Agent-to-Agent GTM and Machine-Native Markets

Julia Nimchinski:

And it’s gonna be the panel of the summit, Agent to Agent GTM and Machined Native Markets. How are you doing? Mark, you’re on mute.

Mark Organ:

Thank you, I’m doing great, very excited to be here. This is such a groundbreaking topic, and so… Awesome. Yeah.

Julia Nimchinski:

Take it away.

Mark Organ:

Alright, let’s rock! Do I have my co-panelists here? Let’s see.

Amanda Kahlow:

I think we’re here!

Mark Organ:

All here? Alright, hey there, Amanda.

Amanda Kahlow:

Hey, how are you, friend?

Mark Organ:

I am doing great, so, today we are gonna talk about, HDM and machine-native Markets, and we’re going to explore how leaders are redesigning pricing, procurement, retention, expansion, and buyer engagement from markets that are moving towards agent-to-agent transactions.

So I’m Mark Oregon, I’m currently at Category Knots, I help companies design their categories, and this is pretty exciting because, you know, this really feels like it could be at least powering a new category. Or several categories.

So maybe we can just start by everyone introducing themselves quickly, and just kind of their perspective, where they’re coming from.

I already sort of introduced myself, but, you know, where I’m coming from here is that I really want to explore, you know, what agent-to-agent is really for, what’s the, you know, what is the killer use case, because in my experience, it’s not like, you know, agent-to-agent’s gonna replace everything, it’s going to replace some things and do some things very well.

And the company that figures, or the people that figure this out, earlier, where it fits with everything else, are the ones that are going to win. But maybe, Amanda, quick introduction, and, you know, your 30-second perspective.

Amanda Kahlow:

Yeah, yeah, hi, good to see you again. Amanda Kalo, I am the CEO and founder of OneMind. I was the founder and former CEO of Sixth Sense. My one line to the market is I started Sixth Sense to find buyers, and One Mind down to close them. At OneMind, we are building agent-to-agent… agents and superhumans.

We call them go-to-market superhumans, so they can engage with agents just the same that they can engage with humans.

that cover the life cycle of your customer journey, from first touch, on top of funnel on the website, all the way through to our superhumans join calls as solutions engineers can give live demos, and when I say live, I mean truly live, not just interactive. Precanned scripted demos, they can solution with buyers.

And then all the way through to being in product to help people onboard, guide them, show them how to use products. And then cross-sell, upsell, and deal rooms, etc. So, we want to support the lifecycle. We think there’s a new world with AI, that there’s no longer roles and silos, that everything has collapsed into one, in a AI-first world.

And at the end of the day, we’re here to create better buying experiences, so that is my North Star, is how can I serve the buyer and not serve… not be as focused on serving and creating for the seller, which I believe most SaaS companies, you know, even my past company, a lot of them were… we created this, at the time, the right thing to do was around a seller versus around a buyer.

Mark Organ:

Yeah, I agree. A little more than 30 seconds, but that’s alright. I totally agree, I totally agree on the buying experience, so from… in my whole career, that has been my touchstone as well. So cool. Hey, Prem, are you here?

Prem Parameswaran:

Yes, I am.

Mark Organ:

Alright.

Prem Parameswaran:

Okay?

Mark Organ:

Yeah, let’s do it. Introduce yourself and kind of your overall perspective on this topic.

Prem Parameswaran:

Perfect. I’m Prem Parameshwaran, Chief Technology Officer at Gainsight, and I’m also the General Manager for Atlas, which are… which is our Agentic suite of products. So for those of you who do not know, Gainsight, we’re basically the leading, customer success platform.

So we’re part of the GTM stack for a lot of SaaS companies, partly because if you are selling SaaS products, you never stop really selling, because you have to resell your customer every year, or through the buying cycle.

And what I’m really interested in here is, I want to second what Amanda said a little bit, which is that I think done right, this is going to create a better experience, both for the buyer and the seller, because the asymmetry that existed prior to this is going to go away with agents, and we’re all going to be participants in that journey, so I’m excited.

Mark Organ:

Cool, thank you! Tuba!

Tooba Durraze:

Hi, thanks for having me. I’m Tuba, the founder and CEO of Amoeba AI. We are the world’s first neurosymbolic AI decision intelligence layer. Essentially, we solve for a very basic question, what’s going to help your business grow, essentially?

And, like, from a C-suite level, the right observability into what decisions you need to make at what point to solve it.

And… I’m a little bit more on, like, I guess, like, beyond the future side, where I’m like, at some point, let the agents handle it, we’re gonna end up on a beach, so my viewpoint is a little bit in understanding, what does the world look like in agent-to-agent architectures look like for leaders in the transitionary phase, which is what I would consider us being in right now, and how is that going to change, and how fast in the future?

But in the ideal world, to me, it’s like, you’re either telling an agent to solve it, or you’re letting the agent solve it? So, kind of that paradigm.

Mark Organ:

Got it. Thanks. Deepinder?

Deepinder Singh Dhingra:

Hi everyone, I’m Dipandur, I’m the founder and CEO of Refshore.ai, and what we are building is the only enterprise-grade context layer for B2B go-to-market across sales and marketing. Essentially, what we do is we enable enterprise B2B go-to-market stacks, which are highly fragmented. We work with customers that have over 75 systems.

right in their go-to-market stack, become major intake AI ready. Right? The current… GTM tech stack, right, the SaaS tools that Amanda talked about, or the CRM, marketing automation, other tools, were never built for agents, they were built for human beings. Right?

And so what we’re trying to enable is the cold infrastructure that’ll make the GTM tech stack agent ready, whether that’s for agent-to-agent go-to-market, or for agent-to-human go-to-market. That’s what we do. My background, I’ve spent over a couple of decades in the big data and AI space.

I was part of the early leadership team of MuSigma, the first and only big data analytics unicorn. out of the Bangalore ecosystem, and probably the first unicorn out of the India ecosystem, right? And, my DNA is enterprise, and what we’re trying to solve for is enterprise go-to-market problems.

Mark Organ:

Very good, thank you, and Hudson?

hudson:

Hey, I’m Hudson, founder and CEO of The Hog. It actually stands for the Head of Growth, in case people are wondering why I named my company The Hog. And I was an operator as a head of growth for startups for the last 10, 15 years, and last year I was building agents to automate as much of myself as possible.

And then during that process, got into YSC, I’m a newly, freshly-minted founder.

And I’m actually living this pivot about this panel right now, so I’m super excited to talk to everyone in their perspective, because we were building agents and command centers to execute go-to-market sales and marketing, and realized The best agents are very subjective, but the best data is not.

