Under the Radar

260: Subscription School


00:00:00   Welcome to Under the Radar, a show about independent iOS app

00:00:03   development.

00:00:04   I am Marco Arment

00:00:05   And I'm David Smith.

00:00:06   Under the Radar is usually not longer than 30 minutes,

00:00:08   so let's get started.

00:00:10   So the topic for today's episode is going to be, I guess,

00:00:15   subscription school with underscore,

00:00:17   and talking through some recent things

00:00:21   that I've learned in my understanding,

00:00:23   and the way that a subscription business works.

00:00:28   Now is the school itself subscription price and can I get a free trial?

00:00:31   I suppose everyone is on a free trial right now.

00:00:34   If you listen past five minutes we'll automatically bill you $25,000 for the year of school.

00:00:41   There you go.

00:00:41   Oh god, school is much more expensive than that now isn't it?

00:00:44   It is.

00:00:46   Thankfully under the radar is, so far at least, if there is a trial limit we haven't quite reached yet even if it's episode 260.

00:00:55   So, who knows? Maybe at some point we'll reach our trial image, but we have not gotten

00:00:59   there yet.

00:01:00   Yes, we – this podcast itself is not a very good business. Hopefully you'll learn some

00:01:06   lessons here to make your business better.

00:01:08   Yes. You need a better LTV than we have.

00:01:12   Yes.

00:01:13   But yes, so it's like subscriptions and I think what's interesting with subscriptions

00:01:17   is that they are – in this context, in the context of like a subscription as people sign

00:01:23   up, give you money, and then agree to give you money into the future.

00:01:28   And the kind of subscription model that we're talking about here is mostly going to be obviously

00:01:31   like app subscription, where there's some subscription inside of your app and you use

00:01:35   it and that's great.

00:01:36   But I think it applies to functionally to any kind of business where that's the case.

00:01:40   So it could be doesn't have to be just software subscriptions, but the general terms will

00:01:45   apply to that.

00:01:46   I think is increasingly becoming the predominant model for monetizing software applications.

00:01:54   You and I both use it in our applications.

00:01:57   It's increasingly just kind of becoming the norm and that is for a lot of really good

00:02:01   reasons that it provides a kind of consistent, reliable input stream to your business that

00:02:10   you can predict and plan for.

00:02:12   It lets you align the kind of value that people are receiving from your application with the

00:02:18   income you generate from it, which is great.

00:02:21   There's lots of things that are great about it, but the kind of logistics of subscriptions

00:02:27   are more complicated than, you know, if you go back to the kind of old, more classically

00:02:33   used paid-up front model, that was straightforward.

00:02:36   Like user gives you money, you give user app, and that's that.

00:02:40   There's no kind of ongoing calculations and predicting and forecasting and things that

00:02:45   you need to worry about.

00:02:47   The only thing you could really look at is like how many new purchases you have each

00:02:50   day.

00:02:51   Like that's really all there is.

00:02:52   And so, but with subscriptions, things get complicated.

00:02:56   And recently I've gone on a journey of this to try and better understand the kind of the

00:03:01   fundamentals behind subscriptions and to understand really what's going on there because I was

00:03:07   running into some weird issues. But before I get into my specific thing and the discoveries

00:03:13   I had, I think it's probably useful to just start off with just a quick overview of some

00:03:17   of the terms I'm going to use. Because unless you're coming into software development

00:03:23   from an MBA, a business background, or kind of that side of things, you may not be as

00:03:28   familiar with these terms. So just quickly run through what these terms are and what

00:03:32   they mean. So there's a subscriber. So a subscriber is a person who has signed up for

00:03:37   a subscription that is, you know, and then they've agreed to give you money on a regular

00:03:42   basis. You often will have something called a trial, which is an initial period where

00:03:48   someone will sign up for a subscription but not get billed until the trial duration has

00:03:53   elapsed. And related to that, you will have a trial conversion rate, which is the proportion

00:03:59   of your users who start a trial who then go on to actually make their first payment. So

00:04:05   So that's your trial conversion rate.

00:04:08   Following on to that, you have your retention rate or your churn rate.

00:04:12   And these are exactly the same things, just sort of one minus the percent the other way

00:04:17   around.

