Under the Radar

71: Apps as Annuities


00:00:00   Welcome to Under the Radar, a show about independent iOS app development. I'm Marco Armet.

00:00:05   And I'm David Smith. Under the Radar is never longer than 30 minutes, so let's get started.

00:00:10   So today we wanted to start what is largely going to be a kind of a two-part series about,

00:00:16   I guess, more of the financial business side of being an independent app maker. And specifically

00:00:25   we're going to be talking through viewing your apps in two different lenses. One we're

00:00:34   going to talk about today is kind of viewing your apps as though you're creating annuities

00:00:39   that pay you back over time. And then next week we're going to be talking a little bit

00:00:44   about viewing your apps as sort of assets that you're building up for the purposes of

00:00:49   selling down the road. I think this is one of those topics that is very much outside

00:00:57   of the typical developer-y kind of topics and the kind of things that we would normally

00:01:03   discuss in the sense that we're not talking about Xcode, we're not talking about design.

00:01:07   There's going to be a WDC session about how to sell your business.

00:01:11   Yeah. This is a topic that is very much, this is like a business topic. We could be making

00:01:17   anything. The fact that it's an application is in some ways kind of immaterial. But I

00:01:21   know for myself, when I started off doing this, I gave no thought to these types of

00:01:29   topics at all. I was focused entirely on the development side of things. I was focused

00:01:33   on the design side of things and the marketing to some degree. But thinking of these kind

00:01:38   of broader business things wasn't really something that was on my radar at all. And now, whatever,

00:01:44   eight and a half years later, it is something that I think about more and I realize the

00:01:48   importance of thinking through, mostly just because depending on what you want your business

00:01:55   to look like, how you want it to pay you both now and in the future, what your ultimate

00:02:02   goals are, will impact the types of apps that make sense for you to build.

00:02:08   Different apps have very different revenue curves. If you make a paid up front app, you

00:02:14   are far more likely to have a big burst of revenue at once and then it will fall off

00:02:19   fairly dramatically. If you have free with ad or free with in-app purchase application,

00:02:26   you're more likely to have this slow, hopefully growth of revenue that is fairly stable over

00:02:33   time or at least shifts in these kind of more seasonal waves. Depending on what you like

00:02:43   there will make a big difference.

00:02:44   For the purpose of today, what I'm going to talk about to start off with, and this got

00:02:48   started for me by, I've been doing a lot of work updating my app audiobooks, which I think

00:02:53   I've mentioned a few times over the last few episodes. As part of this update is the realization

00:03:01   that I haven't done any work on this app, Realist, really, for about 18 months, which

00:03:07   is not a good idea, isn't great. It is because other opportunities have presented themselves.

00:03:12   I've been doing much more work on the health and fitness side of things recently in this

00:03:15   app that for a while was like my bread and butter with the thing that was paying my mortgage

00:03:20   each month. It kind of had fallen a bit to the side. But as a result, I got this opportunity

00:03:26   to have some really interesting data about, of course I did, I made a spreadsheet. I took

00:03:32   a look at just one, two or three, it's fine. I don't have a problem, I can stop anytime.

00:03:41   I took a look at the revenue of this app over the last 18 months because one of the things

00:03:46   that I've always been curious about is what happens to an app when you stop working on

00:03:51   it, when you stop maintaining it, when you stop marketing it, when you stop doing anything

00:03:54   with it. What does that actually look like? Because in the back of my mind, I've always

00:03:59   had this thought of sales won't immediately drop to zero. That's just not the reality

00:04:06   because I go for weeks or months between updates typically. You usually don't let them go out

00:04:12   to 18 months, but you kind of get the sense that over time, sales will likely fall. If

00:04:18   they don't, stop working on it. If doing nothing makes your revenue go up, do nothing, that's

00:04:23   awesome, but that's not realistic. What I did for audiobooks is I took a look over the

00:04:28   last 18 months and I looked at what the revenue curve looked like for this app with no maintenance,

00:04:35   no anything, it's just been sitting on the app store. The app makes its money from advertising

00:04:39   and to a very small degree, some in-app purchases for some kind of premium books. What was interesting

00:04:46   to me is that the curve that the revenue took matched pretty closely to kind of an exponential

00:04:52   decay curve, like I said, around with some best fit lines and things. Essentially, what

00:04:57   it meant is the sales went down by half a percent, I think it was, per day, sort of

