Martech’s Separation and Return to Data Infrastructure with Scott Brinker of HubSpot

Eric Dodds 00:03
Welcome to The Data Stack Show. Each week we explore the world of data by talking to the people shaping its future. You’ll learn about new data technology and trends and how data teams and processes are run at top companies. The Data Stack Show is brought to you by RudderStack, the CDP for developers. You can learn more at RudderStack.com. Welcome back to The Data Stack Show. Today we’re going to talk with Scott Brinker. And I actually have a huge amount of appreciation. And I owe a lot to Scott, he is sort of the king of martech, or marketing technology, something that we definitely need to cover on the show. But he’s the one who first got me thinking about the data stacks. When I was in marketing. In my very early days of my career, he started writing about this 15 years ago, marketing technology and has written 15 years of content on his blog about that he now is leading a bunch of awesome product initiatives at HubSpot and has a huge amount of insight into the data stack from the angle of marketing, and sort of the needs of marketers in the technology stack. Kostas, I want to ask Scott about the difficult first, let’s say storied relationship between engineering and marketing over the past 15 years, because he’s been writing about that a lot. And there have been a lot of changes in the way that companies operate. And there have been a lot of changes in the technology landscape. And I think both are to blame. And so I envision a future where there’s mutual love and respect, and fully integrated technology. So that’s what I’m gonna ask about you.

Kostas Pardalis 01:50
Yeah, I have like two things in my mind. One. First of all, like, I think he’s the perfect person to share with us words, like these Martic landscape use, like, what’s out there, like, oh, but what is Martic? Like, what technologies and what categories of products we have there, too. That’s definitely one of the things that I would like to hear from him. He’s like, probably the most qualified person to like, talk about these things. And the second is, like the connection between marketing dense data. I think it’s almost like a cliche to say that’s okay. The big source of innovation comes from marketing in the data infrastructure, like the needle marketing has been said many times, but it would be awesome to hear from him how this relationship has evolved. Like, how marketeers today operate, like with data, like how technical savvy they are, what does this mean, how they work with, like fit engineering themes, and all these things. So let’s dive in. And like, I think we have the right person to answer all these questions.

Eric Dodds 03:09
Let’s do it. Scott, welcome to The Data Stack Show. We are so excited. I’m personally very excited. You know, having followed us for so many years, I guess I can say it’s over a decade now.

Scott Brinker 03:23
Time does fly in the tech world. Well, yeah. Thanks for having me here on The Data Stack.

Eric Dodds 03:29
Show. Okay, well give us your background. So you actually were a software engineer, you did a lot of Mar tech stuff. Now you’re HubSpot, but give us the abbreviated version of the journey. Yeah, so as you said, I started out

Scott Brinker 03:41
as a software engineer and kind of an entrepreneur, and then was building software for marketers and marketing teams and got really fascinated about it, you know, with the explosion of the web, how dysfunctional the relationship was between marketing and IT software organization, that’s, you know, back in the Dark Ages there. But yeah, as these worlds converge, I built a business called Iona interactive, that was a SaaS company. Let marketers do all sorts of cool interactive content. And then part of doing that I realized, oh, man, I was one of hundreds, eventually, 1000s of companies that were building all these martech products. And while there’s all this great innovation happening, oh my goodness for like the poor marketers, like, why won’t please so I want that. But how do I get these things to talk together? That’s kind of become my cause. And, yeah, that’s what led me to HubSpot about five years ago to lease with one Mar tech company, you know, help them build out more of an integration ecosystem.

Eric Dodds 04:44
Yeah, very cool. And you have a very widely read site, Chief Martech. Do you want to tell us a little bit about that? How long have you been riding on that site?

Scott Brinker 04:54
In just a few weeks, it will be 15 years. You know, so the gray hair is starting To show, yeah,

Eric Dodds 05:03
very consistently.

Scott Brinker 05:05
So, ya know, it’s been fun. And you know, the original Genesis of that was, again, more on the human side of this, you know, how do you get, you know, people who came from an IT or technical or software, technical background, and folks who had come from a very different world, you know, in, you know, classic, you know, marketing backgrounds? How do you get these collaborations? How do you find these hybrid individuals who sort of have a foot in each world? Yeah. And here, 15 years later, that’s actually still the most interesting part to me is like, how the organizations and the human side of this work, but yeah, we keep getting sort of pulled into the undertow of like, okay, yeah, but there is actually all this technology, and how do you get this stuff to work together? Yeah, we helped my goodness, the data explosion?

Eric Dodds 05:54
Yeah, no, it is kind of funny. Like, you know, Mark, DAG marketing technology is almost a little bit of an unfortunate term, just because it really is more about the teams. I want to dig into that. But first of all, you know, as a software engineer, who has built, you know, software to, you know, sort of leverage data that is used by marketers to do X, Y, or Z. And having written about this for, you know, well over a decade, a decade and a half, where did things go wrong? In terms of the relationship between sort of IT and marketing, right, because I think, you know, if a lot of the data engineers, you know, sort of data team members that I know, that I work with, right? Like, marketing can be kind of a burden, because a lot of times there, you know, it’s not always this way, but a lot of times, you know, they’re sort of interaction with marketing is like a ticket for an integration or like, you know, a ticket to like, pull some piece of data or something, right. And marketers are extremely data hungry, and generally need to move really fast, of course. So how do things kind of go wrong? And how did it almost become a meme that, you know, the engineering work in the marketing we’re gonna get along?

