The Post-Martech GTM Stack—Built for Buyers and Agents

Julia Nimchinski:

Second… And… We are alive. Welcome to Day 3 of the Agentic Harness Summit, and today we’re exploring the FDE model, recursive marketing, the company brain, pattern-breaking in B2B, the next AI thesis. and so many more. As always, we’ve got an incredible lineup of CXOs, analysts, VCs, and founders, and can’t wait to get into it. We’re super excited to kick this day off with Scott Brinker, the godfather of Martech, analyst and Advisor, Chief Martech, and Kerry Cunningham leads research and thought leadership at 6sense. Kerry and Scott, how are you doing?

Scott Brinker:

Good!

Kerry Cunningham:

Yeah, doing great, thank you. Glad to be here.

Julia Nimchinski:

Awesome. Kerry? Take it away.

Kerry Cunningham:

Great, let’s get into it. So, one, Scott, great to be able to, talk to you again, and particularly about this, because this is, clearly… you are the godfather, so, this is…

Scott Brinker:

Easy with that!

Kerry Cunningham:

I know, it’s basically… I get the same… I get the same thing in summons, it’s really just another way of calling me old, but I get it. But we’re talking about how AI agents and all are helping transform a Martech stack, or what it means to even have a MarTech stack, and you’ve done some really great and fascinating writing about that. Recently, of course, and so there are a couple of really important ideas of yours that we want to explore. I want to start with what I think is some really important context around this. So, in the description, for this, specific conversation.

The description says, explore what replaces that architecture, the MarTech stack, as buyers increasingly research through AI, agents mediate discovery and evaluation, and go-to-market systems must adapt in real time. So… Go-to-market systems have to adapt in real time. But, as you and I discussed, as we were preparing for this, we’ve got some research, other folks have some research, that really call into question the first premise of the conversation, which is that As, buyers increasingly research through AI and agents mediate discovery. And there are some real reasons to think that that’s not the legitimate frame right now for how buying works.

And how it will probably work in the near future, at least at the enterprise level. At certain other points, we could talk about it, but I think we have to be nuanced in how we think about this. So… You know, our data shows that, one, buyers for most of their Most of their buying, 75-80% of what gets bought. are not discovering vendors for the first time, they’re evaluating vendors that they’ve known for years and years. And so the typical B2B buyer who’s involved in the decision-making process, somebody around 40 years old, been in their career from 12 to 20 years, and the companies they’re evaluating have been, too.

Now, certainly, there are a long list of companies that, if I’m evaluating 5 vendors, I may have 15 or 20 to choose from. And will I know them all? No. But in most cases, in most segments, what buyers are saying is that they actually do know the 4 or 5 that they’re going to evaluate, and they don’t go out to do discovery around them, they just gather the ones that they already know, and that’s the set that they’re going to evaluate. And that’s incredibly important because that’s a very important context, and it tells you a lot about what you have to be able to do with your marketing if you’re going to be one of those companies that they do already know.

And if you’re not one of those, then that’s a job that you have to do as well. Just the last thing, and I’ll kick it to you to comment on that, but, buyers right now say, you know, about 27% of buyers use AI for discovery of solutions. I think if you were to read LinkedIn, you would assume that would be 100%. Huh? And 27% is… it’s not inconsiderable. It was 0 3 years ago, probably. So it’s… it is an increasing number, if you want to think about it that way. But that the way your buyers are actually finding you is that they exist inside a context, already, and their professional contexts are really what bubble up new… categories, new solutions, new vendors.

And as, you know, as anyone reads what you’ve been writing, you talk a lot about that context, and specifically, golden context. So anyway, I want to throw that out, get that started, maybe put a monkey in the wrench in the description right away and get your thoughts on that.

