May 3, 2023

Episode 263: Kirill Potekhin, Co-founder & CTO of Adapty

Kirill Potekhin started programming at 12. He made a website for his father's business and his father paid for his classes and work. At the age of 17, he co-founded Easy Ten, a mobile app to study foreign languages. They were #1 in App Store by revenue among educational apps in many countries and went through 500 Startup acceleration. Over the years they had more than 20 million users.

After that, he and his friend and colleague from Easy Ten founded PotehaLabs. They did various natural language processing and machine learning tasks, such as conversational bots long before ChatGPT (at a much smaller scale of course).

And then together with Vitaly again they started Adapty. It was a combination of their previous experience: running and growing a mobile app and lots of analytical and predictive ML. Adapty helps developers to implement subscriptions in the app and grow revenue by running A/B tests, receiving reports and insights, and other analytics. They bootstrapped it for the first year but then went VC way.

Kirill was CTO for the last 5 years, but since 2022 he combine that with being CPO. They grew Adapty from 13 employees at the start of 2022 to over 50 now.

His interested in politics and economics and like reading books/articles on how the former affects the latter. Kirill is a big fan of Wait But Why, Hoover Institution, Kurzgesagt, and RealLifeLore. Kirill is good at strategic board games and huge Lego Technic sets.

Julian:Hey everyone. Thank you so much for joining the Behind Company Lines podcast.Today we have Kirill Potekhin co-founder and CT O of Adapty. Adapty is arevenue management platform for all applications. Kirill is so excited to chatwith you and, and not only get to know you a little bit more and then shareyour story with our audience, but also really get dive into the cool thingsyou're doing at Adapty where.

I think a lot of,companies or individuals may think they know or have an, an idea of how to kindof source revenue from applications that they're building, especially if theyhave a huge user base. But it's a challenge to really find the right modelthat, that, not only adds value to the users, but also helps them, it helps thecompany kind of grow and expand and application grow and expand its revenue aswell.

So I'm excited toto dive into how you're helping companies do that and solve that problem. Butbefore we get into what you're doing at Adapty, what were you doing before youstarted the company?  

Kirill:Yeah. Hey everyone. Me and my friend Vital, one of the, like other founders ofAdapty, we was developing mobile lab called EIT 10.

And it was likefour, five years. After. I finish in the college, well, it was like during thefinal year of the college actually. And we grow it to several millions of likerevenue per year. So it was quite successful. And then we face some challengeslike grow in it even more.

Especially when wewere like dealing with in-app subscription as it was quite new at the time. Andso we developed the system to kinda run tests effectively and analyze ads likeads from Facebook and the different platforms in conjunction. And this is howthe idea of Adapty. Was born it was like several years ago, but back then we decidedthat we'd like to experiment like a little bit with different AI things, and westarted our own agency that was focused on different AI implementation, mostlychat bots before like chat GT was a big hit.

And yeah, then wejust accumulated the money that we earned over the years and decided to startAdaptys the way we visualized it like from the start.  

Julian:Yeah. It's so exciting to think about, kind of that, that when you experimentedwith the previous company, you kind of stumbled upon this idea of, ofexperimenting with other applications and revenue models and what a particular,did you see at that time when you were building this?

I guess, it'salmost like an insight engine to identify what the best way to, charge your,your subscription or, or your audience. What were some ways that you saw wereuseful in terms of gaining more traction? What kind of revenue models for theapp were actually successful?

What did you findout along the journey while you were kind of developing that, that kind ofinsight?  

Kirill:Yeah, sure. So, back when we started implementing an app subscriptions into E10, it was quite new. So we were like one of the first step to do that. And wefaced lots of technical difficulties, obviously both on I and Android side.

But overall like,the market was pretty free, not so much like not so many competitors back then.So, star, like to become profitable was not really a problem for us. Theproblem was like when we started to scale, it was we started to buy like paidtraffic from Facebook and later on from Instagram and some other platforms.

And it was verydifficult to figure out which like, ad networks, which sources work well andwhich do. So it's called attribution and mobile world. So that was like thebiggest problem. And for us it was like the break the breakthrough. Then wefinally realize how to connect the data from Facebook, for example, with thedata in our app, like the revenue.

So this is when westarted to kinda run weekly different app campaigns, and this allowed us tokinda analyze them efficiently in terms of return on investment and some othermetrics. And that's how this like idea of Adaptya was born. So basicallyconnect data from the different sources and put it in one place and make iteasier for developers.