And that’s kind of the world we’re betting on, and we’re betting that most tasks, if not all tasks, 80-90%, will likely be either agent completely ran, or mostly agent-supported, and that’s the world we’re betting on, and that’s why we actually pivoted recently from an agent product to an API data layer product.

So I’m really happy to talk through my live pivot with some industry experts. Spurs right now.

Mark Organ:

Yeah, very interesting perspective. Well, I’m gonna pick… I’m gonna pick some, people that I think are relevant for different questions, but if you really feel like you got a point, you can jump in. But first, Depender, I’d love to hear from you.

You know, people hear about agent-to-agent go-to-market, they imagine bots buying from other bots, but what’s the most realistic near-term version of this in enterprise B2B?

different places, you know, there’s a lot of different places that, different areas of buying, What’s, you know, what’s, what’s the most, you know, what’s most likely to happen first?

Deepinder Singh Dhingra:

Yeah, I think it’s… I think what’s more realistic is AI-mediated buying. So, essentially, your agents, whether on the buyer’s side, are going to be doing a lot of discovery, a lot of evaluation, ranking, prioritization. We see that daily when customers come to us.

their agents, or they have kind of used some kind of AI to essentially research us, right? So that is the most realistic right now in the enterprise, right?

My focus has always been on the upper mid-market in the enterprise, maybe in the consumer, in the B2C space, there’ll be agent-to-agent transactions directly, but we’re not seeing that in the enterprise space yet.

So I think the most realistic use case is AI-mediated and agent-mediated buying, where there could be an agent on the other side that is serving information, that’s serving, content to the buying agent.

So if I’m a SaaS vendor, or I’m an AI SaaS tool vendor, etc, right, I want to kind of build agents and build my content in a way that can help buying agents consume information much better, whether that’s pricing information, whether that’s content. reviews, etc, right? So I think that’s the most realistic use case. As of today, yeah.

Mark Organ:

Yeah, so kind of discovery and vendor shortlisting kind of right up at the top of the buying process. Yeah. Any… does anyone disagree with Deepinder on this one? Nope, I think…

Tooba Durraze:

I think it’s, like, agreed with, like, machine-native markets is kind of where we’re heading towards, and, like, composition of those markets will… change, like, what is, like, kind of the machine-native portion will change, depending on how the markets change, how the rules and the positions change.

So that shape will look different the same way job descriptions look different, right? From, like, olden days to these days, but I think that, broadly, that sentiment is, like, accurate.

Mark Organ:

Okay, sort of at the top of the funnel. So everyone, everyone, seems like everyone’s in agreement that that’s kind of where we’re going to see it, first, in sort of discovery and vendor shortlisting, kind of like the, the, early phases of the buying process, it sounds like.

Prem Parameswaran:

I think you might start seeing it at the other end as well, like, anything that is verifiable and arbitrary rule-based, because those are the things that were harder to automate with plain language. So, for example, you could think of, like, a Vanta-like service that can verify the things that that what a vendor says, in their material.

So, I think on the procurement and the verification layer as well, it’s likely to happen. I think the middle is what stays sort of out of reach, because the real-world verification of whatever you claim to the agent, is that really true?

That actually crosses the boundary that is programmatically accessible to an agent, so I think that still has to be done with humans. But I think the both ends, I feel like, are likely to get impacted, by agentic.

Mark Organ:

Yeah, interesting. Yeah, something else to jump in around that. So, what part of the buying journey still is sort of human accountable? You know, Prem makes a case it’s in… it’s in the middle of the buying process. What do you guys think? Is that… does that sound right? Is that where… is that where humans are gonna remain, sort of, the longest?

Amanda Kahlow:

I think… I think I would say that it’s the type of deal, and who, like, what you’re selling. So obviously, I think for, like, commercial or SMB business, both, you know, from top of funnel all the way through to close, and onboarding and supporting will be supported fully through an agent.

But for your enterprise and strategic deals, like, there’s still the relationship aspect. I don’t think that’s going away anytime soon. I do think we’re gonna move into a world where where we no longer have siloed jobs, and the human job is not going to be what it is today.

So the human job that we see as, like, an SDR or an AE or a sales engineer, a customer success manager will all collapse into one, and so the human will maintain the relationship, and the agents will be doing the jobs and the tasks and the information gathering that happens across the life cycle.

that’s what I believe we’re going to see in a very near-term world, that these, like, the world of humans and the way we think about the role of the human will evolve to where humans are best, where we can make connections and have empathy.

However, I wouldn’t put it past, like, AI is moving to the place of really understanding and having empathy, so when you think about being able to solution and understand a buyer. I actually think AI will do, and is doing already, much better job than most humans are doing.

Mark Organ:

That’s interesting, because that really is at odds with, kind of the… the way things have gone. I mean, generally, humans have become more and more specialized over time, not… not more, sort of, I guess, maybe you say that if they’re just relationship builders, then that is a specialization unto itself. But, that is interesting.

Anyone take issue with Amanda on this? This is kind of a very bold statement that she’s made.

Tooba Durraze:

I don’t think it’s as bold. I feel like she’s… like, it’s, like, the equivalent of… you can walk, you can ride a horse, or you can take a car, right? At some point, when there were no cars, like, we were like, people are gonna walk places, or take a horse places.

So it’s like, if that vehicle, it’s like, how much compute you would need to solve a problem. eventually, even for shorter distances, we started taking cars at our convenience. So I think that’s kind of the human-agent interaction, where there’ll come a point where things are so easy that, why wouldn’t you?

And these models are obviously exponentially going to get better versus, like. how fast, like, human intellect is progressing. And humans are smart at application, right? So, I think that that level, we’re, like, at the… not everyone has a host, but most people are still walking and getting a host phase.

I think eventually it’ll get to the car phase.

Amanda Kahlow:

But I think that… and I think the real big… the big difference here is that humans have massive capacity limitations. They have all of the best intentions, but we have capa… like, if you think about, like, Databricks, or somebody that sells, like, multiple products to every industry and every vertical.

You can have specialized humans that try to understand that vertical and truly don’t even do it well enough, or you can have AI that understands everyone and every role, and every function, and gathers the context from first touch all the way through to close, and onboarding, and brings that into the conversation in a way that I get my prep doc to a call, a prep doc to a call, and, you know, I read it, but maybe 10% absorbs?

And it’s the best of… I want to, but I can’t. Like, I have so much going on that I can only absorb so much. So, I actually think that that’s where the AI is so much better than a human, so it will go to a place of a human being more of a generalist.