00:05:07   compounding. Say for argument's sake, the app had been selling for $100 a day and then

00:05:16   I stopped working on it. The next day, it would lose 50 cents in revenue and the next

00:05:19   day 49.8 cents and then 49.5 cents and so on. It loses less each day because there's

00:05:28   less from it. That curve of about a half a percent a day seemed to match fairly closely

00:05:35   to it, which meant that obviously as it falls down, you start losing less and less per day

00:05:42   and it almost starts to stabilize and level out. Obviously, it's still losing as it

00:05:48   goes but it gets slower. That was really interesting to me because that number and that multiplier

00:05:55   is I'm sure very variable based on a lot of things, but it's instructive to think

00:05:58   about how an app's revenue will almost certainly have that kind of a decay curve. The slope

00:06:08   of it and the steepness of it will change, but that's a really helpful tool for me

00:06:12   as I think about apps because now I have some data to back up this notion that I've always

00:06:19   had that in many ways, the way business that I like to build is I want to have a steady,

00:06:26   even income from my applications that in an ideal world requires a relatively small amount

00:06:33   of constant maintenance and input. I'm coming from a background where I used to do consulting

00:06:37   where all of my income was prepaid in the sense that I have a contract with someone

00:06:44   that says I'm going to work for them for 80 hours and they're going to pay me $100

00:06:51   an hour, so they're going to pay me $8,000 and that's the contract and that's the

00:06:55   most I'll ever make from that project and it's all prepaid. And shifting to a world

00:07:00   where everything's postpaid, where I do all this work for nothing essentially, I release

00:07:07   the app and then hopefully it pays me back over time. And in many ways, that's the

00:07:11   same kind of thing that in the financial world which you'd call something like an annuity,

00:07:16   which is where you take a large sum of money and you give it to a bank or an investment

00:07:21   institution and they pay you an amount of money back each month that you're shifting

00:07:29   the risk for them generating that money onto them and you're getting a steady stream

00:07:33   of income. And that model for what an application is, I find is really kind of nice because

00:07:39   you get the sense of my goal is to build up a broad diversified pool of these little income

00:07:45   streams that each individual one is kind of slowly decaying over time and my job is to

00:07:52   I can increase or reduce the degree to which they're decaying or in some case hopefully

00:07:59   grow them by putting in work, but if I don't, if I want to go on vacation or if I get distracted

00:08:05   or things happen, my income will very slowly decay over time, but in general we'll keep

00:08:11   going and we'll have a pretty long runway. And I think this was actually before Under

00:08:18   the Radar, but back when I remember talking about it on my old show, Developing Perspective,

00:08:22   where there was a period of time a couple years ago where I had to take for a variety

00:08:27   of personal reasons, I had to take about six or so weeks off of work basically, and it

00:08:34   was lovely that during the course of that my income didn't fall off dramatically.

00:08:39   It wasn't like all of a sudden if I'd been a consultant that would have been a really

00:08:42   big problem to suddenly just not have any income. And so instead my income fell and

00:08:47   it kind of slowly decayed, but I still had this kind of annuity that was paying me back.

00:08:54   And so I love that, and so what I do now is I think I structure my business around that.

00:08:58   That's why I have so many apps, that's why I have so many of these little streams

00:09:01   coming in in the hopes that kind of building that layer cake that will average out to something

00:09:07   that is somewhat dependable and somewhat reliable.

00:09:09   I love the idea that you're living off of background noise, like the asymptotic tails

00:09:17   of all these curves and you just pile enough of them up and here I have a background radiation

00:09:23   income.

00:09:24   That's kind of wonderful.

00:09:28   And that's the thing that is kind of nice about it though, is that it is so often in

00:09:34   software development we kind of get the thing that you hear most often about is the big

00:09:40   impressive first month of the sales of an app or a new update or anything. It's not

00:09:51   easy, but it is more easy to focus on that big exciting part of it, where you do a big

00:09:58   update and everyone loves it and the sales go up and everything's exciting.

00:10:03   But what's more interesting from a lifestyle perspective is what happens six months down

00:10:10   the road with that and is there a way to make your living on that noise, on that slow easy

00:10:18   part down below where you're not having to put in the effort that you did up front.

