Name: #MartechDay 2025 Keynote (Part 2): Evolving Martech Stacks
Uploaded: 2025-05-06T17:08:08.765Z
Duration: 52 min 4 s
Description: #MartechDay 2025 Keynote (Part 2): Evolving Martech Stacks

Transcript for "#MartechDay 2025 Keynote (Part 2): Evolving Martech Stacks": Alright. We are back. And, boy, hopefully, you you enjoyed. Got a lot of insights, out of the the latest data on the landscape, and our AI and Martech survey. But, actually, now we get to, like, even some of the real meat of what we've been looking forward to share with you. Right, Brian Brian? Yeah. Absolutely. I can imagine after, you know, almost drinking water from a fire hose in the previous sections, it's, you need a deep breath. So I hope you're catching your breath, but there's so much more we want to share, so many ideas because this is the big question. Where is it all going? What do you need to do differently tomorrow back in the office, so to speak? So, yeah, we're trying to do our best here and and and give our perspective. Alright. Well, let's, let's start by talking a little bit about the structure of the stack. You know, I mean, the AI survey, we covered, obviously, there's a lot of AI technology that it's entering into MarTechDay. You know, if agentic and agents are the things that everybody is talking about, what does this actually translate into the evolution of the MarTechDay stack? Is it just like all these existing applications are going away and it's all moving to agents? And the short version in our opinion is no. We are entering what, is sometimes called a liminal period, where you have these overlapping, errors of technologies, capabilities, as you're going through a transformation. And so what we are seeing in more and more of these MarTechDay stacks by the way, if you come to the stack, you stick around at the end of the day, you'll get to see some real world stacks, that echo this, you know, is the core MarTechDay applications and platforms are actually still as relevant as ever. In fact, for a lot of the things that you see agents doing, they are looking to those systems as either the sources of truth, the sources of knowledge, that they're able to draw upon, or even in the case of more of these classic engagement platforms, that they're really the systems that help orchestrate and govern what is possible with these agents. So, this there there there's a lot in this diagram. We're not gonna go through it all now. This is one of the things that's deep into the report. I think the thing that I would wanna share with you is we've started calling these things instead of systems of record and systems of engagement, really try to reframe it as, like, systems of knowledge. Right? Because with AI, it's this ability to synthesize knowledge, structured data, unstructured data, streaming data, you know, that really is the fuel feeding so many of these AI. So even have making sure that these systems, you know, are solid is it it's just essential to be able to really implement the AI capabilities you would want. In our conceptual model here, your stack may differ. Your stack may have fewer products. Your stack may have more products. This is just to place things in a way to show how they relate to each other. We do see the cloud data warehouse lakehouse is increasingly that foundation, you know, for storing data, but also taking responsibility for being able to distribute it and hydrate it, you know, across the rest of the stack. When you get to systems more like classic systems of record like CRM or CBP or DAM, you know, they're still systems of knowledge, but they start to take on more of that responsibility of being arbiters of truth. They're the ones that are putting the rules in place like, okay. Is this data proper? You know, what is the, you know, if there are competing, you know, records, record elements, like, which one wins out? Super important. But then these systems also start to blend into the systems of the classic systems of engagement, in the MarTech stack. We're thinking more as systems of context. And the reason why you start to reframe it in systems of context is, certainly, you still have employees as well as customers who interact directly with things like your DXP, you know, with your marketing automation or customer engagement platform. And they're in the business of helping to provide the right context, you know, to those users, be they, internal users or external users. But it's the direction where things are headed with more and more of these agents, particularly agents that can be spun up on the fly, that they really adapt to the particular context in which a user is trying to get something done. You know, for years, we've talked about the jobs to be done. Framework thing is a really great way to be able to understand, okay, what is someone actually trying to do here? Whether they're employee, a marketer, or, you know, a prospect or a customer. And one of the things that's really exciting about these agentic interfaces, these agents, is they really are having more and more ability to adapt to the specific context, of those users. So the way we sort of see the structured framework is you have the core systems of knowledge, which are mostly existing, platforms, systems of record in the cloud data warehouse. Above that, these levels of, you know, like, the core systems of engagement that still are serving as largely the centerpieces for orchestration and governance, of those engagements. Some interesting things happening between those two with AI decisioning, and, Taja is, from High Touch, when we interview him, a little bit later here, we'll, pick his brain a bit about the AI decisioning as well. And then these agents and AI processes that surround it. And I think this is just what makes things really interesting here is you're going from a world where we've always thought of these Martech applications as relatively fixed in their interface and their operations to now also augmenting that stack with more and more of these dynamic, AI agents and process