And the AI will be able to, like, go into the nitty-gritty and truly understand buyers and solution with them, and buyers will demand it as a result.

hudson:

Yeah, I want to agree with Amanda.

Mark Organ:

Hadza, Hudson, yeah.

hudson:

Yeah, where I think the context is solved by AI, and all that information, and even reasoning, I think buyers are going to be able to make more comprehensive and just better buying decisions in general. But if I think about first principle of where the human has to stay in the loop, this might be a very general version of it.

It’s like, where does someone get fired? Who… what decision… does a machine make and get someone fired is exactly where the human will always stay. Right now, we’re in that… that, like, basically, like, accountability aspect of it, right?

I think this is where the… is interesting about this panel, is, like, where that trust continues to evolve and grow, and what are the guardrails we’re gonna set there.

But until we figure out that kind of, like, high-level context of what is the liability that an agent decision makes that impacts the rest of the company, I think that’s where the finite, like. Pain point is that we’re trying to solve.

Deepinder Singh Dhingra:

Yeah, I think one thing I wanted to add on to this conversation is that if you abstract out, let’s say you’re having a buying agent. Right? There are a couple of aspects. How much coordination is required? in that buying process, right? So, for example, if you go to an enterprise buying cycle, right, you have buying committees, right?

We see buying committees of 12 to 15 to 20 people, right? Each of those buy… all of those buying committees are trying to verify, to Print’s point, or to discover and rank certain pieces of information, and it just goes… and you try… you go deeper and deeper as you get into the middle of the buying process, right?

So there, there’s a coordination tax, right? Because Although buying committees are, you know, it takes time to go through buying committees, but research from Google and from Harvard is also talking about how there’s a coordination tax. Agents are not really good at coordination.

So if I’m trying to orchestrate a buying process through using agentic GTM if I’m a buyer. The amount of coordination that I need to do from agent to agent, and sometimes it is sequential coordination, that actually has a huge tax associated with it on agents and even multi-agent systems. are not very good at it.

So I think the processes where you need less coordination, right, is… are going to get more agentic, right, and going to be more AI-mediated, versus the processes that require more coordination. And if you think about agentic performance, that’s now… Getting through whether from the buyer side, from the seller side.

We believe in a team of agents approach. at RevShore, right, not one single agent. What you’re finding that agent performance is no longer about intelligence, because agents are much more intelligent than human beings. They can absorb a lot of context.

The earlier capacity constraints of human beings, either of one individual human being, to absorb a lot of information, become an expert, is no longer there because agents can do it. But now it is actually shifting, the frontier of performance is shifting to context. How much context can you provide the agents? Right?

So because agents might not be able to… the agents are only kind of given context versus the frame that you provide them, so if you don’t have the right context in the harness, they’re not going to have that, although they can absorb a lot of context, and then the coordination. So now it’s all about where do you need context and coordination.

And they’re still… agents are not that great, right? Because there’s a context and coordination tax that one has to pay.

Mark Organ:

Right. I think that…

Deepinder Singh Dhingra:

way to kind of abstract out where that agentic GTM, or agentic buying, or agent-to-agent will kind of progress versus not progress.

Mark Organ:

So, yeah, good point. So, both on the buying and the selling side, context and coordination is a place where humans are likely to stick around longer.

Amanda Kahlow:

I think there’. a couple companies that are doing it really well, like Sierra and Decagon have really figured out how to do multi-agent orchestration extremely well.

I believe we have as well, but, like, I bring them up, like, taking in multiple, because that context layer, that is where it gets expensive, especially if you’re having a… communication, latency is really important. The more context you put in, the low… the slower it gets.

So, you have to maintain, like, the low response times with high context, and I think that’s, like, those are really hard problems. That some of us are solving, and I actually think once you do solve it, it’s exponentially better than the orchestration from coming from a human.

Deepinder Singh Dhingra:

Yeah, I think the two… the only difference there is whether it’s a sequential coordination or whether it’s parallel coordination. So parallel coordination, you know, people have kind of figured out agents are much better at multi-agent systems, but when it’s sequential coordination.

in the buying process or in the selling process, that’s where the tax is pretty high, currently, yeah. And that’s a problem to solve, I think we all are trying to solve that, yeah.

Mark Organ:

Manning. Yeah, I’d like to jump into a related question that’s around trust. Big thing that I’ve been involved in at my last company is how to generate more trust in order to improve the buying process. But things are changing now in the agentic world, so who owns trust in you know, in this new market.

I mean, in human-first markets, we had references, we had, you know, brands, analysts, security reviews, you know, 5-star reviews online, I mean, all those sorts of things were generating trust. What’s the machine-native equivalent, and is this even a relevant thing to consider? In a more agentic buying world. I’m happy to jump in.

Amanda Kahlow:

Yeah, I mean, I think humans… first of all, humans hallucinate. More than agents, and they do so knowingly. And I say that, like, yeah, it’s a joke, it’s funny, but it’s also, like, incredibly true that humans are doing the best they can, but they skirt over a topic because they don’t know the answer, or they, worse, they give a wrong answer.

And so, I actually believe if you build agents, like, with the right guardrails. then buyers will demand to talk to an agent because they will trust it more. So, Hudson, to your point that that people will get fired, as you said before, you’re gonna want to talk to a human.

I actually think the reverse, that they’re going to demand an agent, an AI, because they’re going to trust it more than they trust the human once we do it right. But there have been so many bad actors that have done it wrong, and have gotten, you know, had some major consequences because they didn’t put it on the guardrails.

But we have evals and scorecards and the ability to do that now, so you know that the… that the… we call them superhumans or agents, whatever it is, is staying on the rails, and it is more trusted.

So, I think, you know, as we… as buyers start to realize that my answer from that agent is going to be better than a human, they’re not going to want to talk to a human anymore.

Mark Organ:

Okay, but yeah, I mean, I was really talking more about some of these external things that we’ve… we as humans have used this for trust. Is there an equivalent for… for the machine world?

Deepinder Singh Dhingra:

Yeah, I think…

Tooba Durraze:

I think that layer doesn’t exist. Sorry to jump in. Go ahead. That layer, it’s, like, because there’s a shared responsibility for trust on both sides. Like, as an example, like, a seller can hack the system by having a bunch of, like, unlisted blogs that get picked up in AI visibility to say they’re the best, they’re the best, they’re the best.

So then, when, like, the buyer is, like, going by that, like, it’s going to kind of create this, like, mistrust, right? I think that that layer in the way that you’re speaking of doesn’t… necessarily exist yet, but I think it will exist.

