Analyzing Your Subscription App Business

Get an over-the-shoulder view as we analyze real-world data from plant care app, Greg

February 22, 2023

There are two main challenges when trying to understand the health of your subscription app business. First, which metrics to track and where to track that data. Second, what those metrics actually mean for your business. In this conversation with Greg CEO Alex Ross and RevenueCat senior product manager Dan Pannasch, we’ll explore the RevenueCat dashboard with real-world data from the Greg app. We’ll talk through what each metric means and how it impacts the health of your subscription app business.

Topics we’ll cover:

  • How to think about realized LTV vs. LTV estimates
  • Trial Conversion and other funnel dynamics
  • Cohorting and segmenting subscriber retention
  • MRR/ARR and other basic metrics
  • Transcript

    Hello hello? I can’t tell if anybody has joined us. Oh, Wow. Yeah 30 people have joined us. So I was explaining to Dan and my colleague Corey that I screwed up big time. All of our marketing went out for the webinar to start at the top of the hour. And when I was configuring this event in live storm, I said it for 10 after the hour. So what we’re going to do is we will start the actual event in 10 minutes. I apologize to those of you who came on time and set your calendar event to the marketing and not to the marketing we did and not the invite email. So what we’re going to do is start the event in 10 minutes. But for these next 10 minutes, we can just do a quick Q&A, hopefully make it kind of worth your time, too, to be in here. So if anybody wants to jump into the chat, ask a question. Dan and I can also just ramble on. I mean, we love talking about this stuff. Alex from Greg, I think aligns his calendar to the 10 minutes after and we’ll probably be popping in, so we’ll probably have to do a Mic check and that sort of thing. But my son’s, like, showing me messages through the window. It’s like you got a producer inside, like, super professional today. So I think he was showing me I had unplugged my ethernet cable and plugged it into his PS5 and I told him last night, remind me to get my ethernet cable. So that I’m hard wired for the webinar. So it was thoughtful of him to do that. So any quick questions he’ll want to ask as we’re getting going here? It looks like Ben’s typing just in case we run over starting a bit later. We’ve access to the recording of the event. Afterword Yes. And so what we’re going to do, and especially since I screwed this up, is that life storm will record the event. So it’s already recording will edit out this first 10 minutes since I screwed up the start time on the calendar invites and we will post the edited video in full and we’ll well you know what I typically try and do is for the event to actually only last about 45 minutes and then leave plenty of time for Q&A. So if you miss anything, it will hopefully just be the Q&A, if you do have to leave it at the top of the hour. But again, all the Q&A will be available as a full video on YouTube, when probably later this afternoon or even tomorrow, we can know, since we have to edit this one, I’ll probably take a little longer to get turned around. So you’ll get an email. Yeah Cory just posted in the chat that we will get it on YouTube. And if you subscribe on YouTube or just watch for it there and there will be an email going out as soon as we get it on to YouTube. So again, for those of you just joining, I screwed up big time today. The calendar events invites went out for 10 after the hour. All our marketing went out for top of the hour. So we’re going to chill and hang out and answer any questions for the next few minutes and we’ll start. So if you have any other questions, drop them in the chat. For now and I am giving Corey a deadline, I am also going to hit up Alex and make sure that he is on his way. Oh OK. Alex is actually here. I need to promote him as a speaker. So part of me screwing this up was also that I told Alex to come 10 minutes before and. He came 10 minutes before. Except that was top of the hour. All right. I’ve promoted you to Speaker Alex. I was telling. I was telling Dan, but before we went live, I still get a little nervous at these things. And Dan was saying he was a little nervous. And I told him, you know, for these webinars and even the sub club podcast that I do, I really think that overly polished, overly perfect things sometimes are just boring and not as fun. So you’re going to overhear us just chat about Alex’s business, and it’s going to be a really fun 45 minutes. It’s not super polished. We don’t have it scripted. We’re just going to talk through the revenue dashboard and how Alex thinks about their subscription business. So, so. And Thank you for indulging me these 10 minutes. I know it’s probably a little frustrating too, to show up and have me just rambling on, but I appreciate it. And hopefully once we get rolling in 4 minutes, it’ll be worth your time. So Alex yeah, we. I screwed up the invites went out for 10 after and all the marketing was for top of the hour. And so we’ve got about four more minutes and then we’ll kick it off and I’ll do the normal introductions at the 10 minute mark. So they’re in the final YouTube video. People will get the full experience. We get a quick question and there will be Q&A afterward. So you can also save questions to ask more specific things. What should I handle on my app is being developed very new to app development world. That’s a great question. You know, there are so many great resources out there now. YouTube channels I code with Chris is somebody that we collaborate with on videos of how to use of revenue as you can install into your app. And I know Chris actually has quite a few other videos. If you’re still kind of ramping up on coding, there’s some incredible resources, like 100 days of SWIFT and other places like that. So yeah, YouTube would be a great resource to just kind of search, you know, building. My first app I actually saw there is a YouTube series I saw recently of like taking you step by step through the whole process of building an app. So, yeah, there’s just so many great resources. Daniel, you want to jump in? Yeah one thing I would add to that, too, is I, I bet for most of the people watching this right now, the idea of what to build probably comes first and is top of mind. And I think the problem of how do I go distribute what I built probably comes a few steps later. And I think getting ahead of that and starting to think about like, all right, I’m making this thing, how am I going to go put it in front of people? Like those types of resources I think would be invaluable during the building process. Yeah, that’s a fantastic tip. Yeah you don’t leave your marketing plan till the very end. You really want to be thinking about that through the product development process. All right, just a couple more minutes. Alex, anything you want to add to fresh air? App developer can already hear me. All right. This is. This is my Mic check. Yeah, I. Yeah, you sound great. Cool Yeah. No, I just plus one, the distribution thing. I’d also mentioned that I think a lot of people start diving straight into code and building like a really, you know, well-thought out native app. Takes a lot of time before considering, like, less time intensive ways to validate their idea. So maybe like a website, honestly, even sign up forms, waitlist things like that. I’d spend a lot more time there before really building out an entire mobile app just to validate the idea and maybe even to test your distribution strategy. Yeah I have a favorite app right now called kindred, which is for home sharing. It’s kind of like a modern Airbnb. And they did a modern airbnb, you know, as they did their first 500 nights of stays without an app, just Google, spreadsheets, forms, stuff like that. So I think there’s something to be said for, you know, taking time to validate your idea and then being sure about what you’re building. Yeah, there’s actually a great blog series by Zach’s acad where he actually posted on Reddit an idea he had and got press before him belly up. He didn’t even end up building the app, but it was a pretty incredible kind of product discovery process. And he kind of journaled on the blog, his thought process of how he came to that. All right. We hit the top of the hour. Quick intro. I screwed up the timing of this. This this will be on YouTube starting now. But I scheduled this webinar for 10 after the hour. That’s where all the calendar invites went out. But we marketed it for the top of the hour. So we’ve had a little Q&A and chat that we will edit out and make checks and all that fun stuff. But we are going to move for it. So I’m David Bernard. I’m going to be hosting this little chat. We have Dan, who is a PM revenue cat who actually oversees the dashboard in the team that works on our front end. He comes to us from teltech, which got acquired by IAC. They’re an incredible and like Tinder where Alex Ross cut his teeth. I mean, Alex cut his teeth in this subscription app space very early on and incredible learnings that Dan got there at Teletec and with it being acquired by IAC even got even more exposure to a whole portfolio of subscription apps. And then Alex was incredibly generous to building in the open. It’s a scary thing, not a lot of founders who are willing to do this, but he is going to share real live actual data from the app. He’s building, Greg. And Greg is actually a Plank killer app. And we’re going to get a little intro to the app to understand a little bit better and kind of have context around discussing the business. But before we do that, let me just do a few more little housekeeping items. During the session. If you have a question, there’s a Q&A button and go ahead and drop your questions in there. And then upvote the ones that you want answered. And I’m terrible at being able to read the chat and engage with the guests and I’m actually managing the dashboard as well. So I’m going to be clicking around and all that. So I won’t be able to moderate and watch the chat and ask questions live. So go ahead and put those in the Q&A. And I’ll put the ones that are most interesting to you. And we will leave time at the end to answer as many of the questions as we can. Get to another couple of things. We will be demoing the revenue dashboard today. Most of you are probably familiar, but in case you’re not, in addition to the dashboard, we have a native speakers and even Flutter court over a bunch of client side SDK that help manage your transactions. They’re open source battle tested and they go into your app and manage the transactions. And through that process we also then pass the receipts and the data to our server side, which becomes your single source of truth for your subscription status. And once we have that data server side, there’s a lot of really cool stuff we do with it, so we normalize it across. I was an Android to where a free trials and free trial of billing issues. Of billing issues. So when you’re sending out wind backs or doing analysis in amplitude, it’s all normalized across IOS, android, the web, Amazon app store, and we’ll be expanding to other platforms over time as well. So we have integrations, we have the dashboard, a lot of stuff, server side as well, and the events that we send to our integrations are incredibly powerful as well. So quick overview of revenue and then, of course, today we’ll be talking about the dashboard. A couple of other things. In addition to these live sessions, we also do a podcast called sub club. Corey can drop a link to that in the chat. And as. So we also have a community. So the sub club community, we chat about all things subscription app and there’s a, there’s a waitlist that you can sign up for that Cory can also drop in the chat. It’s in the profile Twitter profile of sub club HQ. And so a couple of other great resources. If you’re here, you’re probably thinking at home and running, thinking about subscriptions and running a subscription app. That’s another great resource. The recording of this will be available on YouTube probably the next few days since we have to edit out the early chats. So yeah, that’s all the housekeeping and I’ve introduced the guests. Alex, let’s kick this off. Thank you all for indulging me all this time and tell us a little bit about Greg. Give us kind of 3 minute overview of the app, why you founded it, and then kind of, you know, where you’re at as a business. Yeah, sure. Thanks for the introduction. So first, Greg, as you mentioned, is an app that helps you care for your plants. So it’s mostly geared towards people who are growing plants indoors right now. I think it became a very popular hobby with younger generations. During the pandemic. A lot of people want to create a really nice, wholesome Space for themselves while they’re working from home or even offices, et cetera. But a lot of people don’t know how to care for plants. I think there’s this concept of a green thumb, right? And that like only certain people know the magic of like creating a, you know, oasis. But actually plants are entirely just little beings of physics and science, and there are very deterministic ways to keep plants alive. And so that was kind of the truth that my co-founders and I had stumbled upon as around three years ago now, we started going deep into the world of plants after killing many, many of them yourself. I’m talking thousands of worth like plants are it can be very expensive and b, they can die really easily if you don’t know what you’re doing. And so we decided to build that brand and to build that functionality, those guardrails that can turn anybody into a