automations that spin up on the fly. Yeah. And and this one, just like the the previous one, if you look on the left side, it's more fixed. It's more data. On the right hand side, it's more dynamic. It's more services. And I think the more you go to the right, the more interactions with the customer, with the world outside world you have. And it's so important to explain to the board that this is different in Martech because they are used to finance systems or ERP systems, and they have an API and that integrates with the outside world. That's it. It doesn't have to converse, you know, no conversations whatsoever. So that makes this this a completely different game. Yes. It's software just like ERP, but different completely because you need some kind of, you know, layered approach like you showed in the previous slide. I love that, and this one as well. It's a layered approach, and we don't know if this layered approach is even a roadmap for the next ten years, how AI will replace everything or maybe not because some of the people said, you know, down there, this is your content data repository, and that's where you will have your somebody called it DeepSaaS in one of our conversation online. Remember? I love that. I love that too. So that might end up, you know, fastest commercial software. So not everything has to be bespoke or, you know, why why build something yourself if there's brilliant stuff out there? And on the green side, on the upside, on top side, you see more what people some people call micro SaaS. So you build your own stuff or you build you buy something from the long tail. And and this is yeah. At the heart, you see composability working like this, and you need to have some kind of picture like this, and it's never done. Let's be very clear because we keep get getting the question, okay. What is the best today is this. Tomorrow, it might be something else, and you don't have to completely overhaul and maybe not today and tomorrow. It's more like today. Get it right the next one or two years, and then the world changes. But if you get it right in two years, you're already outperforming competition. So I love this, this overview a lot. No. And it's exactly this is the reason for composability. Is because we know it's gonna change. We just can't predict exactly how it's gonna change or when it's gonna change, you know, but if you invest in an architecture with, designing for change in mind, that's really the best way to set yourself up. One thing to note it here, because, yeah, you're talking about more and more of the conversations, you know, that, we have not just inside our companies, but also with our customers. We're so used to thinking of the MarTech stack as applications and products we own. Oh, yeah. This is our stack. We bought this. We built this. We plugged them together. Way up in the upper left corner of this, you know, the thing we even put in yellow to, like, say, there's something a little bit different here, is you're starting to see more and more AI agents that are being run by buyers, and they're having those AI agents interact with sellers on their behalf. And this starts to get us into a really interesting world where, okay, it's not just, human sellers and human buyers, but it is these AI programmatic sellers, AI programmatic buyers interacting. And while they're not technically a part of your stack, they're much more likely to interact with your stack in programmatic ways. And so, this is gonna be one of the things we're gonna cover in a little bit more depth because it really is like it it turns on the head, you know, the possibilities, of what marketing is and how it works. You know, if you take that sort of seed of the Google bot and the SEO and, like, alright. We always had a certain amount of optimization for that. It's like orders of magnitude of, like, oh, wow. A whole universe of potential software agents and programs that now become a conduit in an interaction an interface, for us to our customers. Yeah. When when we talk about this, we get some pushback from people saying, yeah. But, that that we don't wanna live in a world where AI agents from the buyer side speak to the customer facing or the seller side. That will be mayhem. That's not what we're suggesting. We're basically saying marketers become more of moderators rather than operators, so we don't have to do automation flow plumbing. We can now orchestrate and say, okay. Within these boundaries, you can talk to an AI agent from the buyer side. And those, you know, boundaries or guardrails are never fixed because language changes, needs change, markets change. So you will be moderating your way through the years and making sure your LLM is still answering within those boundaries and those agents as well. And I think that will be the game. So it's maybe less tech and more, conversation, but maybe it's wishful thinking. I think it's coming pretty fast. I think so. Yes. And, yeah, stacks are composable. That's not the wish. It's it's what we found. I I I shared it earlier in our first report or report last year. Sorry. We saw stack composability is a thing. Tool composability, yes, but it's also stack. It's LEGO blocks. And moreover, we saw that companies on purpose duplicate features and capabilities because in one case, they need to test out new things and don't want to harm the internal core stack, and they also have the same features in the core stack because those are serving a different purpose. They're still sending out maybe an SMS, but here it is generating revenue in an existing, customer journey, whereas we're trying out new stuff, and you don't want to, harm the the the workflows and the customer journeys that are bringing you the revenue. So what we saw is what we started to call something like a solar system. We spoke to a lot of people, and then people said, yeah. It's like a solar system. So you have a center platform. Like we said earlier, in b two b, it's more CRM and marketing automation. In b two c, it's more cloud data warehouse and CDP, as a rule