Scott Brinker:

Yeah, no, it’s fascinating, and your point that what appears in one’s LinkedIn feed is not necessarily truly representative of reality on a larger scale. you know, if you think about AI, let’s just sort of map it to the funnel. I know the funnel’s not in fashion as much anymore, but, you know, it simplifies it. You know, so much of what’s been talked about has been the very top and the very bottom. This idea of, like, okay, yes, AEO and GEO, and oh my god, how do we get discovered, and, you know, there’s importance to that, but we’ve waited a tremendous amount of effort and calories and thinking on that piece of it, even though, of course, just like Google. you know, everyone’s like, listen, you just focus on doing your stuff well, and we’ll figure out how to surface it.

And then there’s also some really fascinating talk at the very end of the process of, like, oh yes, and then the agents are actually going to be able to conduct the transaction. You know, and some of this is happening in the consumer space, but it’s a very small percentage. It’s a very science fair project, you know, at the moment. Not to say that it won’t be here in the future, but just here today, it’s just not representative. But where you see a lot of AI truly being leveraged by buyers is in that middle process.

It’s like, okay, well, I’ve actually got a set of vendors, you know, I’m not trying to evaluate for a particular solution, who’s going to offer me, you know, the best capabilities, here’s… let me describe my problem and have AI help me, like, work through that, not necessarily the way the sales reps at each of those companies might have presented it, but the way I want to control it crawl it.

You know, this throws all sorts of fascinating things into challenge, like, You know, pricing, but particularly in enterprise, you know, used to be part of the negotiation lever, you know, in talking with reps and, you know, the exchange of, you know, tit-for-tat information. increasingly, you see buyers, they’re just turning to AI to be like, no, alright, tell me, based on everything you know and everything you’ve scanned on the internet, what should I be expecting about pricing and discounting, and then when it comes to actual terms, you know, hey, listen, don’t review this, you know, I’m not going to send this whole thing to my legal department yet, I want AI to look at it first.

And what’s fascinating is I feel like the buyers are actually much more sophisticated right now on leveraging that middle process with AI to their advantage. then you and I were chatting when we were preparing for this, that a lot of sellers haven’t yet adapted to taking advantage of the opportunity that, you know, documents they’re sending, you know, decks and this sort of stuff, you know, that they’re still designing it assuming that it’s being consumed purely by a human on the other side, and perhaps missing the opportunity to influence that middle process by Providing more content.

We’re the buyer’s AI.

Kerry Cunningham:

Yeah, and I love that, because I think, It’s probably quite true that a lot of people who might have read your documents start to finish before as part of their due diligence that they absolutely know they have to do. don’t read them, now. They get them read for them and summarized for them. Part of the context I think you put on top of that is In most categories that have existed for 10 years or so, which is most categories where there’s money being exchanged right now, that’s where most of it is. Your buyers already kind of know those vendors, or really know those vendors.

They enter the buying process with, not just some knowledge, but opinions. So, in a research that we haven’t published yet, gonna be published in, a couple of months. We asked buyers, on day one of your buying journey, how many opinions were there about which vendor your company should go with? 2 and 1.5%. Said there were no opinions. Two and a half. 70… roughly 75% said there were two or more, opinions on day one. So this is a set of buyers who’s not entering the buying journey going, oh. Please, tell me about accounting platforms in my space, are there any that I could look at?

And is there anybody who’s good? They’re going there, and there may be 10 people in the buying group, and 4 or 5 people with, you know, really strong influence in that buying group, finance person, IT, or whatever. They go there saying, hey, I’ve used these 3 or 4 in the past. Other person says, yeah, I’ve used these 3 or 4. They take the overlap of those things, and that’s the shortlist on day one, and there’s 3 different opinions about which one’s the best. And each one of those people is going to spend the buying journey trying to convince the rest of the buying group about doing the thing that they wanted, you know, going with the vendor that they like, and why should they do that?

And you think about that middle part of the buying journey, where buyers are using AI, is not an objective process of handing content to the LLM and saying, please summarize this nicely for me. It’s… I really like Vendor A, can you give me a table that shows how vendor A compares to the rest? Right? And lo and behold, Claude says, yeah, vendor A, wow, awesome, here you go, you know? And out that goes, and then that goes as, as, content that goes to the rest of the buying group, who takes that and puts it in there, Claude, and says, yeah, but I like, and can you, and then on we go with the discussion.