To actually go back to one of Amanda’s earlier points, when buyers become, sellers become a little bit more buyer-oriented versus them building it my way, then there’ll be some sort of unified, like, trust dimensions that get built. It’s not… it’s not… that responsibility is not shared with the same incentives on both sides yet.

Deepinder Singh Dhingra:

Yeah, I think this whole idea of trust also falls to, I think, Brain’s first point, which is verification. So let’s say I’m trying to enable an agent-to-agent GTM perspective, right? So there’s a buyer agent, how does it kind of trust the information that it’s getting from the seller’s site, right?

Now, there are a few things, like, as human beings, we use review sites and peer-to-peer networks. And we use brand. Brand becomes an aspect of cross-sex sectors, and that’s what human beings use. what are the elements that agents do you use to establish trust in the seller’s information on the seller’s expertise, etc, right?

Especially if you’re doing agent-to-agent, then remember, there is no… there is no human being on the other side, right? So if an agent could just concoct a demo of a product, right, right, assuming there are guardrails, but how does the agent trust that the demo I’m seeing or the information I’m getting is really verifiable.

And so I think what that will kind of lead to is validation approaches that agents will have to apply in terms of the information. Obviously, they’ll have to kind of look at external information, right, on reviews, etc.

But also, what are the validation frameworks on the information, the content being served through the seller’s, let’s say if you have the seller’s website or the seller’s agency. Right? And that is not yet figured out yet, right? Because it’s all about, obviously, one part is about delivering content in a way that agents understand, right?

Obviously, what are the right formats of information, structuring it in a way that agents like to consume it, performance of the information, how are agents kind of… the agent responding to the request from a buyer agent? And I’m not necessarily talking about just like, hey, there’s a video that… a demo that a user is seeing, or an FAQ answer.

I’m talking about the multiple layers of detail of information that an agent will transact with in another agent. So I think that aspect is not yet figured out. I know the question is, like, what establishes trust, but I think some of those aspects are going to come in, like, how do you do verification? Right?

That’s important, because I, as a seller, can’t just expect, as a buying agent, can’t just expect that my seller’s agent is, like, perfect, and is giving me the perfect demo and the perfect information. Right? I have to verify it. Now, how do agents evaluate products?

Maybe there’s an agent that… there’s an evaluation agent that will evaluate the product as well. I think so those aspects and layers will need to be built in to establish this trust framework more strongly.

Prem Parameswaran:

You know, I think, you know, one way to kind of… Bifurcate the answer is… It’s possible that there is a… Almost like a power law distribution, where a vast majority of decisions that, become low consequence because you have strong verifiability, the consequence of getting that wrong is not that high, and it’s recoverable.

Those might actually become entirely machine-to-machine, and you can see an entire ecosystem that opens up that, make it easy to, easy to trust them, and if there is a mistake, then it can be recovered, pretty easily.

So, the interesting perspective that I have is, but for decisions that have consequences, where I define accountability as the ability to bear a consequence. So if you’re a CISO who’s making a decision, or if you are buying, like, authentication product where the consequence of getting it wrong is very high, the human is finally accountable for it.

There, as well, I do think that a lot of the rote work that we spend a lot of time on Will get automated by either machines or augmented machines, so all of that gets taken care, and… it then surfaces the three people that I want to really engage with, and I think that is where the relationships will actually become even more important, because right now, we don’t have the time to build these relationships, because you’re spending a lot of time in these, rote things.

And the last point I’ll make is that The key here, at least, is that the asymmetry that existed prior to agents is that the buyers had a lot of advantage. Meaning, we were always targeted as a consumer, right? Like, you had a lot of information about The, the, who you are selling to.

And this is the first time now the buyers have the ability to do the same thing, because they can spin up an agent, there will be companies that will come up and figure that out. So, I like to call this the cognitive asymmetry, or the intelligence asymmetry that’s going to go away.

So, attention is no longer going to be something sellers can target, because agents have infinite attention, they have all the time in the world. So they will find things out. So eventually, you’ll be forced to compete on something that removes that asymmetry.

And I don’t know what exactly it is, but one for sure is, can you build a relationship with the person on the other side?

Amanda Kahlow:

Or can you build the best product? What if we move to a world where actually the best product wins? And the best product for whoever they’re selling to, right? Like…

Mark Organ:

But what is the…

Amanda Kahlow:

wait.

Mark Organ:

what is the best product, though, right? Like, how do you… I mean, how does an agent determine that? Like, how do you prevent an agent from, you know, flattening the differentiation in your product to just a, you know, a bunch of feature checklists?

Tooba Durraze:

But you, you have to, you as the producer of the product, that’s the responsibility, right? So to Prem’s point, like, I think there is an attention problem even with agents, because people hack agents by frequency and recency, right? So it’s like, that’s the trust piece, importance of the trust piece.

Which, by the way, like, to a certain degree, maybe it doesn’t even exist now, in the sense that you will buy more from people that you know, under the assumption that you people, that you know, you trust may or may not be the best product. you’re more willing to try from people that you know.

And it’s like, in order to move agents towards a world where it’s not as subjective.

Prem Parameswaran:

Like…

Tooba Durraze:

you have to make sure that the agent’s world is evaluating on not just, like, very… also very quantitatively, but what is, like, the qualitative. And the qualitative is determined by whatever, like, the trust-building exercises on both sides. But how do you frame that to, like, a machine?

hudson:

Yeah, and it will summarize, like.

Mark Organ:

Absolutely.

hudson:

what you said, like, Prem, Amanda, and Tuba, like, I think it’s basically always buyer-driven, and what we want, and what that is, is changing. We want more information, we want it faster, we want to trust things better. ideally, that takes out all the fluff in sales and marketing, right?

Like, LLMs, they literally know how to categorize context and words, they’re like, that’s a fluffy sales word, versus that’s an actual, like, constraint I’m looking for. And I think so, like, with Prem’s, point about, like. building that trust, like, is what the buyer wants, like Amanda said, like, the agent is gonna tell us what they want.

And on the seller side agent, seller side marketplace, we’re gonna start producing those type of protocols, right? Like, Google has their A2A protocol, Visa launched their TAP protocol, so they’re starting to build these trusted trust identification. They’re just not adopted yet.

I think when we start equipping our buying agents with the certain things they want to look for, the other side has to, like, produce that for them. It starts to become, like.

table stakes in how we structure our website and our products, where I think like Prim was talking about, we’re gonna get to a point where every single company’s gonna have a sandbox environment, probably an API, MCP, like, all that documentation is out there that’s gonna be very different from what us as humans see.

