How FlipCX grew FlipCX to $12.5M revenue and 250 customers in 2026.

Nathan Latka (00:00) Hey folks, my guest today is Brian Schiff. He's the co-founder and CEO of Flip, formerly RedRoot, a verticalized AI voice assistant that automates customer service calls. He and his co-founder, Sam, originally started the company as a ride sharing app at Cornell before pivoting to voice AI in 2018. Brian, you ready to take us to the top? Brian Schiff (00:18) Let's do it. Nathan Latka (00:19) I gotta talk about that first. How do you go from ride sharing to void? You're not just jumping to the hot trends, are you? Brian Schiff (00:24) You know it's funny when we started this certainly right Sam and I met a decade ago my co-founder we were freshmen in college at the time and back in 2015-2016 no question the hot thing in startups was uber and ride sharing and all that stuff but they were banned in upstate New York and we were going to school at Cornell So I think we were just looking to build and that was an easy place to start. It was sort of a product that you knew everybody wanted and it was a gap in the market. So we dove right in and then you just keep on running, right? Nathan Latka (01:00) This is great. You've had serious consistency. I interview finish all the time. It's like one year at this startup, one year at this startup, but I'm looking at your LinkedIn now flip ⁓ January, 2018 to present going on eight years and three months, huh? Brian Schiff (01:12) Yeah, yeah. And I don't think it was even officially flipped until like 2022. We were still operating under the Red Route name for some time there. But it was sort of, you know, the earliest seedlings of what the product and the business became. Nathan Latka (01:28) Yeah, very interesting. Well, take us into that. Let's just fast forward to the product today and then we'll go back and get your and Sam's history. So the website header says automate your customer support calls with voice AI. Is this broad or you a specific niche? Tell us what the product does today. Brian Schiff (01:41) Yeah, so I think when people, AI is the technology of our lifetimes. And when you look at the opportunities that people are zooming in on to use AI inside of a business environment today, there are two big use cases people are pointing to. The first one is AI coding, and the second one is AI customer support. So we are in this massive opportunity space of AI customer support. And it's And there are two approaches that exist right now. There are these generic horizontal platforms that are really going after companies in any industry. And then there are these hyper-focused, verticalized industry solutions. And that's the bucket that we're in. So we started in the transportation space. We have added retail and health care over the last couple of years. But it is a very deep, verticalized solution for companies in those industries. Nathan Latka (02:33) Let's talk about transportation. I give me a sense. Can you name a transportation customer you work with and then give me the specific use case on how they use you to cut down support time? Brian Schiff (02:41) Yeah, it's the largest ground transportation companies in the world. As you mentioned, we started in ride share. for example, A2B transportation is the conglomerate that ⁓ organizes most of the ground travel across the entire country and continent of Australia. And they receive tens of thousands of phone calls into their contact center every day with people looking to schedule rides, modify their rides, connect to their driver, all the things that you would expect. And we're able to automate all of those routine calls. So we automate somewhere between 85 % and 90 % of the calls that they receive, helping those customers move about their day. in the way that they need to and helping that company stay at the cutting edge. Nathan Latka (03:26) So I use A to B, I'm landing in Sydney, I'm flying to give a keynote, I want a nice little limo, I wouldn't do that by the way, I like to stay lean, but if I wanted a limo, I'd call A and B and what you're saying is if I called that line, it used to be some guy I'm getting or gal somewhere, you're automating most of that. And it's not one of these things where I'm like stuck trying to press the right number for like 20 minutes. This is like, how do you measure a resolution rate? Brian Schiff (03:47) Yeah, ultimately you want to, as quickly as possible, establish credibility with the caller that this is not your 90s era automated assistant and it is something that understands who is calling and it has a level of, ⁓ you know, intelligence and cleanliness to it. ⁓ And then as quickly as we can figure out what it is they're calling for, we're going to have the integrations on the back end with the systems that are required to schedule the trip, cancel the trip, change the pickup address, get them a price quote, all of the above. Nathan Latka (04:21) When a new