Scott Brinker 07:16
Yeah, well, you know, the genesis of that problem is also quite frankly, the genesis of why there was this explosion of all these marketing technology tools, it’s that if we go back, you know, 2025 years, marketing was a relatively well defined set of things you did, there’s the stuff we do with the advertising, the clever all we’ll manage the pricing before PS, you know, it was a very well, box profession, and relatively predictable in its career arcs. Then with the web, everything changed, right? I mean, on so many different dimensions, you know, first of all, the whole relationship with the market, you know, I mean, like this thing, where companies could sort of dictate narratives and sales processes, completely blown up thinking, I like the individuals, the buyers have so much control and transparency on this, you know, but then we also got these explosions of all these different channels, all these different touch points, you know, just all these things changed. And so many of them changed in the context of marketing teams like, Okay, well, what are you responsible for? How do you do it? What are the practices? And it ended up chasing all these consumer technologies, that me if we think back to just the massive innovations, right, so certainly, there was the web to begin with, you know, what’s coming, you know, like, you know, social, you know, mobile, it’s like, all these different things. And so, marketers have been in this incredibly accelerated pace for now, two decades of just trying to even keep up with all this stuff and understand it, and engage with it and manage it. And that’s created all this market for all these different tools. But you know, like, oh, well, you need to do this, or you want to do that, or there’s, there’s no emerging channel here. How do you manage, you know, these engagements and TikTok, or things like that? This obviously created a great stress in the relationship between marketing and IT from the beginning, partly because, although marketers knew they had to work with this stuff, for the most part, they didn’t have a technical background on this. And so all the lessons that, you know, technical professionals have learned over the years of, you know, how do you think about scoping things and managing this and lifecycle and SLAs and all this stuff? Yeah. All right, good. You didn’t have that background, but they needed all this stuff, to get ads on TikTok. Exactly. And so, you know, like it very early on, like, put all these brakes on it. And that, I think, was the origin of this, you know, really incredible tension between the two. Now what happened was, for better and worse, thanks to all these martech vendors, marketing largely took control of their own destiny. We don’t necessarily need it. We’re just gonna get a Credit Card, we can buy this with SaaS, we’re good, thanks anyways, you know, and so the MAR tech stack, in many companies sort of grew independent of the IT department. Yep. Now the great thing, probably the only rational thing that could have happened at that moment in time, but that what’s happened is we fast forward to where we are today is the rest of the organizations become incredibly digitally mature, much more advanced, you know, everything’s getting connected, you know, these layers, you know, particularly the, you know, the data layer. And so now we’re trying to like, converge, the two are trying to like, Okay, well, that’s having this Bartek stack in these isolated silos, actually, isn’t good for anyone. It’s not good for marketing. It’s lack of progress coming, not good for customers. And so we’re trying to now get these things re integrated. And yeah, I think once again, that creates some interesting tensions of like, okay, wait, who’s doing this? Who decides this? What’s the priority? Yeah. Yeah. Maybe the short history of the world

Eric Dodds 11:01
know that. I love it. Yeah, it is crazy. How many marketing tools, you know, are out there. Now. I mean, what’s you because you do this every year? Right? You sort of publish like a landscape? And what’s the latest? You know, sort of, like the latest count of tools? 1000s and 1000s?

Scott Brinker 11:21
Yeah, it’s over 10,000 at this point. And I always have to disclaim that, you know, this is just giving a number like that doesn’t really reflect the industry? Well, because the truth is, it’s totally a long tail, you know, and at the head of the tail, like the majority, if we were to order these companies based on, you know, a bar, you know, market share things like yes, yeah. And that kind of gives a fairly consolidated industry that, you know, there’s maybe a dozen massive, you know, like public companies, maybe a couple 100, you know, like real leaders in different categories. But then you get this really long tail out to the horizon. And these different specialists are these new challengers or regional leaders. And I actually think I’m probably biased on this, but I actually think a long tail is a great thing. I think that’s where a lot of the energy and the innovation, like, keeps that industry moving forward. But yeah, if you just give it like a count of like, well, how many are there?

Eric Dodds 12:21
It’s a lot. Yeah, super interesting, how much so we think about sort of it, like, there’s this explosion of marketing tools, you have IT and engineering works. One thing that’s always been interesting to me is that, you know, around the time this explosion started, it was actually interesting that, you know, let’s just, I kind of frame it, I kind of think about, you know, the iPhone, as, you know, a convenient inflection point to talk about this timeline of the explosion of both, like digital data that’s being produced by consumers, by the web, by, you know, all this stuff, all these digital interactions, then you have the explosion of the tooling of the marketers who are trying to take advantage of this, like you said, take advantage of mobile. But that’s a really interesting time period, because you’re actually right before some of the, like, technologies that have defined what people now call the modern data stack, for lack of a better term, seen as 2007. iPhone, you have this explosion of mobile, and you don’t see sort of what we consider, like your standard, you know, modern Cloud Data Warehouse, for years out, you know, after the iPhone. And so on some level, it seemed to me, you know, and even sort of living through part of that, that enterprises who had the, you know, sort of the ability to afford taking advantage of all this from an infrastructure standpoint, did that right, we recently had someone on the show, you know, sort of built the first big data system at Facebook in 2008. Right. But they’re, you know, so that their product teams and marketing teams can drive better growth. But, you know, very few companies have the resources or the ability to sort of attract talent to do that. You agree with that, sort of the need for this, you know, sort of more advanced tooling and data usage, like on the marketing side, actually sort of sped ahead of where the infrastructure side of things was, at the time that inflection point started to take off.