I think there’s a lot more context, in a way, than there’s ever been, and how do we understand that? And I was hoping you could explain your concept for people who don’t know it, of this… the golden context, and the components of that, and how we think about AI is helping us there.

Scott Brinker:

Yeah, no, no, and thanks for that. You know, I joke a little bit at the beginning of the year, like, okay, context is going to be the word of the year in MarTech here in 2026, you know, and I don’t know, my litmus test for something being the word of the year is, like, by the time that we get to the end of the year, are we all just terribly sick of hearing that word in, like, every possible context, you know, so I think we’re well on our way there. You know, part of the challenge is, like, just the word context is used in so many different, I’m asking, use the word context?

That’s a self-referential thing. You know, but it, like, we’re talking about, like, oh, context windows, and, you know, LLMs, you know, we’re talking about, you know, the context, like, the traditional way we used to think of it is like, oh, yeah, someone is making a decision here, you know, what is the context they as a human are thinking of? when we looked at, in our State of MarTech report back in May, trying to… provide some sort of framework for thinking about context more systematically, we broke it into three lenses. The first lens being customer context, which honestly is, I think, what most marketers have spent the most time thinking about, and for good reason.

Like, we’re just trying to understand, you know, the buyer, you know, what are their motivations, what’s their job to be done, we’ve… spent a number of years here with all this infrastructure for signals and listening to signals that, you know, can help us have a hint of what that context is at that moment. You know, and this is great! I mean, there’s a lot we’ve achieved there, there’s a lot more we can do.

The second lens of context, though, is the context about the business itself, the company, you know, this, sort of just understanding of, like, okay, well, what are our goals, our priorities, you know, it’s not just about understanding our products and services, but, you know, where are we making money, where are the incentives, you know, for the people in the process, you know, how do we talk about these things up against competitors? There’s a whole ton of this knowledge That one of the things that’s been fascinating is it’s very… It’s not very programmatically accessible in a lot of companies.

It’s stuff that, you know, we’ve learned, you know, you go through trainings, you listen in, you know, you develop it over time, you know, it’s there, it’s very institutional, but a lot of it’s not very actionable, you know, by systems, and certainly by AI. And we’ve started to recognize this, that’s a lot of, you know, these efforts you’ve seen this year, is people are like, oh, yeah, can we take the context of our business and start to put it into, you know, markdown files that we can feed into the AI?

Kerry Cunningham:

Could I ask a question about that? Do you think even about, because one of the things that we’ve seen in a couple of our surveys of, say, BDRs, of marketers. Is how incentive structures within organizations are misaligned to what the functions are doing, and often haven’t caught up. And, you know, so there’s an incentive structure over here that’s really built for something that nobody in the company is doing anymore, that really doesn’t match… it doesn’t match what needs to be done, so that… that context is… it’s part of the context also. All of those different systems and processes in your organization that either militate for or against having the right, approach for your… right go-to-market approach for your customers.

Scott Brinker:

Yeah. No, exactly. It’s funny, it’s like, when people actually go to, like, write this stuff down, and like, okay, we’ve now got to capture this in a way that, you know, AI can leverage it, that’s actually when they find themselves like, oh, actually, we weren’t really clear on who makes these decisions, or why do we do this? And oh, wow, this thing actually does conflict with this other thing here. Yeah, there’s a lot of discovery of the company context. And so this third lens is actually all around that. It’s saying, like, listen, The customer’s context exists. whether we fully understand it or not, they’re operating in their context, and same with us, you know, we operate in a company context, whether it’s actually programmatically actionable or not.

You know, so this third lens, which we call systems context, is really like, okay, out of that. how much do you actually have in a form that you can operationalize it, you know, with software, with AI? And so, you know, I’m a nerd, so I think of this in a VIN diagram of how much overlap between the customer’s context and the company context. And how much of that overlaps with the third circle of what’s actually available to our systems. In a systems context. And at that intersection of those three circles, which… in some cases is very small, but we would like to see larger.