00:04:18   So the retention rate is the percent of people you keep at each renewal period.

00:04:23   And the churn rate is the percent of people that you lose at every subscription period.

00:04:28   Essentially, this is very much the kind of like cup half full, cup half empty kind of

00:04:32   a difference.

00:04:33   actual functional difference here. And these rates can be applied at various points in

00:04:39   a subscription. So you may have like your overall churn rate, which is kind of the average

00:04:44   of your various, at various renewal periods. Or you may say like, what's your first month

00:04:51   churn rate? What's your three month retention rate? These are just terms that you can do

00:04:56   to kind of look at the data in slightly different ways. But ultimately, you're just trying to

00:04:59   work out how many people are you keeping or how many people are losing, depending on which

00:05:03   side of it you're trying to understand. The next term is MRR or ARR, sometimes referred

00:05:10   to, which is monthly recurring revenue or annual recurring revenue.

00:05:14   It's not like a pirate thing. Arr. I'm sure a million people have made that joke

00:05:19   in various like growth hacking and marketing meetings in every app company everywhere.

00:05:24   I am pretty confident that is true. I think there is even actually a thing in this where

00:05:29   people refer to pirate metrics.

00:05:31   Oh no!

00:05:32   Where there's – it's like ARRR and retention rate. So ARRRR.

00:05:36   Oh no, I'm so sorry to the world.

00:05:42   Yes. So I've run into pirate metrics as a thing. But anyway, these are just – essentially

00:05:49   if you take how much money you could expect to make in a month based on your number of

00:05:55   subscribers and the amount of money that they're going to give you. And so this is just a way

00:06:00   to kind of predict what this is going to be. And the nice thing about MRR and ARR is they

00:06:06   tend to be somewhat predictable because they're based largely on the size of your subscriber

00:06:16   database. So you can sort of predict what these values are, and so they tend to be things

00:06:21   that are nice to use for forecasting and planning purposes.

00:06:27   The next is LTV, or long-term value. So this is how much income you will expect to receive

00:06:35   from a particular user over the lifetime of their use of the app or the lifetime of their

00:06:41   subscription. And typically, LTV values have some kind of a window to them, because obviously

00:06:48   you can't predict what is the long-term value of this customer on an infinite time scale.

00:06:56   That doesn't particularly make sense or is logical. So you'll tend to say something like,

00:07:00   on a 24-month time scale, how much income would you get from a user? What's the 24-month

00:07:04   LTV? So that's just something to keep in mind. And it's useful for getting a sense of how

00:07:10   valuable is a new subscription to you. Because while they may, you know, say you have a subscription

00:07:16   that's $1.99 a month, their value isn't $1.99, it's, you know, $1.99 plus the $1.99 times

00:07:25   essentially how many times the average user will renew. And so you kind of can do some

00:07:30   math to work that out. And that's useful for another metric, which is your CAC or your

00:07:36   AC, your consumer acquisition cost, which is if you have costs in order to acquire a

00:07:41   user, so say through advertising or affiliate marketing or some method of acquiring them

00:07:46   that costs you money, it's you can that is how much money you are spending to acquire

00:07:51   a user, which ideally you obviously would expect to be ideally less than the long term

00:07:55   value of that person so that you are making money rather than spending. If you're spending

00:08:00   $10 to acquire a user and they are worth $5 to you, you are losing money on every subscription.

00:08:07   So that's not a good place to be.

00:08:10   And lastly is just price.

00:08:13   It's just worth sort of mentioning that in all of these things, you tend to have some

00:08:16   kind of price that you assign to a subscription.

00:08:20   In Apple's case, then there's also the proceeds rate that will get applied to that.

00:08:24   And so your revenue or your proceeds is the thing that's actually valuable.

00:08:29   I think some people, depending on where you are, I know RevenueCat's MRR is before Apple's

00:08:34   cut, but I always think of MRR as the after Apple's cut version just because that's what

00:08:40   actually matters to me.

00:08:41   The fact that there's this bigger number that maybe if I was trying to put it in a pitch

00:08:46   deck and get funding or something would be useful, but for the purposes of what's actually

00:08:50   going to hit my bank account, I tend to think of MRR on a proceeds basis rather than on

00:08:55   kind of a gross basis.