of thumb. And then around that, you will have some specifications, some orchestration platforms, you might call it, some planets. You know? And those are also tools that you, need to have an owner for. We often don't see that enough, and somebody owning not only the features, also the data inside and making sure they have the right integrations, the right data integrations. And then finally, you have your activation, which is the outer rim, so to speak, of your galaxy or universe or solar system. And and that's where we see a lot more experimentation, replacements. And once it works, now we're gonna look into the planet or the sun to harness it and to embed it forever and make money, you know, almost like money printing machine. I'm exaggerating, but that's that's the idea of a solar system. So we looked at how does that work? How how do these capabilities basically, enter your stack constantly? And and it's a lot of experimentation that we do in marketing, which is basically hacking and growth hacking, and those are a lot of one off experiments. And and once we do them, we see some patterns and then we can standardize them. I call it packing. So hacking, packing, and then finally stacking, and and that's where IT comes in. So the hacking is marketing, the packing is marketing ops, and the stacking is IT. And we need each other. That's so often that I enter a a kind of a boardroom and then people are saying, marketing is using these point solutions. Sorry. We need to. We have to find the revenue of the future while you're protecting the current revenue. And this is how we see, Stacks working based on all the research that we've done. Now AI comes in because we did the research, in this, survey, and we were looking at, okay, how is AI affecting the fabric of composability of a stack? Well, AI is eating the stack. It's basically, you know, eating its way outside in. We're not saying it's replacing everything. Just like the first graph we showed in the perspective section here, it's not. We don't know. It could be a road map. It could be something hybrid forever. But we will see. And, we're really, interested to see where it is going. So with the Gen AI use cases last December, we had to look at what are people using, more the outer rim, and now we're looking at AI composability, the inside. And this is how we think AI is kind of eating the stack or contributing or helping. Sorry. I was just channeling Mark and Andreessen and trying to look cool. No. No. Hey. Software ate the world and, yeah. AI, I AI is eating software. Although, I mean, again, to your point, it's it's interesting because it's not that it's, like, replacing all software. They're definitely new AI technologies. I mean, AI is software. But it's also yeah. You you saw it in the AI, and composability survey, data that we were sharing earlier that, like, yeah. Actually, many of your, like, major martech platforms, they're some of the leading implementers of these AI capabilities. So, certainly, AI is gonna proliferate every facet, of what the MarTech stack is. How that shakes out over the next, let's say, five years, you know, in the balance of, okay, what platforms are playing which roles within that? Yeah. I think this is one where we're like, okay. We we sort of see the current state probably gonna be relatively stable for a while, but what this looks like five years from now, yeah, the, the future is not yet told. No. And and that's regarding the technology we're showing here, how the stack looks like, how it's supported, what type of technologies. But what does it mean for your job? And that's a question we asked ourselves. So, I I just mentioned hack back and stack, and and we've shared it, in previous, MarTechDay days. I just wanted to highlight that we have to maybe think differently about customer journeys and and experimentation and, experiences. We have to prioritize journeys. It's not about tools. It's about, you know, the experience we create for people. So maybe we should kind of prioritize marketing as we call it. So just acknowledge there are some customer journeys that are making you money. That should be in the stack. That should be harnessed. And mind you, I see a lot of companies that haven't done that. So if you replace your credit card, you sometimes end up in weird processes where you have to do strange stuff online, offline, create extra logins. That shouldn't be happening. That should be humming and buzzing in your stack. If that happens in your hack phase because you're trying out something new, it makes sense. You fail 10 times, and then maybe one time you you struggle, and then you make sure you pack it and put it in the stack. And and I think these are more the capabilities, the skills of the future that we should work on. Anything else you would like to add about the the customer journey, Scott? Well, okay. That is actually a segue to something that actually isn't in the state of MarTech '20 '20 '5 report. So this is an exclusive, for you who have actually tuned in, to this keynote. We did some research, earlier this year, sponsored by the folks at PathFactory. We're really looking how generative AI is changing the buyer journey and the buyer experience. And one of the things that is fascinating, I mean, talking about these journeys is for so long, marketing has kind of thought of itself as the owner of the journey. Right? You know, we have journey orchestration engines. We have these maps we do with nice flowcharts, you know, of, like, how we like, this is the journey. We are the owners of the journey. And I think at some level, we always recognize that was a fiction, that, but, actually, we're not in control of the journey. The buyer is in control of the journey. But one of the challenges for us was, okay, well, the buyer is in control of the journey, but we're the ones in control of the experience and the tech. So we're, like, always trying to triangulate how do we get, like, you know, the definition and the structure of the experiences and the engagement we can deliver to