plant growing superhero. Anybody can create a really nice, like, verdant space in their home or wherever they are. And so the app today is mostly geared around that utility. And so it tells you when to water. It tells you the light that your plant needs. It helps you with fertilizing. And there are lots of other tips around seasonality and what to do as the sunlight changing in your region changes how you need to care for your plants. These are all the common things that draw people up, and there’s just enough that it can be done with a mobile app. So, you know, people are pretty happy, I think, at this point. And a lot of people are. I saw just yesterday there’s someone with 550 plants in the app, which just blows my mind. Most people have closer to like, you know, five or 10. But yeah, it is being used to grow that many plants. And then another really cool aspect of what we’re doing is, I mean, you mentioned the app’s name is Greg. That’s short for gregarious. And gregarious in botany is a word that is used to describe plants that grow in communities, which you might not think of when you think of plants, but trees grow in communities. And so we wanted to build a communal app. We wanted to build an app where it’s not just about growing plants, but it’s also about growing plants with other humans. Right? because I think we’re at a point in time in the world where caring for nature or caring for things besides humans, caring for the ecosystem around you is increasingly important. And so we want to bring people together to do that through our app. So the app also has a very large community component where there’s profiles people are asking and answering questions about how to grow each other’s plants. They’re sharing wins. There’s communities, there’s local communities. People are starting to organize in-person meetups on the app to like swap plants or to just come together and socialize. And so that’s basically what Greg is today, is a utility app to help you grow plants and then also host the increasingly large community for connecting with other people who also grow plants. Very cool. And then as a bit of background too, and this is all Google abl, but Greg has raised a seed round, right. And so CR funded company. And how many people are on the team now? Yeah, that’s right. So we raised a seed round in 2021 may, so maybe a little bit over a year ago from Index Ventures and first round capital. And right now the team is around 10 people, mostly engineers. You know, there’s one person who does marketing. We’ve got one product manager, one scientist, one botanist, and then the rest are engineers. And the app is available on iOS and Android natively too. They’re amazing. And I mean, just out of curiosity, because I haven’t asked you this before, did you use Flutter or a cross-platform solution or is it both native to each platform? That’s a great question. Yeah, we started native from day one. Know, my background, as you mentioned, I was a director of Engineering at Tinder and before that I started another company and I’ve been doing mobile development for a while. And I personally believe that when you know what you want to build. And when you know that having a really, really good experience is crucial for your business. So your special sauce then, I think building natively is a pretty good idea. And so we went that way and don’t look back. I’ll say we were one of the very first apps, you using UIKit throughout the entire app. That was a journey. Maybe not the best timing for a swift swift UI, maybe. Yeah I’m sorry, swift. You are not. You are? Yeah Wow. So do you have horror stories? Yeah, we’ve got some. We’ve got some. OK, now we actually we just we write most new features just in slabs because was it’s when you’re doing what you should be doing with it, it’s great when you’re trying to do more. It’s really, really bad. Yeah all right, well, people then come here to hear about that, although it’s super fascinating. We could talk another 30 minutes just on that. But let’s go ahead and dive into the dashboard. So the first thing I wanted to discuss is. R.N. and so just as a quick overview, we’re not going to go through every single chart in the revenue dashboard that would have the kind of conversation I’d like to have around these things that would take hours, actually. So we’re going to pick the highlights from both what Alex died into day in, day out, in understanding his business, and then some charts where he might not have spent quite as much time in because there’s several kind of new charts that have just been introduced. And we actually redid the visuals of charts just in the last like month. So a lot has changed. And so we’re going to also dive into some charts where maybe Dan and I actually even have a little more insight into how to think about it. But then, of course, Alex, having been at Tinder and running this business, can then reflect on how he thinks about those things. So R.N. is a interesting topic. So Dan, I wanted to kick it off with how revenue calculates a R.N. and y And then I know Alex has some thoughts on the benefits and drawbacks to how we do that. So go ahead. Takeaway, Dan. Yeah, definitely. So the good news is that the calculation is pretty simple, right? So R.N. at a literally all we’re doing is we’re first calculating your MRI, which we’ll do in a second and multiplying it by 12. Your MRI is basically taking your active subscribers right now and normalizing them down to monthly recurring revenue. So for example, if you have a monthly subscription, that’s just a single payment on that monthly subscription. If you have a yearly subscription, that’s 1/12 of the payment on your yearly subscription, all normalized down to your MRI to get a sense of for your active subscriber base, if it stays exactly the same size and mix and price mix that it is right now, here’s what you’ll earn in revenue over the course of a month. The reason why R.N. is especially useful then is because so many subscriptions are annual instead of monthly. R.N. is telling you that over a given 12 month period. So for your active subscriber base as it exists right now, here’s what you would earn from them over a 12 month period at the exact mix that you have today. Right? so exchange option gets multiplied by 50 to a quarterly subscription gets multiplied by four, semiannual, which I’ve been surprised as we’ve started doing benchmarking how many apps are actually using quarterly and half year and weekly? You know, to me, it’s always just been, you know, monthly and annual. But all those do exist and they are factored into this calculation. Non-recurring, both consumable and one time in-app purchases are not included into this, correct? Yeah, that’s correct. And in the line of thinking there is that though those are potentially valuable revenue contributors to your business because they’re not of a recurring nature. They’re not necessarily. That you can predict going forward, right? Like you may have trends in your business in terms of how non consumable purchases do get made, but it’s not really recurring revenue. So it’s not necessarily something that you should be expecting over a 12 month period on an exact recurring basis. Right and then so Alex, that’s actually a perfect segue where investors love to talk IRR and and B2B SaaS areas like the number. And so I imagine it’s a very important number for Greg when talking to investors, when sending out investor updates, when thinking about the business. But there are challenges with consumer, so why don’t you share your perspective on R.N. and the benefits and drawbacks to using it in the consumer subscription space? Yeah, definitely. So this is the graph that I probably spend the most time with, and the way you can think of it is when you’re raising capital from investors past your first round. So like when you raise a seed round or pre-seed or whatever you call it, most investors are just looking at, you know, do they believe in you? Do they believe in your team? Do they believe in your first idea? But they know you might pivot anyways. So in any case, the first round you raise doesn’t look too much at rr, but when you’re raising a series a, series b, things like that, there are two things investors look for. The first is how much revenue are you making on an annual basis? And the second is how efficient are you at acquiring that revenue? And the first one, what it basically is, is it tells the investor this team is capturing a large opportunity. Right because investors don’t want to invest in like a McDonald’s franchise most of the time. They want to invest in the next Snapchat or Tinder or whatever it is. So this basically means that investors end up with benchmarks in their minds because they’re looking at a lot of different companies. And so they need a way to compare those different companies and they even need a way to compare those companies across spaces. So you can kind of broadly group businesses into business, to business or B2B as I’m selling your products to another business like revenue is or B2C, business to consumer. So I’m selling a product to an individual person. That’s what Greg does. And then within that, there’s also kind of B2C direct to consumer. And that’s usually like I’m selling a physical product to somebody like allbirds, et cetera. So what’s interesting is investors will basically look at ways to identify this business is a really good one to invest in. That one is not. And so they do that by saying, OK, once a business has crossed, you know, 1 million annual recurring revenue, that’s an indicator that they may be ready for some growth capital to grow. And then later on for, you know, like a series b, it’s OK. They’ve crossed call it like 5 million or more than that in annual recurring revenue. And so you end up with these as you’re going out looking for capital, it’s really important to know what the benchmarks are at that point in time. They change over decades because, you know, business changes. But that’s why reaching a certain R.N. is typically almost like a bit of a checkbox and raising capital. And so that’s why it’s important to watch. The catch, though, is that there’s a big difference between business to business. R.N. which is really where R.N. is like the primary focus and B2C. And that is that business to business software companies usually have much higher subscription retention than, you know, businesses that sell to consumers because businesses, once you get in and they’re using their product, they’re too busy. They’re not going to cancel their subscription. Right? whereas consumers are constantly thinking about, what can I cut out of my budget? What can I stop paying for stuff like that. So like a business, a business company is usually looking for net retention of above 100%, which means that not only are there customers not canceling, but they’re actually buying more because they’re adding more seats or because they’re using the product more so to pay for it more. Whereas in business to consumer, I mean, revenue catches published benchmarks, but even the best of the best I’m talking like Netflix, spotify, et cetera even those companies have 75% annual retention. So they’re always losing customers to some degree, which means by bring it all the way back to R.N. and revenue cap, the key is that the R number does basically include subscriptions that are canceled but have not expired yet. And I think the reason for that is that those subscriptions, those people are still using the app and they may actually resubscribe. But that just means that, you know, knowing that retention is less than 100%, it means that your actual cash collected is probably going to be less than the R.N. number if you acquired, you know, more subscribers. Yeah John and I have. Yeah Dan and I have already been talking about this, and this is something I’ve wanted to address for years. No promises on dates, but know that it’s something, you know, we’re thinking about is that and when we get into the subscription retention chart, we do have ways within revenue card to look at the percentage of users who do still have auto renew enabled. But at some point, I think we really do need to figure out a way with this chart to either have a dotted line or something else that would indicate, OK, here’s your R.N. as we normally calculate it, and then here’s a R.N. minus anybody who has already turned off auto with you. So it’s something. We’ll figure out in the long run, a smart way to do that, but it is really such an important thing to bring up. With retention, what it is in consumer subscriptions. Before we leave this one, I did want to show a few things and then you can probably talk to this if I segment my first purchase month. The cool thing here, this is the line chart, super messy, but I actually do like looking at this because you can really see these different cohorts over time where you have a ton of revenue coming in and then you can see a year later where it drops precipitously at that annual renewal. So these are probably a mix of annual and monthly. And so that’s why you do have some loss of air as a monthlies churn, and then you have a steeper drop. But another great way to look at it, and this is I don’t know if it’s an industry term. I mean, I’ve just heard it more and more lately, but a layer cake chart, I guess it’s just a visualization term, but this is a really great way to look at a subscription business looking at these cohorts. And so what you see here with Greg is what you want to see in a subscription app business is that you have these cohorts in and we know and again, you know, we share benchmarks. You can look up these posts on our blogs. We know people are going to churn. You just can’t