So on a hot take on… what I think the world’s gonna look like, there’s gonna be a web interface that’s designed just for agents, and there’s gonna be one designed just for humans. Even, like, what’s a parallel AI, when you go on their website, like, are you human or are you agent?

I think that’s gonna be standard practice going forward on how… how things navigate websites.

Amanda Kahlow:

Yeah, but I think at the end, though, I do agree that will be, like, the content will be delivered to a human or to an agent, but actually, I think the context layer and everything below it is very similar. Like, you can have a conversation with a human or an agent, you’re just… the way that you spew it back, the way you give it back is different.

But when we, like… Yeah, when I was thinking about, like, the different examples that you guys were sharing with, like, the agents… actually, you know what? I’m not gonna even go here. I don’t want to go there. Sure, keep going back, sorry. I was gonna share an example, but I’m gonna hold it.

I’m gonna hold it, because I’m not supposed to say right now, we’re not quite allowed to share quite a.

hudson:

I kind of have a question for kind of, like, this panel, because you guys are so well-versed, and we… I think all of us have seen the evolution, like, market go to market. You know, when SaaS first came out, it was, like, really sales-led, like, people were, you know, that was kind of how the OG SaaS market worked.

And then all of a sudden, the PLG motion happened, right, with the Slacks, Dropboxes of the world that came out. And then it kind of moved into the last couple of years, the four deployed engineer, deployed, people, that kind of took off.

now when we’re thinking about the new buyer, new generation, like, in my perspective, it’s kind of swinging to a new kind of a PLG motion, but the PLG is more agent than it is, like, PLG back to humans, and that’s kind of… I don’t know the answer, so I’m actually… sorry, Mark, I’m taking over, I just want to ask them to handle the same thing, because I… that’s the world I’m thinking I’m building, like, the new PLG motion is gonna be agent-led, and then they’re gonna loop in the human when the agent thinks, like, hey, like, I vetted, like, 3,000 companies that you gave me constraints for, here’s your short list of 5, and here’s your pros and cons, you still have to push the million dollar button, kind of thing.

Amanda Kahlow:

We call it AILG, right? So it’s…

hudson:

I swear…

Amanda Kahlow:

Right? So it’s like, that’s what we… it’s where PLG, a self-serve, free trial, start… go through the process, but you actually need a conversation, and you need to have answers, and you need to be solutions to your…

hudson:

Oh, gee, I like that.

Amanda Kahlow:

AILG is, like, basically, okay, I can actually have a conversation, an enterprise-grade level conversation, I can get the demo, I can be solutioned with. And, like, to the last conversation, it’s all about the outcome, right? So what is the outcome that we’re trying to drive towards?

And that’s what the AI will choose, like, what’s the best outcome for my business based on all the information that’s out there?

Right now, so much subjectivity comes into buying, and Like, even, like, when we’re going head-to-head with some of our competitors, I know for a fact some of it is because they have relationships, they’ve been there, they’ve used it before, and they don’t want to rock the boat or try something new, and they’re afraid.

That, like, the fear mentality is… is… is sad sometimes, right? So, you don’t want to try the next thing, because you just want to try what’s safe.

But we’re not living in a world of safe anymore, like, as people are moving, and the AI will allow us to make decisions based on what’s solution to my business and what’s going to create the best outcome, and then AI will allow us to actually close bigger deals and more, and have the enterprise conversation at, like, the AILG, like, that level where you can’t just get a free trial.

You actually have to pay for it, you have to put 50K, 100K, whatever that is. in as you’re, like, moving forward to move forward with a product. So I actually think that’s the new world, is that we can actually start selling, and sell enterprise deals through AI.

Deepinder Singh Dhingra:

Yeah, I think… I think… I think what… I think we might have veered out from agent to agent. I think what… there are a couple of things that we’re talking about. One is the… how does the agent-to-agent transaction, or the agent-to-agent communication be more trustworthy?

I think what we’ve kind of… we’re talking about how we can use AI to sell better. right, and serve questions, and those questions might be, you know, by… from human beings, right? So if I’m taking from enterprises, I might have an AI solution engineer you know, helping serve questions to me, right?

So, I think, still, the paradigm is still human being, right? So, how does a human being buy, right? And how do we use AI agents to kind of thing? But what… how does the interaction change when agents are buying? Right? And so, are we… are we just saying that the agent buying process is similar to the human buying process, right?

That’s what it seems like. And I don’t know whether… how you were actually referring to that, right? Are you referring to… are we saying that the agent buying process is the same as the human buying process? Does the buying process change?

Amanda Kahlow:

No, of course.

Prem Parameswaran:

Does it?

Amanda Kahlow:

Like, you can bring all the context into one, and you can have all your answers and go back and forth and be done in 5 seconds, right? So when it’s agent to agent, like, you’re getting to the outcome versus having to go through the steps and bring people along a journey. You don’t have to go on the journey anymore.

You can actually just start with the outcome and then find the solution.

Tooba Durraze:

Yeah, it does change…

Deepinder Singh Dhingra:

Assuming no coordination is required, assuming no coordination is required, essentially. Would there be an agent buying committee?

Amanda Kahlow:

Of multiple agents, you mean?

Deepinder Singh Dhingra:

Yeah, I mean, like, that’s why I’m asking, like.

Amanda Kahlow:

I don’t think…

Deepinder Singh Dhingra:

I ain’t committing anymore.

Amanda Kahlow:

a human context paradigm. I think we’re moving into a different paradigm, where I don’t look at it… I don’t see it, like, happening like that.

Deepinder Singh Dhingra:

No, that’s what I’m saying, right? So, what I’m saying is that today is… so, what is the agent buying process of the future? Is the agent buying process without buying committees?

Tooba Durraze:

they have agent buying committees, like, I think there’s a thesis out right now about agents basically negotiating with, like, a… council of other agents, and then those agents bidding to see what agent comes out on top to the buyer agent, right? Essentially.

So, council of sellers and a buying agent, and then the theory becomes that, like, is it… the best winning, best being subjective here, is it by way of, like, agents that are the best negotiators, the best problem solvers, versus agents that are, again, sitting on top of all the information, have the right information?

So, I think it’ll be a new paradigm, but I think some of those things as humans will translate, actually, unfortunately, to that paradigm as well.

Amanda Kahlow:

But why do you need multiple… why is it a committee? Why isn’t it just one agent that can represent multiple roles and multiple…

Tooba Durraze:

So 5 solution agents, let’s say. Like, five different vendors. Agents, and then this is the.