00:08:56   Yeah, it's probably better to think – it's not good for your psychology and mental health

00:09:00   if you see how much Apple's making more frequently than you need to.

00:09:05   No. Yeah. And so that's – so hopefully that's a helpful kind of just like a quick

00:09:11   overview of the terms involved here because I definitely didn't know all of these when

00:09:15   I started and I would be reading something or hearing something talking about subscriptions

00:09:19   and these terms would come up and I didn't know what they meant. So hopefully as I kind

00:09:23   of dive into where I'm going from here, that's a good place to kind of lay the groundwork.

00:09:28   We are brought to you this episode by CleanMyMac X. Make sure you can trust your Mac. The Mac

00:09:34   is a crucial tool for work, education, and life. And MacPaw is on a mission to help machines

00:09:39   help you, so they developed CleanMyMac X, an ideal decluttering app for the Mac that

00:09:45   can keep it in tip-top shape. So what exactly does it do? CleanMyMac X includes 49 tools

00:09:50   to find and delete invisible computer stuff.

00:09:52   Junk, files, you know.

00:09:54   And it helps to tune up your Mac so it

00:09:55   runs at its maximum speed.

00:09:57   So it organizes your disk space.

00:09:58   It shows you large, hidden folders

00:10:00   so you can free up tons of space so your Mac never

00:10:02   runs into issues with storage.

00:10:04   It also fights Mac-specific malware and adware

00:10:07   and protects your computer from them.

00:10:08   And it prevents a Mac from cluttering, lagging,

00:10:11   and slowing down.

00:10:12   CleanMyMac's all new menu app helps

00:10:14   you take care of your Mac's health with six detailed

00:10:16   monitors that provide useful information about your Mac's

00:10:18   storage, state of protection, CPU performance, RAM,

00:10:22   battery and network speeds.

00:10:24   And CleanMyMac X is notarized by Apple

00:10:26   and is also available in the Mac App Store.

00:10:28   So it's been checked for security

00:10:30   and it really stands out in its design.

00:10:32   In 2021, CleanMyMac was honored with the Red Dot Award,

00:10:35   UX Design Award and has become

00:10:36   a Webby Award nominee this year.

00:10:39   You can get CleanMyMac X today

00:10:41   with 5% off at macpa.app/radar.

00:10:45   This discount is only valid for two weeks,

00:10:47   So go now to macpaw.app/radar for 5% off.

00:10:52   And one final note, despite the war in MacPaw's home country

00:10:55   of Ukraine, the team have worked hard

00:10:57   to make sure there's no disruptions in the support

00:11:00   and development of CleanMyMac X.

00:11:01   The product is stable, safe, and secure.

00:11:04   So our thanks to CleanMyMac X for their support

00:11:07   of under the radar and all of Relay FM.

00:11:10   So the next thing that I want to talk about--

00:11:12   and this is the discovery that I made, which is not a new thing.

00:11:17   This is one of those, it's not, you know, I went on this long kind of journey of discovery

00:11:22   and ended up discovering that it's a well established body of work and is a very known

00:11:26   thing in subscription management, but I didn't know it. And so I'm sort of want to talk about

00:11:31   it here in case other people are in the same place. And this is sort of the concept. I

00:11:35   have heard many people refer to it as carrying capacity, which is an ecology term for, you

00:11:41   know, if you imagined, you know, sort of how many birds could live in a particular forest,

00:11:46   there's a limit to that where the resources of the forest can only support so many people

00:11:52   or so many birds. And in this case, in the context of subscriptions, it's this concept

00:11:56   of what was strange initially when you start a subscription business, the first, say, year,

00:12:02   which you will often see is this wonderful kind of growth in your revenue every month.

00:12:09   And it's delightful, I got to say. It just it feels super cool that the first, say you

00:12:14   had 100 new subscribers a day. That's your thing. So in the first month you're like,

00:12:22   "Wow, I have 100 subscribers." And then the next month, you know, I have, say, 190 subscribers.

00:12:27   Wow, that's 90. I'm increasing by 90. And then you go up again and again, and it feels

00:12:32   like your subscribers are always growing, and there's this wonderful growth period that

00:12:37   feels really awesome. And then what I was finding in my absence, specifically in Widgetsmith,

00:12:42   because it's now over two, it's like two and a quarter years old with a subscription.