transportation company signs up to use you, I imagine one of the key things you must do to activate them and keep net dollar attention high is figure out how to help them train your system with their brain, their context, whatever it is. How do you do that? And what is the thing they're usually giving you to train your system? Is it a bunch of Google Docs? Is it a bunch of call history? What is it? Brian Schiff (04:40) Yeah, so the whole benefit of a... industry specialized solution is that 95 % of what any company is going to need we already have in the platform out of the box. So when you think about those key integrations with the phone systems and with the dispatch platforms, when you think about all those types of topics that we just touched on that people might be calling for, we have those workflows out of the box battle tested across hundreds of customers and millions of calls. So when the next company up it is this dynamic of standing on the shoulders of giants and they can come in and they can very easily configure it to the specific business logic and brand preference ⁓ of their company but they don't need to go through this experience of building the whole thing ground up from scratch like you would need to working with any of the horizontal players. Nathan Latka (05:37) A lot of folks are saying, man, your enterprise software's in trouble, AI can just rip it out. But you and I both know building what you built here on my screen right now, which is maintaining all these integrations, ingesting the data, running an ETL process on it to normalize the data, then use across companies is like very, very difficult. Would you agree or disagree with that statement since you're doing it here, know, front and center? Brian Schiff (05:57) I agree wholeheartedly and it's one thing to have enough of an integration with Shopify for example that you can put them on your website as a listed integration. It's another thing to, as you said, be deep enough in it with enough customers such that you have run into all of the edge casers and you have made it something where before the issue even arises, your team has seen it before and they can anticipate it and they can navigate around it with the customer. Nathan Latka (06:26) tons of sense. Again, you guys obviously you might know Aircall for phone systems, but Brian knows that for AMP and transportation folks they use this weird thing we've never heard of called Fortivoice and he knows he's got that integration live and it's required for an edge case, right? Brian Schiff (06:38) Exactly right, and I'm not sure they're running on that one, but somebody is. Nathan Latka (06:42) Yeah, fair enough, fair enough. Okay, talk me about pricing. How do you get people paying for this thing? Is average ACV called mid-market enterprise? How do you think about that? Brian Schiff (06:51) Yeah, so inside of these industries that we operate, one of the beauties is this works for companies of all size. And because of how easy it is to get up and running, it's not prohibitive and it's not just for the large companies in a space. So. For all organizations, we're able to offer no money up front. We handle the setup and the integration at no cost. And then we do what we call listen mode, which is basically intaking ⁓ a couple thousand of their phone calls to understand across, if we use a retail example, there are 200 different call topics where we have automated workflows out of the box. What are the ones that are most important for them, for that company, which then establishes a roadmap for what the prioritization is going to be for turning on new automations week over week. At the point when we do that, we'll enter into a proof of concept period. And the customer is primarily just paying us $1.50 per contact that we're able to resolve end to end. Nathan Latka (07:52) $1.50 per inbound contact request that you resolve end to end. Okay, interesting. And are those so so there's obviously paid proof of concepts then what how do you do you move them on to a plan where they're buying a bulk number of resolved tickets so it's $1.50 times a minimum of a thousand and they're paying for your how do you move past that? Brian Schiff (08:09) There's nothing worse than over complicated pricing models. So we try and keep it really simple and aligned with the outcomes that we're driving. So it's just a monthly invoice that is capturing the amount of usage that you had at the billable rate. Nathan Latka (08:26) Okay, so I'm now a CFO at A &B Transportation. I'm going, Brian, listen, buddy, we love your software, but you are resolving so many cases. I cannot predict what our bill is gonna be to you every month. Give me something predictable. Set a limit, set a cap. How do you respond to that? Brian Schiff (08:40) Yeah, so it's funny. This sort of usage-based billing model is all the rage with AI startups right now. When we started doing it, people thought we were...