00:10:27   You need either enough apps for them to decay at a slow enough rate that it still sums up

00:10:36   to a reasonable living, but if you can, or at the very least if you can get close to

00:10:42   that and be hoping that you have some big bursts that build up your reserves and then

00:10:48   you have this noise in the middle that you can live on. It is an approach that seems

00:10:53   to have worked a couple times for me and is something that I think the math checks out.

00:11:00   And that's why I was kind of really excited to have the opportunity to sit down and really

00:11:03   dive into it with audiobooks because now I can kind of prove to myself that the experience

00:11:09   I've been having of this seems to be working for the last couple of years, the math checks

00:11:14   out and it kind of follows a pattern that is somewhat predictable or at least reasonable

00:11:21   to base a livelihood on.

00:11:22   Yeah, it's funny. In other areas of web-based work, things like blogs, the idea of building

00:11:31   up stuff up front that then has a very long tail of revenue over time is not a new idea.

00:11:38   In the blogging world, you have straight up blog ads and things like affiliate links where

00:11:44   you can write a blog post and then it might get a certain number of views from the main

00:11:50   subscribers and active audience that day or that week, but then over the next 10 years

00:11:56   it could be getting search traffic and then you can be slowly building up more and more

00:12:01   ad money over time from this big back catalog that you have. And for me, on my side, I never

00:12:08   made a ton from the ads, but I actually started in the last few years making a pretty good

00:12:13   amount of Amazon affiliate revenue because I have, over years, I have written a pretty

00:12:19   good number of product reviews, things like headphone stuff, coffee stuff, the baby stuff

00:12:24   guides and those have slowly built up search relevance over time.

00:12:31   And at first when I started out with Amazon links forever ago, like Amazon affiliate links,

00:12:37   I would make effectively nothing. It was like maybe like 20 bucks a month for a while and

00:12:42   I stopped looking for a long time because it was so low I kind of stopped paying attention

00:12:47   and then eventually it became real money. Like it became, after a few years of doing

00:12:52   it and slowly building up from like 10, 20 bucks a month, it eventually became a few

00:12:57   hundred dollars a month and then now it's like over a thousand dollars a month and that's

00:13:02   all from stuff I wrote years and years ago for the most part, not a lot of new stuff

00:13:07   yet. Just over time, it's kind of like where each post is one of these little like decaying

00:13:12   asymptotes like over time I have enough of these and it does build up to something substantial.

00:13:18   And it's interesting to think of apps this way because I don't think most of us think

00:13:23   of apps that way. I think most of us, the idea of doing blog posts and links that way

00:13:28   I think we are accustomed to by now, but the idea of doing apps that way almost seems like

00:13:34   wrong or like you're getting away with something because I think people expect that apps need

00:13:39   to be the more high profile launch kind of things and the, you know, have to be constantly

00:13:44   maintained over time and have to constantly be getting updates and new versions and keeping

00:13:49   everything fresh and new while it's totally okay to let a blog post sit there for 10 years

00:13:52   unmodified. I wonder like is this kind of a temporary thing? Is this always going to

00:13:59   be the case? I mean what the causes might be of why apps can behave this way or why

00:14:04   they might not behave this way. Like we have this coming fall it's looking increasingly

00:14:09   likely that Apple's cutting off support for old 32-bit apps and that's going to cut off

00:14:15   a lot of people's long tails and let's take a look at them. But I wonder like it seems

00:14:20   like it seems like this is probably okay for the long haul overall as a model. Like it

00:14:25   seems like this is probably going to keep working because it turns out that keeping

00:14:30   apps up to date for the most part really isn't that important for whether or not they will

00:14:36   have, you know, a long tail of success.

00:14:39   Yeah and I feel, I think it speaks to, I think there is something there like it's why the

00:14:47   reason why I thought it would be a good topic is because I feel like it also is important

00:14:50   to think through the type of applications where an approach like this would actually

00:14:55   work because the reason this approach works for me is because the types of apps that I

00:15:03   am making by and large don't have large ongoing costs and expenses and infrastructures that

00:15:12   I have to have in the background to support them. You know like the, I think all the sum

00:15:19   total of the infrastructure I have for audiobooks is I think it's about a hundred dollars a

00:15:27   year at Linode. Like I have a little server that serves thumbnails and a few other things

00:15:33   but it's essentially nothing in the scope of the app. And so that allows the fact that

00:15:40   the app isn't, as its revenue decreases slowly over time, its income, the expenses that I

00:15:50   am having to put out for it, they become a higher proportion perhaps of its income but