Deepinder Singh Dhingra:

Am I going to… am I going to money? And these are open questions, I’m not… I’m not insinuating one versus the other. Am I going to spend a million dollars on an agent buying committee that has evaluated tools, had, like, 4 or 5 I don’t even want to call it meetings anymore.

I had ways of kind of evaluating the product in terms of questions, solution, depth, performance, scalability, as well as pricing, and then put a million dollars behind that.

Mark Organ:

Yeah, there’s no way… that’s happening.

Prem Parameswaran:

Yeah. I think there’s also, like. In some ways, anthropomorphizing… I always get that word wrong… anthropomorphizing agents is definitely useful, because the agents are trained on human-like data, and they will exhibit things that resemble emotions.

I feel like where it starts breaking down is when you take it to, like, groups and sociology, and the reason for that is Like, negotiation as a word, we use it because we think slowly and we are biased by a lot of things, but if you put a bunch of agents to interact with each other, we have seen it over and over again, that they will find an equilibrium.

Which is why Maltbook went crazy after a while. So, I’m not saying that agents buying agents is not going to happen. What I am saying is, firstly, we are already buying from agents. Sellers, they don’t look… they’re not traditional… I mean, agents are, at the end of the day, software, right?

Like, there are bots that are running on Amazon and on the buyer side that are actually making these decisions. I think the first thing that’s going to change is what does the buyer get now? And if you put them together to negotiate, I feel like For a certain set of decisions.

They will reach an equilibrium where spending time on compute becomes a useless activity, so you’re just throwing compute at each other to figure out, like, your agent is better than mine. I think that will just go away, and the discovery and the marketplaces will just become more transparent.

I think there was a question somewhere in the document which talked about transparency. I think they’ll just become transparent. And you could in some ways, argue, as an abstraction that’s already happening today, when I use AWS or any of these on-demand tools, they are agents buying and selling on our behalf.

ad exchanges are agents buying and selling on our behalf, and what you will notice is they don’t form committees, right? Like, you basically are responding to external demand, and not just on compute power.

So I would say… so I at least don’t feel like, that world is gonna happen, and I think this is the point that Mark was trying to make, which is that where… so it’s not… the human involvement is not going to be based on capability, it’s going to be based on consequence.

If it is of consequence, then regardless of who ultimately makes the decision, whether you use an agent or not, it is… it is your job that’s going to be on the line.

hudson:

Yeah, who gets fired?

Mark Organ:

I’d like to jump in, hang on, I just got, I gotta, I gotta…

Tooba Durraze:

clarify my point about the negotiation thing, because I think I… when I use the word… like, counsel, it, like, kind of derailed the conversation a little bit. Negotiation is a mathematical concept as well, right? Game theory, etc. Like, you trade in machines, you trade packets, you negotiate based on the objective you’re trying to reach.

Where do things fit? When I say this agent to… one agent to multi-agent interaction that I’m talking about, where multi-agents is what I’m terming as a council, and that’s the wrong word for it. It’s basically, like, what is the information? Even if the human’s on top, I don’t think agents are signing off million dollars on products right now.

We don’t know if they will in the future. But even at the end of the day, if there’s a human on top, the question that you’re trying to solve is, how are you getting the best information to make that consequential decision, the million dollar decision?

So your buying agent is, like, responsible for giving you all the information that you would need to make that decision the best way possible. When I say it’s negotiating with other agents, it’s… right now, we think of things as very binary.

Like, one agent researching a product is going to go and find bits and bobs about those products, like, one-on-one, one-on-one, one-on-one.

The new thesis is that… they’re going to do that in that kind of a dimension, where then, like, rebuttals, like, if I write about a battle card about a different product, obviously my agent’s gonna write it in the best way possible for me, versus a different agent.

So, doing it together in a room this is, like, Andre Coparty’s, like, concept, right? Doing it together in a room basically means that the buyer agent ends up with a better understanding of, like, what it’s… what are the differences here? So then, what they surface to the human.

Mark Organ:

We gotta… hang on, I got two, but I’m gonna cut you off here, because I want to rustle this back onto, back onto the list.

So I wanted, actually sort of answering one of Hudson’s questions and move on to one of my most, passionate topics, which is that of pricing, because I do believe that the era that we’re in is actually one of, tremendous innovation in the way that we price. Pricing historically in the last 20 years in SaaS has been actually pretty dumb.

So I’d love to get your, yeah, I mean, even, I remember seeing this quote from you know, from, Sam Altman, how he picked the pricing at OpenAI, like, he just did that. There is, like, no science really behind it, but how does pri- how does pricing… how should pricing change when the first evaluator is potentially a machine?

What should we be doing with respect to our pricing in order, in order to win in this new world?

hudson:

I think it’s gonna be adaptability. Like, like you mentioned now, like, pricing is a mo… pricing is a business strategy at this point, right? Especially for AI agents, where they’re not tied to the same kind of constraints as a SaaS platform.

you can do so much more in so many different ways with AI as kind of, like, this new frontier of technology. Where it is either outcome-driven, usage-based driven, you can still have seats. I don’t think that world’s completely going away, but I think it’s gonna be very different. Because, like, we’re talking about agent-to-agent buying.

One of the things is agents are gonna be buying SaaS. So SaaS, I don’t… that’s the other thing, too, my take is that SaaS is not going away, there’s no SaaSpocalypse.

I actually think there are gonna be more people using SaaS, but the way people are gonna be using SaaS is likely through a terminal, and they’re going to be calling it, just like how Salesforce has gone headless. There’s… imagine a world where all SaaS goes headless, what that looks like. So the long… the… My answer is that I don’t know.

However, it has to be adaptable to what the buyer wants. Whatever the agent wants is gonna demand it. If it’s demanding usage-based, credit-based, outcome-based, I think the market’s gonna adapt to that.

And I think we’re in the phase of the industry right now, is that we’re all testing this theory, if usage and outcome-based makes more sense than seat-based. Most trending is yes, but I think we’re about a few months, a year out before people say, like, it is the right answer or not.

But everyone’s testing it, so I think it forces us to have to compete in that type of a model.

Mark Organ:

Yeah, I agree. Anyway, who else is doing interesting things around pricing?

Deepinder Singh Dhingra:

Yeah, my perspective is that there’ll be multiple models, which is whether that’s usage, work throughput, which could be transactions, number of leads, number of deals, what is not going to work, and which is not going to be sustainable, and where I’m taking a very contrarian view as outcome-based pricing is not going.

hudson:

org.

Deepinder Singh Dhingra:

Right? And it’ll not be sustainable, because the first year, I generate outcome. And by outcome, I mean real value, right? I don’t mean outcome, like. Right? Means, like, improve revenue, reduce costs, stuff like that, right?