expect it to be a B2B business where these lines are flat. But you can see the shapes of these curves in the stacked area chart, you see the churn that happens. But what you see over here is that there are these retentive cohorts. So there’s this orange cohort here from May 2021, you know, 18 months, almost 18 months later, there are still a bunch of people subscribed. And so the key to building a great subscription business over time is stacking these retentive cohorts is that you want that retention curve to flatten out and you want to stack these cohorts. So I love this stacked area chart, this layer cake chart too, to look at that and get a visual representation of stacking those cohorts and you will see apps. I should probably switch to my own app because it’s embarrassingly bad, but if you’re headed the wrong direction, you build these cohorts and then you don’t have that stack toward the end that are just building up. It just drops off a cliff when your retention is really bad and you’re not acquiring those new cohorts. Dan, anything else you wanted to add to that chart? No, I think that’s exactly the right point. And I think it illustrates what Alex was saying perfectly, that like your R.N. as it stands today, if you acquired no one else, right? If you don’t get another one of those pancake layers on top, it’s going to drop. And so like keeping a close eye on, Yes. Both the retention trends in your business, but also the acquisition trends and how you’re adding on top of those cohorts is so critical. Yeah, awesome. So in talking through Ar, we’ve been talking so much about retention. So let’s just dive into the retention chart because this is actually probably my next favorite. Alex, you were saying since this is so new, you told us like I have to. Sorry well, I’ll let you talk while I undo whatever filter is making this not load. So, Alex, you were talking about actually haven’t spent much time in this retention chart. Yeah, right. Because it’s so new. Then why don’t you give us a little overview of our thinking on this retention chart? Sure Yeah. So basically, the way subscription retention is working, it’s showing you for a given period of time, right? Whatever time period you select, what is your average retention for all of your cohorts through that period. And then breaking it out by the individual cohorts? So once you’ve got it pulled up, like you’ll see that there’s a total line. That’s the main visualization. That’s the combination of all of your cohorts from that period that you’ve selected. And then you can look at each of the individual cohorts in that period and see how they compare against the average. To understand whether your most recent cohorts are behaving similarly to the average above below, and understand where your retention is trending at the moment. All right. We get it pulled up. Always fun to live demo with real data on production. I have no idea what happened there for like two seconds. So a couple of things to look at in this chart is that because the display of monthly and annual is so different, you actually need to select these retention periods separately. So we’re going to start with the monthly retention and switch to all time. And so another interesting thing about this chart is that we We are cohorting not by like month here. We’re actually in the chart showing you the combination of all cohorts. So what is the first monthly renewal rate? So that’s right. Here would be the first monthly retention across all the different cohorts. And so this kind of gives you that kind of overview of of, you know, is your know, is your retention curve flattening out over time? And what you see with Greg is that it is very nicely flattening out around 20% And one thing that Dan had mentioned that’s really fun to do is actually go back through and then highlight a few of these more recent cohorts and then scroll back up here. And so, you know, what we can see is that here the May 1 to May 31st cohort, the first retention was actually significantly lower than the average. And so something happened in May and spoil the ending. We know from having talked to Alex about his business strategy is that around that time they really shifted strongly to annual subscriptions. And so whoever is seeing that monthly subscription is highly filtered now because they’re really promoting the annual subscription. And so by selecting these more recent cohorts and comparing them against the average, you can actually see that those cohorts are behaving differently. And then you look back and try to understand what did I change or what went different to see why these cohorts are performing differently than in my average. Yeah Alex, anything you want to add in thinking through the retention curve and how you think about this? The only thing I’d add is I think a lot of apps are thinking about should I offer monthly, annual or even quarterly, weekly, like you mentioned. And I think. It’s really important to do the math on, you know, what does I will actually get into those, I guess like the LTV of like OK so you know, when I, when I look at this, I want to know, you know, how much does this monthly user bring in and you can only really do that once you see your retention flatline. Right? so in the early days, Greg was only around for maybe six months and we had to make some time decisions around like, all right, like how, how long does it look like a monthly subscriber is going to stick around? And so I would just point out that it’s important as time goes by while you’re developing your app, to keep a really close eye on how retention is developing. And so I have a habit of looking at our retention graphs at bare minimum monthly just to see how that new cohort has stacked out. Yeah and then what I’m going to go ahead and do here to illustrate how this can even be helpful for annual is to switch to the yearly retention. So so Greg only recently started shifting to heavily towards annual subscriptions. So we don’t have, you know, long term data where you’re going to see the annual retention curve because there’s only been, what is this like six months of eligible conversions to see what that first year retention rate is. But again, this is what’s great about being able to get a little bit of a peek into the future. And so what I’m going to do is switch from absolute to relative, and that’s going to give us a percentage. And so when you look at this chart, what you’re seeing is the year one renewal rate for the six cohorts who have had an opportunity to renew. And so, you know, as we shared in our benchmarks, I believe the median is 28% So so some of this early conversion is actually in the top quartile. And then I think the top quartile starts at like 41% So you’re kind of pushing the above median and into that top quartile according to our benchmarks. Of course, you know, to build a great business, you really want to push this as much as you possibly can. But again, understanding those benchmarks is that, you know, consumers subscriptions just aren’t going to retain, you know, except for the best of the best of the best anywhere near, you know, B2B or even, you know, a Netflix or something like that. But what gives you a little peek at the future here? Is this dashed year to rate for this initial cohort tells you, OK, 52% stayed subscribed and retained across that first year. And then the 33% is how many still have auto renew turned on. And so and this is relative not to a 33% drop from 50 it’s how many are still retained from that original cohort. And so that’s where you can already see that it is flattening out because if you had 52% retention in year one and you only had 50% retention of those who remained, you would end up much lower. And so that’s where this is going to be, that kind of flattening out curve. And then you can see even some of these newer ones, the curve appears to be flattening. So again, because they haven’t hit the renewal yet, some people can still turn off, order new and not actually renew before that period. But this seems to be a good indication that these later cohorts, even though they are renewing in the first year at lower rates, that it really is flattening. And we could talk for hours about how to think about these things. Sorry, my camera dropped out, but we’ll just focus on the video, the dashboard anyways, so we could talk like hours about how to think about this because you know, what? What users were you acquiring during this period? You know what ab test, what product experience, where they’re coming in to. And so with a super early app like Craig, this initial cohort, we’re seeing a completely different app. Then if we go down here to the more recent cohorts, you know, that’s when community was introduced, additional features, onboarding was changed, like the product has evolved so much over time. So again, we could talk for hours about this, but you constantly need to be thinking is looking at these retention numbers, looking at patterns of how people turn out to renew on and off and trying to kind of get to the reasons why an understanding your customers and the patterns of the customers. Yeah any, any kind of top of mind things for you, Alex a things that changed from 2021 to 2022 that for now anyways know there is a high auto renew status on what have you have been experimenting in the product. Yeah, that’s a great question. First, Greg, Greg has a very new app. As you can see, we launched subscriptions well over a year ago. The two big things that we’ve done in the app are first, we focused a lot on just adding deep functionality. So we didn’t have anything around the light that your plant needs, the fertilizing, things like that. There’s new types of reminders, there’s new tips. So there’s just maybe four times the surface area of just the RGB functionality of the app. At the beginning of this, this year, our team’s motto is basically like value worth paying for, right? We just wanted to increase the depth, the utility to people. And then the second major thing is that the community. And I think our community is comparable to Strava and that people come for the utility, right? Like if you’re using strava, the running app, you use it to track your runs, but then you stay for the community, right? Because you’re there because your friends are on it or because you’re making friends on it. You like the competition. So I think that our community, as it is today, is mostly a retention feature. It’s not like a growth feature. Yeah very cool. And again, those are the things you want to think about in your product development cycle. And then that’s what’s so challenging about annual subscriptions is the feedback cycle is so slow and that’s where this chart and seeing the auto renew status at least gives you a peak that you’re headed the right direction, even though you don’t know the final conversion until it happens. We’re hoping to do another benchmark analysis soon to look at auto renew status over time. So creating a benchmark timeline of what the typical pattern is, you know, are people turning off auto renew you? I would imagine there’s a pretty big spike soon after the first auto renew where people get charged, they see it on their credit card and they’re like, oh, wait, I probably don’t want to pay for that, or I want to make sure it doesn’t auto renew without me knowing. So it’s probably spike there. And then there’s kind of a trough where people trickle in turning off auto renew and there’s probably a little bit of another bump toward the end where they get the notification from Apple that the auto renews coming up. And so I’m hoping we can do a benchmark on that. But you can see in these numbers that, you know, the July cohort, 79% still have auto renew on. So is that the July cohort is just especially strong? Maybe, but you do kind of see a slow decay here, kind of indicating that there is going to be a trickle of users who turn off auto renew over time. So it gives you a peek at it. And again, once we share those benchmarks, maybe it’ll kind of give you a better framework for thinking about these numbers and how it might play out across the year. But it gives you a little peek at what’s going to happen when the feedback cycle is so slow. Another quick thing I’d love to hear your thoughts on regarding retention and so many more. So I wish we had like 5 hours, but I think this one’s really important. How do you, Alex, think about subscription retention versus usage retention and do you. So we don’t have usage retention charts and revenue cap, but imagine you use like amplitude or something similar to make sure that people are still engaged. And then by using our I imagine you have our integration set up with whoever you use for analytics. And do you do charts where you’re looking at auto renew status versus like engagement in the app and things like that? Yeah, that’s such that’s exactly where my mind was going. I think this decade from August 20, 21, 39% Then it goes up to July 20, 20 to 79% You do see month over month, there’s a very steady decay working back from July 22. I have a feeling that’s like 100% just like basic app retention, right? We’re basically over time, fewer and fewer people will keep using the app until you get to you’re very baseline app like open retention or whatever you define your activity metric. As for us, it’s like watering plants or something. I would love personally because I know, you know, revenue cap could graph like app opens against like subscription retention. I think that’d be a really good idea. We do do it internally. I’ll just say, like we used mixpanel for a long time, we never considered amplitude cause it’s pretty expensive. We were very technical team. I write a lot of SQL and so