best match with the buyers, who yeah. Their actual buyer's journey. I think the scribble you did here on the right is probably, an a simplified version of what that journey looks like. Our our good friend, Tom Fishburne, hopefully, everyone here, like, subscribes to MarketDoonist. Yeah. It is it is the humor that helps find the sanity and all the insanity that is modern marketing. I think this is one of your favorite. Right, Frans? Oh, I love this. I use this one, every now and then because on the left hand side is really how we have been doing stuff so far. The linear line, rule based, top down, and and somewhere lingering understanding it's not my journey, it's the customer journey, but we'll call it customer journey, but I will tell you what to do next, and here you go. So you're here, so your next step should be this or that. And then we miss the point completely. You know? On the left hand side, you see more of what I would call, the context. You know? We do understand how old she is, what she's doing, but but then the actual need or intent is something completely else. So, actually, I'm just looking for the bathroom right now. I'll be back and then maybe following your funnel or whatever. And I I just this is so hitting the nail on the head, exactly of what is going on. And I you might even make a cut and say on the left hand side, you see what we have been doing so far, and on the right hand side, you you see what AI is adding because you saw that graph that we shared earlier where we have all this semantic analysis of all the different, inputs and data sources, and I love that. And, yeah, recently, we had a a a webinar together where I said, it feels to me like Jenny and I switched on sound in a silent movie, and I think that's also what you see reflected here. Yeah. Well, so if our mission had been to define the best possible journey as we could, because we had to be the ones to implement it, and then try and, you know, however we could decode where is this particular individual at to map them into the experience that, yeah, they should be here. Yeah. You know? And, hopefully, we had a place for them to go to the bathroom. And, you know, so this whole field, you know, I we've sort of largely come to refer to this as intent. And we've been at this pretty much from the beginning, of the web. You know, the digital marketing journey, you know, early early days, it was about page views. There was a phrase quite a decade and a half ago called digital body language of, like, okay. Well, when you can see which pages people are going to, that's, like, revealing some sort of intent of what they're interested in. You know, with form fills, right, particularly in the whole landing page explosion that started, you know, in the mid two thousands. Like, oh, yeah. We can ask them fields, but turns out people don't like filling out forms. So, you know, they only reveal so much. So just even sending in a form, we almost treated that as, like, this, epic moment of, like, submitting a form. This is a massive intent unlock. Some of the work that I did, you know, before I joined, HubSpot, I was a cofounder and CTO of a company called Ion Interactive, that build a platform for interactive content where we'd actually design experiences to let people make choices, you know, whether calculators or the expert help systems or things. I'm, like, trying to understand. Sometimes we call this sub zero party data right now where the customer is telling us very explicit choices. And, again, that got a little bit closer. There's a richer level of intent, but still, right, we as the marketers had to be the ones to design those experiences. And so they were constrained by the way we viewed the world, not necessarily the way the customer did. And then probably over this last ten years, you know, there's been a lot of innovation around combining first party, second party, third party data. You know, hey. Can we look at the signals of all the other different places where we can get some visibility of what the customer prospect is doing and try and triangulate that for intent? And don't get me wrong. There's been a tremendous amount of innovation and value created through this maturing of intent capture and content intelligence. But still, it was largely guessing. It was us trying to figure out based on these, you know, the digital exhaust touch points, what is what's the customer actually want? And this is the thing that if we were diving into this more with, you know, these, buyer facing agentic interfaces, you know, and that's where the chat bots on steroids, if you will, is oh my goodness. We no longer have to guess. Like, the customer is telling us explicitly, oh, yeah. This is what I'm looking for. Okay. Well, it's coming this color. I need this thing. How long is it gonna take to get this? Oh my goodness. Like, the questions, the, requests these customers have. This is the buyer's journey. And what's amazing here about where we're headed with these AI things is, first of all, just being able to, like, get that direct feed into the brain of the customer. What is it they actually are looking for? Where they actually have questions when they have actual concerns? And being able to use that as the intelligence that we activate on. I mean, again, we're just barely getting started with this, but, incredible opportunities. But it's better than that because there isn't necessarily as much of a delay. You know? The more we are able through our composable stacks to make sure that these buyer facing agents have access to the right data, unstructured data as much as structured data, that they have access to the services from a composable stack to be able to take actions that the customer is looking for. You know, what is the status of my order? And to be able to, like, feed those up so that, boy, not only is the customer able to express directly what it is they're interested in, at an increasing percentage rate, we will be able to fulfill that request immediately, instantaneously. And so I don't know. I