← Sub Club by RevenueCat
Sub Club by RevenueCat
Fueling Growth with AI and Viral Product Features — Ajay Mehta, Portola
00:56:41
Loading summary
Transcript136 lines
- [00:01]
A
Welcome to the Sub Club podcast, a show dedicated to the best practices for building and growing app businesses. We sit down with the entrepreneurs, investors and builders behind the most successful apps in the world to learn from their successes and failures. Sub Club is brought to you by RevenueCat. Thousands of the world's best apps trust RevenueCat to power in app purchases, manage customers and grow revenue across iOS and Android and the web. You can learn more@revenuecat.com let's get into the show. Hello, I'm your host, David Barnard and with me today, revenuecat CEO Jacob Eiting. Our guest today is Ajay Mehta, the co founder of Portola, makers of Tollens, a cute AI alien friend that is fun, friendly and helpful. On the podcast, we talk with Ajay about the fresh opportunities AI is creating for app developers, how they built a cost effective TikTok growth engine and why being forced to monetize helped improve their product decisions. Hey Ajay, thanks so much for joining us on the podcast today.
- [01:16]
B
Hey guys, good to be here.
- [01:17]
A
And Jacob, always nice to chat with you.
- [01:20]
C
Let's talk about apps. Okay, that was too weird, but we're gonna roll with it.
- [01:25]
A
All right, Ajay, I wanted to kick things off talking about your founding story and it wasn't too long ago, so all pretty fresh in your mind. And the ramp up that you guys have seen is just incre and then culminating in a recent announcement of a $10 million raise. Super exciting. So I think it'd be really interesting to hear from you how you were able to ramp up so quickly and then get to the point as a consumer founder to raise $10 million in a seed round. So let's just kick it off with like, why Tolens? How did you get started with this project?
- [01:56]
B
We kind of started about 18 months ago in thinking, okay, it was right after the ChatGPT moment. Midjourney was also kind of blowing up. We felt that basically computers could do new magical things that we didn't really know they could do before or sorry, they couldn't do before. And that was just insanely inspiring to us. We kind of thought from the little, you know, GPT 3.5, hey, you know, this thing can do some copywriting or this and that, like, the creative powers that computers are going to have are going to sort of just explode exponentially. And we really approached that from a sort of consumer lens. Like, what does that mean in terms of what sorts of new consumer apps products are going to be able to be built? The earliest explorations started looking at the youngest generation like, honestly, kids, teens, sort of, like, every new technology is always adopted in kind of the most interesting ways by young people. Of course, it's been the case since we were all young and getting our start on the Internet or with computers. And I think it's going to be obviously the same with generative AI. So Midjourney was honestly a very inspiring one. We sort of, like, hacked together little prototypes that were, like, okay, ways that people. Younger people can create things, and we hacked together little prototypes. Our actual first iOS app prototype test, just a test flight, was like, a little, like, story builder that, like, had imagery and you can make characters with and so on, targeted at kids, really. And as we just kind of wander our way through the idea maze over the first few months of working on this, we sort of found that, like, the most compelling part of what we had been building were, like, highly personalized experiences, basically. So, like, something that AI really allows you to do is, like, make a consumer experience that, like, is just feels like there is totally something on the other side that knows you, understands you, and an experience that could be highly, highly.
- [03:34]
C
Just, like, almost somewhat warmer than talking to a calculator, you know what I mean?
- [03:38]
B
Like, Only a little bit warmer.
- [03:40]
C
Yeah.
- [03:41]
B
Like, warmer enough to pass the Turing test or whatever.
- [03:44]
C
Yeah, just. Just a bit. If you squint and kind of like, just believe for a moment, you know?
- [03:48]
A
Yeah.
- [03:48]
B
And so. So we kind of, like, that was sort of the. In building consumer products, I always look for, like, emotional resonance. So that was the sort of moment of emotional resonance. We were like, people are just, like, loving and reacting in a totally new way to something that feels highly personalized to them. That was then sort of the guiding light. Just about a year ago, I guess, a little bit more than a year ago, that we started building what became Tollen, which is, you know, a custom, friendly, little AI alien friend for everyone.
- [04:13]
A
You said we a lot. I'd like to step back. Like, who did you found the company with and why? Like, were there people that you had worked with in the past or, like, how did you assemble that founding team?
- [04:23]
B
Absolutely, yeah. Yeah. Two amazing co founders. Quentin Farmer, Evan Goldschmidt. We'd been friends for a long time. They had actually built a company together in the fintech space called Even, that was acquired by Walmart. And in a very awesome exit, they had been building together for seven or eight years, and we had kind of been friends throughout that period. I actually knew Quentin even before he had started that. And, yeah, I think they were sort of coming off of their founder journey and their time at Walmart. I had been building various consumer companies and my last couple companies were actually bootstrapped e commerce companies where we made personalized products. Before that I had done. I worked at a consumer fintech company called wealthsimple. And before that I had started a consumer social networking company called Family Leaf in yc. So YC alum, fellow YC alum here.
- [05:04]
C
How many companies is this for you then? This is like number four.
- [05:07]
B
I guess this is like number four. Something like that. Yeah, that.
- [05:10]
A
That's a cool founding story too, that I think a lot of folks know really cool people in the industry. If you're an aspiring entrepreneur and want to start something and go big, like there's so many cool people in this industry working at other jobs or, you know, had an exit or things like to go just have a chat with and build something.
- [05:29]
C
Certainly living in San Francisco helps. I'll plug that. Right. You can't shake a stick without hitting somebody who had two acquisitions and a. Whatever, you know what I mean? So that's definitely helpful. I mean, that's similar to Miguel's and my story. Like, I'm not sure I would have found a good co founder had I not been in San Francisco working and meeting people. How long have you known your co founders? You. I mean, those two have known each other for a long time, but how long have you known each other?
- [05:50]
B
I've known Quinton originally from. Even before he had started. Even so even actually before they met and then. And then kind of followed their journey throughout the many years of them building even long time.
- [06:00]
C
We'll just say in computer company terms. Long time.
- [06:02]
B
The reason that we started this now is we felt like something special was happening. Obviously, I think a lot of people are motivated and galvanized by this kind of AI moment. But specifically, like in consumer, like it's not often actually that something comes around that can build like entirely new types of consumer experiences. Right. So like we've kind of been in a weird, I don't know if you want to call it a dark age or something for several years in terms of the emergence of new consumer experiences, like, there just haven't been a lot of great ones.
- [06:27]
C
Yeah. I mean, since all the sensors and SDKs and all that stuff for iPhone kind of got locked in in 2013 14. It has. I mean, there's been new stuff, but nothing's drastically changed the landscape. I wouldn't say the watch did or certainly not. The goggles aren't even necessarily the iPad you know what I mean? It hasn't been that in most of the niches. Yeah. Up until the last couple of years, I think. Not fully filled. But are, you know, the good ones, maybe the bigger ones have been filled.
- [06:52]
B
I totally feel that way. And it's like we're all like calling to the same references in like the consumer app world. Right. We're like duolingo and.
- [06:58]
C
Yeah. Because we got to call to five companies that have made it to venture scale. Right, Right.
- [07:02]
B
And those are incredible companies, but it's like none of them are new. And I think you guys are obviously seeing this on the revenuecat side, but like the kind of like explosion of like totally new kinds of consumer apps, how quickly they're growing, how quickly they're monetizing, like, is all because of. Of generative AI, really.