Scott Brinker 14:30
Yeah, no, absolutely. And it’s led to actually why the problem today is so challenging is basically we got used to the fact that we’re going to have these stacks of dozens and dozens of different tools. And each tool is kind of going to have its own little silo of data. And so we’re using this data in this tool like this. And, you know, maybe there’ll be ways we throw pieces of it over the fence here and there. You know, but the sort of the de facto way in which people were working with these tools is that they were incredibly siloed. You know, and unfortunately, well, fortunately, or unfortunately, depending on how you want to look at it, it turns out actually, people could get a lot done with those things. Because again, we’re like pushing these frontiers of capabilities that just never existed before. So getting this capability, even if you were, quote, unquote, disadvantaged, you know, from beyond island with it, you still were able to, like have, like, a really significant impact on the business. And so that allowed this proliferation, you know, of the modern tech stack, you know, it was just all these different disconnected tools to grow to the level it did before finally, like, you know, we sort of hit a tipping point where like, Okay, now actually, the cost of having these things just connected is now starting to exceed the benefits we get of just having, you know, these isolated capabilities. And so it’s perhaps a wonderful convergence of things that that tipping point has been hit here. And also the exact moment that this idea of Yeah, like the modern data stack, to this infrastructure that could say like, Well, okay, technically, we probably could now pull all this stuff together and start to really connect it, you know, with a common data fabric. And so yeah, I think that’s the exact moment in history we’re at here. Yes. Yeah, we needed and now it’s becoming actually practically possible.

Eric Dodds 16:33
Yeah, for sure. It’s super exciting. I want to jump back, he mentioned the cost. Can you just describe some of the costs of having, you know, sort of data silos? I mean, I know the symptoms of that, you know, sort of upstream, if we think about a data team, right, there’s, you know, the classic pain of like, you know, we need to export data out of this system and get it into another system, right? I mean, there’s so many people in the world who still export CSVs, and upload, you know, to transfer data between systems, you know, but then there’s also sort of, can we build a custom integration for this, etc. Right. But when so that’s one part of the cost. But can you describe the, you know, some other flavors of the cost and how it’s greater than the ability of sort of an individual team to say, like, get really efficient on paid spend, or email marketing or whatever tactic that they’re executing? On the marketing side?

Scott Brinker 17:28
Yeah, I mean, there’s an internal and an external version of this, like, says the external one, you know, is the customer experience, which is, you know, when customers are, you know, touching us on these different, you know, interfaces that are run by separate products. Right, from the customer’s perspective, they don’t know that they’re separate products, I don’t care that they’re separate products, all I know, is like, Hey, I had this thing on the app, I had this thing on the website, you sent me an email with this thing. I called up the call center, I got, you know, like, it just looks broken. You know? And, yeah, we’ve one thing has been consistent expectations of customers just continually, you know, rise. It’s like, Shuddh. You know, and so when we have these failures and customer experience, more often than not, you can actually trace it back to like, oh, yeah, well, there actually systems and data barriers, you know, that allowed that disconnect to happen. I think internally, the costs, you know, range, I mean, again, certainly there’s this cost of like saying, Oh, well, I need to do x, it isn’t happening automatically, or it isn’t, the systems aren’t designed for this. And now I have to put in all this manual effort, you know, to do it all I need to run a report for x, you know, right, I’m gonna pull this together, I’m gonna get into a spreadsheet and, you know, spreadsheets, my goodness, Glue that’s holding all universe together. You know, but the truth is, the bigger costs, or you could argue is actually the opportunity costs is, you know, How many things did people want to optimize, you know, a, you know, broader flow, or they wanted to answer a question, you know, about like, Hey, how is a particular segment performing better than other this? And when they looked at the cost and time and effort it would take to answer that question. They like, Alright, now, it’s great. It’s not worth it. And so that we just didn’t have, you know, the data to make those, you know, things and I think that’s a huge,

Eric Dodds 19:26
huge cost. Yeah, yeah. That’s, I think that’s really well said, the hidden costs are often, you know, sort of the most pernicious, you know, because no one there’s not like a lot of physical weight to like, it’s just going to be too hard to answer that question. So I’m not going to write it’s like, well, if you did, what would that have done? You know, for the business. Let’s jump over to the people side for just a minute because I love what you said about you know, sort of people want to talk about the technology but you started writing 15 years ago because As you’re passionate about the people side, one thing that I’ve seen is that, you know, before maybe where marketers were, you know, I’m gonna do my, I’m gonna execute my marketing tactics, I’m gonna do this and, you know, I don’t like, get, you know, get it out of the way, like, I don’t, you know, I don’t want to be, you know, sort of weighed down by that right or, and maybe it had this attitude, not across the board, of course, like, this is more playing on the mean, but it is like, Please don’t, you know, please don’t another marketing ticket. But I think in a really healthy way, marketers are becoming more and more technical. And I think a lot of data teams are getting closer and closer to the business, because, you know, they’re working with data that’s driving the business. And in order to understand the context of what they need to do, they need to understand the business, which I see, I think that’s a really exciting sort of almost like convergence, you know, are sort of meeting in the middle of their sort of marketing ops roles, or like analytics, engineering roles that sort of blend these concepts. Are you seeing the same thing?