00:12:47   Wow.

00:12:47   I'll get to it. Was it this? Yeah.

00:12:49   Time flies during a global pandemic.

00:12:51   It sure does. Yeah. But so it's like it's two and a quarter years old. And what I was running into,

00:12:59   and if you've been paying attention to Under the Radar for a while, I've been doing a lot of work

00:13:04   to try and increase my revenue from that app. I've been doing optimizations to my paywalls and

00:13:10   and changing things inside the app and moving features around and doing lots of work to

00:13:17   have it continue to grow. Because what I was finding is it was incredibly stable and anything

00:13:24   I did seemed to have almost no effect. I would make these changes, I would increase my conversion

00:13:31   rate and my paywall, increase my trial starts, and it was just not doing almost anything

00:13:37   to my overall, like, how much, you know, my MRR, my overall revenue that I could get,

00:13:41   and it was driving me crazy. Like, what is going on here? How am I, how is all this work

00:13:46   not actually translating? And so what I found, though, is that's like, okay, what I need

00:13:52   to do is I need to work out kind of modeling into the future what's going to happen if

00:13:58   I make improvements or changes to my app. And, you know, so I built the tool, and a

00:14:04   a little brief aside here. So I built this tool. It's this terrible little SwiftUI thing.

00:14:10   It works, but it's not great in any way. It was made purely for my own purposes. But because

00:14:15   you, under the radar listeners, are a special group of people, the world-exclusive URL to

00:14:20   download this tool, if you wanted to play with it, will be in the show notes for this

00:14:23   episode. This will be nowhere else. And if you want to try it and play with it, understand

00:14:27   you are doing so completely without warranty or any expectation of it working or being

00:14:32   accurate in any way, but because I like you and I want you to be able to play with it

00:14:36   while I'm having this conversation, it'll be in the show notes in case you do. But anyway,

00:14:41   so what I built is a tool –

00:14:42   It has a really great app icon, by the way.

00:14:44   It does. It's quite fun. Anyway, so I built this little tool, and what it does is it lets

00:14:50   me say, "Okay, how many subscribers do I have today? How much do I cost? What is my

00:14:56   price for my subscription? How many daily new trials do I have? How many of those trials

00:15:01   do I retain? And then what is the retention rate that I have in my first month, second

00:15:05   month, third month, fifth month, sixth month, and then sort of thereafter? And it'll draw

00:15:09   a graph showing what your subscriber base will be in the future. And the thing that

00:15:15   I discovered is that for a given set of those values, there is a ceiling, like a maximum

00:15:23   subscriber base that you can ever have, that you cannot grow beyond, which was very counter-intuitive

00:15:31   to me, that there is this sort of natural limit that you would think, well, it should

00:15:35   grow sort of indefinitely. But the reality, and this is the insight that wasn't intuitive

00:15:42   to me, but now that I see it, I can understand it, is sort of as you're going, every subscriber

00:15:48   you have will have a retention rate at some sort of applied to them, and it sort of continues

00:15:53   on in the future. And the longer you have a subscription in the market, at some point,

00:16:00   the weight and kind of the size of the existing subscriber base that you have. Even if you're

00:16:07   retaining most of them, like I think my long-term retention rate is like 95%, which feels pretty

00:16:13   good.

00:16:14   That's great, I think. I mean, that sounds amazing.

00:16:16   Yeah, I thought I had the same thing. It sounds great. But it meant that there is still a

00:16:19   limit where after two and a quarter years, retaining and losing 5% of those people starts

00:16:28   to dominate any amount of growth that I could get from new subscriptions. And you end up

00:16:34   with this ceiling. And you can see it in this tool. You put in your numbers, and it will

00:16:38   kind of show you where that limit is. And that was surprising. But it also was kind

00:16:45   of reassuring as I was going through this, because it meant that I wasn't losing my

00:16:50   mind and that all this effort I was doing wasn't actually making that much of an impact,

00:16:55   was, you know, the changes I was able to make were relatively small. You know, I could increase

00:17:00   my daily trials by 10% or 20% or something like that. And that feels like it should be

00:17:06   a meaningful change, like, wow, I'm getting 20%, you know, more people starting a trial

00:17:11   now than I used to. But the reality is, is like the weight of the existing people who

00:17:16   are subscribing, that is where, you know, that is what's determining my overall subscriber

00:17:21   base far more than any change that I could make in the short term. And so it's just

00:17:28   kind of an interesting difference. And I don't know if that would have been intuitive to

00:17:32   you, Marco, but it was definitely not intuitive to me that there would be a ceiling like this.