Julian:Yeah. And how are those necessarily, I guess, track mechanically? Is it sometype of landing page that, or some particular page that funnels into, whereyou're, you're seeing the interest grow from your audience? What, what inparticular? Enables you to track and also identify the, the, I guess, thechannels that are more successful for your app?

Kirill:Yeah. So before I 14.5, it was rather easy because there was a thing called I GF A, probably like lot of you heard about it. So this was like a universalidentifier of your device. That like Apple and Google they would allowdevelopers to access this information. And then when you retrieve the data fromFacebook, for example, it was easy to measure.

But a couple ofyears ago things changed drastically in the like mobile world. Because Applebasically restricted access to a dfa. It is technically still possible, butvery limited. And right now you'll have to do like much more things to makethis snapshot of the customer who installing there and to figure out who, likewhat source did he came from.

So, right now weuse several different like internal tools to do that. And we also partner withsome bigger services like Adjust or SLI to do that. But yeah, in general,that's how it works.  

Julian:Yeah. And how have transactions changed in terms of, how customers are able topurchase in-app and, and, for companies to really identify whether it's aprescription, a subscription model, or, identify whether it's a, a featurelimitation model in, in terms of whether people gonna earn the value of thatapplication to pay and things like that.

Have transactionschanged in terms of, implementing them within the application on the phone or,or within a certain service or making sure that you can kind of transact indifferent areas? How have transactions changed in, in kind of progress overtime?  

Kirill:Well, I, I, I mean, we currently work only with mobile applications, just likefor clarification.

Sure. And here wesee like, good growth every year. So sometimes it's gonna be like 20% sometimesit could be even more. Especially then covid like started several years ago wesaw a huge bump of in transactions. But overall, we just see that much moreapplication rely on subscription model right now.

Usually it doesn'tmean that customers cannot not use the application without subscription. Sothere is most of the time there is a premium functions that are available foruse, but like to get. To reset the limits or like find some additional features.They have to pay for subscription.

And there's alsolike some cases, especially in utilities where you can just purchase one time.Also called lifetime access to the, it's also popular among some MO utilities,but like Ilities have like, I dunno, like, complications for Apple Watch orsomething similar, but most of the time it's gonna be after renewablesubscriptions because well it's just much more profitable for the developersand they want to earn the money.

So that's how itworks.  

Julian: Yeah.Yeah. Is it easier to maintain kind of an application that is on a subscriptionmodel if you have kind of that recurring revenue? Or is, are there somechallenges with, with certain models of, kind of how you're charging your, yoursubscriber and, and have you seen any consistent success with one particularmodel?

Kirill:Well, well, I, I mean, as I said the, the most common and the most profitablein, in many cases would be just this renewable subscription model. Most of thetime you have to provide something. New, so, new during the subscriptionperiod. So, like the good case could be if you develop an for example, fitnessapplication or dietary application, you would provide a new contact contact fora customer every day.

But sometimes it'salso possible to do it like we, when you provide access to the services, likeCharge P or something similar in mobile, this is also a good scenario where youcan like leverage app subscription, renewable one. But if you provide just like,one one off solution for something, then it could be better to just go withlifetime purchase because this allows your customers like not to worry aboutRenewing it all the time when they only need it like once or several times peryear, for example.

Julian:Yeah. Yeah. And in thinking about just like how applications are typicallybuilt, I is, is there anything that you know, is broken in the model in, inregards to, I think for mo most apps, right? If it's the freemium model, maybeit's a trial period and a certain amount of days, or if it is, feature locking,it's a certain amount of feature sets.

Maybe it's accessto more activity in the app or more things that you can do, more functions. Isthere anything that, that you see in terms of what's kind of common practicethat you think can be improved in terms of how people are building apps and,and then charging their customers?

Kirill:Yeah, I, I, I guess there are several things like the biggest thing that couldcan be improved actually is subscription management itself, which is currentlycontrolled by Apple and Google. And customers I mean, developers, they, they donot have, like, really control over it. So, for example, if the customer asksfor a refund or subscription cancellation developers cannot do it on Apple.

And there are somelike workarounds on Google, but it's still like not really convenient. And wesee it like a lot that complaints from the customers saying that why don't youallow us to cancel subscription or make a refund? And this is something thatcould be improved obviously, but developers do not have access to it right now.