Scott Brinker 21:02
Yes, I mean, this has been the thing about these marketing technologist marketing ops people, is when you look at their backgrounds, either, wow, it’s interesting, it is kind of a 5050 split. But like, at least 50% of it is like people who basically worked in IT software, the technical people by background, they came to marketing, you know, because they saw an opportunity there, they were passionate about it, but they bring enough of the actual discipline, you know, of, you know, solid technology management practices that, yeah, it’s not the total Wild West. And then I think you’ve got another set, actually, interestingly, is these people whose careers initially started in marketing, but they just got really fascinated by the technology. And they put in the hours, and they put in the effort to really understand it. And yeah, I do think these hybrid professionals, this is what makes it fascinating, like, because this is, you know, there’s a bunch of literature out there that you could argue it either way of the benefits of specialization versus hybridization. Sure. But no matter what, even if you say like, yes, there is value to having a deep specialist in X, we realize now that the world has gotten so complex, that having people who are these bridges, you know, that span two or more disciplines together, can just bring so much value in like connecting and translating between them. And so, yeah, I do think really good marketing ops people are really good marketing technologists. I think that actually, they helped create good relationships between the marketing department and the IT department because they understand where each is coming from. And they can do the translation.

Eric Dodds 22:45
Yeah, yeah. Yeah, it’s so funny, I think about even 10 years ago, you know, sort of work, you know, trying to do data driven marketing stuff. It’s like the idea that, it’s like, well, yeah, the marketing ops versus going to, you know, jump in and write some SQL queries to like, figure out what. But that’s pretty common today, actually. Yeah. Which is really cool. Okay. I’m gonna hand the mic over to Costas. What one more question. So, you know, and maybe this is a bad question for, you know, the name of your blog, Chiefmartec. But based on everything we just talked about, it seems like, you know, there was sort of, you know, a technology organization, you know, sort of, like marketing sort of leveraged stuff that got from them to do their stuff. And like you said, it was a pretty well defined, huge explosion. You know, there’s sort of a marketing technology stack that separated and now it’s coming back around. So do you see the future as it’s just sort of going to be the data stack? You know, does the MAR tech stack go away? What does that architecture look like? As for those integration problems, both on the technology side, but also on the team side? You know, continue to converge?

Scott Brinker 24:07
Yeah, that’s a great question. And I yeah, in all humility, I don’t know exactly how this will go. I think it’s a very fluid thing. And it’s also one of those things that it might not be one pattern that’s just universally like Qlik, but we’re gonna see for quite some time that different companies with different cultures, different talents, different maturity, different industries will just draw these lines in different places. But if I were going to take my best guess at like, all right, you know, squint, here’s what I think the future will generally look like. Because I think the data layer absolutely has to be centrally owned. I mean, this is the foundation on which everything is built. But I also think, like the specialization of what data means. I mean, you guys need more data experts than I am but am I. My understanding is that the hardest problem in data still remains, like modeling this stuff and getting the definitions right. And it’s just, you know, the pipes are one thing, but agreeing on what the heck it is that we’re putting through it is a very hard problem. And it is very domain specific. And so you really need specialists, you know, not just in marketing, but in sales and customer service and product ops and finance and all this to really be able to have the mechanism to influence you know, those definitions. But then there’s a layer above that, which is okay, now, the actual operations. Okay, well, if we all agree, we’ve got the common data, and it’s piping and flowing the way it should, you know, what are we doing with that data? You know, what’s the web experience we’re delivering here? How does it change? You know, the level thing? Like, are we tracking these, like, cohorts? If we’re doing that, what sort of actions do we take with them that are different from, you know, the other cohort? You know, and all this sort of, like operations and service layer on top of that, I think that increasingly like, yeah, that’s still gonna have a lot of domain specialization. So I think you’ll have a lot of martech marketing technology that’s focused on delivering you know, these experiences and executing these campaigns and programs. But if you look under the cover, the data that they will be working with, and the data that they will be, you know, collecting and sharing will filter and work directly with that common data layer. Yeah. Fingers crossed.

Eric Dodds 26:24
Yes. super interesting. I could keep going. But Costas, please, please jump in here. Yeah,

Kostas Pardalis 26:31
yeah. So I might sound a little bit like, a naive question for people who are, you know, like, in the marketing industry, but you mentioned, while you were talking with Eric, that’s like, the landscaping martech? is like, more than I don’t know, like, 10,000 products out there. That’s massive, right? Like, I don’t know, like, to me, at least it feels like, wow, like, how do you navigate? That’s right. Can you give us a high level of like, the basic categories of products that they fall under, like the Martic. Industry? To get a better understanding of like, also how, because, okay, obviously, like the industry like mob somehow marketing itself, right. So by seeing the different product categories there, we can also understand what marketing is doing, especially for the more technical people like in buildings that we have.