That intersection, is what we call the golden context, because if you’re really in a position that you can have alignment between, you know, the customer’s context and your own, and it’s actionable within systems, I mean, that’s the… that’s the golden ring that we’re… golden context, you know, that we’re all.

Kerry Cunningham:

You’re not. Well, that’s kind of the difference. So, you know, one is… so, like, virtually all of the work, not all, but most of the work that I do in the research is around the customer context. How do we understand what that is? And I think that there’s a huge disconnect between we think we understand the customer context, but I would argue that for most organizations, your company context is 10 to 12 years or so displaced in time from your customer context. So you’re saying, you know, the customers in a lot of ways are more advanced in certain ways, and I think Customers have moved on from the way they were buying 10 or 15 years ago.

They’re different people with… they’re the same people, actually, but with a lot more experience, context, process. But our systems and processes on the seller side, on the company side, have not caught up with that. our views of that are wrong, so the overlap of the Venn diagrams between those two things and what gets captured in systems, I think, is… It’s really small, right now. And the question is, how do we… How do we get a big… how do we increase the overlap in that size? What’s the… what are the things that… that need to be done?

Scott Brinker:

I mean, you know, this is the work that I think a lot of people are focused on right now, is trying to, like, first just actually write it down. You know, that process of writing down the company context side of things, it’s going to take a tremendous amount of work, but… for so many of these things that people want to leverage AI for, you just can’t do it if you don’t have it written down, and If I’ve written down, one of the key things is having it written down in a way that’s shareable across everyone who’s wanting to refer to that, because when we… how you opened up this, like, you know, on the buyer’s side, you know, like, well, okay, the CMO, you know, is using their version of Quad to, like, evaluate something, and the CIO, maybe he’s using ChatGPT, or she’s using ChatGPT, you know, and getting it in their view, you know, and that’s fine when you want to bring it to, you know, a fun, mealy between stakeholders and a committee.

But, you know, from the company side, yeah, we don’t want actually, you know, marketing to have its view of what the company context is, and sales to have it different, and, you know, operations, customer, delivery, another. So it’s about, you know, writing that down, which then often means working through those disagreements about what that context is, because we’ve been sort of able to sweep some of those disagreements under the rug, you know. resurface them, you work through that. That’s gonna take a lot of effort. I see this as a multi-year journey of us really getting better at being able to programmatically capture, you know, company context, but the good news is. it’s… it’s not rocket science.

It’s not, like, a particularly hard technical challenge. It’s a management leadership, you know, decision to lean in on doing this.

Kerry Cunningham:

And it’s a, you’re pointing at this, but it’s an investment that has to happen, because it’s not, it’s… it’s time intensive, all right, to document processes and where you have differences in what… differences of a perspective on what those processes actually are, even today. I mean, that’s enormously valuable to uncover, but probably at least as enormously painful to work through and try to correct.

Scott Brinker:

It actually reminds me a lot of, you know, the challenges we went through over this past 10 years around, you know, customer data, you know, management is, for a long time, in marketing, like, you know, customer data was not well maintained, you know, and then when GDPR and, you know, all these sorts of compliance requirements, like, forced people to, like, alright, we gotta really start to clean this up. They were reluctant about it, because it is… it’s hard work. But honestly, the compliance benefits from that were always the smallest actual real benefit that people got out of that.

Like, as they got their data to be in a better shape, and it’s more accurate, you know, their ability to better leverage it in actual marketing and sales increased tremendously. And I think there’s a similar thing here in, like, you know, defining and writing down and making programmatically accessible, you know, your company context is. Yes, it is going to be able to help you do more with AI. You know, that is the journey that probably catalyzed why we’re doing this. But yeah, actually uncovering these things like, oh, holy crap, this thing is, like, 10 years out of date from, like, you know, what we’re actually doing, or we’ve got some major desync between, like, you know, the sales org and the customer service org on this.

Like, uncovering those things and fixing it actually makes your business better, whether or not you actually apply AI to it at all.