I actually we have our own whole set of databases. And then we use something called NetApp base to do the graphing because it is just like an order of magnitude less expensive than mixpanel or amplitude, right? So I just want to share that because a lot of people say, you know, go to amplitude, go to mixpanel like there are options that are a lot less expensive for people who are like indie developers. But also. Yeah, just curious, do you use our Etl functionality or do you use our events to stream it into your database web hooks? Yeah one of the first things we did. And we set up revenue was integrated the webhooks just so that we had that flexibility and it’s fantastic. Yeah I hope over time that we’re able to solve more and more of these use cases in our dashboard. But so many of our customers, especially sophisticated customers like you who’ve done this before, who kind of know the ins and outs, very technical CEO, but so many of our customers do have their own internal data warehouse where they use our charts for, for what our charts are really good at. And then they have their own data warehouse to do some more specific analysis that we don’t yet have in our dashboard. But that’s an exciting thing about working in revenue, and it’s a whole other topic. But like we have a data scientist now, we have a whole team doing this stuff. And so in the coming months and years, you know, you get a data scientist for free as we build out charts that are more and more sophisticated, more and more tailored specifically to running your subscription business. So that’s something I’m always super excited about because I still run my own apps. And so when we do something internally, I’m like. We eat and it’s like, I’m going to get that chart for free. I’m paying like $8 a month. You know, I don’t have a data scientist. I can’t run school. You know, I’m not going to set up a data warehouse to do this analysis. So as we kind of expand our charts, it’s super exciting to me. All right. So I did want to touch on a few others. Golly, we’ve come so, so long already. We really could talk about this stuff for five hours. Maybe we’ll have you back on for a follow up in six months or something. We can do another six charts or new charts that we introduce. Yeah, I think another really important one to hit is just kind of basic funnel analysis. What are the top ways that you think about and look at funnel analysis in revenue? Catch our topics totally. So I think one of maybe the third most important number is actually exactly this. I had to rank my metrics that I keep in mind, it’s R.N. and then subscription retention year one, year to year three projected and then it’s conversion from app install to paying user. And so anytime I’m talking with other CEOs or other people who run paid mobile apps, I’m always asking them for like these three numbers basically. And this one, it was just total coincidence that was the order I picked. We didn’t, I had well I mean, yeah, it’s best practices, right? Yeah so, so this one, in my opinion, is the most interesting because it is the most difficult to compare across mobile apps because it depends heavily on how freemium your app is. Right right. So you have apps like duolingo, which are like heavily freemium that have very low single digit conversion rates from as far as I understand, from install to paying user. And then or Tinder for example, is also in that bucket. And then you have apps that are much more heavily paid like these days, you can’t actually use much without paying for it. And so they have, I believe, much higher rates of install to subscriber, but then they have net lower app retention, right? So like the general way I’d frame this is if your app needs people to be more valuable like tinder’s only useful people are on it. Right? then you kind of need to be more freemium. Whereas if you as a company are providing all of the value yourself, like Karma’s out there recording LeBron James to put you to sleep, right? They don’t need people using I don’t care who’s using call. Right then they can push a lot more people to subscribe. And if people churn, it’s less important to them. Right still important because you want people using your app to get word of mouth growth. But in any case, yeah, this is a metric that I watch. And in the last year, actually, we put a little bit of effort towards improving it. So you kind of see like we were on a Downward slope and then and then we did some pretty major updates in April and May and we boosted that way up. And it’s still something we want. We want to keep moving up. But of course, we are a partially freemium app. We have a community, so people being in the community is important. So we do care about free users as well. Yeah and one thing to note, this August we saw is a better way to represent a charge is hard. But this final conversion is lower. Not necessarily because a lower there is lower conversion. It’s that you have a lot of cohorts that have come in and haven’t had a chance to finish the free trial. So over the course of the month and then as we end August, this likely will go up. It’s not that you’ve seen a precipitous drop in August. Yeah Yeah. And then I think the. Dan, anything you want to add and then I want to move on to a couple of our other funnel charts. Yeah the one quick thing that I would add is I think tracking a chart like this is so useful for exactly the reasons that Alex was saying, right? Like this is sort of like the lifeblood of your acquisition business. Almost always, I think when you see some kind of a change in this chart, whether positive or negative, the next step is usually going to be some kind of segmenting or filtering. Right it’s going to be all right. If if my conversion rate is going down, what’s going on there? Is that a country, driving that is at a particular store driving that like I think starting with what question is exactly right. When what changes when you’re conversion rates moving up or down, you’ve got to immediately then jump into the y to figure out what’s actually driving it. Yeah, Yeah. I actually hadn’t segmented this. That’s pretty cool. Like segmenting by country and segmenting by other things and then filtering the country. One is going to be overwhelming at first because of how. Yeah, there are, but it’s useful data. Right well I’ll just move on and switch to. Well Yeah Yeah but I could probably unselected and just select few that we would want to look at. All right. Let me. Looks like we. I live the next one I wanted to move on to. There we go. Is initial conversion and trial conversion. Actually, let’s just go to trial conversion. Oh, one other thing to note on conversion to paying is for those of you who do have lifetime or one time purchases in your app. Conversion to paying will actually take that into account. And so that’s conversion to any form of paying, not just subscription. It’s a good thing to keep in mind. So let’s hit this quick and then we’ll move to realize LTV, because that’s such an amazing chart that I wish we could have in a half hour just on that. But yeah, this, this is the trial conversion chart. Dan, why don’t you talk us through, you know, how this is calculated and how to think about this. Sure Yeah. So basically what this is visualizing, which also to if you get rid of the product duration filter, it’ll give you the view of all subscriptions. But basically what this is visualizing is for particular cohorts. So that’s important. I’ll get back to that in a second. For particular cohorts, what is their trial conversion journey from actually starting the trial converting to paid? And then the couple of other cases that happen in between that the way we’re visualizing this is we’re breaking it down into cohorts based on when you were first seen in the app. So right now this is the monthly view. This is grouping customers by the month that they first showed up in your app. At least that revenue cap site. What that allows you to do is then compare performance of those cohorts over time. So like for example, for your trial conversion rate, what you would be hoping to see is that for a cohort that started a few months ago, if you’ve been working on improving your trial conversion rate for your most recent months, that rate is going up over time. If it’s not a cohort of you, this can get really messy because trials and payments don’t happen at the same time, right? The payments happen some amount of time later. So viewing it in a cohort ad level is really critical to understanding how this rate is moving over time. Then the other brief thing that I would add to if you scroll down to the bottom of this table, what you’re going to see is that there are a lot of trials that are in some kind of a pending state for your most recent cohorts. That’s a critical metric to track as well. You’re not going to pay quite as close attention to that as you would the converted number. But what this is telling you is the number of trials that are still eligible to convert. It’s like going back to what you were mentioning on the previous chart, David, were like, you know, that conversion to paying looks like a drop, but it’s not a real drop. They just haven’t had a chance to convert yet. This is basically our way of showing that. And saying, all right, like August is lower right now, but that’s because you’ve got a lot of trials that are still pending that are going to have a chance to convert over the next few days and weeks. Yeah And then this is cohorted by first thing. Correct so. So what’s interesting here is then you have 10 pending conversions, which would mean 10 users who were first up in the app in March are actually currently pending in trial. Is that correct? That’s correct. And it could mean that they did start their trials recently, right? You could have started you could have opened the app for the first time a few months ago, but actually started your trial recently. What we see for most apps is there’s usually a pretty strong correlation between the time that you start the trial and the time that you first open the app. Yes, it’s often, you know, over 50% of the trial start you’ll ever get are within the first day and even sometimes the first hour of app open. Yeah Alex, I wanted to ask and I do kind of have some context on this, but I know you all really ramped up ad spend and so this must have been something you were really looking at closely. And there was a lot of customers coming in, a lot of trial started in April, may and June and July. Seattle tell me your thinking and what you think about when you look at this chart. Yeah, definitely. First, for context, do you see this? This whole chart has like a camel shape. There’s two peaks, right? There’s two reasons for that. The first is that our business is seasonal. As you can imagine, we help people grow plants and plants grow best when there’s a lot of sunshine, which is spring and summer. So we and all of our competitors see some pretty big increases starting in March and running through the end of August. And then adding to that, both years we ran reasonably large advertising campaigns. I think at peak, we were running something like 50 or $60,000 per month, ramping up from like maybe $5,000 or 10,000. And so that those two things combined created this really big upswell in New subscribers, right? And basically both years were mostly post the removal of the EDFA on, on iOS. And we were running, you know, Facebook, Instagram, TikTok. So we were basically testing the, the efficiency of those channels both years with like, like medium ad spend basically. Yeah and so then looking at this, what did it tell you about the cohorts you were bringing in and were you happy with the results you were getting? Yeah, I have so many lessons here. We were not able to make mobile advertising work. That’s kind of like the old bury the lead. Yeah, it’s super challenging. I’ve talked with a lot of different, like mobile and consumer companies and I think, you know, maybe some teams are making it work right now, but I think the vast majority are not. And so we pivoted fully away from paid mobile, and we’ve got some other channels that we’re working on. I will say that one of the most important things that we did as a mobile business was adding a set of surveys at the beginning of the app to basically understand who these people are. And so the really important one was, where did you find us? And so we asked them, did you find us on Instagram or a friend or partner? Then we have other ones around. Like, how many plants do you have at home? What’s your experience level? What are you interested in using the app for? Like what functionality? And you know, some people think like, oh, those types of things will cause people to drop off the app. We haven’t found that we see almost no churn because we weren’t answering really simple surveys. If anything, I actually think it almost May like increase a person’s interest in the product because it’s like, oh, it’s asking about me, it’s going to understand me better. But those are really, really important because those gave us the insight to understand that the users we were acquiring from paid are like one half or one quarter as likely to subscribe as users who come through either friend referrals or we have partnerships with plant retailers. So if you buy a plant, you can get a subscription to Greg, be introduced to Greg. Users who come to those two channels are like two to four times as likely to retain, to subscribe, et cetera and so that’s what we found. And it doesn’t show up a ton and well, it shows up in both like the initial subscription rate and does show up a bit in the trial conversion rate. In general, trial conversion rate for us has been like