's... yeah.
Brett House (59:06)
Yeah, and the innovation life cycles are...
Just like the go-to-market life cycles for consumer brands that BDB brands are being compressed the the the innovation life cycles of AI We talked about this one of our other episodes. I think an AI year is like 56 human days Now that's totally supposition. It's something that I read somewhere, you know I don't think that's factual but it's it gives you an estimate of how much more rapid the innovation life cycles of AI is when it's self-learning You know, it's super interesting ⁓
So should we wrap this up with, because we certainly wanted to jump on the ⁓ perspectives and opinions around the Magic Quadrant. I think we've all come to the conclusion that they're really looking at two different things. It's the platform play versus the agentic play and these competing models, which maybe calls into question the definition, the foundational definition of what a CB is and whether it's even required to have the category. Does this, and let me ask
I your perspective on this too. Like the DMP wave from Forrester that went the way of the dinosaurs, like they had a marketing cloud magic quadrant from Gartner that went the way of the dinosaurs. And why did it change? Because those categories became less definable and less relevant to, you know, sort of client use cases. Do you think that we're facing the last year of the magic quadrant for CDPs? Do we think that this categorization?
Matthew Niederberger (1:00:36)
Maybe not the last year, but if we look at the effects of tools like G2, the Magic Quadrant, how effective are these feedbacks now? People get paid by vendors to, hey, can you recommend this on G2? And if you look at qualitative research that I did in the past, is a certain amount of not the opposite of objectiveness. Sorry.
Brett House (1:01:05)
Subjectiveness, subjectivity.
Matthew Niederberger (1:01:06)
There's a lot of subjectiveness in this data that you will not be able to ⁓ differ. You will not be able to make that split when you start analyzing that data because a lot of that influence will be invisible. So it's very hard to tell how trustworthy these reports are. And I think at the end of the day, what you need to do is learn yourself, get out there, talk to people within your industry, within the same roles that you do, go to these
I think these events will become more important to bring people together to share their knowledge and do your due diligence any way that you can. And I think that is going to be the most important. If these reports will disappear in a year, I don't think so. Maybe a few more years, but I think there are better ways to get started.
Rio (1:01:54)
Yeah, look, I agree. think they're going to continue for a while because Gartner makes money on them and they're probably, I don't know if there is that the argument is just things bad PR. They're getting a lot of PR, a lot of it negative, some good, some people are saying it's a good report. So maybe they're enjoying this. I don't know. My guess is these will continue generally because it's part of their business model and you know, just human inertia, you know, it's part of their business. People are doing these, so they'll continue. I do think though, has, look at how long these things, Brett, mentioned they come and go, CDP has been remarkably durable, right? It's gone on for a while. We've had almost 10 years of CDP magic quadrants and they're gonna continue. My guess is they could do continue on for a bunch more years, but I mean, looking at like broadly like...
Should we even have these reports? I Matthew, I agree with what you're saying. I think it's almost the better question. People don't buy based on these reports. And I think it's an increasingly smaller market that does. Even procurement is getting better reformed in these things. think people should be going to conferences, talking to colleagues, going to... LinkedIn is the arena. Go there. can ask the MarTech therapists. You can ask any of... You can ask all these people and they will tell you what they're thinking. You can jump in a column. I think that's the way I'm seeing people start to make decisions and buy.
Brett House (1:03:03)
Yeah.
Rio (1:03:12)
I think maybe the analyst report is one data point they look at, but it's increasingly a less relevant one. So I think the way people buy software, the way people make decisions has fundamentally changed the last five years. That's really accelerated. And I think that is a trend that's not going to reverse.
Matthew Niederberger (1:03:30)
Increasingly less relevant. Well said, Rio.
Brett House (1:03:30)
And I'll end with, and I think. Yeah, and I think that lends itself to the buyer education, I think is the point that you made, Matthew. CMOs, that functional change, you have to be significantly more ⁓ technical ⁓ than you did in the past, potentially, in terms of understanding how these things actually work, three steps down, four steps down.
Rio (1:03:54)
You don't need to know every tool, but you better know people and have people on your team and know people you can call who do.
Brett House (1:03:59)
Yep.
Matthew Niederberger (1:03:59)
And don't forget that there's a generational shift happening every day of people moving up in these ranks, in these companies with more knowledge from on the streets, bringing it up onto the sea level, and hopefully we'll be making better decisions in the future.
Rio (1:04:15)
So, Matthew, someone wants to get a hold of you to the, someone needs MarTech therapy. How do they get a hold of you? Maybe you can tell the audience about where to find you.
Matthew Niederberger (1:04:22)
Well, I wish I could say Dell 911, but that's a little bit, no, just go to martechtherapy.com or ⁓ find me on LinkedIn. Always happy to do a quick chat with you because I'm really interested in learning the problems that people are experiencing within marketing technology. And there is never, never a case where the stories are the same. Everyone's unique and I look forward to hearing people's ⁓ problems and challenges that they're running into.
Brett House (1:04:49)
Thank you Matthew, thank you of course Rio, this was terrific and for everybody that listened to this entire episode, subscribe YouTube, ⁓ Apple Podcasts, Spotify. We need more subscribers, follow us on LinkedIn, we're publishing weekly, even sometimes bi-weekly, twice a week, ⁓ episodes, articles and more and stay tuned for signalanoise.ai which is launching in the next couple of weeks. So thanks everybody.
Matthew Niederberger (1:04:50)
Thank you, Brett. Thank you, Rio.
Comments