Kerry Cunningham:

Yeah, so maybe we should talk about what organizations would do to start that, or to know where to start, maybe. Know where to start, and then what do you do, if you want to really understand that company context?

Scott Brinker:

Well, I think, you know, let’s start where we started today’s discussion, which is like, okay, you know, go down from the funnel. You know, like, what… what are the next 2 or 3 touchpoints after there’s a quote-unquote initial discovery, you know, with a customer? You know, and how do you approach handling that? Like, what is… what is the conversations, or the emails, or the content that’s being provided with them? What sort of actions are you trying to, like, you know, pull out of them? Usually, people have a pretty good plan of what they think they want, but they actually haven’t very well defined the company context around that.

Like, the actual rules You know, of like, oh, how would we want this context? Like, when a salesperson, you know, is sort of modifying the way in which they’re going to talk to someone, or the sequence, you know, of how they handle a set of discovery calls, you know, these sorts of things that are, in a lot of companies, they’re sort of done by a certain amount by, by feel. You know, sort of learned feel within a team and a group, but, you know, not actually written down, and I think going through and actually, like, writing those things down, it’s… a way to basically even focus on just improving and optimizing those specific touchpoints, how you align it within what we think we know of the customer context, even if we aren’t applying AI to it yet.

Kerry Cunningham:

Yeah, I mean, I love that, and I’m looking at the time. We’ve only got about 7 minutes left in this session, and I want to dive down a rabbit hole here, because, I mean, there are so many of these places that you could think of, and I’ll just mention one, and then I think we need to kind of move to… I’d love to talk about your notions of transformation before we finish, but if you just think about, like, what are the reasons that you’re losing deals? What do sellers say? What happens in calls? You can use AI to understand much more about that now.

And then take that and go look on the marketing side of the house and say, do we have anything that actually helps here? Or that could… could stave off those objections that come later. And so I think, like, my hobby horse right now is, do you have any content that would help the CFO, or the CIO, or whoever those supporting functions are, actually understand the value proposition, not just for your solution, but for the category of solution. We’re going to be talking a lot more about that in coming… months, but do you have anything there?

You’re losing deals because of that, I’m sure of it, but do you have anything in your organization? Is there anything in your company context that could help there, or no? And the answer to that question, I mean, you know, if you do and you’re not using it, then easy. If you don’t, there’s work to be done.

Scott Brinker:

I’m so glad you mentioned that, too, because I feel like, you know, actually providing good, objective insights on a category is something I see so many businesses resisting, because so many people want to be the category of one. And I think you’re right, it misses a huge opportunity, you know, for these adjacent stakeholders, you know, who actually are less interested in the specifics of the brand, and they just want to understand, even, like, that capability in that category, how to think about it. So, great opportunity.

Kerry Cunningham:

Yeah, I’ll just throw out 30 seconds here. If you’re, in our buyer experience study, what we’re seeing for this year is that a substantial part of the reason that… not that you are losing a deal to a competitor, but that everybody’s losing a deal to no decision, is that the buying group, finance, IT, don’t get the value of the solution that your primary buyer… your primary buyer persona wants. They want it. They run everybody through a buying process for it, they can’t justify it, it doesn’t happen. So if you’re not infrastructure, if you’re not something your buyer has to buy.

Then you’re losing a lot of deals to that right now, and you’re probably not actually doing anything about it. It’s probably a big black hole, is what it looks like.

Scott Brinker:

Good advice.

Kerry Cunningham:

So, in just a couple of minutes, you can take this a different direction if you want, but you’ve got this great, framework that is the shape of transformation, using the caterpillar to butterfly, metaphor, which I really like. Maybe just a couple minutes overview, because I think it helps people understand, like, you know, where you are, where you gotta get to. With your Martech stack in the age of AI.