Apart from that,yeah, I guess. Since introduction of the trial periods this was a huge boom forsubscriptions because now you can give your customer, like, dig in into the,like all the features that you provide in there for a three days or for a sevendays. Yeah. Like these are the most common.

For the trial andthis greatly kinda increased conversion from install to the paying customerbecause before you had to kinda make a leap of faith when you were subscribingand currently just activate the trial. You can try the features yourself andthen decide whether it works for you or not.

So I guess that's,that was a big improvement. And yeah. My advice here is to always try to uselike in your product, use the free trial. It could be free or seven days. As Isaid before, you can actually compare it to subscriptions without trial aswell. But to our data subscriptions with trials in general work much betterthan those without it.

Julian:Yeah. And, and in regards to most applications, what would you say is, some ofthe, one of the biggest challenges, is it, acquiring new customers or users? Isit retaining those users? What do most apps and most developers kind of thinkabout day-to-day? Is it, is it the acquisition part or is it the retentionpart?

Kirill:I, I guess acquisition is more complicated. I mean, it's there are a lot ofinformation, really good information on how to build like the good product andspecifically applications, but there's not so much good information on how toacquire the customers, how to do marketing, how to do sales sometimes, and thispart is definitely a trickier.

Julian:Yeah. Yeah. And thinking about, kind of, app development is how much do youthink companies are missing out in terms of their revenue with, incorrectrevenue models or, or just I guess, revenue models that that they've testedand, and maybe aren't maximizing the amount of value that they're getting and theamount of value that consumers are seeking out.

I don't know if youhave a number or anything like that, but yeah, I guess what would be your bestguess for how many apps are, not doing or, or not maximizing the revenuemodels?  

Kirill: YYeah. Well, There's no like, concrete number here, that's for sure. But like,wild assumption would be that I, I guess, modern 50% of the apps like, mostlikely they could earn like at least 20% more.

Because that's whatwe see from like our experience. After, like starting working with ourplatform, we provide ab test, we provide some extensive analytics, and we havelots of cases where within two to six months, the revenue increased. Like, from20 and sometimes to 300% just by running lots of ab test and figuring out theright like products and not, not only products, but also like, the time whenyou show this paywall it's like the screen where you can.

Start thesubscription. So lots of criteria there. And the best case, I mean, the besttransfer to it is just to run as many tests as you can because like you, Idon't know what will work. In your case there are some best practicesobviously, but still like, gets in the right price gets in the right moment toshow the payroll is pretty unique to every app and so you, you have to test it.

Julian:Yeah, yeah. Tell us a little bit more for the audience. Give a little bit morecontext. How many users, do you have working on the application? How many appsdo you have on the platform now? Give us a little bit of insight into thetraction that you've seen up to this point. And then what you're pretty excitedabout in terms of the next phase of growth for Adapty.

Kirill:All right. Yes. So, currently we have just over 3000 applications on ourplatform. Those are applications of like, well, not all sizes, but like manysizes. Some of them they don't have pretty much any revenue at the moment atall. Some earn like millions of dollar of dollars per month.

So like, reallydifferent scenarios here and we grew a lot during last year. Because like weintroduced several pool features and we expect to continue do it. We justreleased revenue and LTV prediction which, which will be kind of hugedifference because before that you had to, you you had to wait for like, fromone month to three months in order to understand whether your.

Like acquisitionwas successful or not. And currently we can like predict with a 90% accuracyafter three weeks, what would be your 12 or nine months outcome, like in termsof the revenue and we already seen like, Good numbers there. So yeah, we expectto kind of get much more applications as we grow, but overall, yeah, the marketis big.

It's still like,it's obviously not in the like early stage. So there, there are lots ofcompetitors there. But if you do the right product and if you know how toacquire customers, if you know how to do markets, and this is like a reallycool market to be in.  

Julian:Yeah. Yeah. I guess whether it's external or internal, what are some of thebiggest risks that you think Adapty faces today?

Kirill:Yeah, I guess, so, talking about risk, I the biggest risks that we are well notfacing, but we just think about them from time to time is still since we are.Pretty committed to the Apple and Google infrastructure. It's if they'd like tochange something. And so, for example, at some point they may come up withideas that they do not allow like third party services to manage subscriptions.