fine. You can see it ranges from like 50% to 75% and I think that’s reasonably good. And so it hasn’t been a focus area for us, but it does change dramatically. Like for example, another impactful thing we did is we put a paywall in onboarding as like a lot of our wins. I’ve been putting stuff in onboarding because that’s when everyone is experiencing it. When we did that, trial start rate went way up, but trial conversion rate dropped from like 65 70% down to 50% So it was really important for us to look at like, was that actually a net when it was for us? But there’s definitely a trade there. Yeah, I was going to ask you about that because you definitely see you see the trial start rate jump up quite a bit. So the trial start rate being 2% So that was when you didn’t have a paywall in onboarding. And I imagine it’s something I’ve talked to Jake more he runs super while we’ve talked a lot about is that it’s actually a really interesting metric to track is what percentage of users even see your paywall. And when you don’t put it on onboarding, it can be surprisingly low. Ours ours was around 20% Just to give you an idea. Yeah, like only one fifth. The way that we put it when we did this analysis in January of this year was only like one fifth of people ever even saw a cash register in our business. So one of the things I mean total tangent, but one of the things I’ve been thinking about as well after talking so much with Jake and about paywalls and thinking about it, is that when you walk into a store, you expect for there to be prices on things. And I’ve been a little hesitant, as you probably were, and this is why only 20% of people saw the cash register is that you feel like, oh, it’s being aggressive. You know, you don’t want to, you know, overshare the paywall. You don’t want to, like, scare people off. don’t want to. So you want them to have this great experience. You want them to like, you know, be a part of the community and really enjoy your app before you like, you know, slam this offensive paywall. But the more, you know, I’ve thought about it and it’s shifted my perspective to thinking it’s kind of just putting a price tag on things. And that’s actually, for consumers, a positive thing. don’t walk into a store with no price tags and feel comfortable, like, you know, the average person doesn’t walk into that kind of store because it doesn’t have price tags. It’s going to be very expensive. And so showing the paywall can actually be a good thing to do, just kind of like you said, like put him in front of the cash register, let them know that they can pay them. Yeah, this is really interesting to see that, you know, you did have that 69% conversion rate. And then by just moving the paywall, I was wondering if this shifted and there’s probably a couple of things going on. So you made the shift at the same time that you were also ramping up marketing. And so. Wondering if some of the trial start rate increase actually had to do with a higher intent user coming in from marketing. But it seems like it’s probably more the paywall change that drove this trial start rate increasing. Well, yeah, no, it wasn’t. Actually, those users were less likely to subscribe, but it was both. It was both. We did change probably too many things at once. It’s hard to move fast as a start up, do a lot of things, but then also know exactly what’s going on. So we put a paywall on onboarding. We also paywall more new functionality and so, so like and that was also, I think, very important to be honest. We actually don’t know yet which one was more important. We did both at the same time and now we need to go back and like A/B test that. Yeah, but it was probably both. Yeah yeah, that’s great. And one thing I hope, you know, kind of on the subject of us kind of becoming your data warehouse, it sounds like you, you know, run some of these analysis on your own. But at some point, I think it just makes sense for us to allow you to send those answers to where you came from, the source as a subscriber attribute. And then to be able to segment this based on subscriber attribute would be really fascinating because then you could see more directly right in our chart. So that’s something Dan and I will, I’m sure be talking about over time and figure out how to do that and make things like that more accessible. Last chart I wanted to talk about and then we’ll start taking Q&A is the realized LTV. So Dan, I mean, we it’s another topic we could literally talk about for an hour, but give us the overview of how to think about LTV. And even you know, there’s a lot of predictive models around LTV, but there’s a lot of hairy stuff around that. So the first chart we introduce, your revenue is realized LTV. So why did we do that and how should people think about this? Yeah, it’s a great question. So I guess first realized LTV that’s realized lifetime value, that’s the acronym there. A lot of folks have probably heard a very similar metric just as our RPU and said average revenue per user, it’s very much the same idea that the reason why we’re calling it realized LTV is because of exactly what you were saying, David, that predicting LTV is extremely common, extremely important for just about any subscription business, right? Like getting a sense of when a subscriber comes through the door, how much in lifetime value can you make from that subscriber over time? It’s a really difficult and a really important question. Right like you can’t you can’t know how much you can spend on customer acquisition costs without answering that question and understanding the liquidity of your business on the other side of that. So predicting LTV is really crucial. It’s really difficult to write like getting that prediction right if you get it wrong, if you assume your customers are going to subscribe a longer than they actually do, that could be extremely dangerous for your business. So with revenue cap, we’ve started with realized LTV because one, it’s just the more straightforward thing for us to measure. But two, it’s, it’s the thing you should be comparing all of your LTV predictions against, right? Like over a certain time horizon, what you should basically be doing is going back and saying, all right, I thought that in a 12 month period my subscribers are going to be worth x. Were they actually right? Like the realized LTV chart basically lets you go back in time, set a lifetime for your customers, right? You can customize that to 30, 60 days a year. You can have it be unbounded if you like and understand over that time horizon how much revenue is each subscriber generating? Yeah and so I want to talk to you a couple of things real quick. And this chart is brilliant. I hadn’t gotten to dive into this chart because my apps suck right now. And so until Alex shared the Gregg made me a collaborator on the Greg app, I hadn’t really got to understand the power of this. And so it’s so fascinating, especially looking at the real life LTV by different customer lifetimes, is that when you go unbounded this is kind of relatively flat. So interestingly on average per paying customer. So far. Greg makes between $30 and $40. They’ve realized $30 to $40 per paying customer over that lifetime. When you switched to the 30 day lifetime, what you see is exactly what we discussed earlier is the switch in strategy from promoting the monthly subscription to the annual subscription. So over 30 days, these monthly cohorts are only making on average, you know, what is this like 5 to $10? And then when you switch to the annual of course, the annual subscription is, you know, I’m guessing in the $30 ballpark. And so you just see that you see that flip. And it it’s so fascinating how some of these things normalize out. So looking at the older monthly cohorts. On average, they stay subscribed long enough to kind of equal out. So far to the annual subscription price. But what’s going to be really interesting is that as Greg switched to the annual subscription, this line will likely lift over time because when we say unbounded next year, when these cohorts renew, it’s going to shift that line up. And then if, you know, two years from now there’s another renewal and a certain percentage stay renewed, it’s going to lift that line up. And so you’re going to realize more LTV for these cohorts over time. But again, it’s really tough because you don’t know what that’s going to look like for your annuals. But Greg is such a fascinating example of this chart since they did switch from monthly to annual to see that the monthly really is so far still giving a similar LTV. So Alex, I’ve kind of gone on and on. How do you how do you think about LTV and how do you use these charts? There’s actually an additional wrinkle here that I want to point out. If you go back or if you went back to the bounded version, that jump was actually not the switch to annual, but I believe that was we switched from a 30 day trial to a 7 day trial. And so like and so any trials were excluded, right? Because when we switched fully to annual, we were already at like 60, 70% annual. Interestingly, we used to show monthly, annual and lifetime just completely next to each other, very little bias. And we found that 70% of people chose annual on their own. And so we just bumped that by like another 25% So that big jump was actually because we shortened the trial duration from 30 days to seven days. Yeah, we did. And we did A/B test that we did 37 and 0 days trial and we found that the seven day trial was for sure a winner. I think revenue cap may also have a blog post on trial duration, but that was in all of our ad testing paywalls. Like we tried stuff like video and paywall, blah, blah, blah. None of that really did much for us, but shortening the trial duration did quite a bit. So that that’s kind of interesting that yeah, that’s fascinating. So, so I switched it to three months, which takes out a little bit of that bias, but you still do see that kind of increasing over time. Yeah any other thoughts on LTV, Alex. And then we’ll, we’ll switch it to questions. Well, I’ll just point out that this is they’re thinking about an app business, whether you are bootstrapped or raising capital or whatever. LTV is really important to think about in terms of just the product that you’re offering, right? Because it’s not like LTV is, you know, more something you discover, I think, than something that you create because different products have different values to people, right? If you’re selling as an example of masterclass like their LTV, this is like a platform for watching videos with Gordon Ramsay, the shop or something. Their LTV is something like three, $400 or $500 because each of their classes is very expensive, right? Whereas tinder’s LTV is closer to like, I don’t know, $60 or something. A lot of apps are in the like 50 to $70 range. We did and analysis. And visit really important because your R.N. going back to that, that really important number is basically just what percentage of people that you acquire pay for your app and how much do each one of them pay. And the reality is that like businesses that can charge more make more money, right? Like Apple is a really successful business because they charge a lot of money for their products. Right and so I just want to point out, like, this is very hard to move. I think it’s very hard to move LTV without like sure, you can improve retention. That’s probably one of the best ways. But in terms of changing how much you charge, which really makes a big impact, it’s hard without making a brand new type of product, right. My favorite example actually recently a speech of I where last time I checked, there’s a mobile app that lets you do text to voice. They’re charging like $120 per year. I still want to know how, but like, I feel like that really makes a difference to the business model. Right and so I just want to point out that LTV is kind of like the meat of your business, right? How much is someone paying over time? Yeah well, one interesting thing on that front, too, and you kind of touched on it, was I talked to Zach. She said he has an app called expert, and he did quite a bit of experimentation across weekly, monthly. Oops we lose. David Thank you. Last minute. We’re not going to know the end of that experiment now. I know. Whatever I can, I can share a little bit more. I got to go. Oh, I was just going to mention that if I was like the PM for gray and I’m looking at that unbounded LTV chart. I am so excited to see how those back tables come. Thing is almost certainly that. Yes, now you’re get. OK my ever optic cable is actually dangling next door as they construct a new house, maybe somebody kicked it or something. So what I was going to say is I talked to Zach and he did quite a bit of tests and he found the LTV to be pretty consistent. So if you used a weekly subscription at like $3 a week or whatever it was, people would stay subscribed almost exactly enough to equal the LTV of his monthly subscriptions and his annual subscription, which was just a fascinating and one of the things we talked about on the podcast is that there is a certain have to, for your own app, kind of find that threshold of consumer willingness to pay for the value that you’re creating for them. And so as Alex was saying, like, the more value you can create, you will be able to charge more and bump that line. But that’s like the hardest. So hard to do and it’s also hard to experiment with. And then again, it’s like back to a lot of the stuff we talked about. When you’re doing annual subscriptions, it’s even harder to understand how those curves shape over time. So, yeah, it’s running out, running a subscription business in the consumer space. There’s, there’s a lot to