Scott Brinker:

Yeah, no, well, thanks. Yeah, this was from our State of Martech report back in May. You know, we wanted to talk about transformation, and being the relatively uncreative people we are, we’re like, oh, what’s a good metaphor for, you know, transformation? Oh, well, it’s, you know, the caterpillar to the butterfly, which has been overused as a metaphor a billion times. But as we were, like, sort of digging into it, it’s like, well, wait a second, it doesn’t just go, from caterpillar to butterfly overnight, you know, there’s this really fascinating, and honestly somewhat disgusting, chrysalis stage, you know, where the butter… the caterpillar melts down, and the butterfly emerges, and, like, if you were to pull that thing apart, man, it doesn’t look like a… Caterpillar or a butterfly, it’s just, you know… Yeah.

It actually felt like, you know, that is the better metaphor of, like, what’s happening right now.

Kerry Cunningham:

Yeah.

Scott Brinker:

I guess we sort of have the vision of where we want to get is, you know, these AI butterflies and go to market. We know the caterpillary stage is, like, it’s losing its efficacy, you know, we’re seeing competition that’s, you know, outsmarting us, you know, with some of these new capabilities. But most companies that are on that journey, like, they’re in that middle mushy stage, you know, of figuring this out, learning this, and I guess part of our message was, like, that’s okay. Like, this sort of transformation does not happen at the speed of a prompt.

This is stuff that takes time to organizationally shift, so, you know, I guess that message is like, yes, you want to keep on that But the fact that it, you know, doesn’t happen in a week, yeah, no matter what you read on LinkedIn, yeah, don’t be.

Kerry Cunningham:

Yeah, and everybody should look at this report. You know, I’m in a long conversation a little later on today about the ROI of AI efforts in go-to-market, and part of what, I mean, everybody, we want and need it. But part of it is, you’re doing experiments now, and some of them are going to fail. Maybe a lot of them, maybe most of them, are not going to produce ROI, and you’re going to think it’s a failure. But you have… you’re gonna have to go through some of these things, or you’re gonna be very late, right?

If you wait until it’s all settled, you might be able to do that, and that might work okay for you. But if you try to skip through this chrysalis phase without actually going through the process of figuring out what’s gonna work for your org. You may not get very far.

Scott Brinker:

People are just leaping off the branch. It doesn’t end well.

Kerry Cunningham:

It doesn’t, and I think a realistic expectation is you’re going to have to go through some things here, as you’re, experimenting and finding where those… what projects are going to work for your organization, where do you start from? So I think that company context that you talk about is a great place to understand where you’re starting from, how does your org… and then your buyer context, customer context, is partly where you need to go. You’re not going to change their context. You can understand it and align to it, but you can’t… you’re not the one who’s going to change it.

So, you know, if you go from where are we today to there, and how does AI help us get there, I think that helps you understand that you’re going to go through these stages, and one of them is this chrysalis stage where You know, it may not feel great, all the time, but you’re making progress.

Scott Brinker:

Well, I have to say is one of the things that makes me happy, is I feel like this year, we are starting to see a bit of the pendulum swing towards… it isn’t just pure cheerleading about all AI all the time. It is like, okay, let’s be a little bit more realistic of, like, what are we doing, what’s working, what’s not working, what are the economic benefits of, you know, spending, you know, a ton on tokens, you know, like… We’re starting, you know, to find a way to, like, balance how do we really make this work.

Kerry Cunningham:

Yeah, and there’s, like, 20 other things we wanted to talk about there, but I think we’re out of time. Scott, it was so great talking with you.

Scott Brinker:

Oh, thank you so much, Kerry.

Kerry Cunningham:

My pleasure.

Julia Nimchinski:

Phenomenal conversation. Thank you so much, Scott and Kerry. What’s the best way for the community to support your work?

Scott Brinker:

Well, I will say at the moment, I am running a survey on contacts, so anyone who would want to support my work by taking that survey, I think we can plop in the link, or I’ll put it on my LinkedIn page. Thank you.

Kerry Cunningham:

Yeah, you can find me on LinkedIn pretty easily. I think I’m the only Kerry Cunningham, or certainly one of the only ones, and I’m posting there pretty much every day about research, and a lot about that company context as well.

Julia Nimchinski:

Thank you so much.

Kerry Cunningham:

You got it. Thank you.

Scott Brinker:

Had back.