The Exchanges

Q So let's, we're gonna jump around in time, but one of the things a lot of, we're all thinking about, in two years or so, what does the AI enhanced sales executive look like? How is it, from your perspective, in the trenches, no Sony, no baloney, but so close to it, what does sales look like in two years with AI?

A Yeah. So first, no drudgery. Reps spend about 75% of their time on non-selling time. If you measure the hours that their customer facing, it's 25%, sometimes even less. Sometimes we see two hours a week, right? The rest is spent on filling in forms, one-on-one meeting, forecast calls, sifting through leaks, all of that. Much of that, well, ideally all of this goes away, but if not, then even if we cut half, that's huge. The second, I think, is blending of a role. Uh, there'll be no more SDRs, CSMs, AEs, BDRs, XDRs, all of these. Uh, I see them converging into one role, essentially. As SaaS companies, the last 20 years went hyper-specialized, where there are very specific roles. Everybody has their own unique thing, which has its advantages. First, it's easy to train people on a small sliver of their job, and it gives focus. But it also creates a not so great customer experience where they're being like handed over from one person to another. It creates inefficiencies in the company because it's hard to align the capacities of one group versus the other. But if you arm people with AI that can do a lot of the things, right? If you take a black cab here in London, right? It takes four years to get trained or something, right? It's insanely hard to do that, right?

AI assessment note: “first, no drudgery... The second, I think, is blending of a role.”

Q But implicit in that, and then I want to come back to the present, there, Isn't that head count reduction? Isn't that where you get that, that access at 70%? It's doing, it's getting, it's doing away with 200, a 1002 thousand people that do rote or mechanical work, isn't it?

A Yes. We pay people like six figure salaries to do mechanical work, but people want their roles like punching cards or filling expense reports. And I don't think anybody should be filling in forms in a system, whether it's like after a job interview or a customer call and all of that, it's not why we pay people. And I think like in. Now I could just say, hey, it was like a good conversation and the AI kind of understand what's going on and the world is moving towards applications that are like a fly on the wall that understand what's going on and actually act on your behalf and remove the drudgery. So that's like a way bigger potential than everything we've seen with cloud applications.

AI assessment note: “Yes. We pay people like six figure salaries to do mechanical work”

Q Is there anything that you think, obviously you're a bull from the beginning, but is there anything you think We're too optimistic on in the near term in the next 24 to 36 months that we think AI can do that it's not going to do, that we're too bullish?

A Yes. Uh, of course. First, we don't know exactly the, uh, uh, this is evolving at mind blowing speed. Like every week there are like new developments of the technology. Uh, that are significant, not like small iterations, substantial. Each one is like more promising than the other. Uh, so it's hard exactly to predict, but we need to, the AI is not a hundred percent reliable. Everybody loves talking about like hallucinations or accuracy, which is a real thing. A lot of talk about agents right now that still don't work despite all the things I think that they will, but don't work. The autonomous engine, not, not quite yet, but. We need to look at the things that is actually that an 80% accurate job is, is incredible, right? I wouldn't put it like for things that are mission critical that have to be a hundred percent like, like a Mars mission, right? But there are so many things that if it works like it, it's way better than, than anything out there. Actually, the one thing most people don't understand is actually, it's not very smart. It's, if you've read kind of book like system one syncing system, Two thinking, right? So this is like AI generated value system one thinking, right? It's not, it doesn't reason. It doesn't understand what it's telling you. Just get the piece of text and generate another. It happens to have like incredible accuracy, but it doesn't reason. It doesn't…

AI assessment note: “A lot of talk about agents right now that still don't work”

Q All right. Last thing I want to come to before we run out of time, even though we already hit it, I think it's so important. The 37 seats, the changes, Into next year, twenty-twenty-five, is it gonna be, is it getting a little bit easier in this core B to B to B? What is your, what are you seeing for next year? What are you seeing?

A Yes, we're definitely feeling it, and we're seeing customers that are buying and are growing. If they're adding seats, that's always, like, a good, we measure, like, how many companies add seats, right? So that's, that's a very good, and we also look at, uh, cohorts, like, when they bought. So far, and again, we live in a very volatile world, so I'll be cautious. There are a lot of things that are happening that are with mixed conflicting forces. But overall, we think that, uh, we're seeing growth in our numbers and our customers, uh, signals and, uh, as founders, obviously there's always like a good amount of optimism, right? You always, but it's hard to predict exactly how long this will take, right? But one thing for sure, like in It's just a matter of time until the market bounces back. The AI is a massive opportunity, a little fool will feel a tremendous growth for many companies.

AI assessment note: “Yes, we're definitely feeling it, and we're seeing customers that are buying and are growing.”

Q Once you're looking back on it, any learnings from that? Um, not we can't all do that.

A It is, it's a legit strategy and it's working. I know some companies that have, that have started building something massive. I don't have data, but it's an interesting question. I think that most of them have not succeeded. Now, this is a counter example, a very good one. It's hard to, the key to app something is like, you need really to nail it, right? You can, there was so much stuff, and then you have a user, it doesn't exactly work, it's not exactly, you need a lot of, like, fine-tuning narration, right, and Our approach, which I think most, it's not unique. Like most companies that you start with one product, like really narrow, make it really good and keep building from that. So one success follows another. If you nail it right, we did learn that we're now at a scale. It's not just a meat in a lawn. Hey, why don't you add this feature? And tomorrow, like our customers are using it. It's more, it's way more complicated, difficult to get like our sellers that they have a product that they know how to sell. And they love, and it's working well to get them to sell new stuff, right? And to equip them the materials. So the whole expansion of multiproduct is a learning. Now we're at our third product and we see that it's actually getting better. We're still not great at it and you need, there's a good learning. So you should think about not just how to get the product out, but …

AI assessment note: “So the whole expansion of multiproduct is a learning.”

Q you said a couple minutes ago that Some of this convergence in platform is not primarily about cost savings. It is, it's number two, but it's not. So unpack that a little bit. What's, if it's not to save money, is it really just vendor fatigue or simplification? Because we've always had best of breed apps since you and I started. What did you really mean by it's number two?

A Yes. So people call it the swivel chair. If I have a role, let's say I'm a customer support person, which apps The people like use, okay, my reporting is this, but my dollar is here, and my forecasting is here, so that's, It creates some productivity loss for the people in the front. It creates like a big headache for the people that are in like in their operations or enablement or IT, which personally, I don't think it's like a, it's a strong enough reason to do this, but these are people that own the budget. So I can't argue with that. It doesn't matter what I think, right? This is how they feel. Oh, I'm managing like 20 apps. I don't have time. I'm going to three now, right or wrong. It's, it's a brutal decision. I do think that there's something that people aren't thinking about today because it's still Out there. But again, it's the AI revolution. People still think it turns, oh, I'm going to click here and click there. That's according to me, right? If I'm right, it's going away because I'm gonna say it's all going to be like speech. I'm just going to talk. We're going to have a conversation. The apps are going to run in the background and AI will orchestrate. You can see somebody like Microsoft pilot. That's our approach. There is no more like user interface. This is all chat GPT that so It doesn't really matter. There's still some operational headache for people, though…

AI assessment note: “It creates some productivity loss for the people in the front. It creates like a big headache”