Scott Brinker 27:30
Yeah, yeah, happy to do that. And I will say, just as a preface to that, like, we think 10,000 products and martech is a lot. Right, it is a lot. But like, if you go to a software review site called G two, you know, and they review all, you know, all kinds of software, not just marketing, I was talking to their CEO there. And this was a few months back, they had 103,000 different products that they were tracking reviews on. And they were the first to admit, like, oh, yeah, we don’t have them all. This is like just a fraction, you know, of what’s out there. So I don’t know, this problem may have first exploded in the context of marketing. But I think this is now actually a universal challenge we face as the world is just full of software. But alright, so in the context of marketing, like what, what is marketing technology? What are these categories? So in our map, we have like six main categories. And then a bunch of subcategories are the main categories or tools for managing advertising and promotion. So this is a lot of what you would think of as ad tech, you know, solutions, we have a whole thing around content and experience. And so this is everything from like, you know, our, you know, web platform web experience platforms, what we do for like marketing automation for like delivering email, there’s things like, you know, digital asset management systems, you know, that feed the, you know, we’ve got, you know, interactive content tools, like the things that I used to do with high on interactive. So there’s all these things about, like, oh, how do we create and distribute content and experiences, then there’s a whole world around what I would call social and relationships, you know, obviously, social media marketing was a big part of that, but it’s also like software we use for managing community, the way we’re doing reviews and reputations. This is where like CRMs you know, sort of generally like looking at the relationship view of the customer or there’s like events, you know, that we run influenced by social relationships, all bunch of stuff there. Then commerce and sales, you know, so if it’s like a b2c business, of course, you’ve got your ecommerce platforms and a whole bunch of E commerce marketing tools. If you’re more in b2b sales, you know, you certainly have these tools around like, you know, sales and get Each band sales enablement, you know, there’s sales, yeah, just sales automation of it’s unkind. And then we enter what guy is actually tracking the column for data, that you could probably push back and say like, Okay, well, this isn’t martech, this is just tack. But I try to look at it through the lens of like, okay, well, if I’m in marketing, I’m, this might not be just for me, but this is something that’s really essential to what I’m doing. And so it’s everything from like, you know, like business intelligence tools to some of the IPASS CDP technologies of how do we get the data, you know, to the right place? There’s folks who are doing all sorts of stuff about, you know, second party and third party data, you know, how do we share and manage that there’s attribution, so all those fun things. And then the last category is, again, like data, probably not specific to marketing. But marketing is up to its eyeballs. And, and this is what I think of is more of like, management oriented tools, like, you know, how are we using these collaboration tools? How are we dealing with projects and workflows, there’s, you know, agile approaches, or all that, you know, marketing’s embraced and, you know, parents in their own unique ways. So I know that’s kind of like this overview of when you break it down, you like, you start out and you’re thinking there’s no way marketing needs a whole bunch of tools. And then they actually start to like, go through all the different things marketing does engage you like, yeah, no, wow, actually, yeah, there’s your daily wide open.

Kostas Pardalis 31:33
Absolutely. So if you have to define, let’s say, the Blitz, okay, I’ll come up like, we won’t care. But like the minimum viable marketing stack, for a marketing team, like what the like, let’s see the how to like the most basic and required stuff that like pretty much every team work starts with, because obviously, like, there’s no end of like, options to adds functionality to reading and processes and all that stuff. But someone starts like a big new young company, they have to build their marketing departments, like the first people like, what’s the main set of like tubes that someone needs?

Scott Brinker 32:15
Yeah, you can get away with a pretty small set of stuff. So I wouldn’t say like, sort of the three things you need, is alright, so you need a CRM or something like a CRM, that’s basically gonna be like to keep track of the EQ we’re dealing with. The second thing you’re guaranteed is guaranteed some sort of CMS or dx be like, Okay, well, I need a presence on the web, and I need the ability to manage that. You know, and then the third thing is probably what you would call marketing automation. I mean, it could be as simple as just saying, it’s email marketing, but it’s basically like, okay, you know, how do I, like, you know, run these campaigns, I have a subscription list, you know, I’m engaging people, if they do come in, and I get them on my subscription list, how am I nurturing them? When do I determine that they’re qualified to be turned over to a sales team? You know, and if you actually, if you get your CRM, your CMS and your marketing automation platform, I mean, you can do like that if there is a lot you can do with just that. And often My advice would be, well, if you can get really good with those three things, before you worry about, you know, all the other possible bells and whistles you could bring to the world.

33:31
Yeah, that makes total sense. And what about, like, interaction points between like, Martic and like, the data technology, right, like you mentioned, you have, like a whole category of tools there that are, like more part of, let’s say, the technology landscape of like, the data infrastructure? Or like, from all these pillars that you mentioned at the beginning, do you see that some of them like, rely more on data, or at the end, like the data infrastructure is only like very horizontal, that’s, you know, goes across like that as we fold like the different tooling that marketing has, and if you take it like to another level, I think it translates into how much data driven at the end like marketing itself is as a practice right? And I’m asking that because to be honest, okay, I’m not coming from a marketing background, but having liked to build data infrastructure products so what I really in very early started appreciating is how much of a driving force marketing is for these technologies, right? Like, there’s a lot of work getting done even today after like, I don’t like 15 years since I started wanting a laptop.

Kostas Pardalis 34:50
That’s the reason that we are trying to improve the technology and even the algorithm that we have is because we’re trying to accommodate the needs of marketing. Right. Yeah. So I’m trying to understand myself as a technologist. Do I just see only one face of marketing? Or, like, what I get through the interaction is actually, you know, like, across the whole spectrum of marketing?