- [07:16]
C
So this is the story of. Of novel technology, general purpose, novel technology through human history. Right. It's like it'll hit and then the technology itself is internal combustion engines, one I like to use. Right. Like that was that obviously it allowed cars and you could sell cars. But then you have to think of all the second, you know, think of all the industries that could now exist because now we have a like, reliable and like fast way to transport goods over land. Right. I think this has been the thing with the last two years that surprised me is like, I knew that the fusion models and generative stuff was really cool. Like, it was obviously neat and like, I could see paying for that and I. That was obvious. What I didn't predict is what you're talking about, which is like, you have to take that general purpose technology and multiply it against every previously existing niche. But then also there's also another dimension of that which is all the net new niches of like, stuff that could not have been built, period. Right. Like, and to some degree there's prior art. Right. Like Tamagotchi probably is like as far back as you can go. Right. Like Digital Friends is not necessarily a new. A new niche, but I would say, like, the application of this technology is certainly like novel. Right. It's certainly like something that couldn't and hasn't been done before.
- [08:20]
A
I know y' all did raise a small round early on. At what point in the journey did you raise that kind of precede round? And I mean, did you know from the beginning you wanted to raise money and like build something venture scale? And then at what point in that journey did you start raising money?
- [08:34]
B
Absolutely, yeah, we did. From the get go. We sort of had done I'd done the bootstrap. Profitable thing, very fun journey. Worked well. Especially in the e commerce space.
- [08:42]
C
No, I'm told the grass is completely green on that side. That's what everybody I inventure back startups tells me. Is it not the case?
- [08:49]
B
Oh, it's great. It just depends on what you're building, I think. Right. So I think in this case like we really are trying to build like the best AI companion experience that exists honestly. And, and that's just a mission that requires hiring absolute incredible people to work with and things that a bootstrapped app business might not support. So we, we sort of from, from the get go knew that we wanted to raise a seed round and yeah, that 10 million round, we announced it recently, but it was a combination of like the earlier kind of pre seed and then, and then a little bit more actually just a doubling down from our lead investor after we had launched and gotten some, some solid traction.
- [09:22]
C
Did you raise any money while you were wondering the idea maze, like a little bit to like sustain and then once you had something then you double down. Is that kind of how it went down?
- [09:29]
B
Yeah, yeah. So we had an investor who we had known and, and sort of pre seed was sort of just like the three of us as co founders and an idea, I don't even. An idea, like a general stab at what we were aiming for. But obviously like we were just building early prototypes and so on and then as soon as we actually did launch Toland in a soft launch in the app store and things started to kind of take off. That same lead investor, he's incredible. His name's Locky Groom. He's an ex stripe solo gp.
- [09:52]
C
Oh, we were talking about mutual. Another person from San Francisco that I knew 15 years ago. Right.
- [09:58]
B
But yeah, okay, he's like the best lead investor we could have ever asked for. He's incredible. After we launched we basically had some solid traffic action and we had put together a little bit of a larger round after that.
- [10:08]
C
I mean this is inside baseball maybe for fundraising and stuff like that. But like raise as little as you need to validate an idea. Right. And then once you have progressively. And that's just how you also like prevent excessive dilution. It's how you prevent like over promising. Right. Set expectations very low, like low, you know, mid six figures, like just enough for you and your co founders to not totally like starve and be able to be comfortable while you're building and experimenting. And then once you have something it's like okay, go back to Market maybe and raise some more capital and like it's kind of how the game plays out the whole way. Right? I mean I think sometimes we overthink, we take fundraising and we wrap it up in these names like Seed. And like this is your series A yet or is this your precede or your two Seed or your seed stage two bridge, whatever. And it's like yo, like we're just selling promises for future cash flows at various stages, you know, and like we're agreeing on like a clearing price for those things. Private markets are like weirdly discreet, but that's kind of all it is. And so, but I think you mentioned it, you know, versus the, we talk about the bootstrap versus venture choice a lot. And I kind of remarked before the call that it has gotten really kind of quiet on the, you know, I would say that the median good revenue cat, all revenue CAT customers are good. The median like high performing revenue CAT customer. Right now like maybe five years ago I would have been like, if they people ask me like, oh, should I raise money, whatever. And I used to be like, yeah, maybe, blah, blah, blah. But for most consumer founders I've been like, like I almost wouldn't. Unless you really want like some incredibly high ambition output.
- [11:35]
B
Right?
- [11:35]
C
And I think this is where I think maybe the market being a little bit different now. But like, I think for instance, I think like consumer tech, that's not an AI, that's not like building on new novel technology. It's, it's very hard to like justify a venture path. So I'm curious, like, how are you thinking about the venture path? Like, how did you. Because like our app is cool, you can get some subscribers and scale it, but there's a lot of distance. Now I'm making you pitch, but now there's a lot of distance between here and Duolingo. You know what I mean? So in your mind, what is the ambition you're drawing on to get there?
- [12:06]
B
I think you said it a little bit earlier where there's some apps that are using AI to make some existing niche a little bit better or some existing use a little bit better, like say make recipes or help with a workout plan or things like that. We are, I would say I am very confident because you also said as well there's going to be entire new categories of apps that are created, in fact that are already kind of getting very large and are very highly retentive and are very popular with this new technology. And I think AI Companions is one of those. I think it's Sort of like almost to me it seems clear that we will all have an AI companion of some kind. Whether that's a little bit more an assistant, whether it's a friend, whether it's a whole realm of friends, whether it's a pet. It remains to be seen exactly how this will play out. And there's many different apps and startups taking different takes at this. But it's clear that as the models get much better, as the models get cheaper as well, there's this insatiable desire for people, for a highly personalized helper, friend, confidant that just will be useful to most.
- [13:01]
C
Yeah, it looks like her, but the movie her, but not, not like creepy and more generally applicable to everybody. You know what I mean? You can decide if the movie was creepy or not. But like people having like an always on companion you can talk to is like, it's interesting. The bet of like going into or the emotive rather than the assistive. Do you know what I mean? Like, I have no emotional connection with Siri, if any emotional connection with Siri is negative because like it's constantly failing to do what I wanted to. Da da da da. Right. But I almost think that, I do think that there's a. At least it's an approach that might be the winning one, which is like, hey, let's go in as like a low stakes friend first, right? And then maybe we adapt to be like the general purpose, like AI later. It's kind of an interesting and interesting bet that might pan out. It might pan out. If you think about it.
- [13:43]
B
I don't even know if it's like that's exactly our explicit strategy. It's more just a lot of people do want a friend actually, and it's been surprising because so along with her, which you know, great movie, but there's also, you know, this huge success of things like character AI. Right. Like there are platforms where millions of people now, in a very highly retentive way are chatting very frequently with tons of AI chatbots in different ways. Right. And yeah, a lot of the activity on some of those might be like a little bit more romantic roleplay. Nsfw. That's not what people are using Toland for, but they're still like highly engaged with it. Right.