00:17:37   I mean, I've always seen my subscription revenue plateau off like this. I've never

00:17:44   really understood why it did that. I just kind of assumed, "Okay, that's just what

00:17:47   subscriptions naturally do.

00:17:49   And I mean, honestly, you see the same thing

00:17:52   with audiences, with things like how many people

00:17:54   listen to my podcasts, how many people subscribe to my blog,

00:17:59   or follow me on a social network.

00:18:02   I see the same pattern in all of those things,

00:18:05   which is when something first launches and is getting

00:18:08   traction or is first, for some reason, getting attention,

00:18:12   you see a nice upward movement.

00:18:15   and then it slowly curves off and seems to approach

00:18:20   a certain natural limit.

00:18:22   And then it just kind of stays there, ideally.

00:18:25   I mean, sometimes it goes down over time

00:18:26   if things aren't going super well.

00:18:28   But it just kind of stays in that kind of upper limit range

00:18:31   and never really gets to the next level,

00:18:35   the next order of magnitude,

00:18:36   or never doubles from that point or whatever.

00:18:39   And so it seems like this applies to pretty much anything

00:18:43   where you're kind of looking at ongoing attention

00:18:45   or ongoing money, where you do at some point, you know, you launch, you go up, and then

00:18:51   at some point the rate at which you're losing people or churn, you know, like the rate of

00:18:56   churn is at some point kind of keeps things in equilibrium with the rate of new people

00:19:03   coming in, and you just kind of, you have this kind of natural limit, or at least this

00:19:08   kind of range that you stay in, and it's very, very hard to ever break out of that unless

00:19:12   unless you have some kind of dramatic change in why people would come in in the first place.

00:19:18   And so this is how the total subscriber numbers or listener numbers or followers, this is

00:19:27   how it looks for everything I do.

00:19:29   Yeah. And I think that's, it's just interesting to me to confirm with math that that observed,

00:19:37   because I had the same observation, that it seems like you always end up in this steady

00:19:42   state and it's very hard to get out of.

00:19:44   And I think it's interesting to start to understand the math behind it.

00:19:49   And in some ways for me that was like a relief.

00:19:50   That was like, "Okay, so what it means is that for all I can, you know, I know the inputs

00:19:54   and I can determine what the output of those inputs would be.

00:19:58   So if I increase my daily trial starts by 10%, I can predict almost, you know, sort

00:20:03   of mathematically what the end result is going to be, what kind of a change that's going

00:20:08   to be, you know, is that where I should be putting my effort or should I be putting it

00:20:11   into retention, which one of those things is going to give me a bigger long term value.

00:20:17   And I think something here that's worth mentioning is the ACE, like my tool, which

00:20:22   you can, you know, I said, you're welcome to play with, will do this in a much more

00:20:26   detailed way where, you know, I let you break down this retention rates, you know, by the

00:20:30   first six months, and you can do input a bunch of different variables and try different models.

00:20:36   But a good equation, if you sort of want to left this conversation with a single equation

00:20:40   in your mind is that if you take the monthly subscription starts that you have and divide

00:20:48   that by your churn rate—so say you had 3,000 people start in a month, so you have about

00:20:54   100 a day, and your churn rate was 25 percent. You know, 3,000 divided by 25 will give you

00:21:01   whatever your maximum—that is your ceiling. So in this case, it'd be like 12,000 is

00:21:08   number of subscribers that you would expect to be able to have. And that's just math. Like that is

00:21:13   just that version of it is very straightforward. And like, what my model is doing is make, you

00:21:19   know, sort of letting you tweak rather than just like one bundled churn rate, it's letting you

00:21:24   apply different churn rates to different cohorts, you know, so this is something that I see very

00:21:28   much in my own app that you have very different retention rates for different months, like my

00:21:33   My month one retention rate is about 65%, but my long-term one was 95%, which sort of

00:21:39   makes sense intuitively that there's a group of people who start the subscription, start

00:21:45   using it.