think about. And so just to kind of, you know, wrap put a bow on all of this, you know, I think our charts are a great way to look at it. And hopefully this session has helped everyone attending and people watching the YouTube video afterward. Just think more deeply about your app business and then how to use revenue cut specifically and then other analytics broadly to think about your business and to really dig deep on why things are happening to be able to be on top of these movements. So that when you make a change, you see the result and can react. And so, you know, if you’re running a business and trying to grow and I kind of made a joke about it in the promotional material but like up into the rights are always good. And you know, when you looked at early on when we were looking at air for Greg, it’s up and to the right. You know, it looks great. But we sat here for an hour talking through all the different ups and Downs of making that go up and to the right. And I’m sure, Alex, you would love for it to be going up to the right quicker as well. So there’s always just so much to think about. And so much to analyze. Alex, Thank you. Is so incredible that you were willing to the latest look at your real life data. In this session and talk through how you think about your business. I just can’t express how grateful I am. I think it’s just going to be so incredibly helpful to others and maybe anybody in the audience who appreciates what Alex did and want to do it yourself. You know, I would love to do more sessions like this where we really dig deep into specific businesses because it’s just so valuable to look at real numbers. And it’s even valuable for us internally to be able to talk to customers like you and understand how you think about our charts and how you think about our data. So Thank you so much for doing it. I’m going to start taking questions. We’ve got 17 minutes over. So, Alex, if you need to drop it at any point or if you want, just drop. Now, feel free to just wave and check out. Dan and I will stick around and answer some questions. But Thank you guys so much. Yeah Thank you, everybody. I’ll hang out for 10 minutes just in case I can help. OK, cool. All right, so I’m going to go to the question section and we will just sort by upvotes. So Phelps question, what do you think of grandfathering PED users versus offering them long trials? What have been the best practices when converting from pay download to subscriptions? Are there any technical tips that lead to a seamless transaction and good user experience? I’ll take this one. I don’t think Alex has much experience on transitioning a paid app. I did this with my once in a pro app. I’ve talked to quite a few developers who’ve done this, a great resource on this topic. Dan if you’re able or Corey if you’re able to search the sub club episode with Madge, tabby and Jasper housing the darkroom co-founders, we talk pretty in-depth about their transition, and I think they did it really well. And just as a quick overview of their thinking and kind of how my thinking has evolved on this is that it depends significantly on your business goals and the current state of your business. So if you have a very active user base that’s, you know, hundreds of thousands of users and those users are love your product, find a ton of value. You know, finding a way to capture some of that revenue could be very important to your business. So like what darkroom did is they did paywall new features that they were adding, but they grandfathered all existing features in, and I think that’s potentially the best way to do it. I’ve talked to other businesses where they were just in a situation where the business demanded that they capture more revenue and so they did pay great features that had previously been free. That is really tricky to pull off because once you’ve given something away for free, it’s really hard to convince people to pay. So if for business reasons, you need to go down that route, I would talk to a lot of customers. I would really hone in on your messaging, on how it’s presented and find really careful ways to do that. I think the ideal scenario is what darkroom did. What I did with launch a pro is that you draw a line in the sand and you grandfather all existing paid features that they paid for or free features. If you’re changing the paywall to everyone who expected to already have access to that and then try and create so much more value on top of that, that you can convert those users as they need access to these new features. And that went really well for dark room. Again, the podcast episode is great on that. And then Dan, I’d love it. Yeah, I’d love to hear your thoughts on that as well. Yeah, I would just add to. It’s really difficult to put a number on this, which I think is why people hesitate with it so much. But those initial customers who are still using your product that are potentially going to see the paywall that you’re thinking about putting up, those are probably some of your strongest advocates, like those are probably some of the people that are like most excited about your product. And almost certainly there are other ways that they’re contributing to your revenue over time. You know, whether it’s their friends that they’re talking to, the feedback that they’re giving you. So it’s hard to measure, but that is not just purely lost revenue if you grandfathered them in. Yeah no, that’s a fantastic way to think about it. All right. We have an answer for our question for Alex. I see Greg’s subscribers can get three free months for referring a friend. What do you use to operate and run the referral program within greg? Yeah, I want to give a quick shout out. Actually finally got confirmation from Apple because there has been some app review uncertainty around referral systems, especially incentivized referrals. So just like 90 seconds of history games used to do these really shady incentivized systems where it’s like get 100 free gyms and it created some like perverse loops that weren’t good for the App Store. And so Apple became really sensitive to incentivized referrals, incentivized reviews and things like that. And so they have rejected referral systems in the past for subscription apps or any app that uses in-app purchases. But I did confirm with them just in the last week that they are allowed now their app review. I can’t guarantee you what’s allowed, what’s not exactly how you do it, but it is really important to know that you can do it. And then I’ll hand it off to Alex for like how we actually implement that. And then I actually love to hear if you ever got any kind of pushback from Apple on it. Did we get pushback from apple? No, never. Every week. Now I’ll just say, even at tinder, we had times where our apps could not be released for an entire month or two because Apple was holding up on some copy on a paywall or something. It was insane. Even tinder, we had like a direct line to Apple. Anyway, it’s their job to incentivize referrals at Tinder while you were there. No, we. We consider ongoing. Never ended up doing so. So we were rejected once because we have had an incentive system where if you refer someone, you get free Super Greg which sacks and the person you referred gets free Super Greg. Apple rejected us on the basis this was like a year ago. They said, you can’t give the person who’s joining something for free. We could give it to the person who’s doing the referring, but it couldn’t be like joining could not be incentivized. I don’t know if that’s still the case, but I will say the more interesting thing that we did is we implemented this advertisement for revenue cap, but we did it with revenue cats promotions. So we have a little mini in-house. We do a lot of things in-house, but we have little deep linking things we can recognize. When someone who’s signed up came from like a person who viewed a web page, you can use like branch I/O for that kind of thing. We just did ourselves. It’s fine, but then we, we grant a person free like super Greg with revenue cats promotional feature, which is a really handy way. The previous question about grandfathering existing functionality, we also use promotion rules for that. So when we switched from a free app to a paid app, we gave all existing users a promotional lifetime. You get free subscription thing through revenue accounts, promotional system. Very cool. I’m curious how that has been informative for you. When I get asked about referrals, it’s often from very early stage apps, you know, probably even earlier than Greg, but maybe even in the ballpark of Greg’s growth trajectory. And what’s always challenging for me to think about, especially as an engineering project and the hassles and getting it in the paywall and figuring everything out is that it’s a numbers game at the end of the day. And so where you put that call to action of, hey, refer a friend, you’re going to get a certain number of views. And so if you’ve got and Greg is probably a little different than a lot of apps in that you do have a lot of free users that you can kind of promote this functionality to and you have a larger user base at this stage compared to your revenue than a lot of apps do, but especially for smaller hops. So you don’t get that call to action seem enough. Yeah, maybe you get 10,000 views of the call to action and then, you know, an amazing, you know, a conversion rate on the call to action might be 10% And then an amazing conversion rate from having shared the link to people actually using the link might be 10% or you know, just using round numbers here. And so you just have to have so many views of the call to action for it to filter down to being meaningful revenue for your company compared to the cost to build that out both in time and using a service like branch or we’re just building it. And how so? Yeah, I’d love to hear how it’s performed and how you think about that investment over time. Yeah, I’ve gone deep in this because we are pivoting from paid acquisition to other channels and one of those is friends and referrals. It’s a blanket statement. I don’t think referral programs are they cannot be a primary growth channel. Right any sort of incentivize, any sort of anything around that. And these are structured referral program is I have never heard of it contributing more than like, you know, even in the early days at Uber I think it was like 10% ish when they had those $20 credits. But I’ve just I’ve talked to people on airbnb, they had a program, uber, et cetera even for the greats that are known for word of mouth growth, it’s not the referral program. It’s building a product that people want to share with their friends. And we even see this at Greg. We have that incentivized referral program and it contributes like 2% to our installs. Like it’s, it’s like infinitesimal. Whereas friends, word of mouth is closer to like ten, 15% So most people who are telling their friends about it aren’t even using the referral program because like all they really care about is it just come up and conversation, et cetera. So yeah, I would say anybody who’s looking into that, like I would highly recommend not thinking about the mechanism for offering your product more. Think about how do I make a product that either is worth sharing or is like social by nature, like an app that you, you know, clubhouse became very big because like, yeah, you invite people to it, you know, blah, blah, blah. Yeah and then how are you thinking about how that kind of fits into your long term strategy? Is a referral program something that you’re going to, you know, create a banner in the app and be pushing it more consistently and/or have you already been doing that and you still only saw that 2%? Yeah, we don’t push at that. I believe that’s definitely true. It’s kind of hard to find right now. As I mentioned, we are pivoting away from paid acquisition as our primary growth channel into we’re testing both deeper partnerships with retailers. That’s just a good fit for our business. And we’re actually focusing on can we make Greg like a social first product? So that matters a lot more to us. Can we make it a products that you kind of need to or really want to invite your friends to in order to use it? Yeah, which is very notable. Operate and then a referral program is just cannibalizing your word of mouth. Yeah Yeah. And I mean I think that like, you know, there’s actually a really good blog post. I, if anybody is interested in this topic, Andrew Chen, who’s kind of like a growth guru dude, just wrote a blog post about referral programs. He ran Ubers back in the day. He breaks us down a lot, but it’s like the referral program can accelerate it, but it’s not going to create the word of mouth growth. And so I think we won’t focus on the referral program for another six months. First, we need to create the basic human reason to invite your friends, right? Yeah Dan, anything to add? The only quick thing I would add there is that when we experimented with this at teltech, building out the referral program was way more complex than we gave it credit for at first. And so that coupled with the like moderate value prop on the other side of that, like I definitely look back on that with a bit of regret that it probably wasn’t worth their time and energy. Yeah Yeah. Like, and I don’t know if I’m sharing a big secret here, but we were working on a public API for revenue cat and this is the sort of thing that somebody could potentially build on top of revenue cat a drop in referral program and if it really is. A few lines of code and very simple to implement. It might be something we’re just doing because it’s easy. And so when it’s really easy, then maybe it’s worth experimenting with. I