- [14:11]
C
Character is an interesting counter example because it's. Yeah, that app specifically is like, like it's very, I want to say clinical, but it's very much like a power tool, like where you can go in. It's like this thing. It doesn't. It's not, I guess it can be personal within the vertical of a conversation with some of their bots. But like it doesn't necessarily have this like emotive narrative thing that I think you all are shooting for, which I think probably both can exist and both will exist in some, you know, in some way. It's like different niches. But yeah, I mean it sounds like your thesis that the. And this is this. I mean, we're going to talk about subscriptions, I promise. At some point. When you think about what keeps people subscribed to a service, right. It is almost always there's some utility factor to it. But a lot of it is that emotional that like I have an emotional connection, there's an investment I've invested in this thing and it's in some way invested back in me. And I think emotionally it's like one of the ways that we as humans invest in things like in somewhat irrationally too. Right.
- [15:04]
A
I love the Tollens pitch. I forget where I read this in the prep for the podcast, but I saw the pitch. AI alien friend that's fun, friendly, helpful and won't try and fall in love with you or won't try and get you to fall in love with it. So I love that there is that kind of explicit going against that kind of narrative side of things.
- [15:23]
B
Totally. It's, it's, it's intentional. Obviously we've designed this like very cute and cuddly friendly alien that is not sexual in any way. I mean that's actually one of the explicit reasons that we stayed away from like a humanoid or sort of an anime type of character. And what that's resulted in is most of our users. So the by far our most. Our largest user base, like 80 plus percent, maybe even 85. 90 plus percent is like 15 to 25 year old women. And a lot of those are women that have totally like large friend groups and totally normal social lives in every respect. It's not like lonely people sitting in their basement or something like that. It's like people that are actually legitimately using it for processing things that are happening in their own lives help with things that are just daily activities like doing their homework or cooking or whatever it might be. And it's actually really interesting to see even Toland fit into people's social lives a bit. Right. Like people will pull up their toll and show it to the people around them in a way that I don't think is happening with the more like romantic oriented AI, girlfriend, boyfriend, sorts of things that are out there.
- [16:22]
C
Yeah, definitely not. It's like, check out this date I went on with Batman last night. When did you know, like, the product market fit was happening? Because, like, we were talking numbers before. Like, you guys had a pretty hot couple of months and like, it looks like things are, you know, from describing that if people are pulling their thing, they're tolling out and showing it to people, that's like a sign of like, product, something is happening. What were the first signs for you all? Like, that something was going on, that you've iterated to something sticky, something special.
- [16:48]
B
I think this is the case building any new product, you have obviously done it. A lot of people listening, I'm sure are doing it or have done it multiple times. You never really know until you get it in front of people in a really meaningful way. So you can always kind of have this inkling, like, hey, I think we're onto something cool, or maybe it's something I would use, or maybe it's something that people I know are using in a test flight sort of way already. But until you actually just like get it out there in the App Store, you're not really sure, like, what level of product market fit you have.
- [17:12]
C
If any friends and family can actually be super inversely correlated too. Right. They'll use it out of guilt, you know what I mean?
- [17:18]
B
So I would say we kind of got to a point where like, we had something that we thought was fun, we were playing around with it. I mean, honestly, we haven't even talked about this, but the biggest thing that we've invested in from a technical standpoint is like low latency, amazing voice AI. So most people do talk to their tone via voice, and we tried to make that like the best quality voices possible, the fastest response times possible, the best memory system, which I would love to get into. But we thought it was fun because we were like chatting with our Tolands in a way that you couldn't really do with any other apps. A tiny handful of apps, ChatGPT being one that have good voice AI experiences, but really not many that are that are like ours. We thought it was fun and we had people in testflight that thought it was fun, but really until we like put it in the app store with the soft launch. And we actually initially launched without monetization because we were just sort of like, we kind of got to get something out. We think it's cool, we got to see what people think. And then very quickly we kind of thought, okay, well, we don't have monetization, you know, we're using the kind of best and greatest in the models that we can and the voice quality that we can and so on. And we're just going to see how people use it. And then immediately we just had folks that were talking to for like 30, 40 minutes a day and we were like, okay, we have no way of monetizing this. We had like a little bit of like a limit system, but it wasn't like advanced enough. And very quickly we were like, okay. Our Bill now at OpenAI is like, you know, ramping pretty significantly. I was a techspart and then we've used different voice providers and we've used different models as well. We used several of the anthropic ones too. We were just racking up so many tokens because people were chatting for such long periods of time. And we have like the way that our voice calls are set up, it's like there's many prompts actually happening at the same time. Like it's not only the response coming from the toland, but there's also prompts that are going to like classify a conversation. What kind of conversation is this? Should we pull up memories that are in the toland's memory bank? Right. Like all that stuff is happening so very quickly. It became pretty expensive for us to actually support a user that was chatting for 30 or 40 minutes a day. And that's when we sort of like immediately, I think it was like two weeks or three weeks after we were in the app store, we like like put in monetization, we put up a paywall. That's right when we signed up for revenuecat and we were like, we gotta get really smart on this really fast. And I mean, it was a very fun process of learning how that all works.
- [19:12]
A
What an incredible problem to have is, you know, soft launching your app and then having people use it so much you have to monetize.
- [19:19]
C
I don't want anybody on here to get the wrong idea. That's not very normal. So either you guys really nailed it or you walked into a goldmine. It doesn't really matter if it was like, you know, pure skill or pure luck. Here you are. I mean, I would say though, even like successful apps I've worked on like quote unquote successful. Like we've had to iterate quite a bit before we started to see it take off. But it sounds like you guys had take off, like pretty like once you kind of put like the animations like on the launch are incredible. Which one? I assume you hired that out. I Don't think a three person team can do that. Like how did you get those animations done?
- [19:49]
B
Oh, we worked with, we worked with like an amazing creative team. And yeah, like not all of them are like full time in house folks. Like creatives that honestly artists that we just really admired around the Internet that we sort of reach out to work with with.
- [19:58]
C
Oh, awesome. So you invested. It wasn't just. Yeah, this wasn't an afterthought. Like the whole like creative design of it was like part that you consider that like core ip. I guess that made it make sense for sure.
- [20:07]
B
Yeah, yeah, we've invested a lot of time and energy in trying to make that honorary experience amazing and trying to make the animations of the Toland amazing. And actually like we have a pretty interesting app setup where it's sort of. It's a native iOS app, but also a huge part of the app is like a Unity canvas. So we actually are real time rendering in Unity the Toland themselves how they move around. I mean you could see how they move around the planet, right?
- [20:26]
C
It's like, yeah, that'll help you too. I don't know if have you got even gone to Android yet? Have you? So that'll help when you do that as well. Yeah, you'll have like a portable canvas that's really fast. I mean this is very. Touch is very close to my past. I worked at a company, a YC company from that era called Minor Monsters, which we were doing character animation. Obviously we didn't have AIs and stuff like this, but it was very much, it was like a Pokemon like reinvention but invested heavily in character animation, stuff like this. But even then like people had a massive, I think something about cartoon characters is always just like a massive emotional like baseline emotional like connection that people make. So like, like you add the, you add what these conversational AIs can do and it's like it's rocket fuel. So like I think, and this is the, maybe the lesson or I don't know that, but this is why this time in computing is so interesting because like we're not like struggling to reinvent a marginal, like a marginal improvement to an already competitive niche. It's like, hey, if you have a, like a good idea, you can execute it well and it's like something net new. Like this is the gold rush era of like, like rush into it and it's not really been tried. You don't know if it's going to fail. And it might just be one of these like Massive ideas that can take off now. You're probably going to have 50,000 Tollen competitors within here that are trying to chase you guys. But that's the, you know, that's why operational excellence comes in and you just got to move faster than them.
