00:11:39 · George Bonaci
[00:11:39] S0: Yeah. We do this a lot with new products. So we... We'll have like a new product, and generally, like, the product manager is responsible for for proving product market fit. It's not just like you build something and throw it over the fence. So there's a nice overlap between, like, when the PM is kind of in charge of everything, and then like when growth gets involved. And, like, that overlap is I I actually think where a lot of the magic happens. You really understand who the ideal buyer is and how to speak to them and what sales enablement is needed and what marketing channels work. So so that's that's kind of like a Goldilocks phase that I think is probably very fun. But to give a a more tactical example, I mean, one of our our most successful growth leads, the company and growth lead is some them might lead one of these zero to one pods or or one to 10 pods, was someone that took on our our accounts payable product back when it was very, very nascent. And we're effectively doing no marketing for it. It was brand new. And so there were no leads. And it was like, okay, well, if you can get this to 10, 20% of our total leads, that would be a pretty big success. We're kinda making a number up on the spot because there was a pod of one, so the rigor could be low. And then like after a couple months, she was saying... Learning some things and showing some tractions like, okay. Let's go and like, let's I I think the opportunity is a lot bigger. Let's really scope out like what this can be long term. And then that turned into like a... An over one year project. I mean, she ended up building a pod where she had like dedicated SDRs to it. They were excited to sell this product. They saw the opportunity as like an entry point versus some of our like more established products. She had like designers that wanted to work on it. She had like... Our stunt team got excited and wanted to run a stunt around it at one point. It's like if there's that much energy around something, number one, you must be doing something right because people want to be associated with things that are going to be successful. Number two, becomes like a flywheel. So there's almost like positive and negative reinforcement loops with like these types of projects where like success like begets success success.
00:13:18 · George Bonaci
But at the same time, the failure, like if you're failing at something, like it's generally just you're a very small team. And so like you can correct very, very quickly. And again, think that's why having like true believers on that pod is is so important where they get the resilience to get over that that failure or get through like whatever tough part is going... Is is going on with that project and still like get people excited as you define success. So I think accounts payable is a great example, but honestly, every product we launch has like some element of like PM obviously like taking it to to to product market fit and then like growth trying to figure out how to pour fuel on the fire. And you don't always succeed. I think there are like a number of times where, I don't know, we thought like a certain vertical was going to be like a huge success for us. Then we saw... We found some like major product gaps or some major like distribution hurdle. And we're like, okay, the ROI actually isn't worth the effort anymore. And we pulled back from it.
00:14:04 · Guillaume Moubeche
[00:14:04] S1: Do you struggle sometimes like to allocate the right resources to the right project? Because as you mentioned, you know, like people need to get excited to join like a specific project. Sometimes it feels, you know, like internally you have like a lot of freedom to come up with your own ideas, come up with your own project, and it also feels like tons of mini startups within ramp. So how do you make sure that competition stays healthy and that your best resources are actually like working on the best project or where you have actually the best conviction.
00:14:37 · George Bonaci
[00:14:37] S0: That is ultimately, I think, like the leader's job. So it's ultimately, I think, my my job. Like resource allocate... It's like, what is the job of like a leader? It's like, okay, it's to have some general direction you want people to go in, to it's have the right people on and then it's to make sure that they're resourced. So I think the resourcing aspect is is ultimately on me. Now, if there's like a project or an initiative that people are just not... No one's excited to be on, or there's a critical function to your point that is needed but is not excited to be on that, then then that is ultimately a failure of the leader of that pod or of me to clearly articulate why this is important to the business. And I think there will be incredibly rare cases where the individual like understands why it's important to the business, still just doesn't wanna do it. And I think that probably means we may have hired the wrong person. Because ideally, you're hiring people for these types of roles where they have really strong business acumen, and they would be excited by working on something that is of strategic importance to the business. If you're not excited by that, then you shouldn't be at a startup. And we're still trying to run as... Run ourselves as a startup. So anyway, it's like, I think putting thought into those types of, like, self reinforcing loops is is very, very important. Whether it's like very tactically how you run a channel, but also just how you you build an organization and hire the right people and have the right operating cadences.
00:15:45 · Guillaume Moubeche
[00:15:45] S1: Nice. I wanna talk about, you know, like, models war and also like AI expenses. Because obviously, like, you can see which models companies are using and also how much you're spending across different providers. So what are you learning that the rest of the markets can't see yet?
00:16:03 · George Bonaci
[00:16:03] S0: I would say probably three things. And to be honest, these are all pretty obvious, but I think it's nice to have data to validate sometimes the obvious. So in this case, I think number one is there is no, like, model loyalty. I think maybe there was, like, earlier on when when AI was becoming a a a thing. Like, I remember I was, a huge Cloud Code loyalist. And then you realize how quickly, like, the technology is adapting just, like, how fierce the competition is between all of the Frontier Labs and even some of the the the open source models. And the data really does show that there is no model loyalty is like number one. I think number two, which again is very obvious and we've talked about a lot is is people just default to the highest intelligence that wastes money. To be honest, I think the user experience of a lot of the harnesses and a lot of the model... Like the products in the Frontier Labs where you choose the model and then you choose the effort and there's like six gradients of effort. It's like the average person has no idea what to pay. What's the difference between medium and high and high and extra high? So they just choose the best because they're not paying for it. They don't actually even really know what it costs. So they're just gonna choose whatever the quote unquote the best is. And I actually think that's why it's really, really important to have some sort of gateway and router that is not forcing that decision on people that have more important things to worry about. But that's that's very a separate conversation and tangent. And then I think like the third that the data is really showing that we've published some research about is you just have to be part of the race. You have to be part of the game. You have to be using AI. You have to be forcing your employees to adopt it if you haven't already. And you see that the companies that are adopting AI are are they're growing more quickly. They're hiring more people and it's like very counterintuitive. But it's just a great example of of Jevan's paradox. So I don't know. The data backs that, but I think if you sat down and thought about what was going on from first principles, you come to very similar conclusions.