I I I really could not be more excited about what I think is, like, a fundamental shift, in how we think about managing marketing experiences with this technology. Yeah. And what what is so great about this graph is it's showing that intent was always a thing. It was just so hard to understand what a client really wanted, go to the bathroom. So that is what you see here. And I think with Jenny, I I agree with you. It's it's boosting that because it makes it so much more simple. You need an expert data science to figure all that stuff out, and then finally, you know, you you're trying to create an automation flow rule based to answer. That's very hard. And right now with agents interpreting and responding almost real time, that is a game changer. So we're really turbo boosting a lot of what we're doing in in marketing. And, again, it was all about intent, what do people really want, but now we're starting to boost it with more meaning adding meaning to to the analysis. Yeah. This is one of the graphs we shared in the last report, in December, and it's it's showing kind of the the infrastructure, the the foundation of communication. How do we interact? So what elements do we have in terms of capabilities? But also you could put logos in here, or maybe teams or skills or what have you. What in general, we were talking about, okay, what is the skill of the future that you should grow? We already are really good with data skills, and that is hard enough. That's the blue, line here where you see data sourcing, quality analysis, activation. Those are data skills. And still, I see that many marketers struggle with data skills, trying to figure out why are they going from here to to there. I I worked with a company and they said, we know how our customer journey works. People first go, go to an event, then they go to a webinar, and then they go to the product page, and then they buy. When we looked at the data, and it was just marketing automation data, so really like tags and and it was Salesforce that we found they got their journey upside down. People first went to a product page, and we didn't understand why. We didn't understand the intent. And it took a lot of data crunching to find out that they wanted just an upgrade of the previous bought product and just find a product page and then, you know, oh, and then they found out they missed out on a lot of developments and decades of of innovation. Let's do a webinar. And then they went to an event, which is sounding weird, but they just said, I just want to know if it's the same company, same type of people, same length, and that was it. And I heard this many, many times that they said events is the last step before the purchase. Interesting. That is intent right there. So they had all their communication backwards, literally. And I think the intent skills is something we have to grow. So what is intent? It's understanding what the customer wants to achieve, And you cannot only do this with circumstantial evidence like clicks and age or or geolocation. You need more than that. You need to understand language based signals. And that was very, very hard, because it really had to go into the database and then look on some wordings and keywords and what were those. This is now so easy with Generative AI. So what are the data sources? Intent data sources are very different from context data source. By the way, mind you, I probably disagree with myself in a couple of months, weeks, hours, no, years from now. But we have I'm I'm trying to figure out, you know, where is this headed. So you have the data layer and you have the content layer. They both have a different dynamic. So the context is more where I think you have cookies, pixels, first, second, or third party data, whereas in the intent data source, and we saw the graph we presented earlier on, the input through chat, dialogue boxes, forms, like Scott was talking about, the comments. Oh, boy. I I saw one of my customers having a form where you could fill out stuff and leave a comment. They didn't need to look at the comments. When we turned looked at it, we found that twenty four percent twenty four percent was straight asking for a offer, an estimate. They haven't seen it. So and I think that is not going to happen if you have better tools to interpret that data, process it at scale, and then send it to the right, sales rep. So I think this is what we see, will change a lot. It is exciting stuff, and I feel like we moved from one exciting topic to, the next. So we we prefaced this earlier that, yes, there is the MarTech landscape, and the MarTech landscape is massive. You know, and over 15,000, hundred x growth, you know, over the past decade and a half. And so many different things in there. Right? I mean, you have, you know, major platforms that are public MarTech companies. You have hundreds of, like, leaders in particular categories or verticals or regions, you know, and that's that's incredible long tail, you know, of more specialized tools that might be tied to a particular ecosystem. They might be a product that a service company has put together. You know, early stage startups, late stage startups. It's just a incredibly diverse, set of amazing products that get created, but still, it's almost like the tip of the iceberg. It is the tiniest fraction of the amount of technology that actually exists in MarTech stacks. There's always been this notion that, yeah, there's commercial software, but then there's also custom software. Now for a while, the commercial software was really taking the lead on that. Right? I mean, let's face it. With 15,000 things to choose from, you know, hey. I wanna do x. Somebody's probably already done x. I don't have to build it myself. I'll just buy it. Because building your own custom software and maintaining it tended to be pretty hard. Right? You know, like, you need to have software engineers. It wasn't just about, you know, building the initial product. It was like, you know, is this gonna be maintained? And so while, you know, the