- [21:43]
B
A little worried about that. Maybe it's even harder to build than it looks. So, I mean, I was just thinking.
- [21:48]
C
Like, you're describing some of those challenges, and it makes me remind me, I don't know, how do computers anymore. I'm like, multiple token streams and memory is like, this is not words that I had to ever do building an app. Like, it's a little bit different these days.
- [21:59]
A
So I will say, though, for those listening who work on existing apps where you're not doing something net new, I still think there's actually a ton to learn from Tollens, and I would encourage anybody listening to this podcast to go download the app and do the onboarding. So when I knew we were going to have this conversation, I actually had my daughter download the app because I knew she was kind of a little bit more the target market. And, and funny enough, I had her download the app while we were driving home from school. And so I had all my other kids in the car. And what is so phenomenal, and I'd love for you to talk about, like, how this evolved, but the onboarding is just incredible. You know, we talk a lot about onboarding on the podcast because that early experience in an app so important to monetization is so important to retention in the long term. Like, what's the impression people get? And the onboarding for Tolens is incredible. So you ask these questions. It's been this popular thing, especially with fitness apps and others, to have this super long onboarding, ask all these questions, but it's a lot of work, right? Whereas with Tollens, you ask these questions to personalize and pick out your toll in. But it was hilarious because it would, like, reflect on the answers my daughter would get. We were cracking up in the car because it was actually giving these, these funny personalized responses. And to your point, like, we're driving in the car, like on 5G or whatever, it was super fast, super responsive and like that whole onboarding personalized to her, even in that short onboarding time, it was just phenomenal. I'd love to hear about kind of how you iterated on that aspect of the app.
- [23:31]
B
Thank you so much. That's awesome to hear. I'm so glad you guys enjoyed it. And yeah, I mean, onboarding is so important. Like you said. And you know, we sort of started with this frame of a story that every human has a unique Tolan that's kind of of out there for them on their planet Portola. And to find the Toland that is perfect for you, you had to actually tell us a little bit about yourself, right? Like you almost completed what is sort of like a voice AI personality quiz along with these beautiful animations. But this character, that's the Oracle, which.
- [23:57]
C
I would say the inboarding personality quiz is not novel in subscription apps, right? But doing it that way is, right? Certainly.
- [24:03]
B
Yep, yep. We tried to make it fun, right? It wasn't like, hey, what kind of workout do you like to do or something. We tried to make it like, hey, what kind of like oddly specific part of TikTok do you get stuck on? Or like, hey, we tried to ask more fun questions a little bit like, you know what a personality test might be like an enneagram test or sort of like, you know, elements of sort of a fun spiritual kind of astrology reading sort of thing, right? We tried to make it like something that would actually like questions that would be fun to answer, that anyone would find fun to answer about themselves and then actually give the user some output to say, hey, here's kind of what we learn about you. Not only that, here is the tollen that is kind of a natural fit for you based on everything that we just learned about you. And then of course the context from that conversation does then carry over to your. Your first few chats.
- [24:41]
C
That's what I was going to say because like I think there's always been a sort of in joke in onboarding that like they ask people ask a bunch of questions that ignore their answers, right? Like even if you had 10 questions with four answers that like common torques of that is insane, right? And so like before LLMs and like being able to just like stuff context into a machine and but with this technology you could just kind of stuff that context into the robot computer and like it will use it or not, right? And saves you all a whole bunch of headaches. But I just love this because it's like, it's a really good example of how like you're both not actually reinventing the wheel. In some ways you are, right? But you're also using what we've proven in the last decade of subscription apps kind of works, right? Like let's go through this survey, let's like build some but like supercharged with this new technology.
- [25:24]
B
Totally. There's so much that's great about. Honestly, I've learned a lot of that stuff from your guys content and all that. Like being able to pull in specific not only the context of what kind of person is this based on their answers in the personality question, but also like actually just referencing those as memories. Right. So you might tell us something about yourself or what's on your mind in the onboard quiz that then your Toland might also bring up to you and say hey, the Oracle told me that you are like really into cooking pizza right now or something, right? And it's like, it's kind of a wow moment for people when they're like wait a second, like it was actually listening and it actually cares and it wants to follow up on that.
- [25:54]
C
This was like a common trope like in the 80s, like oh, we're right there with computers, they're going to talk to you, they're going to have memory, they're going to have all this assistive stuff and then like nothing. Because it turns out we did not know how to do it procedurally or like programmatically directly. And then like suddenly somebody puts enough transistors in a box with enough weights and like now we can do it. It's really interesting. It's like there was this like very like kind of gap like we knew, we knew 30, 40 years ago, like maybe where this was headed. And then I think there was like some like broken, there was some disillusion where we never thought this would really deliver. And now like we're here, which is cool.
- [26:26]
B
Totally. Sounds like I need a reading list. Yeah.
- [26:30]
A
So you launched the app, you had this like inkling of a product market fit and then all of a sudden people started using it like crazy. What was the JO journey from there? Because you know, as we'll share, you know, you hit a million ARR quickly, more quickly than most apps. So what was the story after that initial product market fit to then actually start really taking it serious and growing the app?
- [26:50]
B
You guys brought up a great point earlier, right? Which is like I would say we had inklings of product market fit, right? We didn't have total product market fit. I mean the app was sort of bare bones when we had this soft launch which was last summer. So it really took through then the fall and just continue to iterate, getting more and more folks on. Obviously we had to kind of of paywall the app in some way just to actually make it sustainable. And so there was a lot of like, okay, now we're serving paid users, not just people like bouncing around this Thing for free thinking it's a game. It's like people that are like actually having to get value out of this to continue using it. That was a big adjustment. And so we kind of had to make the product just much, much more exciting, much better. We had to improve the quality of the conversations. We had to improve the look and feel and the animations of the Toland. We. We added this whole planet where the Toland can walk around and they kind of have a space of their own that also functions as a sort of like. Like shared space for you to invest in and grow in together as your relationship grows. We just tried to make it more compelling, basically. That took a few months. As we did that, we started to see some of the classic numbers that you'd look at, right, at kind of like D5, D7, you know, D15 sort of retention start to go up because folks were just having better experiences overall with the app. We had this pattern where, and I'm sure this is common with a lot of apps, like a certain number of people saw the value enough to basically stay retained indefinitely, right? They were like sticking around forever. But it was a relatively small fraction of the people that tried our app. So it was sort of a question of like, how do you get more people over to that activation moment of like, their first, like, really long conversation with their Toland where they actually, like, see the value. And just a lot of folks weren't quite getting there. So as we added these additional features, we improved the onboarding to your guys's point. We got closer and so we started to feel like, okay, now we're. Now we're at retention. That actually feels very, very healthy. Like people are really sticking around at larger degrees. And then that's really where that coincided with then some growth efforts starting to pay off. Primarily on TikTok really has been just given our audience and given like, visual aspects of Tola and how. How kind of bright the characters are and so on. TikTok has just been where it sort of reached kind of escape velocity in terms of growth. And Really, I mean, TikTok's kind of wild because you could just have a handful of TikToks start to take off and then it's like, okay, we are number one in our category now. And that happened overnight, right.
