How Segmetrics CEO Keith Perhac grew Segmetrics to $6.5M revenue and 5K customers in 2026.

Nathan Latka (00:01) Hey folks, my guest today is Keith Perak. He's the founder and CEO of Segmetrics, an analytics platform designed to give marketers clarity on the long-term value of their leads. Another way to say this is, you spend money today, a customer signs up, well how do you look back and go, that was a good customer, a bad customer, et cetera. We're gonna jump into it today with Keith. Keith, you ready to take us to the top? Keith (00:18) I am, thanks, looking forward to this, this is exciting. Nathan Latka (00:21) Me too, you I do want to start off with obviously the AI trends. There's so many people, you know, you see all the one shots on X and LinkedIn going, we don't need metrics tools or interfaces anymore. I can just put it into Levelable and I have a reporting dashboard. What's your mode? Why can't people one shot segment metrics? Keith (00:36) Oh, because it's so hard. It's so hard. We've had so many. So even before AI, like we were talking with we have 130 plus integrations, something like that. And we would talk with some of these integrations before the before we built them. Be like, hey, we're building out this integration. And invariably they come back and like, no, no, no, don't build it out because we're building analytics internally. I was like, OK, sure thing. And then I'll go back six months later. I'm like, hey, where's that analytics that you were building out? It's like it was really hard. So we stopped. It's you can have. AI come in and look at some data, but how do you know it's not hallucinating? How do you trust that data? And then even if you have it, look at that data. Marketing is so complicated. Marketing is A happens, then B happens, then C happens, then maybe D happens. But really there was a B plus and maybe a B minus, and there was all these sub things. There's so many different pieces going on in your marketing that it really confuses AI. We've tried AI because When AI first started coming in, we were like, this is going to be the Holy Grail. And it's really dumb. It's you can do it for simple things, but man, no, it's attribution does not work in AI as far as we've seen. So. Nathan Latka (01:50) I do believe, mean, cause you know, my main job at founder path is investing in software companies. So we get pitched all the time. And when I, where I see real competitive motes and advantages right now, if they have a page like this on their website, it's generally a moat because you and I both know these integrations, they all push their own updates every Tuesday or every week they break. You got to go in and fix them. It's not just magic. And that is, that is something you'd have to go build 133 separate agents yourself just to manage all these integrations. So you'd agree that's one of your motes. Keith (02:07) Hmm. Agreed, agreed. think the integrations are remote. The attribution engine we built, we built a full attribution engine custom for the way that we work and the data that we filter and do all the stuff. it's, it's 12 plus years of information at this point. It's 12 plus years of learning about how this all works together. And we had another 15 years of marketing agency experience before that. So this is it's a quarter of a century of knowledge. into how this all needs to fit together for people to actually use it. And it's just not something that's out there in the ethos to be able to just say, hey, Claude, build us a analytics platform. Nathan Latka (02:57) So with this, and we'll dive more into the product, but with what you've shared so far, help us understand what's the average customer paying you per month or per year to use the technology? Keith (03:06) It really varies. So we have two kind of main customer segments that we're looking at. So there's a lot of creators. So people who are using ClickFunnels or ConvertKit or NowKit, ⁓ that type of customer are using us. They're paying us around 150 bucks a month. They have smaller lists. They have smaller needs. They don't need really in-depth data analysis. So we have plans for them because we want to support those types of bootstrappers, those types of creators, because that's who we were. Then, of course, we have the big companies. and they come in and they have very specific data needs. have very specific filters. They have really complicated things that we're helping them understand their own data. also they have millions of people on their list and they're paying us anywhere from a thousand to ten thousand. I think we got a twenty thousand at this point per month. And so it really runs the gamut. Yeah, yeah, it really just runs the gamut on how many people you got on. Nathan Latka (03:56) Is that your largest customer? 