So when you actually do outcome-based pricing, the first year you’ll get that, but it’s sustaining that benefit year on year. will not work if you are on an outcome basis, so that is not going to work. But yes, I think whether it’s usage, whether it’s unit of work.

and unit of processing, base pricing, and whether you could use it tokens, you could use compute, whether it is… whether it is number of deals, you know, executed, number of transactions executed, that… those pricing models will definitely continue to, continue to operate.

Amanda Kahlow:

Foundation models serve you and serve us as the vendor, and I get it as somebody who is like, okay, I have to look at my gross margins and, like, okay, what am I getting charged to, you know, fund the foundation models? But on the other side of it, I think outcome-based models have proven to work for the support type of companies.

So for, like, for Agent Force, for Sierra, for DiekaGon, who have transactional outcomes that you can Easily done.

Deepinder Singh Dhingra:

Transactional outcomes, yes, but not business value outcomes, but not business value outcomes, because the business value outcomes don’t serve the buyer either, right? So if you say the business value outcome is I’m reducing your cost by 30%, year on year, try that.

Amanda Kahlow:

efficiency outcome, but what about a revenue outcome? So, I mean, we’re…

Deepinder Singh Dhingra:

Or even increased revenue by 5%?

Amanda Kahlow:

It’s not an increase by… it’s basically, if I generate revenue for you, if my superhuman, which they do, my superhumans close business, if I close business for you, I want a piece of the pie on the other side of it, just like a salesperson. And I think there is absolutely a world that we’re moving to, and it’s already working with our customers.

It may not work.

Deepinder Singh Dhingra:

Younger over, that’s not going to sustain.

Amanda Kahlow:

Enterprise.

Deepinder Singh Dhingra:

customers that are.

Amanda Kahlow:

We’re doing this with us, so…

Deepinder Singh Dhingra:

It’s not going to be sustainable year on year, that’s all I’m trying to say.

Amanda Kahlow:

Because you’re looking at trying to say an incremental, but I’m just saying, if I’m generating revenue for you just like a human.

Deepinder Singh Dhingra:

What’s gonna happen year two, year three, it’ll be the new normal, and there’ll be other players that are going to undercut that pricing model that you are just selling. It’s not going to be sustainable.

Amanda Kahlow:

You don’t have a… you don’t have a mode in your.

Deepinder Singh Dhingra:

How to miss that.

Mark Organ:

Okay.

Deepinder Singh Dhingra:

For your SaaS company, right, outcome-based pricing, never sustainable, only works in high-value concerns.

Amanda Kahlow:

Sorry that you’re so definitive of what works in my business model. I wouldn’t do that to you.

Deepinder Singh Dhingra:

No, no, I’m… it’s based on.

Mark Organ:

But that’s…

Deepinder Singh Dhingra:

I used to sell… when I used to…

Mark Organ:

I’m sure you did. I mean, you know what, you guys, you guys could both…

Amanda Kahlow:

I have a pretty impressive resume as well.

Mark Organ:

I’m sorry, I should…

Deepinder Singh Dhingra:

Alright, jump in.

Mark Organ:

You know what, you guys could both be right here. You guys could both be right. It could be worth…

Amanda Kahlow:

I’m not saying he’s wrong, I’m just saying he can’t tell me what works in my business.

Mark Organ:

No, I agree. It’s not what he’s saying, though. It’s not what he’s saying. It’s not what he’s saying. You both could be right, that it’s working for you now. Depender is making a point that it may not work for you in two years, that’s all.

Amanda Kahlow:

not definitively, none of this can be definitive. Nobody knows what the future’s gonna be, so I wouldn’t say.

Mark Organ:

It’s…

Amanda Kahlow:

And, you know, he started the conversation, he was the only one with a context graph, just… that’s just not true either, but a lot of us have context graphs, Tuba, myself, Hudson, etc. So, it’s just no definitive here. We’re all trying to figure it out and get to an end that I just don’t.

Deepinder Singh Dhingra:

I’m useful, right?

Amanda Kahlow:

that my business doesn’t work.

Deepinder Singh Dhingra:

I’m just providing an opinion on the pricing model, not on your business, right? On the pricing model. That’s what I started with, right? That’s number one. And number two, I said, based on my experience, obviously, we are all here for our experience, based on our experience of having tried such models. Right?

These outcome-based pricing models don’t work on a sustainable basis, except for high-value consulting.

Mark Organ:

Yeah.

Deepinder Singh Dhingra:

And that’s my experience and my point of view, that’s all.

Mark Organ:

Yes.

Deepinder Singh Dhingra:

like you said, humans might be hallucinating, right, more than agents, and so… but if I’m hallucinating, so might we all…

Mark Organ:

But I think one thing we can all agree on, though, is that buyers would like at least closer to outcome if they can, right? I mean, buyers would like.

Amanda Kahlow:

Absolutely.

Mark Organ:

buyers would like to take less risk. That’s always been true. Buyers are very risk-averse. So, you know, certain markets may be able to go all the way to outcome. Others, cannot go all the way to outcome, but I think one of the things is that at least getting closer to outcome, at least from what I can tell.

is the direction that software is moving to, and I can imagine where buying agents get pretty sophisticated about demanding from sellers, I want pricing that works for me. Absolutely. What do you, what do you, what do you guys, what do you… Hudson, what do you think, what do you think of that?

hudson:

Mark, I think you nailed it right on the head in a really clean way. I don’t think you did it on purpose. I think what everyone.

Mark Organ:

That’s probably an accident.

hudson:

No, Mark, you’re too modest, too modest. Like, honestly, I think it’s all about de-risking. Like, the buyer wants to de-risk. What that looks like to them is gonna be up to them. Like, every company has a very different way, too. So, up in, like, your earlier question, like, how this changes.

It depends on what the company wants and what they think is risk-tolerable or not, and no company is gonna have the same. We’ve been in business for SaaS for, like, what, like, 30, 40 years, and no company has the same buying process either, right? I don’t think AI is gonna all of a sudden make everyone have the same buying process either.

We’re just supercharging our own buying and selling abilities. What it does do, like, with the amount of context and data that we have out there, is de-risking, right? Like, Mark, you just know, like.

what to us as a company, an agent, not even just an agent in my role, like, the IC media buyer versus my CMO versus my CEO, they’re gonna have their own buying journey and risk tolerance for each product and service as well. And I think what companies have to do is be adaptable to what they’re selling to, like, who that buyer is.