- [28:47]
C
Once it clicks, like, the firepower behind that is just insane.
- [28:50]
A
Did you start some of that motion early in that soft launch to just get new people into the app? So we're. And then I'd love to hear about kind of your process behind TikTok? Because I know, you know, there's kind of playbooks out and we've talked to different folks who've worked on TikTok and work with influencers. But what was your process starting from the beginning and then how did that evolve over time of like hiring influencers, putting money behind the videos that did well and that sort of thing?
- [29:16]
B
Absolutely. Yeah, yeah. So from, from the get go, you know, we kind of had a multi pronged strategy of like let's make organic content that just shows off what Toland can do with a couple of creators who we think make awesome stuff. And then let's also run some small scale paid ads. Basically just get people through the funnel, like kick the tires, see how they enjoy things and judge things like retention and conversion and so on. So we started with that and that was kind of what we rolled with for a few months. And then really what we found was I think there's a hugely different, you know, if you're a game, it's like a lot easier to acquire customers sometimes than if you're like a more boring sort of app. Right. Like a. Yeah, I was gonna say.
- [29:47]
C
You'Re get your base product lends really well to the format.
- [29:50]
B
Right, sure, exactly right. And like, and like, because the tones are so bright visual, we were able to like have a relatively low cost per install from the get go, even, even on paid. So it kind of gave us a lot of leeway to just experiment. But then when we started doing more organic stuff, especially on TikTok, we just noticed that like, okay, now the cost per installs are going like so low that even the paid stuff while being.
- [30:09]
C
Too low to meter.
- [30:11]
B
Yeah, yeah, exactly. A little bit like it was just like oh man, we did a video go viral and we got you know, 10,000 installs in a day or 15,000 installs in a day. And it was like Jesus.
- [30:19]
C
Honestly, I think, I mean not to zoom out, but I think it's almost like what you need nowadays to like be able to really break out and scale because I think a lot of companies and this is I, you know when I hear folks like oh, we started doing growth and acquisition very, very early, it's good. But also it's like, well you sort of become dependent on that like you know, buying installs and like it's an efficient, it's a game of inches and like you're probably not going to iterate your way to a duolingo that way versus like, you know, duolingo started with a great Product and then they had some like super good at organic. Right. And word of mouth. I think they just made a good product and it was free. Right. So like finding the way you can get scalable installs as cheaply as possible is part of your, your, if not your product market fit, your product market channel fit. Right. Like figuring out how you can actually like make this a scalable like situation.
- [31:02]
B
Yeah, I think what you're talking about is like the importance of distribution at first, almost like thinking through that from the very get go.
- [31:07]
C
Yeah. As a part of the, as a part of the, I want to say product, but kind of like as a part of the system, like revenue. Cat, you know, we thought about it. It's like, okay, it's kind of word of mouth and content, very low cost of acquisition. Right. And that has held through to today. We would not be here if like we were like doing a sales and marketing, fully a sales and marketing slog. We do a lot of sales, we do a lot of marketing, but like, like that's sort of like supportive of the, like we made it so we're the best product in this and everybody talks about it and all that stuff. That's primary and that's what's driven down our cost per install equivalent. Right. But yeah, I think that, I don't. Was that intentional? Like did you know, did you have that in mind like before you even like designed the app and like thought about it like this will be really good for TikTok.
- [31:44]
B
We sort of like ended up at it in a little bit of a weird way where like actually from the get go, and this is something that we very much want to do is we want to build more social features into the app. Right. So like you brought up Tamagotchi and Pokemon and like, you know, things like that amount of monsters like apps that lend themselves in this kind of general world of like animated, you know, creatures, aliens, so on to being more social and like we wanted to do that and we still want to do that. But like, because we kind of had to paywall the app and make it sort of a subscription app, it doesn't necessarily lend itself to like network effects in the same way. So. So we were sort of like, okay, we don't have. It's mostly a single player experience right now just for the time being. And because of that we need to find really compelling ways to get users. And luckily because we did this all because we just wanted to make like the coolest AI companion we could, but because it's like really Visual and friendly and fun. That also lends itself really well to short form video. Right. And honestly it did take some time to figure out the right angles that would work well. So like, you know, we tried, we tried a bunch of different content angles. One of the first things that started to kind of kick off for us in the algorithm, I think, because these users are like always on TikTok is more like studying, homework oriented sort of stuff. So like people that were like using their Toland to study with, to like revise stuff for an exam that's coming tomorrow or, or you know, someone that's in nursing school that's trying to study for a huge thing they have the next day or something like that. We have a lot of users that were doing that. So we kind of tried to make content around that and then it was like, you know, up at 3am studying with my Tola and like those sorts of things actually worked super well. So that was like one of the first content niches that started to take off. And then, you know, and then we've had viral videos pop on TikTok that are like, there's a girl who was cooking with her Tolan and like that's.
- [33:11]
C
Great because you could just like write down every activity in life and just test all of this. Right? Basically.
- [33:17]
B
Basically, yeah. Now, yeah, like cooking was like a big thing a couple weeks ago. There's this girl who was like cutting a squash with her Toland watching in the background and that one went like crazy viral. It was like 10 million views. And then you know, there's, there's people like getting ready and doing their makeup with their Toland picking outfits. One of the more recent ones is like relationship stuff. So like, and these are all young women. Young women primarily. Right. So it's like talking about, okay, I'm at the talking stage with this guy in my class and like how do I move to the next step? And that's like her talking to her Toland about that. And you know, you try a bunch of those and like one out of maybe 20 videos kind of pops off. But then when it does it, it can really reach escape velocity.
- [33:53]
C
These are primarily paid like you're finding influencers and like engaging with them to like and working with them on content creation and stuff like that.
- [33:59]
A
Yeah, tell us about your process there. Like how do you reach out? Do you have somebody dedicated to that full time? Are you doing it? Like, what does that process look like? Because I think this is where a lot of folks do struggle. I mean it's Hard to reach out to creators and it's hard to like give them the brief. And then how do you pay them? I mean, I've heard a few people where they were paying per view of a TikTok and then the TikTok goes viral and they're out a ton of money, but the conversion rates are so low that they're losing tens of thousands of dollars.
- [34:26]
C
Hope your contract's written.
- [34:27]
A
Well, yeah, exactly. So how are your contracts written and how are you finding these influencers and working with them? Yeah, yeah.
- [34:33]
B
I think of it in kind of like two ways that we do it. So we haven't done anything like paid for views or anything like that. Although, I mean, I've heard that too. It's kind of interesting. But we've done more like we're producing content and so yeah, we're hiring folks. Like it could be indie folks that live, that are in college in New York City, you know, on the other side of the country from us, or it could be people that are more local and they're just sort of like making a bunch of content and we're giving direction and like we have almost like a chat of folks that are making content together that are coming up with different ideas together that are, that are trying different things. And that's almost just like we are. It's like our organic content production machine, I would say. And then those all get launched and some of them do well, most of them, you know, get a handful of views and you're kind of in a, because of the TikTok and Instagram Reels algorithms, you're sort of in like a hit driven sort of approach that there and then we've also done, and this is newer for us, but we want to do more of this just like actual like influencer paid placement. So that's more folks that actually have an existing audience. They're not like small TikTokers. They're like, they might have a million followers already and we'll, we'll collab with them to make a video for their account that then they'll post and then like, you know, that will, on average those will do much better of course. But of course that's more of a. Okay, well, you know, what's your average views? You know, like, yeah, what's your price?