Scott Brinker 35:17
Yeah. Wow. Okay, so there’s a lot there. You know, so I think it’s always helpful to keep in mind, again, where marketing started from wasn’t a very data driven industry, there was actually a small slice of it called database marketing for like, you know, the way traditional direct mail used to be done, that was relatively savvy on this, but for most of our thing, there was that famous quote from the John Wanamaker, like a century ago or whatnot, like, hey, half the money I spend on advertising is wasted. The trouble is, I don’t know which half, you know, and that’s where we’re going to be okay with that in marketing for a very long time. You know, again, fast forward to where we are in this digital age, first of all, the number of channels and things that marketing can invest in is just , like, infinite, you know, and so, the requirement, you know, that you have to be able to make choices. Yeah, it’s just a very different game. And, you know, how do you make choices? Well, ideally, you have data that, you know, that can help you make that choice. But also, I think, you know, marketing’s moved from something where, because of these practices and marketing around like demand generation funnels, you know, and conversion rates and stuff like this, the expectation has shifted to, you know, that the C suite expects marketing to be able to very quantitatively define like, Okay, where are you spending this money? How are you measuring, like, the return that it has on that? You know, and so this is what really, yeah, just changed the game where marketing is so hungry for data. Again, the kind of data you need, there is still this idea that, you know, a bunch of the data is contextually relevant? Questionable, you know, at a global aggregation level, how valuable it is, I mean, like, you know, I think about all the social interactions that someone has with me, well, if I’m in the middle of engaging, you know, with a social marketing campaign, or trying to do customer service through social channel, there’s a whole bunch of contextual data that’s immediately important to me, you know, to handling those things correctly. If I extend that backwards to have a higher level, do I need to have all of that like good food? Probably not, you know, what portion do I want? Do I want some sort of signal that this customer engaged with us in this channel? Do I want to have some sort of signal of like, okay, was that attached to a campaign? You know, what was our investment in that campaign? How did we ultimately measure it, there’s this whole thing in marketing of attribution. And unfortunately, it’s a, it’s a Black Guard, because you know, what happens is, people have all these different touchpoints with us, and then at some point, they actually convert in the buy in now. And then people say, like, oh, well, was it that last touch thing that we attributed to? Or was the first touch we had? Or do we, like, evenly distribute these things? And we do some sort of like, you know, statistical modeling, they’ll, like, see different cohorts? Yeah, right. You know, so. And then there’s all the touches people had that we didn’t even get to measure because like, you know, they met their friend for a beer. And Deb said, what do you actually think of this company? But, you know, anyways, this would be this. I think it’s that attribution funnel that if you were maybe, if he would come pick two things you wanted a very universal data view of in marketing one would be well, let’s just make sure the history of the customer is accurate and available everywhere. And then the second thing is, can I have at least as close of an accurate attribution model of how different touchpoints with people like influenced, you know, their value to the relationship we had with them? That would be ideal and wonderful. Last one is asymptotic. Like, when there will never be perfect attribution.

Kostas Pardalis 39:28
Yeah. 100% I mean, I think we’re getting to the limits of physics anyway, without staff and speech and we get consumed but that is great, actually. And okay, my next question has to do a little bit more with the marketing people and not like the technologies but it has also to do with like, the perception of that no one’s just have about the marketing people. Okay. And there’s no. I always like this idea that when you’re building technology for marketers, you have to make a very strong assumption about how tech savvy but yeah, right. Now we shouldn’t expose SQL to them. No, we shouldn’t like that they don’t know how to use Python. It doesn’t look like too much for them. Like, it has to be so easy that like, I don’t know, like, even like my grandfather can use it for a marketing or like that you those. These are like stuff that I have shared. In many conversations around like building products that are, let’s say, used at the end by marketeers. How much of this is actually true? And how do you see like marketeers developing, like themselves, like as a professional in terms of like getting the required skills to become much more, not only iterating take bugs being like, let’s say, getting closer like to use some of the tools that I also like, developer could use, right? Because if you think about it, like many of the things that a market like Delta has required, like developer work, like landing pages, web pages, like all these web work that needs to be down there, like attribution, you talking about statistics, data bi? Sure. Like where is like the, you know, like the points where you’re saying, okay, that’s too much for a marketeer like we need like an engineer to like, get involved. So, yeah, tell me a little bit more about that. Because I think at the end market, there’s more savvy than we think. And I’d like to hear your opinion on that.