larger the company was, the more likely they probably had some custom software, you know, but martech stacks in general leaned very heavily on these composed solutions of, commercial products. But about eight years ago, we started to see this rise of what we now call low code, no code platforms. They just actually were reducing the cost and the expertise and the time required to build more custom solutions. You know? And we're starting to get to the place where, you know, yeah, there were, you know, you've looked at the inventory of these, the stats from these no code platforms and the number of different little mini apps that people were building. You know? And it was, like, millions and millions, and you're like, alright. You certainly see this proliferation. It's not like people aren't necessarily building their own CRM or their own map, but they're building these very, like, you know, custom things that do a particular task, a particular job to be done really well. What's been fascinating here is in these past couple years with AI, it has been an incredible accelerant. There are so many of these tools. If any of you have, we we talk a bit about these in the report tools like Replit and Lovable and Vercel v zero, that you literally if you can describe in natural language in these text prompts, oh, I I want something that's gonna do this workflow like this, or I want something that's gonna build a little website or a little web app with this. You describe it, and it actually builds it for you. There's this very famous, AI computer scientist who a couple years ago, said English is now the hottest programming language. You know? And so we started to even see this thing of, like and not just IT built software, although that's being accelerated with things like GitHub Copilot, but more and more of these citizen developed software apps that, hey. If I can sort of describe what it is I want, you know, this platform can, like, build it out for me. And then you even get beyond that where, oh, wow. You can now have these things where you ask an AI agent or an AI assistant to do something for you. Behind the scenes, it actually maybe creates a Python program to go and do it, runs that program for you, and then gives you the result. And you as the user may have had no idea that any software was actually built at all. You just made a request, and you've got the answer and the outcome on the other side. And that's where we sort of have this, like, break in the in in the x axis here where we're like, yeah. No. We're probably entering a world here where in reality, it's not just gonna be millions. It's gonna be billions or trillions, you know, of these little custom apps that proliferate many of them being spun up on demand to do a particular job and then, you know, blink out of existence. And so I think this Pratham, I've heard turn this back over to you, but it's like, to me, this feels like this is a fundamental change in what is possible with MarTechDay and MarTechDay stacks. And it doesn't mean that the commercial stuff is going away, but, boy, the flexibility of what we can build around that now feels like it's just exploding. Yeah. A %. And maybe we're we're stating the obvious, but, when you see the growth in tools, it doesn't mean that every year, we're adding, like, thousand HubSpots or Salesforce. It's the small tools that I call it atomization. So smaller tools that do one task really well. Well, that's Hypertill territory, I think. There are not many, tools in the Hypertill that say, you know what? We'll completely rebuild the CRM. If they would, and I've seen it sometimes, they only build two or three really key capabilities, scenarios, capabilities. So what I think will really change is that a stack is a set of tools changes into a set of capabilities, and that is the quest. Now it's open. What capabilities bring value to your customer journeys? You don't need to cater for all customer journeys with all data points, with all integrations, that's boarding the ocean. You don't have the resources, the time, the money for that, and why would you? It's a lot of work. So, what I would like to to see or or say is that stacks become really sets of capabilities, and you need to understand the capability can be a requirement feature, but but also a skill. Sometimes you don't need to automate something. And how do you do that? So what I would like to see more and more, and we discussed about this earlier, Scott, is is maybe you don't need a backlog of features. You need a backlog of experiences. And who owns that experience? And then the experience that is profitable, of course, and profitable means good for the customer and good for the company. So who's owning that? And many times, like I I just explained this credit card renewal, nobody owns that process, you know, and it's broken. Fix it. That's your core business. And this is not a personal event that I can start with credit card companies. That's not what I'm saying. But I don't know. You've brought it up a couple times now. I look first I've been really long in your banking experience lately. I'm I'm just guessing. Yeah. You see, I'm wearing my clothes for the last three years. Same ones. No. I'm kidding. But the thing you will find and you will know 10 examples in your own environment, credit card like or whatever or your bike, your car, your insurance. I don't care. So I think that is the set of capabilities. Raise a sharp focus on that customer journey. Make somebody own it, across all the teams, the capabilities, the features, and then make sure it's profitable. Maybe kind of wrap ups, type of thing. I don't know. But it I think that that might be changing in the in the future. I don't know. What is your take, Scott? Yeah. No. I think, it is fascinating to think about it both as a set of capabilities, and in which case, some will be commercial, some will be custom that you build. But then that really does frame this idea of, oh, a lot of these other things that we can whether they're agents or things like we