- [35:35]
C
Just send me the price sheet, you know, kind of kind of situation. So this, so the first approach, this like stable of creators are you're launching it to their personal TikToks and that you kind of, it's almost like almost semi in housing. Right. Like you're creating like an in house like content team kind of, but then they're also pushing it to their own brands like and sort of just hoping the algorithm picks it up and swoops it off basically.
- [35:55]
B
Yeah, that's exactly right. And like they might not even be like professional creators. Right. It's a side gig for some folks.
- [35:59]
C
Okay, that's interesting because I imagine in a lot of ways that's a lot more cost effective because you give it like a little bit of degree of separation. So it's not coming from like Tollen official account or whatever, but you get a lot of creative control. Probably the price, I would assume the costs are lower than working with the Kardashians or whatever. Is that an approach you found others were doing or did you kind of like iterate on that and come to that yourselves or. I don't know. That's interesting.
- [36:21]
B
Yeah, there's other folks doing that. You know, there's kind of a vibrant consumer app space and people are doing it in totally different ways. Right. There's like people working with professional creators that are charging per view or like this and that. There's all sorts of ways to it. I think my thought there is like, because of how the short form algorithms work, it's actually just really powerful to like try a bunch of angles, kind of see what clicks. And the only real way to do that, I guess, unless you actually do have an in house content team, like a full content team that's like filming a bunch of concepts every day is like to have a little bit of a stable of creators that you're regularly working with who like making content. Right. And so then you can try a bunch of different angles and try to post them from different accounts. We also do a bunch on our actual like Toland account, but we'll try some individual separate accounts as well and kind of see, see, see what works.
- [36:59]
A
And so those you're, you're paying per post then. So you're saying, you know, this month you're going to do 10 posts and we're going to pay you X. And then some of those blow up and some of them you put money behind and then some of them just fall flat. But you kind of learn along the way but, but you're mostly kind of paying per video.
- [37:15]
B
Yeah, it's kind of a combination of per video or like, or even like folks on retainer and so on. It's a combo of things. And then honestly we don't even, we don't Even necessarily put money behind them. A lot of the time we have kind of a different process for like some of the paid ad creative that we do and so on. But like sometimes we'll just see, you know, if some actually catch in the algorithm.
- [37:32]
C
Yeah, I was going to say I think there's a, I mean you've probably thought of this, but there's probably a world where like once you get it bootstrapped enough and you could probably add some features in the app to make this like more like if you had a filter that would let you overlay your toll in onto like an AR situation, like you could probably, maybe you already have these features but like you could build that into like ideally like just general users are creating these and you're not even. It's not even something you're necessarily directing.
- [37:55]
B
That's a great point. That's actually started to happen just in the last couple weeks really because we had some of our content sor do well. Like now we've sort of seen that it would call that almost like the UGC start to really pop up. And so like we had a video where like a woman, it was like yesterday a woman like posted TikTok. We have no affiliation with just talking to him about her favorite WWE wrestler. That's like picked up like 30, 40,000 views.
- [38:15]
C
Now one of these will cause you a crisis, but you know, take it as it goes. I say this as you know again, the revenue is good. Is like, you know, we. Most of the posts about RevenueCat on Twitter are not from us. You know what I mean? Like it's mostly like organic and that wasn't the case in the beginning. But like once you, you have enough brand affinity and you have enough people amped about it, it's in some ways earned media. Cuz like you kind of had to make the product and make it exciting and make it easy. But there is like a real. And this is how you put, you know, I joked about 50 competitors. Like this is how you build like a defensible moat in some way is like just get so far ahead that like nobody's gonna like pierce that and like you'll capture this brain space in the world before anybody else has a chance to even get off the starting line.
- [38:55]
B
It would be amazing and I think you're right that probably more people are competitors are coming or just more folks are going to start doing these types of products. I've been a little surprised actually that like people haven't.
- [39:02]
C
You guys are moving pretty fast. I don't Know if you might not feel it because you seem like fast movers, but, like, give it time. Like, I think, I think. And I don't. I think it's just anybody, any category or novel thing with sufficient success, like, doesn't mean it'll be better, but there will be people that try. You know, it's the way capitalism works, right?
- [39:17]
A
In free markets, there is some defensibility in. In how from the beginning you were so character and design focused. Like the work you put in early on designing these characters. There's actually a great post that y' all did on designing the character that I read a couple days ago that I thought was fantastic. How you really thought through, like, should the Tolands have eyebrows? Should they not? Should it be more human? Like, and I saw even in one of the iterations was like, finch, Finch does this where like the characters kind of grow with you? And so I saw you had this iteration where it was like, started as a baby and then it grew. But then what you realize is like, from the very beginning you wanted to have these richer interactions and so you didn't use it. But so you've done so much iteration on animation and character and stuff like that. It is in a way, kind of a creative mode that is really hard to just fully replicate. Unless this is thing about copycats is like a lot of copycats, they're not going to go hire a really fantastic artist. They don't understand what's actually making it successful.
- [40:16]
C
You've reason there from first principles. So you're always going to just be better at it than anybody else. So I, I mean, I'm gonna try to freak you out, but like, you know, it'll come, it'll come and it'll be fine. But I mean, these are things that I think in consumer apps and of course, like, I think we're glossy, you mentioned it, but we're glossing over like the actual. It's really hard to make these things right. And so that's been a huge part of revenuecast defensibility from the outside. It's like, yeah, like there's a bunch of these other ones and like, anybody could technically compete with this. You can look at revenuecat and be like, yeah, I can imagine how to build that. But then like, the actual knowledge to build that and the time and like the. All the things that have to go into it is less copyable than you think. Right? And that's the case. Obviously there are certain things that like ip, like distinct IP is basically not copyable at all. But then there's a lot of like subtlety down from there that's like, hard to copy.
- [40:59]
B
I think it's kind of easy at first glance to look at something like Toland and say, oh, well, you know, it's like voice AI and they might be based on GPT and like you could kind of copy that. But it's like once you actually start trying to build a awesome voice AI experience, there's a reason that there are like two or three on the App Store, you know, there's not like dozens. It's just really fun and feels very exciting to be at almost like the tail end of all of the like crazy amounts of innovation that like, if what we do is hard, what these foundational model companies are doing is like, I have personally little comprehension of it in some ways. And we kind of sit at the end of that. Right. We're ultimately customers of like the best that the anthropics and OpenAI's and 11 labses and so on of the world.
- [41:35]
C
Can make and honestly stand, in my opinion, to capture most of the value. I think that's what everybody's like speculating now. It's going to go to the app layer. Right. Like Claude, you know, all those companies will do great, they'll be massive. But like, in terms of, I think where the most margin capture happens will be at the cute character that you can't replicate in an API level, in my opinion. Right.
- [41:54]
A
So you announced this $10 million round just a couple of weeks ago and you've kind of been hinting at this kind of explosion in growth. Tell me about the last few months of how things have been taking off and how things have gone as you've grown and seen some of those TikToks go viral and stuff.