00:21:46   It's not for them.

00:21:47   They cancel it.

00:21:48   But if you've done that, if you have renewed your subscription for 7, 10, in this case,

00:21:54   some people that have been renewing for 24 months in a row, it's pretty likely you're

00:21:58   going to keep doing it at that point because clearly it's become a part of your workflow.

00:22:02   something that you find value in, and so it becomes useful.

00:22:11   If you wanted to just do a basic number, just take your monthly subscription starts, divide

00:22:14   it by your monthly churn, that'll get you in the ballpark.

00:22:17   Or if you want to get more detailed than that, you need to actually do some modeling to it.

00:22:24   But I think there is something just interesting about understanding that.

00:22:28   And I think what it means, once you get a sense of that formula, you start to understand.

00:22:31   So if you wanted to say double your number of subscribers, you mean to either have to

00:22:37   double the monthly subscription starts.

00:22:43   If you think of the equation, if you wanted to double the other side of it, you have to

00:22:46   double the numerator or half the denominator, essentially.

00:22:50   So you either have to half your turn rate or double your subscriber starts, which is

00:22:54   a pretty meaningful change if you wanted to end up in that kind of a place, going from

00:23:00   100 daily trials to 200 daily trials is not going to be an easy thing to accomplish necessarily,

00:23:05   but that's the kind of thing that you would have to do in order to double the overall

00:23:10   subscriber count that you had at the end.

00:23:12   - Man, that's, I mean, it's good to actually have this be, like, you know, first of all,

00:23:19   codified and understood better, and it's good to have this tool to be able to plug in different

00:23:24   values easily and see it over time because I don't do a lot of very structured or calculated

00:23:32   growth hacking or revenue optimization or marketing really of any kind, to a fault I

00:23:41   would say.

00:23:43   And so whenever I do something or think something about like, "Hey, maybe I should do this

00:23:47   in my business," usually it's just kind of shooting from the hip and I just have

00:23:51   some kind of gut feeling or some kind of idea, "Oh, maybe I should try this." And it rarely

00:23:56   results in what I think will happen. Usually, my successes and failures, whatever they're

00:24:05   going to be, come as a surprise to me in both directions. And I think it's, again, it's

00:24:12   largely because I'm not really that interested in thinking in these analytical ways in a

00:24:19   lot of these topics. Like, it doesn't come naturally to me to think analytically about

00:24:22   business the way this does and the way you kind of need to if you're an indie sometimes.

00:24:28   And it's hard because some of these decisions are, I would say, difficult to – if you

00:24:34   have a – the standard indie mindset that I have most of the time, which is like, "I'm

00:24:40   just going to make a nice thing and the business will come." That – I mean, that does work

00:24:46   to a degree for a time sometimes, you know?

00:24:49   And so you can think that way and operate that way,

00:24:52   and in many cases, it'll work for you.

00:24:55   But in so many times where I've made some kind of small,

00:25:00   to me, tweak to the business model,

00:25:02   and then all of a sudden, my money goes way up or way down,

00:25:06   and it's like, oh man, I probably should be

00:25:10   putting more analysis and calculation

00:25:13   into some of these decisions.

00:25:15   So it's good to, first of all, it's good to have somebody like you in our space here

00:25:20   to be that person a lot of times when we aren't.

00:25:24   But it's also nice to have tools like this and concepts like this to know how to make

00:25:31   more analytical decisions and what to expect and kind of what's possible.

00:25:35   And so if I'm looking at this and I'm saying, "Hey, if I want to meaningfully get this

00:25:40   curve up or raise the ceiling here, I'm gonna have to get more people in the door in the

00:25:45   trial stage or in the download stage or whatever it ends up being. And that's difficult. That

00:25:54   might require paying for more search ads or paying for marketing somewhere. And that can

00:25:59   be very expensive and then I have to price that in. And it becomes so much more complicated.

00:26:05   But this is a way that people like us

00:26:08   need to think sometimes.