just request on demand, like, they sit at a level above that, that they can dynamically generate whether it's an experience or particular, like, workflow internally or whatnot. Like, they don't have to be permanent. They can be something that, like, oh, yeah. I've got all those capabilities and in a particular context, we come back to systems of context. You know, in a particular context, just let me create, get the combination of data and services I need to execute it. I'm done. I'll move on. And if I have the same, thing the next day, okay, we can reuse it. If I have something slightly different the next day, that's okay. We create something, you know, tailored. One would even say hyper tailored, you know, to your particular, context. And I think, correct me if I'm wrong, this now actually, we've been teasing this, you know, for the past hour and a half. But I think this finally sheds some interesting light, you know, on this rise of product management in the MarTech stack. Yeah. And it's it's it's again, it's hiding in plain sight. It's really amazing that this the tool count is so stable. So these tools are really doing a good job in, increasing the rating, but also, sales. In this small book of customer technology, we also did some analysis on revenue brand view ratio for these vendors. And, actually, for product management, it's a very healthy one. So these are well managed companies. We'll maybe someday dig into that why. Maybe we do an internship there, but it's these companies are really top notch. Yeah. Awesome. And it certainly it makes sense that, you know, exactly as you were saying, this, like, sort of backlog of experiences. Experiences are products. You mentioned it earlier, this, like, productizing and marketing. Okay. Well, there are a set of tools that are actually very helpful for managing products, digital products, and increasingly both for internal purposes and external experiences. That's what marketing is. Alright. I know we're we're running out of time here. We've got, like, the folks standing by here on these interviews. We gotta get to the stackies. But I feel like before we hand off to the next, presenters, we should talk a little bit about, like, how the roles of marketing and people's role in this new AI era of marketing, how it's changing. And if if you're letting us sort of set up this this graph a bit, but then I I I know you have some thoughts on this that, you're very passionate about, is I would oversimplify and say that marketing had largely fallen into three buckets. There was strategy and creative, which was always what you know, when people talk about marketing, that's always what they oh, that's what marketing is, the strategy and creative. It's what we really value. We put it up on a pedestal. But the reality is not a lot of the time or resources actually went to the true strategy and creative. You know, most of the resources and time and effort went into the second bucket, which we'll just call production and analysis. This was all the work that you needed to do to actually take that strategy and creative idea and bring it to life. Lot of work. We recognized it was a lot of work, but it was never super valued because, like, hey. You know, it's not like somebody wins an award for, like, you know, best image tagger and cropper of the year. You know? But it was the necessary work. And then, of course, we have Martech and marketing operations that, you know, most people recognize as a part of marketing. I suspect most of our marketing ops and Martech leaders would say, like, yes. Even though people had recognized it, it probably hasn't also been as valued as much as it it should be just because, again, oh, market. It's a strategy and creative. Yeah. I guess the market and market now, that's some of what we need to do to just execute on it. I think with AI, something is really interesting. I mean, obviously, people are very concerned about, you know, the it's this double edged sword of, like, hey. AI can do more work for you. And then, you know, the other side of that is, like, well, okay. I get paid for that work. What am I gonna get paid for next? The fact that a lot of this work was production analysis, and in many ways, it wasn't particularly valued, in a long case, it's not even particularly well enjoyed, you know, amount of marketing, is that is really where AI, automation and acceleration seems poised to take a big bite. You know, it's just able to, like, shorten these production cycles. It's able to democratize these production cycles. It's It's able to do them faster and cheaper. And the question really then becomes, okay. As the production and analysis becomes less of a bottleneck, what do we do with our time and talent? And this is where, you know, optimist on this, but I think it's an incredible unlock for us to turn around and invest more in strategy and creative. In particular, you know, I my defense on marketers here. I don't know if marketers can know in advance what's really gonna work. You know, at the end of the day, the only way we really know is we put in a we have a hypothesis, we put an experiment, we put it out into the market. Some things don't work, we toss them on the scrap heap. Other things, oh, wow. This hits. This resonates. Great. Let's double down. Let's do more of it. And, frankly, because we can't predict in advance what's gonna work or whatnot, it's our ability to accelerate the number of experiments we run. The ability for us to have a broader vision, be more ambitious in the kinds of experiments we can produce, I think is a massive unlock. You know? And so for those companies who start getting these efficiency gains and savings on production and analysis, one option is for them to just pocket those gains and call it a day. I don't think that's smart. I think the really smart companies are gonna say, great. This is now an unlock for us to, like, double, triple, 10 x, on the strategy and creative. And then the only other thing I would say about that is in order to make that magic work, boy, you