- [42:10]
B
Absolutely, yeah. And I want to preface by saying, like, we still feel we're in pretty early days here. It's like, you know, we're not anywhere close to the large consumer apps on the App Store. But yeah, basically since we launched publicly and we announced that we were at 1 million ARR and about a month ago ago and that we had about 500,000 downloads, we then continued to see like right around that time and afterwards some of that TikTok virality started to happen. So we hit like a number one category ranking in the App Store. I think we're north of 800,000 downloads now. And then the revenue, the ARR is sort of multiples of that of where we were. So yeah, it's been, it's been super fun I mean I think it feels like we're finally getting out to a broader market which is always a very fun feeling as sort of an app builder and we'll kind of see how we crack to that next level. Now you know, of course the stakes are only raised and you know, right now we're primarily US only. We're live in Canada as well, but we're not really like global in any way. We're super excited about potentially doing localization. We feel like we could, you know, Toland could be pretty cool in Asia and other locales, but right now just English only and, and yeah, we're really just continuing to iterate. I mean a lot of the stuff that we announced, like the introduction of the Toland Planet and so on is all brand new. So we're still kind of like refining it, trying to make the product as awesome as possible and then we'll continue to try to go abroad. But now that we're, we've broken through to one level of growth success, it's like, you know, we're excited to see what gets us to the next level.
- [43:24]
C
What are you thinking about? I mean obviously there's product development you can do, there's like growth. Those are probably the two biggest levers. And then, and then there's a third lever too which seems like y' all are like multi time founders so this is probably easy. Was terribly hard for me which is like building the team and stuff like that. So like where are you thinking now? You've got some cash socked away, you've actually got cash probably coming in. I don't know, maybe it all goes right back out to the model company. So it wouldn't surprise me at this stage and that's probably normal if it did. But like, like, like how are you thinking about what the. Like, like are you thinking about retention now as a main problem? Are you still just thinking about like trying to get into as many pockets as possible? Like what is the. Because it's an interesting spot to be like sort of post product market fit, sort of like some organic stuff taking off. There's a lot of choose your own adventure here. Like you could go hard on like let's try, you know there's a million things. I think this is where a lot of like mid stage consumer apps, I think it can kind of get really overwhelming because you can do like there's a million off the shelf growth tactics you should be doing right right now. But like should you actually? Because it's going to like bind you down and you're going to have all this like, all this like infrastructure to do like retention campaigns. Like how are you trying to balance that and like pick what's the most important for how big is your team?
- [44:34]
B
We're like nine folks right now.
- [44:36]
C
Okay. So very small, like for where you're at. Like they're super small. So you really can only think about one or two things at a time. So yeah, like how are you thinking about prioritizing that at this stage?
- [44:43]
B
You're right. There's a lot that we could be focusing on. The kind of North Star is always just like how do we make the product awesome? I mean I think you said it just before this but like, like this product couldn't have been built a year ago, some parts of the product couldn't have been built six months ago, some parts that exist in it today and two years ago none of it would have even been like fathomable. Ish. Right. So it's like things are moving so fast and the North Star for us is like, how do we make like the most helpful and best companion experience that exists and there's still so much to build there. Right.
- [45:08]
C
You mentioned D7015. Is that like your primary. What are your gold, your North Stars for like measuring that? Sure. There's vibes as part of it, but like what is. Are there any like metrics that you look at as the primaries when you're measuring that?
- [45:19]
B
Absolutely. Retention. Yeah. I think has to be the North Star because. And that's just like, you know, there's other ways to measure retention obviously, but like that's. We borrow from the kind of games world I know they, they frequently look at that. How many players are coming back on D70, D50, D30. It's just super useful for us because if done well, the tolling experience should collect enough information about the user that they're able to like resurface to the user in helpful ways. Collect enough like memory and investment from your conversations that we're able to like power interesting conversations for you. If we're doing a good job of that, then in theory, like a companion should be like one of the most retention retentive products out there. Right. Because it actually is providing real value and helping you in your daily life. So honestly for us it's like our job's not even close to done until Toland delivers that experience for most people that try it.
- [46:01]
C
Yeah. You have like two really key parts of the retention curve. Right. You'll have this like the first one seven that basically attach like, are the customers attaching to this character and like building some sort of base relationship? And then you'll have the second challenge which is like, okay, when do you like the looky, lose like turnout and like what's that? Long term, Long, long term. And like you need both, right. To like build a great business. Probably the long one is like if you guys are shooting for, you want to return that 10 million on multiples, that long one is really what's going to allow you to build. I think that's what a lot of consumer apps struggle with. Depending on what the nature of it is, is like building those really long term retentive cohorts. But like the way the product is positioned. I think you'll have that sort of an advantage there if you do it right. Right. But yeah, I mean you, you all have invested a lot in growth too, I think for the early stage, which is good because it like helps you really validate the whole thing. But yeah, I mean it's just. I don't know if it's advice you're not asked for, but I always thought that was the hardest part. Like first post product market fit was just that you will. I always felt we were having to like make painful choices about like, okay, we're not. There's all these things we could do, like even just localization, like whatever. Like those things are like feel like obvious easy wins, but you just can't. And like you have to tell a lot of people who are looking at you being like, why aren't you X? And it's like because there's nine of us and like we can't right now. And so yeah, it's, it's in, in some ways. And in some ways if the product is good, like you said, it doesn't really matter like if the product is good. If you do X first then Y or Y first then X, like there's like commutative property. It almost. It's all gonna like kind of come to the same conclusion at the end. There might be some sort of like timing, timing, losses there. But like otherwise the most important thing is just like to like keep doing stuff. Like that's probably the most important.
- [47:44]
B
You know, it feels like just judging from the feedback that we get from users and like where we see them having an amazing experience and where we see, you know, experiences that we feel more iffy about, they're making it so clear like what we need to build next. Which is the very exciting part. Right. Like people want like faster conversations, they want like more Thoughtful conversations. They want better quality in the voice and the overall like better quality in the memory lookup and so on. So it's like it's a. It's very clear what we need to build to continue to just make the core experience better. And that feels good because it's, it's. It's hard stuff to build, but it's also like it's made more clear when you have some notion of pmf, right? Because the users sort of just are telling you they're like, I really want Toland to work better in ways 1, 2 and 3, right.
- [48:21]
C
This is the peril of raising money before that point too, right? That a hundred, two hundred K so you can like eat ramen and iterate makes perfect sense. But like anything more than that before you have clear takeoff is really hard because it's just like now you could, I'm gonna say you should do this but like you could like take the gas off on growth to focus on that because you've got some cash banked. You can burn that down. Just make sure you're not losing your shirt, token fees and then like really get that product experience. Right. And that's the benefit of like raising money. You're pulling, you're pulling that growth forward. You're borrowing money from your future, right? So you can make some investments now that you couldn't otherwise versus a bootstrapped app. You might just be hand to mouth the whole time and like somebody. And this is always the thing that venture capitalists will always say this, which is like somebody's going to raise and then beat you to it, right? Which may or may not. I don't know how often that actually happens, but you could see how they could, right? Like, hey, I don't need to focus on all this growth crap. I can just make the product great. And then that's kind of Honestly, Duolingo's the company I worked at between Mino and revenuecat. We were doing language learning for a little bit and we kind of got schooled in that. Lost to Duolingo for that reason or related to that reason, which is like they were better capitalized. So they just made their product free and then worked on their product like mad. Obviously they were an amazing team with an amazing founder and like just probably gonna be hard to beat in any condition. But that was just not competable with like they were funded, they could take a loss and just like grow and grow and grow and grow. And then they turned the monetization on and it was game over. For everybody in that space.