This will beobviously blowout for us, but we think it's not really. Like, it's not likelythat will, this will happen. And we actually see some pretty good opportunitiesin like, recent months and years that both Apple and Google, they are underpressure to make payments. More kind of transparent right now because currentlythey manage it and you have to pay commission from 15 to 30 percent of yourlike revenue to Apple or Google.

And different like,government agencies right now, they try to like make things work differently.And so if this happens, if Third party payment systems will be allowed on Appstore or play store. This will be big opportunity for us because then we'llhave like all the tools and we'll be ready to serve new customers.

Julian:Yeah, if everything goes well, what's the long term vision for Adapty?  

Kirill:The long term is that we'd like to expand from mobile applications only becausewe we have pretty good BI system right now that kind of covers. Several dozensof key metrics of subscription applications. And it's not only applicable tomobile, but also to most of SAS businesses.

So that's, that'slike one of the visions. And so, as I said earlier, we'd like to ex we'd liketo kind of, instead of. Be just the analytical service. We also want to go likeone step further and be a payment provider if it will be possible from Appleand Google and like, emission for this here, for us maybe not so long term, butalso like a big one is that since we have lots of data from like all the appsthat are using our platform, we may provide some insights on how to properly.

Configure paywallor like, what are the best kinda working couples of different subscription lensand stuff like that. And we'll be doing it automatically for our customerswhich is insured probably will increase the revenue even without like, runninglots of test. So yeah, that's our plan.

Julian:Yeah. Amazing. I always like this next section, I call it my founder faq. SoI'm gonna ask you some rapid fire questions. And we'll see where we go. So,here, first question is, what's particularly hard about your job day-to-day?  

Kirill:Lot, lot. Lots of meetings and lots of things to kinda keep track of during theday and during the week in general.

Lots of area ofresponsibility.  

Julian:Yeah. Yeah, yeah. And, and if. And another question, thinking about, mobileapps and, and kind of how you're acquiring customers, what's been extremelyvaluable in terms of how you've been able to generate kind of, kind of ago-to-market strategy? Wh where do you kind of, being that, app developers arein so many different places, they're on forums or in communities, what's the,what's the way that's been the most successful for you?

Kind of go tomarket and, and business acquisition strategy?

Kirill:Yeah. You, you actually mentioned one. So we we manage several communitiesabout mobile applications and so we have like a loyal audience there, and thisis by far the best way to acquire customers for us. So we're not like pushingour product really hard.

We're just tryingto be helpful in terms of different questions that might appear, and every oncein a while we kinda can showcase some things that we did in Adaptyed with ourcustomers, showing that how they increase the revenue and it works really.  

Julian: Ilove that. Thinking about just apps out there, is there, one or twoapplications that you particularly like in terms of, the way that they'rebuilt, the way that they attract customers the way that their payment structureis, is there any app that that's out there that you're a fan of in particular?

Kirill:Yeah. I can definitely mention doling both in terms of how it's built and interms of how it Like monetizing itself. It's very, I, I mean, I mean it's, it'svery uniquely that you don't kinda try to build to build customers for theusage, but you try to upsell like certificates and stuff like that.

And it works reallywell for them. And apart from that well, I mean, Lin was the app that made somebreak for like break things in mobile subscriptions industry. The paywall iskinda very famous in in the industry. So yeah, I guess good is to them too. Apartfrom that, I mean, I have some.

Subscriptionservice. They're like Spotify or YouTube Premium, but there's no special abouthow they do it. It's just that they have really good product and we're happy topay for it.  

Julian:Yeah. Yeah. Yeah. Is there any app that you're currently not working with thatyou would like to work with?  

Kirill:I, I, I, I guess, I mean, if you'd like me to to pick one, I would go with letme think real quick.

Well, probably someof the games for example, I'm a big fan of brow stars. It would be cool tokinda lend them among the customers.  

Julian:Yeah. Yeah. Thinking about your journey as a founder, what's something thatyou're particularly good at now that you wish you were better at early on in,as a founder?

Kirill:Delegation. So that was like a really big issue for me for many years and Iguess only a year and a half ago I started like doing much better in this areabecause before that I was always concerned that something might go wrong if Idon't kinda go there and check everything in details. And only like when Ireally realized that there's no way around it if you'd like to scale thecompany.

It kind of movedfrom there. And like for during last year, we grow from 14 to more than 50employees. And this was in part like because of I and my partners were able todelegate like lots of important things.  