have talked to too many people over the years who spend a whole month or two months really early in the development cycle when they would be so much better focused on just building out the value prop of proving product market fit and they end up spending a lot of engineering time building this out, thinking it’s going to solve their growth problems when it’s just not going to do that, especially early when you just don’t even get as many kind of swings at bat in being able to show that the call to action there. Great, Mike. The lovely thing about using a mirrorless camera in Texas is that my office gets so hot that the sensor overheats and my video goes out. I need, like, separate cooling in here. All right. We’ll go to the next question, though. Why did you decide to keep most features behind a paywall, particularly as it also seems like you want to build a community? There seems to be a trade there, as free users can’t get much value from the app and won’t likely contribute to the community aspect of it. This is like the perennial question and out of freemium strategy is so hard. Yeah, it’s hard. You doing it. How do you think of that? Yeah, it’s I mean, I wouldn’t say that we’re doing it, like, perfectly. We’re still or figuring it out. So I think that the short answer to this is that it depends on your business goals and your business state, which really comes down to your funding rate at the end of the day. You have to if you have a team, you have to meet payroll and everything. And if you have investors, you have to meet their expectations. And so we chose pretty early on to charge for the app because we wanted we wanted clarity on are we creating enough value that people want to pay for it? And we wanted to know who’s paying for it and really is more about the KPI, that metric of revenue than it was about actually making money. Now that we’re growing into a more mature business, it is more about actually just making money enough to grow, right? Like we do. You’re going to have to make salaries, you have to pay for us, things like that. So I think it’s going to totally depend on case by case. The only other thing I’ll mention it depends heavily on is your acquisition channel. So the real answer why we paywalled more things is we were focusing on a advertising based growth process where that basically boils down to the, you know, how many people can you bring in via advertisements? How much do you pay for each one and how much revenue do you make on each one? And as long as you make more revenue within 90 days or, you know, a year, then you pay to acquire that user. Then you’re in business and you’re good, right? I think that channel is now a lot more difficult. And so I think that we will be probably moving Gregg to more focused on freemium in the next like year as we focus more on growth through the community. So to your point, your intuition there, I think is right and we’ll need to focus on that. And then and that may have been part of what was challenging for you all about making ads work is that you were trying to balance on the head of a needle that you wanted, like you’re trying to create this great free experience and really for a lot of apps. And again, Jake more talked about this in several podcasts and webinars we’ve done is that sometimes having a hard paywall is actually the best strategy for an app that doesn’t get value from the community because you forced that decision. So when you pay, you know, five, $10 to get somebody to download your app and they open the app, you’ve got to take some swings at that. And then there’s also strategies there that when somebody doesn’t convert, you know, send them a follow up a day later with a 60% off. And part of Jake’s theory is like, if I can get any, if I paid $5 I want to get anything I can from that user. So he would even go over the course of a week discount more and more and more and, you know, wasn’t a huge part of the revenue, but it was capturing some revenue that otherwise just would have been lost like it was just swings a path that you’re not taking if you’re not following up on that. And so if you’re trying to get paid advertising to work, but you’re also trying to create this great free experience, there’s a really strong tension there. And so so it really insightful for you to kind of lead talking about it really depends a lot on your acquisition strategy of how you’re going to balance that. So yeah, fantastic answer. There was a question about the primary growth engine and Alex did answer that already. So we’ll move on to this one. Is by failure events a metric you usually look at? Where do you recommend searching to lower this metric? And I’ll do a quick aside and then I’d love to get y’all’s input. Another thing I was talking to Jake about Jake and I talk a lot am running super wall and me being a revenue cat. There’s a lot of synergy there and the people we talked to, but then the different viewpoints we have in how things are going. And so he actually looked at this metric, we don’t track it and we’ll probably have the data in revenue cap, but we don’t expose it in the charts or we haven’t looked at benchmarks related to this. But it was fascinating, and I believe, if I’m remembering correctly so, there’s different layers to a failed transaction. When you tap on Start free trial and Apple’s little paysheet comes up, a large percentage of users once they actually see like, oh, this is like will auto convert and I have to use my face or my thumbprint. To confirm, and this is the real deal. A high percentage of users just drop off right there. So if you’re tracking taps on the call to action button, you know, continue or start subscription or start free trial. A high percentage of those wouldn’t actually be failed transactions. It would be just people not finishing the job with Apple’s pay. And then the next step to that is, then you have a lot of users who do actually confirm and then they go through their seven week free trial and then the subscription is supposed to convert. And then there’s we actually did a whole webinar on this. Again, Dan or Corey, if you have a sector to look this up, is the subscription app accounting webinar we have on YouTube with Ariel from app figures. And so there’s this whole stage that a transaction goes through because Apple doesn’t charge immediately when the transaction is supposed to go through, there’s fuzziness in it doesn’t actually even try the credit card like the exact, you know, seven days to the minute after they started the Free trial. It’s like all past seven days. And so what ends up happening is you have all sorts of cases where that transaction can actually fail and/or go into a grace period. So you can have somebody who has a debit card connected and they’re between paychecks and it just doesn’t go through calling the camera. You can have situations where people use gift cards and they only have a gift card. And the balance on that gift card has run out. You can have situations where the credit card expires. There’s all sorts of reasons for that transaction to fail, and a lot of people freak out when they do track this number. And another thing we really should do, benchmarks out on for the data we have inside revenue cat. But failure rates are high. So when you see a high failure rate, don’t freak out. It’s pretty common to have a relatively high failure rate. So then I know both of you actually have insight into this, so I’d love to hear Alex, you go first and then if you need to jump and feel free. And then after he answers and Dan, I know you probably have a ton of experience from Teletec around these failed transactions. Yeah I’ll stick around for one more question for. That’s correct. So yeah, we do track billing errors and we specifically track the ones that like revenue cat sends us an event or has a report on like this was a trial Apple attempted To bill and then it failed and then, you know, they’ll get an email like update your credit card. We do track that. It’s important. I think I saw a chat. Someone was asking that they’re seeing it higher on Android than on iOS. We don’t see that much higher on Android. I would say in general, we do see about half the conversion rates on Android. So Android and iOS do perform very differently in general. Yeah, I saw this, a Tinder out scale like very different, but billing errors for us and not that different. So I would just say like if you’re seeing like 80% billing errors, like it should never look like that. And the only thing that would come to mind is maybe you’re getting like fraudulent traffic or something like that. Like small chance. There also can be a high geographic variance as well. So if your Android version is really popular in a specific country. And that country has higher failure rates because they use debit cards or gift cards or something. Yeah, if you see it, it’s worth investigating. It’s not always going to be completely out of the ordinary, but look first at the kind of geographic variance. Yeah, I might be wrong, but I think we’re usually expecting like sub like 10% or fewer of trials fail because of a billing. Or the only other thing I’ll mention that is not often on people’s radars wasn’t on mine at first, but is related is refunds. So this is increasingly a thing that you can subscribe to an app and then you can request a refund through Apple support revenue kind of has a dashboard for that, but I would put those two together. Billing errors and refunds are both like things that should add up to less than 10% of your your, you know, cancellation rate, quote unquote, but are worth keeping an eye on. Yeah, one thing that revenue cap makes very easy. And that you should absolutely do in your app is that when you get that refund event, you check, check the subscriber. It’s a purchaser info or maybe we renamed the method now, but when that refund event comes in, cut them off. Like because there used to be a thing for a long time or in the early days of App Store is that Apple would let you keep using a paid app even after you got refunded. And there was like people in the know would just buy an app, ask for a refund and then keep using it. They eventually started clamping down on that. But there’s been similar patterns with subscription apps where. A lot of scripts and apps aren’t cutting off the usage. Once a refund happens. And so people are like, oh, I’ll cancel and then just get to keep using the app for a year. So it’s, you know, it’s probably not 10% of your revenue or anything. It’s not going to make a huge difference. But as a kind of hygiene thing, you should make sure that your refunded users are getting access. And Dan, you probably have a ton of experience with this from Teletec. So what if failure rates look like. And what were you tracking? Yeah so I’ll add a couple of quick tips there. We definitely paid close attention to the failure rates. We did see a little bit worse on Android than on iOS. So that was in particular a big focus of ours. I would say a couple of the things that worked well to mitigate that for us. One is that we enabled grace periods everywhere we possibly could. Like even just if you have a product that’s very expensive to deliver, this might be a different equation for you. But assuming your product is fairly inexpensive to deliver, there’s such little harm in turning on the grace period from the Play Store for each of your products in the App Store and just giving a longer time to potentially get that conversion right. There are all sorts of reasons why a payment might fail that have nothing to do with the customer not being interested in subscribing to your product. So that was big for us. The other was just letting customers know about it. Like Alex mentioned, there’s typically an automatic email that goes out. We would even add to that sometimes with a push notification from one of our apps and an interstitial in the app to update your billing information. If you’re still in that state, those things help. They’re marginal, right? They don’t change your business necessarily, but they help. And then the other thing I would add to going back to the other scenario of not actually a failure, but a customer seeing the patient and then choosing to leave it. You can track that cancel button as well on the patient, and that’s potentially a great opportunity to trigger a subscription offer or something along those lines. Because that your customer has some degree of interest in your product. They got they got scared by the price that they saw. But that doesn’t necessarily mean that they’re lost. That’s that’s potentially a great opportunity for some kind of CRM marketing to them at a later time. Yeah, it was, you know, eye to eye on failure rates is that Apple and Google generally do a pretty good job. By Alex, Thank you so much for your time. Really appreciate it. Apple and Google do generally do a pretty amazing job at Dunning. And so Dunning is, you know, finance term for collecting what’s due. And so when you run a subscription business and you have these failures, you know, so, so like revenue, when a customer doesn’t pay us, you know, we eventually manually have to email them and, you know, ask for payment and things like that. And so different businesses have all different ways to kind of collect these payments. So so what Apple does is that people are highly incentivized to keep their credit card updated for the App Store. So they might have a iCloud subscription. They’ll have an Apple TV subscription. And so collectively, all developers together help keep credit cards updated, help keep balances topped off if they’re using a gift card, help keep. It’s like there’s kind of like community incentive dining. If you’re doing web payments can