Scott Brinker 41:40
Yeah. Well, I think it’s interesting, because, all right, there’s sort of where it has been, and maybe is today, and where it seems to be shifting to. So if we think about sort of recent history, I think this has really been the value of these marketing ops, marketing technologists, people who are essentially technically oriented individuals who work in the marketing department. But basically, they should be able to serve as that training, like, they can do SQL, you know, probably a decent number them can do a little bit of Python, or, you know, JavaScript they may get they’re not gonna build holdings, but to say, like, oh, yeah, I need to arrange a query to like, figure out this sort of thing. Like, that’s within their scope. And that serves a really valuable purpose. Because, again, like, the marketing team, as a whole has an infinite number of demands, you know, very creative people, like, I can come up with, Hey, I’d like to know this. I’d like to know this other thing. Can we do this? You know? And so yeah, like feeding those requests, Ron to, you know, the data. Org is the ID. Org? Yeah, it never goes well, you know, until the marketing ops and marketing, you know, tech people act as a nice translation barrier there. But the truth is, in some ways, all you’ve done is like, shift the problem over because you still see that marketers, even with dedicated marketing ops, and marketing analytics, folks can invent far more things that they want than even the marketing ops or the marketing analytics people have, which is why you now even seeing within marketing ops and marketing analytics, a shift of like, okay, can we make more of this self service? Like, what can be the scope of things that we can say, Listen, mark it or you’ve got this question added? Here you go. And what is really interesting here is this whole space around, you know, forgive me if I put this in air quotes, but you know, no code, you know, this idea of these interfaces, you know, that allow people to, you know, visually or, you know, I mean, my goodness now was chat GPT, we’re starting to think of like people being able to just do natural language queries that translate, perhaps more accurately than we would have expected, you know, you know, down to then what could actually be something like a SQL query. And I think that’s really exciting. And like, even if you assume these no code tools are like, they’re only going to work with the low end use cases. And that’s a safer way, maybe the immediate future we see like, oh, yeah, they’ll only serve the low end, easy use cases, they won’t serve the complicated ones. The truth is, like, so many of these questions that marketers have to come back. So you are in that, like, you know, it’s not that it would be that hard to answer that question. It’s just having some other individual like, go and, like, drop what they’re doing and try and figure that out for you is just a really expensive task. And, you know, for the marketer, then like, you know, oh, well, I have to get a ticket and be in a queue. I mean, that sucks. I wanted to make a decision, you know, tomorrow I was gonna do this. And so, anyways, the short answer is I think today you see the marketing ops and marketing tech teams and marketing help with that. And they are tech savvy, like, I think you can have higher expectations of them. But I think longer term, that frontier of what we’re able to enable marketers to do is really the greatest hope, and being able to deal with the delusion.

Kostas Pardalis 45:08
Yep. Yeah, it makes a little sense. All right, one last question from me. And then I’ll give the mic baktun. to Eric. So in this today, like landscape with all these products, if someone is interested in going in, like building new technology for Martic, for marketing rights, what you would advise them to look into, like, from all these categories that you mentioned, like where do you what’s like, makes you more excitements in terms of like, this is right for either like this eruption, or there are things happening here that are going to be interesting in the next like, two or three years?

Scott Brinker 45:51
You know, the fascinating answer to that question is, I actually see this happening across almost all of these categories, because when you really dig into it, frankly, a lot, the technologies here, they’re pretty long in the tooth, you know, I mean, at least by modern standards, right? They’ve been around 10 years, some of them are 20 years, you know, in the world has changed a lot. And what’s possible, you know, and so like just one example, I’ll give you like, you know, like how we think about sales engagement, you know, so All right, well, we’ll have these tools for the salesperson to run their deal management in these stages. And we’ll do that as well. I just came across a company, actually, they just came across it and have been around a couple years, but like, they’ve created a tool, that is essentially a two sided sales engagement tool. So it has an interface for the salespeople. But it actually has an interface for buyers. And it allows buyers to, like manage the process, and like, you know, set up the deliverables the way they want from the salespeople. And so this is actually this incredibly creative way to like help both the salesperson and the buyers, like run the process of the evaluation and decision the way they both won, you know, under a lot more control on their, you know, and it’s like, nobody did that before they wouldn’t even really possible, you know, and I could go through the entire martech landscape, and you still see just creative entrepreneurs saying, well, listen, we’ve always done it this way. Yeah. But I’m sure we have to keep not doing it that way. I think there are better ways to do this. And that’s one of the things I still, you know, I really love about martech because of the people dimension more than the tech. But if there is an element of the tech that I have a soft spot for it is these creative entrepreneurs who are saying, like, you know, the old way, we used to do this, we don’t have to do that way anymore. There’s a better way. You know, and they usually when they first showed us, like people laugh at like, gods, nobody would ever do it that way. That’s crazy, what you know, search for stuff, you know? And then next thing, you know, so yeah, you’d bring your imagination to the field, there’s lots of opportunity.

Kostas Pardalis 48:06
Yeah, that’s the most important part, I think there is a lot of opportunity in marketing. So Eric, what do you think? Are you starting the new Marketo? for like the next big eight?

Eric Dodds 48:18
I agree with you, Scott, I think I actually think that what will enable some of that innovation is that I think a lot of the burden that some of these marketing tools have had to carry on the data layer side will be removed, right? And so because, you know, if you think about processing, data models, even making recommendations, and all of that happening, sort of inside of this tool, if you have a tool that is, you know, let’s say, you know, trying to decide the optimal time to, you know, send a message or, you know, show an ad or something like that. And that in its own right is a phenomenally difficult problem to solve with software. But then also to have to basically build the underlying data infrastructure to power that means that really, I mean, this is just some speaking from my experience, having used a lot of these tools in the past is that neither one is awesome, right? Like, you know, kind of does a decent job with the data and it kind of, you know, does a decent job on delivery. But it doesn’t do either one of them incredibly well, right. And so if you kind of think about, to your point, Scott, if this stuff gets pushed down to the data layer, and you have the infinite flexibility, you have all these data tools under the hood, and then these tools can just consume from that and be really good at that one piece. I think. That’s exciting. I know having the next big idea there yet costs us but I know Scott, would you agree, disagree? 100%?