Julian:Yeah. And what in particular, when you think about, that, that growth what wassomething that.

One thing that youmentioned in the notes before the call was that, you had been bootstrapped upto a certain point. I think it was, it's either this year or, or later lastyear that you, went the VC route and sought out funding. What was kind of thedecision maker for your team to, need that influx of, of capital to be able toscale and grow?

What in particularwas kind of the signal you saw that it was then time to kind of, you knowinterject funding and, and start going that route? What in particular led youto make that decision?  

Kirill:Yeah, well, the thing is, we kind of, we would not mind to get an investmentfrom the first day. It was just not that easy for us back then.

And so that's whywe decided to bootstrap because we had like, enough money to start. And thenonce we saw some initial direction from our customers we just went to several.Good. Like, investors in our network right away, show them the numbers, explainthe vision, and it was more than enough for us.

So we did not tryto bootstrap it the whole time. It was just to kind of jumpstart to get to thepoint where we can get some good proposals from investors.  

Julian:Yeah. And what did you invest a lot of that in? Was it building team? Buildingproduct, expanding to new markets? What particularly were you excited to, toput that capital towards?

Kirill:Yeah, yeah. Like, Pree was only about product development, so we hired morepeople, engineers mostly. And then the second round, like seed round, it wasmostly focused around building the sales team. And yeah, it, it, it was basicallyhappening during last year. Still lots of things to kind of learn there, but sofar it looks like we are on the right track.

Julian:That's incredible care. I'm so excited to, to, ask you this next question. Ilove how founders extract knowledge out of anything that they ingest. Whetherit's early in your career or now, what books or people have influenced you themost?  

Kirill:Yeah, that's a good one. So, I, I, I'm a big fan of Team Orban.

This is the guy wholeads weight, but why and he has like lots of different articles on manydifferent topics. And I also like read quite a bit of Non nonfiction. And I, Ithink it was like sometimes very insightful. I'm, I'm also really inspired ofElon Musk was a big fan of him like for many years.

I'm not followinghim, him that close right now, but I still think that he's like, one of thegreatest businessmen and like inventors in the world. As for the books, like,from my area, well, I guess, some of the technical books or even courses likeabout the program in c multiple CS courses from Harvard and m i t.

And what else? Highoutput management. That's like, I guess really classic classical book. A goodone from zero to one. Also really nice. And yeah, I, I guess I don't think thatthere were like one book that made the difference for me for, for me, butmostly, actually I will also tell you that like region Twitter it's pretty coolbecause lots of entrepreneurs and top-notch investors are talking there to eachother and just really interesting to kinda read it through and understand howthey see like different situations and so it's, it's really nice.

Julian: Ilove that. And Kirill, last little bit, I wanna make sure we didn't leaveanything on the table. Is there any question I didn't ask you that I shouldhave or anything that we didn't talk about that you wanted to just touch onbefore we end here? Anything that we left on the table today?  

Kirill:Yeah. Well, I guess, just like as a small piece for me that if you work withmobile applications and if you have some like issues scaling or just touchingup don't be afraid.

There are lots ofresources out there and it is a good time, as I said earlier, to be in thisbusiness. I guess that's it.  

Julian:Yeah. Yeah, yeah. Last little bit. Curious is what, give us an audience, your,your plugs where your LinkedIn, what's your Twitter? Where can we find you benot only a supporter of what you're doing at Adapty You, but a fan of you as afounder.

Where can we findyou and get in touch?  

Kirill:Yeah. Well, I, I have like a tech at Kirill Potekhin pretty much everywhere. Soyou can find me on LinkedIn as, or you can just Google it. I will probably popup there or tion. I guess we can just highlight it somehow in Twitter, FacebookInstagram, everywhere else. LinkedIn as well.

Julian:Yeah. We'll put it in the show and make sure people can reach out. But Kirill,it's been such a pleasure not only learning about your early experience, howyou kind of conceived the idea out of another application that you werebuilding, and then how you've been able to help developers really scale theirapplications in ways that, challenge their revenue models, test differentthings that will attract either more customers or help them, their, themconvert their customers.

It's exciting tosee, what kind of capabilities you're bringing to the table for these.Developers and where your, your position is in helping companies scale andgrowth. So I'm excited to see where you go from here. And thank you again forbeing on the show. I hope you enjoyed yourself.

Kirill:Thanks a lot, Julian. It was really nice talking to you.

Other interesting podcasts