20,000 a month today? That's great. Keith (04:05) What are you looking to do? All of this stuff. But, you know, at the end of the day, we built segment metrics originally because Tableau was too expensive 20 years ago, right? So 20 years ago, Tableau was like $1,500 a month. And we're like, we're a small agency. We're not paying $1,500 a month for us to have to hook everything in and then do analytics on top of it. It was too much. And so we were like, well, let's just build something ourselves. And that's we like focusing on are those small agencies, those small creators, those Everyone needs this information and it's so hard to get if you don't have a tool built for it. Nathan Latka (04:39) So how, with that, and then I get your backstory, how many current customers are you serving today? Keith (04:45) We have, we actually just looked at, believe the number was a little north of 5,000. Nathan Latka (04:51) Okay, so you have to have that number. You're an analytics company, so you have to have that number. ⁓ Yeah, interesting. Okay, and go ahead. Keith (04:53) Yeah, exactly. But here, okay, so here's a great, I wanna point in on this, because this is a really, this talks to how difficult analytics is, is because, yeah, we have 5,000 plus customers. What percentage of those are self-serve versus agencies versus agency clients versus tech partners? Like, it starts to run a whole gamut at that point of it's no longer a simple answer of how many customers we have, right? And that's why AI is so bad at this. And so you need something that, is tangible at the end and then you can layer AI on top. Nathan Latka (05:31) Those 5,000 though, those are all paying you. They're all paying customers. Keith, I'll put you on the spot for a second. Can I take 5,000 customers times the 150 per month ACV? That puts you around 700, 800 grand a month right now. Keith (05:33) Those are, yeah, all paid customers. It's... little bit lower than that because we got some partnerships. Yeah, it's not going to be that full. Like I said, some people are paying us 1020. So it's not definitely not an even distribution. Nathan Latka (05:53) Okay. Okay. Okay. Can I put a range on it north of 500 grand a month, less than a million a month? Is that a fair range or are you lower than that? ⁓ okay. Great. Let's see your backstory. Now. I love how committed you've been to this business. You're not one these flashing the pan founders. You've been doing this for over 10 years. When did you write the first line of code and how'd you your first 10 customers? Keith (06:01) I think that's a good, a fair range that we can look at. 2015, wrote our first line of code. ⁓ The engine actually just spit out a text file to Skype. So we would ⁓ just spit out a text file and then copy paste it into Skype with our clients that we were talking with. And that was the first version of the app. then, so we were running an agency back in the day and that's, we built this internally and our clients loved it. DelphiNet, yes, that's correct. Yeah. And so, Nathan Latka (06:37) This is Delphi, is this Delphi? Yep. Just so you guys can follow along, right? So again, another story where a successful software founder, a successful software calendar comes from building an agency. He's got unique insights to his customer base. He knows exactly what to build via the agency here. Keith (06:45) no, no. Yep, exactly. And so the clients were like, this is great. Can I get a login? So we had to make it pretty and then give them a login. They were like, hey, we got some we have some ⁓ of our business associates and our friends and stuff who really need this. Can you give them access? And so we're like, OK, now we have to have credit card login. And that's kind of how it got built. ⁓ We did the whole thing. First version, I think, was done in six weeks and janky as heck. absolute janky as heck. Like I said, spits out a text file, but it was enough to start delivering value. And the last 10 plus years has been us building on top of that, adding in new features, making it more usable, helping normal people be able to see the data that builds their marketing. Nathan Latka (07:41) Let me ask you a quick question. You're uniquely positioned to answer. I see this back here with Ramit, even back before he was big. A lot of people see these creators today, let's look at Hormozi, right? And they see these massive launches on YouTube, right? Or they see Russell Brunson's success moving from internet warrior forum, right? With the beautiful red branding to his books, to then a software on top of it over a hundred million in revenue bootstrapped. ⁓ What is a big mistake you think most people make when they look at the business models that someone like Hormozi or Ramit Keith (07:46) Mm-hmm. Mm. Mm-hmm. Nathan Latka (08:10) our building. Keith (08:13) I think the business models are sound. think one of the biggest problems that people think of is that they're very impatient, right? So they see Ramit...