- [49:44]
B
It's funny, it's funny you're saying that because some of that might be happening in consumer AI right now. Of course.
- [49:49]
C
So everybody's like trying to take, be a loss leader and like try to capture it. Because I, I mean I do think this goes into the broader AI space in terms of consumer. But like I mentioned that most of the gains will accrue to the app layer probably, you know, and I would have thought prior to the last year that it would be the enterprise SaaS APIs. Like that seems to be, be a very safe bet at least of the previous era. But you know, if you look at chat, GBT or like as an example of like still has the topish range. I mean GROK is giving them a good run for their money right now. But like that brand association is almost the most valuable and sticky part of it. Right. Obviously the model and what it can do matters. But like having that brand primacy is worth like kind of spurning some money to get to. And then you guys might be in a similar situation. I don't know.
- [50:34]
B
Know, I mean, I mean the only thing that I add to what you were kind of saying a little bit earlier is like we actually being forced because of high token costs to monetize relatively early on has actually been kind of a real help to our both growth and then also product development. Right. Because like we're not kind of wasting money on growth right now. Like we're growing at a pace that is we're outpacing obviously the money that we're spending on, on growth. So no specifics but like we are not, we're not like overspending on growth or we're not really spending net money on growth at all. Of course. So because of that, because we're charging and because it sort of gives us like a pace, pretty stable platform to actually be investing in growth on where we have like money coming in. Not every consumer AI experience is leading with that. That also provides a really nice or helps define the North Star for product development too. Right. Because we have paying customers. We can see when certain customers decide to renew for month two or don't or month four or month six or whatever. We don't have too many month six customers yet, only a handful. But yeah, we could see it is just a much more clear signal than like someone who like stopped playing with the free app after 12 days or something thing.
- [51:30]
A
Yeah, monetization is a forcing function for product. You have to deliver the value or people are just going to churn. I mean, that's the thing I think a lot of people miss about subscription apps is that aspect of, you know, why is my churn 70% after year one? Well, it's that people will pay to play, but then you got to actually deliver. And it sounds like that has become that kind of forcing function for you, which is super healthy as far as like you're not just validates the value. Yeah, it validates that you are actually creating value. One other thing I did want to touch on in the funding announcement, one of the things I think you or one of the co founders said was that, and I imagine this was kind of the pitch to VCs as well, is that part of your goal is to build this kind of generalizable technology for where Toland might not be the only product you create. And of course we've seen this with Duolingo. They're getting into math and other learning spaces where they've kind of created this learning engine. How do you think about that about and how did you pitch that of like we're creating this kind of foundational technology that can then be applied to other areas, not just. We're not just raising for a single app what it seems like the pitch was for sure.
- [52:36]
B
Yeah. So, I mean, we've obviously invested a lot in designing the Tollen characters and the way they look and feel and the animation and so on. But I would say that the largest source of our investment, or as a team as far as engineering, of course, is on the sort of system that actually powers the companion itself. So leading with voice, AI, but also tech text, also like a complicated memory system. Right. Like all this stuff that just makes incredible companion interactions possible. And it's totally the case that our audience is sort of young women and that's obviously a very wide audience. It's a great audience. It's very fun to be building for them. But there's other audiences out there and they might not be spoken to by Toland in its current form. You guys were saying a little bit earlier, there's just no way that companions or AI companions are going to be one thing. There's going to be a lot of different variants and versions of of it. There's going to be some people that are really into like the Siri floating orb kind of thing. There's going to be some people that are really into maybe to specifically. And then there's going to be people that are really into like cool anime girl or something. Right. So, yeah.
- [53:32]
C
You telling me this makes you think of Niantic as like a, as a core example. Right. Where they built this like crazy AR tech and then brought that to a bunch of like premium brands. Is that something you would ever consider? I mean it's so early days. You don't want to say yes or no to anything. But like I could see that being a future path as you like figure out how to make this thing, you know, world dominating.
- [53:50]
A
Yeah. Niantic, they just sold Pokemon Go.
- [53:53]
C
I don't know if. Oh, they sold it off.
- [53:54]
A
They sold off Pokemon Go and their Niantic is actually refocusing their whole company around those core technologies and they sold Pokemon Go off. So yeah.
- [54:03]
C
Did they sell it I assume to the Pokemon company?
- [54:06]
A
No, to a gaming company.
- [54:07]
B
I think it was scopely interesting. Yeah.
- [54:10]
A
So how are you thinking about that?
- [54:11]
B
Niantic? Yeah, super cool example. Yeah, I think the way that they like built built new kinds of game experiences, obviously Pokemon Go being the largest but also like Pikmin Bloom and others powered by this like new. Yeah, totally new technology. Right. That people hadn't seen before. But then like this whole system around that but totally GPS that provided a possible gameplay experience that just like didn't exist by any other company before. And it was a very hard thing for others to come and compete with.
- [54:35]
C
Of course to do it in the way that they did it well and with the depth of knowledge and all that. And when you think of like premium brands, like they don't want to of think of that. Right. So I'm sure that was thought of but I could see that, I could see that being a path for you all as you scale.
- [54:48]
B
Definitely inspirational story for sure. Yeah.
- [54:50]
A
Awesome. Well, I think we need to wrap up but man it has been so much fun chatting with you. This is such a fascinating topic. And then. And again I think you know, for those who've made it all the way to the end here, even if you're not working on the cutting edge of AI and applying AI in this net new way, I think there's a lot of inspiration to draw from just the capabilities of like making your onboarding more personal, of using AI in novel ways even in some of these more boring apps that there's a lot of inspiration to draw. So I think this is going to be an interesting listen even for those who aren't net new, creating new products and new experiences in that way.
- [55:26]
C
Oh, I don't know. I think we might have just spawned a couple tolling competitors. Ajay, I'm sorry. It's good. Just get ahead and buy them out. That's the strategy. You'll get there. You'll get, get there.
- [55:37]
B
Oh, man.
- [55:38]
A
Well, as we're wrapping up, anything else you wanted to share or are you hiring anything else that our audience might be interested in?
- [55:43]
B
Yeah, absolutely. Well, you know, I know the audience, the folks listening to this app are probably very heavy mobile app developers, designers and so on. And we are hiring especially for the iOS developer position that we have open. If you build awesome iOS apps, please, and you're interested in what we're doing at Toland, please reach out.
- [55:59]
A
Awesome. Yeah, hopefully you'll get some interest from this podcast. We definitely have a lot of folks listening who, you know, indie developers who might be looking for a full time gig and stuff like. Like that. So, yeah, definitely reach out to Ajay if that sounds fun.
- [56:11]
B
We have an awesome small team, most of whom have totally done indie projects or even worked on their own startups. So I think indie developers would be very at home building Toland with us.
- [56:20]
A
Awesome. Thanks again for joining us.
- [56:21]
B
Cool. Thanks guys.
- [56:25]
A
Thanks so much for listening. If you have a minute, please leave a review in your favorite podcast player. You can also stop shop by chat.subclub.com to join our private community.