00:26:10   We need to be able to make these decisions

00:26:11   and be able to look at our businesses

00:26:13   and be able to make educated decisions

00:26:17   and have concrete goals and concrete expectations

00:26:21   that are somewhere in the ballpark of reality.

00:26:24   And so this really helps a lot to look at this

00:26:27   and to think like this.

00:26:28   - Yeah, and I think, I'm glad to hear it's helpful to you,

00:26:31   but it was something, I think for me,

00:26:33   I think what you're saying,

00:26:34   word that is most useful is making it an informed decision. Because I feel like so often, what

00:26:40   I end up doing is I take like the first two months, say of two or three months of data,

00:26:47   and then just like linearly extrapolate that data. And I'm like, wow, this is going to

00:26:51   be amazing. And I can start to like, you know, really invest in something or start to really

00:26:55   kind of go after something expecting it to just like grow indefinitely or kind of viewing

00:27:00   it in that way. And that's useful in some ways. But really what it's more helpful is,

00:27:05   in this case, it's like, understand, okay, given the first couple of months, or given

00:27:09   what kind of industry averages are, where could I expect at a given level? And what

00:27:17   does that mean? Especially, one of the most, sort of my favorite parts of being an indie

00:27:23   is that people will tell me and talk to me about when they go independent. They make

00:27:29   an app, they put it in the store, they think it's going well, and they decide, you know,

00:27:33   I'm going to quit my day job, and I'm going to go full time. And that's awesome. Like,

00:27:36   I love those stories. Anybody who's listening to this, if you ever go indie, let me know.

00:27:40   I love it. It's just like my greatest joy. But I also want to make sure that when you're

00:27:45   going when you're stepping out on the limb, if you're basing that on numbers, that those

00:27:49   numbers are firm, and realistic and reliable and predictable. And you're not doing what

00:27:54   I do and linearly extrapolate like two or three months and then expecting that to continue

00:27:58   to grow and kind of be good into the future. So my main thing here is just be thoughtful

00:28:04   about it, be understanding that there is some math behind it, there's some things going

00:28:08   on here that if you understand can be really helpful to you, can be really powerful as

00:28:13   a way to predict the future essentially. And then the great thing about subscriptions is

00:28:18   that they are kind of predictable. That you also can in the same way that if I say that

00:28:23   I have, you know, whatever, 1000 subscribers, and my daily trials goes to zero. Like you

00:28:28   can predict at your current churn rates, like what will happen and how much revenue you'll

00:28:33   have over a given period of time, which is lovely. Like it's one nice thing about subscriptions

00:28:38   is that they have this, you know, they're both resistant on the up, you know, that if

00:28:42   I make a change and make a 10% increase in the number of daily trials, it's kind of it

00:28:47   has a muted impact on my overall subscriber count. But the great thing is it goes the

00:28:51   the other way. If I have a 10% reduction, there's also kind of a muted, immediate

00:28:56   impact on that. And so that's why we love them so much, and it's why I think I will

00:29:00   continue to use them as my primary monetization scheme going forward.

00:29:04   Yeah. Frankly, I love subscriptions. It matches what we actually need from the business, as

00:29:11   you were saying at the beginning of the episode. Subscriptions give you recurring revenue so

00:29:16   that you can keep developing the app on an ongoing basis, which is what everybody expects

00:29:20   from apps these days anyway. So it's a great business to be in and it's really nice to

00:29:25   be able to understand it better.

00:29:26   And that's probably a good place to conclude subscription school with Underscore today.

00:29:30   And I think it's probably one thing I just wanted to quickly mention before we wrap up,

00:29:33   is that I think Marco and I now are almost exclusively on Mastodon, so we'll have links

00:29:37   to our Mastodon accounts there. So if you wanted to follow up or understand kind of

00:29:42   what we're going on, what we're working on in between the two weeks that we're here,

00:29:45   that's the place to do it, rather than other places that are no longer.

00:29:50   Yeah, there's no almost about it. I'm exclusively there.

00:29:53   Yeah.

00:29:54   All right. Thanks for listening, everybody, and we'll talk to you in two weeks.

00:29:57   Bye.

00:29:58   [ Silence ]