know, that underlying stack infrastructure, that is, like, you need that. You need the data dimensions of this to be right, those systems of knowledge. You need to have all the right orchestration and governance in place so that people can truly leverage these systems of context. I actually do believe that, in this AI era, Martech and marketing operations teams are finally gonna get much more of the recognition that they deserve, but I think also a lot more resourcing to be able to power all that next new, level of, like, strategy and creative. Alright. Sorry. That was a bit of a ramble. Yeah. A lot. Yeah. A %. And and it is it is truly a game changer. And if you look at how, a MarTech budgets, but also marketing ops budgets are distributed, it's it's mostly from a cost perspective. And that, let's overstate this, is gone now. Like you wrote down here with production analysis, that was hard and taking a lot of time because it needed a lot of resources, expensive tools maybe. And now with AI, you can do that in a much quicker way. So the cost focus is really like for a CFO who decides what you get in terms of budget, the cost, the license fee show up on the balance sheet, but a missed opportunity not. So that's where we always have been, you know, in a struggle. It's important. And we were you know, and we do represent future revenue, not the current one. So in times of, decline economic decline, we are hit first. Now with agents and with AI, we can and replicates and what have you, we can build stuff much more quickly. So the cost of experimentation is is much lower. So you can give it a try, and then you stumble upon something. So now maybe these these new ideas make it more quickly, and we still need a lot of ideas. It's it's a lot of serendipity involved, etcetera, but the cost of experimentation is a lot lower. And what we've been doing, what I see a lot with the more traditional companies is that they try to experiment with the core stack. So it takes nine months to implement a new module, give it a try after two weeks. Oh, that doesn't work. You know? That's costly. Right now, you can do it with an AI agent maybe or workflow or combination of commercial and AI and then play around. Once it works, you pack it and you stack it. So I think that that is the true game changer. And, yes, one of the results we didn't share, but we saw a lot in the it didn't make it through the landscape because it's not a category. But one of the the the big, increase also we saw is in, websites that suggest, ideas for a SaaS product or for a new commercial enterprise. Those are, you know, blossoming a lot. So you might argue, okay. You need that one, then you need product management, and you're you're good to go. So this whole experimentation infrastructure becomes democratized, meaning you don't need a lot of expertise, a lot of cost, sorry, budget or resources for that. That is now diminishing. And that is a big, big opportunity that you should grab, and we see the big companies or the successful ones, the outperformers, they are grabbing that. They understand that. And I think that's putting everything upside down if you're creative enough and bold enough to, you know, try new ideas and experiment. Oh, it is, it is the best of times. Some would say it's also the worst of times, but I think it's the best of times to be in marketing and Martech. Alright. We have barely scratched the surface, of the things we covered in our research that's in this report. So, again, just to remind everyone, like, yeah, you should be able right here in Goldcast download the report. But, boy, we've now got a whole another few hours of, like, amazing content for you. Up next, we're gonna have our candid interviews. I interview first Chris O'Neil, who's the CEO of Throat Loop, talking about the concept of compound marketing. How do you accelerate gains with AI based experimentation loops? Taysh Vanohar, the CEO of Hightouch, benefits of an independent AI decisioning layer in a composable stack. We barely touched on that a little while ago. Teshias is gonna give us a much deeper explanation. Greg Brunk, the head of product at MetaRouter, often referred to as a CBP's best friend, is gonna talk about data infrastructure and why that's the key to rationalizing your MarTech stack for AI. Ravi Donna from MoEngage. Very excited to talk to him about marketing, like, being in a time of reinvention, you know, all the way from idea ideation through workflow to customer engagement. Sarah Fats, another returning champion about why AI is still about humans, building for humans. And, well, speaking of reigning champions, Jonathan Moran from SAS, a long, long time supporter of, our research and MarTechDay. Just absolutely love Jonathan. We got a great discussion with him about the critical middle, bridging the gap between the data and the channel. And then after all of that, you know, the, the icing on the cake, will be the 2025 Stackey Awards. We have some amazing entries this year. We're gonna go through them. We're gonna announce some new winners. We've got a few, lifetime achievement winners. And then after that, boy, if you haven't earned yourself a, you know, a beverage of your choice, to sit back and say, wow. There is a lot of amazing things happening here in the world of Martech. You will have certainly earned it by then. So I think, yeah, just, you know, take a quick break, you know, before the next interviews, start up. But just wanna thank you, you know, for coming to listen to, get the report. This time, the report is not gated. So if you like the report and you wanna share the report, feel free to, friends and family, you know, I'll be sure to send it to all my in laws. I'm sure it will increase, you know, their their happiness and joy of what is that, Brian? Yeah. I I share that feeling. I I I will share the with my friends and family as well. But I'd like to thank everybody for for your attention, for showing up, for joining, for sharing, and see you in the next sessions. Alright. See you in a bit.