be a whole rabbit hole. You’re there where payments are at Teletec. Right we did have a small scale, but Yeah. And so if you ramp up web payments. There’s a lot to Dunning that is that the Apple does so well that you don’t realize until you have to do it yourself for web payments. Because when you’re collecting a credit card directly, when that payment fails, you have to get back in touch with that user. They have to be so highly incentivized that they’re going to go back to your website, re-enter that credit card number. And on the App Store and the Play Store with grace periods. As Jim was mentioning, if you can enable them anywhere. You can. That gives time for that user to be incentivized because their iCloud subscription expired and they got an email from Apple and then their Peloton subscription expired and they got an email from Peloton. And so that kind of collective Dunning is way more powerful than I think a lot of people give credit to. And so, you know, if you are doing web payments, that’s something to really keep an eye on both the failure rates and then also the failure rates of renewals, which is even more tricky. All right. Let’s take one more question. This this went really long, but this has been such a great conversation. A lot of good questions. So, yeah, a lot of great questions. I mean, based on how long it took us to answer them, especially good question by the depth of conversation that leads to. All right. So what’s your advice on having free access to premium features for a few days before starting Apple’s free trial? Oh, this is an interesting one. Yeah why don’t you take it for us? I have a lot of time, but it is. Yeah, I would say first it depends a lot on the product. And what I mostly mean by that is what the intent and the state of the customer is when they’re coming to your product. Right so like if you think about most acquisition channels, like getting to the point of installing an app is usually like a really high intent action, right? Like if you imagine that you like just search on Google for some keyword that led the customer to your app. They clicked on that, they read the description, they clicked on that to install like that, customers really motivated right now if that’s the case, that also probably makes it a really good time to pitch your premium features, pitch your best benefits and your most expensive benefits. And so losing out on that opportunity, it can be pretty costly, right? Like the customer’s really motivated right then. But this was very true for our experience at Teletec. Like we have apps that allow you to record a call which you probably have to do right now or to block a spam call that’s annoying you right now. And so like if that’s a really high intent moment for you, it’s really valuable then to put the paywall in front of them and give them a chance to convert on the most valuable features. That being said, I don’t have as much experience on the other side of that, so I’m curious to get your take. David I could totally see a scenario where there are products where if that’s not the case, if you’ve got to do some extra selling before the customer’s is going to be convinced to pay for your product, then maybe it is valuable to add some additional functionality in front of the free trial because just like we were saying, like it’s a free trial, when you pop the pay sheet, there’s a price there. And like there is real friction. Valuable friction. But it’s a real. Yeah and this is so context dependent. And so. There is the key here is you want your potential customer. You know, somebody who’s downloaded your app to understand at the deepest level they possibly can. The value you’re going to deliver to them with your app in. The challenging thing is that depending on what value you deliver, it can be hard to. Demonstrate that value and it can be hard to communicate that value. And so that’s kind of the framework you should be thinking about in analyzing how to use free features to demonstrate that value to users and to explain the value and convince them that it is something worth paying for. So I think dipsy is a great example of one end of the spectrum where there a female sexual wellness app we had him on the podcast is really fun. We talked about sex, lube and retention in the same focus. So there are asexual well, female sexual on this app and they have auto erotica. And what they do is they have stories that are multi-part series. So they might be like a 10 part series where it’s the same characters and things like that. And so what we talked to Fe, one of the co-founders about on that, on that podcast was that people engage with those stories and kind of want to see how the characters develop. So they don’t have a free trial. But what they do is they give away the first couple of episodes. And the most interesting thing here about this particular case is that they give away the most valuable, the ones that are most engaging their best content, because those are the ones that are best at demonstrating the value that people are going to get. And you’re in a situation where they can experience that value and then want more fitness. I on the other end of the spectrum, again had a podcast with Jake and multiple conversations on YouTube and other places. He runs an outfit and say Hi. He faced kind of an opposite problem is that the real value of the app is when you’re in the gym getting to experience these amazing like 3D avatars that show you exactly the motions that you need to be doing in the programming and everything else. Well, when you download fitness, they are sitting at your desk because a friend told you about it or you saw an ad. When you’re scrolling Instagram, you’re not in the gym. So how do you explain and demonstrate that value to somebody? And this is where I think a lot of people get to him on camera. A lot of people get onboarding. Wrong is that people aren’t going to read an essay that you write in your onboarding and people just skip your little like tutorials and stuff. Because ultimately, if those things aren’t delivering and/or demonstrating the value of your app and it’s walls of tack, like you think, oh, well, I’ll write this thing and convince them of the value. People are going to read that. Yeah so what Jake had to do is get creative about how do I demonstrate this value. And so what he ended up landing on that was like tremendous for his app was he shows off the paywall as soon as you open the app and the paywall has a video that shows you what it’s actually going to be like working out with fitness because you can’t get to that value prop until you’re actually at the gym doing it. And so a lot of fitness apps, you know, take the opposite approach where it’s like 60 onboarding steps and there is ways in which that can be a good approach because you think, wow, if they’re looking at my height and weight, they must have some really sophisticated algorithm that’s going to give me the exact right plan for my body and I’m going to lose weight and feel better. And so, you know, those kind of things can kind of demonstrate value. But for Jake, he found that it was much more powerful to put that video in the paywall in front. So when you’re thinking about free features, those are like two pretty opposite ends of the spectrum is you want to figure out what free features, support, understanding the value, but in a way that entices them to actually pay. Because if you give too much away for free, even if it’s like, wow, this is, this is amazing feature that everybody’s going to use and love and use the app. You can very easily give too much away for free. And that’s where sometimes experimenting with a hard paywall might actually be a great way to, like, really understand what is going to engage the users instead of trying this really big freemium strategy. So again, with Greg, there’s so much value in this free users with the community and other stuff. So, you know, that would be the worst strategy for that sort of app. But so there’s just a crazy spectrum with all different like ups and Downs along the way. But I think that’s a great way to think about it, is what free features are kind of contributing to that value perception without taking away from or at least balancing the kind of need to pay. And so content apps have it really easy in some ways where you can give away the first episode in a series. You know, apps like fitness say I have a much more challenging time. And so you just need to talk to your users to, to analyze your onboarding and retention, look at the features that long term subscribers are more engaging with. That’s actually another really good lesson from the podcast episode I did with Sean Ellis, and he was talking about when you do re-engagement campaigns. And so this is it’s kind of a lesson here that applies as well in understanding which free features are the ones to put behind or in front of the paywall is that you find those features that lead to the kind of outcomes that you want and then you greased the path to getting to those experiences. So yeah, I could go on for another 30 minutes, but I think that’s just a great way to think about it. Any closing thoughts, Dan, and then we will go and wrap up. Yeah, no, I’m probably just repeating what you’re saying at this point, but both of those cases that you describe, like what I love about them is it’s such like a transparent way of the developer just thinking about how do I sell my product, right? Like, how do I how do I make the pitch to you that this thing is worth spending your time and energy on? And if you think about it through that lens, does it make the questions easier to answer? But it makes the questions to answer very, very obvious. Yeah all right. That was super fun. Thanks to those of you who stuck around to the bitter end. And if you’re watching this on YouTube, I hope you’ve enjoyed our conversation. We will be doing more of these. Dan, it was fantastic talking to you. We got to have you on and do more things like this. Your wealth of knowledge from your time at Teletec and just the way you think about these things, I think brings a lot of insight that we need to share more. So we’ll have to do more of these. Again, check out the couple of podcasts I mentioned, like five or six episodes. It definitely informs my thinking. So if you want to know how I think about things, a lot of my thinking is really informed by the questions I ask and the founders I talked to, the investors I talked to on the club podcast, sign up for the waitlist on the community. Again, similar to a lot of really amazing conversations we do at Amazon there and things like that. If you stay this long in you’re not a revenue customer. We’d love to chat. You know, I’m obviously super biased, but I’m here at revenue cat as a customer. I use it in my apps and love the product and excited about what we’re building. And we have super thoughtful people like Dan actually building out the features. And so, you know, what you see today is pretty awesome. We’ve got some great charts. Our scores are incredible. You know, I had so many bugs in my apps around subscription renewals and other stuff that like revenue cat just like took away all the edge cases there, solved all the weird identity things where there’s multiple receipts because your wife’s logged in with your account, but she had her App Store account, like all those like weird things. Like we’ve seen it. We process like $1.5 billion in subscription app revenue every year. So we’ve seen those cases, we’ve solved it. So our cases are incredible. Our back end infrastructure, you know, we’ve had apps hit the number one on the App Store multiple times. So if you’re worried about can we handle the scale, we’ve handled the biggest scale the App Store can throw at us. And it was crazy. I mean, I’m not going to lie. It was a challenging week, you know, when widgets met. We’ve talked about this before. We just hit number one on the App Store. It was a tough week like it showed us the weaknesses. We didn’t go down. We didn’t have any major issues. But it we learned a lot about our infrastructure that we’ve since improved and had other apps go hit, number one, and not have the internal firefighting that we had to have. So what we have now is battle tested against the top apps in the App Store. But what’s so exciting to me inside revenue cap, but also as a customer is that we have people like Dan working week in, week out to make it better. So we have some really cool stuff in the works in the short term and then long term plans to just keep adding value on. And so by integrating revenue cap today, you’re not just getting what we already have, which is super valuable, but you’re also now on team revenue and are going to get all sorts of stuff added and more value and more and more over time. So that’s my like bleeding heart. I love urban closing to talking about all these things. So again, like our dashboard, pretty amazing. Dan and I are constantly talking about ways that we’re going to improve that over time. And so definitely reach out to me, to Dan, to customer support. You know, if there’s feedback that you have around a dashboard, if there’s ways you want to analyze your subscription business that we don’t help facilitate, you know, we’re constantly thinking about these things. And with the camera going off, I will just wrap it up. Thanks, Dan, and I’ll see you on Slack soon. Sounds good. Thanks for the invite, everyone, I guess.

  • Subscribe to newsletters

    Watch more

    Want to see how RevenueCat can help?

    “RevenueCat enables us to have one single source of truth for subscriptions and revenue data.”

    Olivier Lemarie, PhotoRoomOlivier Lemarie, PhotoRoom
    Read case study