Scott Brinker 50:04
Yeah, this whole standing on the shoulders of giants. I mean, you know, again, actually, we didn’t talk about it much. But you know, this explosion of all these software tools, part of it was because of the explosion of demand for all this specialist demand, but it was perhaps even more. So the enablement of things like, you know, AWS and Azure and Google Cloud. That’s just my goodness, like, the shoulders that someone can stand on now to build out stuff is incredible. But you’re right, you know where it is today, compared to where it’s now evolving very rapidly on Oh, well, what if he just had access to all of this data, you know, in like, you know, a highly performant way? Like, oh, what could I build with that? And to it just, it’s better on edge? Like, how many things do you come across in life that they get better at, like, all of these multiple dimensions? Like it’s, you know, better from a cost perspective, it’s better from a you know, like specialization of like, who does what really well, it’s, you know, better from a, hey, getting all this stuff unified. So we break out of these, I mean, just, you know, I would not downplay the complexity and the hard work that’s still involved in doing this, obviously, this is what you guys do for a living, you know, better than I. So I know, it’s still a long road ahead. But it feels like the benefits that are going to come out of this progress are just amazing. And I can’t help Sorry, just want to plug this one thing I know, everyone talks about these days, but all the stuff with AI, you know, again, it’s great to see things like GPT, and, you know, sort of this next generation stuff happening here. But all the real power of AI is predicated on us being able to feed those engines with the right sort of data at the right time. You know, and so the fact that we’ve got these AI engines that have really now, you know, crossed the chasm in their capabilities, if we marry that now with like, you know, truly unified data across, you know, the orange, it’s hard for me not to be excited and feel like, yeah, the past 10 years has been a really exciting, innovative time in the world of marketing and technology. But what seems like is ahead for the next 10 years, I think it is gonna make that look like kindergarten.

52:19
I totally agree. I mean, I think, I think many marketers would agree that, like, claims of game changing AI within a piece of marketing software over the last 10 years have been, you know, I mean, I’m not their husband, some cool stuff, but you know, really hasn’t lived up to the promise, I think, because of what you’re saying. Right? It actually was a problem that needed to be solved outside of the tool, and that the tool can, you know, sort of access and leverage.

Eric Dodds 52:49
Yes, I am very excited for Scott, we are at the buzzer here. This has been an amazing conversation. I’ve learned a ton. And I know our listeners have as well. So thank you so much for giving us so much of your time. Yeah,

Scott Brinker 53:03
Thank you so much for having me.

Eric Dodds 53:06
Got this fascinating conversation with Scott. As I said, personal honor for me, because he was the one who sort of originally got me thinking about the data stack, of course, from a marketing angle. One of the interesting things that I really appreciated about this conversation was the discussion around local optimizations, you know, that’s sort of something that you if you work on a data team, you know, and I actually, you know, in my current job, have the, you know, I sort of have one leg, I work on the team that, you know, manage the data stack. And then I also work on the marketing team. And so I get to see both sides, but it’s really easy to see how you can have a limited view on a data team of like, what a marketer is trying to do downstream, and how for them, data can be such a rate limiter. And many times it’s, it can even be difficult for them to know that. Right, they just experience more and more difficulty in trying to do something in their job. And so, you know, a theme that we’ve had on the show a ton that Scott got back to, especially in thinking about that relationship between sort of like technical teams, and downstream teams, like marketing is empathy. And then we also had this really interesting conversation about, are those two things actually just merging together? And what does the future of that look like? Which was really interesting to hear his perspective on your sort of team structure that’s flowing from deeper integration in the technology stack? So yeah, I thought it was a really helpful conversation about you.

Kostas Pardalis 54:47
Oh, yeah, it was first of all, okay. It was great to hear from you Mike. How this Martex landscape looks like. Like,

54:58
if you have seen all these maps out there. He’s the right person to narrate the map, right? So it was great like that. I mean, I understand now I understand like, much better was like Martha keys rights. I found it very fascinating, this whole conversation around the evolution of the relationship between marketing and technology. Right now we have marketing ops people and revealed in like, how does this like change?

Kostas Pardalis 55:29
Like, the belief that we’ve had so far about technical, let’s say, skills, like marketing is hard, right? And finally, one of the things that I found, like, probably the most fascinating one is like, you know, the landscape and the mark like industry of like 10,000, vendors, something like that. So I can have, like, a crazy number. How she convinced me that probably today there’s like, is the time for advance to go and disrupt this industry? So I think, I don’t know, I feel like I got a lot of inspiration, together with a lot of knowledge, when it comes to marketing technology from humans. That was amazing.

Eric Dodds 56:20
I totally agree. And he gave some great examples of ways that people are innovating, like the platform you talked about that allows a sales process to be sort of transparent between the buyer and the seller, you know, some interesting things like that, that are approaching problems in new ways. So definitely a great episode. Thanks for listening. Subscribe. If you haven’t, we’ll catch you on the next one. We hope you enjoyed this episode of The Data Stack Show. Be sure to subscribe to your favorite podcast app to get notified about new episodes every week. We’d also love your feedback. You can email me, Eric Dodds, at eric@datastackshow.com. That’s E-R-I-C at datastackshow.com. The show is brought to you by RudderStack, the CDP for developers. Learn how to build a CDP on your data warehouse at RudderStack.com.