The issue now, I think, is that one company is gonna have multiple buyers, right? They’re gonna have the agent people buying, they’re also gonna have human buys. Like, the humans aren’t going away. Humans aren’t gonna delegate 100% of their buying decisions to agents.

A good chunk of it will, and a meaningful business aspect are gonna be there, but then which part of it is still part of that kind of buying journey? And that’s what I meant, like, they’re gonna have websites and products and services where it’s gonna be tailored for humans and tailored for agents.

What those are gonna look like is gonna be dependent on what the demand drives them, and what the context, what the information, what they want to be given. Even at that time for that buyer.

Mark Organ:

Yeah. Yeah, I think extreme flexibility on the back end, I think, is going to be key to success. You may have to have not… it’s not just multiple buyers, you may have to have multiple pricing models. to compete, and if you’re truly going to have agent buyers out there, agent buyers can come out and say, this is the way I want to buy, guys.

Who wants to sell to me in this way?

hudson:

Yeah, like, you won’t… It’s an RFP, yeah, you won’t make it on the RFP. I won’t even evaluate you if my agent can’t get these checklists. Do, like, TAN mapping and tiering. We’re gonna do the same thing with product and services at scale at this point, right?

And so that’s, like, to me, like, you have to be adaptable, and right now we’re living through it. Like, I literally pivoted my whole company from per seat to usage, now to, like. Pure, like, combination of seat and usage, right?

Mark Organ:

to figure out.

hudson:

out, like, what is gonna find trash and adapt. And that’s, for me, like, the exciting time that we have now, because everything is on the chopping block, until it’s completely proven, until the next technology comes out and we have to switch again.

Mark Organ:

Yeah. Yeah, cool, I could talk about pricing all day, but we gotta move on, and I want to get really practical here, because we have a lot of people in the audience that, you know, have no idea what to do next. They want to know, what’s the next best thing I could do? What’s the first practical move if you’re a B2B company?

What should you do to become agent-ready today? And then, what should you not do because it’s, like, years and years down the road? Let’s go with Tuba on this one first.

Tooba Durraze:

The simplest thing you can do is go ask, like, 3 different LLMs, like, what does your product do, and for whom? and how much. It’s, like, a very basic thing. You’ll at least start to understand how you’re showing up for agents, whether… however many are out there.

The thing you should not do, apparently, based on this conversation, is an agent council of fire and stuff. I’ll leave it at that.

Mark Organ:

Alright.

Tooba Durraze:

Far away for that, but yeah.

Mark Organ:

Nice. Great, succinct answer, thanks. Pram, your turn.

Prem Parameswaran:

You know, Tuba stole my answer, so I have to find something else.

Mark Organ:

Yeah.

Prem Parameswaran:

I just maybe elaborate on that and give some… give a little bit of reasoning, which is that… we all talked about this, which is, discovery is the first thing that’s gonna go, because that was the hardest part, both inbound, outbound, and all of that.

So, making sure that… whatever you put out into the world, it could be a website, make it machine-legible, and the standards are changing constantly. We just heard, Markdown, markdown, and then we said HTML.

The point, here is that you have, SEO is going to absolutely translate into AEO, and it’s going to move to inference time, not what gets into the model, because they’re going to start cleaning up that data so that the models perform better.

What will happen is, There will be a battle that goes on between companies or model providers that are trying to build better models to surface the right kind of information, and there will be… ways in which you can plug things into the model, so you gotta keep watching for what’s happening, and make sure that what’s working this week may not work after a model release happens newly, and stay on top of it.

That’s what I would… I would do, and the thing not to do is don’t think too far out into the future. We, again, said we’ve been building and selling agents for the last to two and a half years, right from when we were just all prompts, right now, building on cloud-managed agents, and the hardest part is always the people.

So, it’s good to build for the future, but there is a lot of path dependency to it. So, retain your people, work on reskilling them, it is super important to get us to the other side.

Mark Organ:

Alright, that’s great. Amanda?

Amanda Kahlow:

Alright, so, I think one thing, just like… Simple, but, like, to me, very obvious thing across, like, when you’re thinking about your go-to-market, is where do you have, like, the obvious question of where do you have the biggest pain across your go-to-market, and that’s where you want to insert AI, and it’s really where there is no business model also to put a human.

So, where you can’t scale, where you can’t get the efficiencies out of the human if you’re trying to sell, or you’re having churn problems, risks, etc, that’s where to focus.

And I think what not to do is try to solve your problems today with a human, a point solution that’s just gonna make what you’re doing today better, I really think we have to go back to first principles and think about, in this world of AI, what would I do differently, and actually think about blowing up the systems that you have to to do something that’s actually going to have dramatic impact and difference in your business.

I think what most people are doing is they’re slapping AI onto existing processes that they have, and they’re expecting a big change, they’re expecting something different, versus really thinking from, wow, we have this amazing technology at our hands right now.

Let’s not build… let’s not take the tools, and there’s so many tools that are just making the process today better, and incremental point solutions. And I just think it’s… we have to think things from, like, a beginner’s mindset.

Mark Organ:

Yeah, that’s true of every technology. I remember the brochureware websites when they first came out, and then it took probably a good decade before we saw companies really take advantage of the full capability of the web. I think it makes sense. Okay, Hudson.

hudson:

Yeah, I think right off the bat, it’s making your website and your product machine-readable and usable. And I think there’s very easy ways to do that, like, elements like text, making your documentation clean.

That’s gonna be… I think that in a few years, that’s just gonna be a standard that everything, like building SEO header section, conical tags, meta tags, like, we think about building good websites now. building good websites where agents are one of the key observers is gonna be one of them.

And if you follow me on LinkedIn or X or whatnot, I’m gonna release a new scanner, so I’ve been building the scanner to make sure our website and our product is agent-rated and agent-native, and you can basically use it to scan different things. What not to do is forget, at the end of the day, the human is still the buyer.

You make transactional decisions, but at the end of the day, of every company and every business, a human is at the end of it, so don’t forget that that is still your ultimate end goal and ultimate end user.

Mark Organ:

Thank you! Well, it’s 4 o’clock, Eastern Time, so I think we’re… we’re up, Julia.

Julia Nimchinski:

Thank you so much for the phenomenal panel. Thank you, Mark, and everyone. Mark, what’s the best way to support you?

Mark Organ:

Yeah, best way to support me, check out the website at categorynots.com. You can also find me on, LinkedIn as well.

If you are curious about your category and how, you know, are you in the right category, should you evolve it, how best to compete within it, I’d love to have a conversation with you, free of charge, and see what I can do to add value.

Julia Nimchinski:

Awesome.