00:15:54   they are still so small to start with that it doesn't really matter. Whereas if you had

00:15:58   an app like, so for example for me, Feed Wrangler and I imagine for you with Overcast to some

00:16:03   degree, like if you have an app that has a bigger server infrastructure that is going

00:16:07   to be a thousand or two thousand dollars a month in server costs, it starts to become

00:16:13   trickier to do this type of thing where that slow, that sort of slow decay will very quickly

00:16:21   cross over and then be losing money rather than gaining money which is essentially what

00:16:25   happened for me with Check the Weather. Where I had a weather app which I hadn't done updates

00:16:31   for in a while too but I had to eventually kill it off because I was losing money on

00:16:35   it because there was this ongoing cost for weather data providers that wasn't being

00:16:41   offset by it. And so I think with software, the key for this to work is the more that

00:16:47   you can make the applications that you're building kind of built in such a way that

00:16:52   they are more like blog posts or they are more like YouTube videos. There's lots of

00:16:56   things that you, like you said, this isn't a new concept at all. This is the business

00:17:01   model of most kind of media companies where a media company makes a show, like they make

00:17:09   Seinfeld and they put it in syndication in, or they put it out on broadcast TV and they

00:17:15   put it in syndication and they sell DVDs and it's exactly the same thing that they're

00:17:19   just kind of reselling over time. And the degree to which an app can be that I think

00:17:24   is it's something that is fairly self-contained, that is built in this way that doesn't require

00:17:29   a lot of constant updates and maintenance. And fair enough, eventually you'll have

00:17:35   to either update it or it will stop working. I think software has a higher barrier for

00:17:42   that versus just a plain text file that you're putting on the web, essentially. But it doesn't

00:17:52   have this intrinsic need that it has to be updated every few months. And in some ways

00:17:56   that's counterproductive, that we've talked about many times, the trap of over-featuring

00:18:01   an application and having the sense that there's this urgency that it always has to be new

00:18:05   and better when in fact maybe it isn't. And you can just kind of let it go. And as long

00:18:10   as you have a customer base that likes the app that you shipped whenever you shipped

00:18:15   it, people still use audiobooks that I shipped 18 months ago and they use it on a regular

00:18:21   basis and it still works. And the updates that I'm doing now are in some ways unnecessary.

00:18:28   To it, I think it's a good thing in terms of trying to bump up that rate a little bit

00:18:36   and sort of start decaying from a higher point again. But if I didn't do anything, my guess

00:18:41   is they would just keep going slowly until at some point it breaks. And who knows, maybe

00:18:47   iOS 11 would introduce something that would force it to be updated, which, fair enough,

00:18:51   that's the reason why it was updated 18 months ago was because that was what, iOS 9 broke

00:18:58   something in it and so I had to update it and I did and I fixed it and then I haven't

00:19:02   done anything since. And those kinds of maintenance updates are so much easier and so much lighter

00:19:07   weight than the big broad things that you'd have to do if you have a perspective of you're

00:19:12   trying to grow, grow, grow and that's your goal.

00:19:15   - Yeah, and it really does kind of impact how you develop everything. If you're thinking

00:19:21   about, how can I design this app and build this app in such a way that it will require

00:19:27   the least maintenance over time and be the least likely to unexpectedly break in future

00:19:33   OS updates or with future changes to the landscape, that affects things. Like for instance, running

00:19:39   off of a Linode server, which actually they were sponsored this week, which I'll get to

00:19:42   in a minute, running off of a Linode server, that is to me more future proof than running

00:19:49   off of a service like Parse or some kind of like managed high end or higher abstraction

00:19:54   layer service because that landscape of those services changes all the time and if you write

00:19:59   an app today on like a higher level abstracted service, in three years you might have moved

00:20:06   on and you're just leaving this app in maintenance mode and hoping it just collects money over

00:20:10   time but then that service might shut down or do some kind of forced change that will

00:20:15   break your app and then you're stuck having to decide either well do I just shut this

00:20:19   app down completely, is it just done now because it's not worth the investment to move it to

00:20:23   something else or do I now lose a big chunk of this investment to move it over. Like do

00:20:31   I invest a bunch of time in it and lose money on it for a little while to do this move and

00:20:34   like that's, if you build from the start to try to build on like boring stable stuff,

00:20:41   both on the server side like just doing Linux servers and also even in the app like if you

00:20:45   just use mostly like vanilla UI kit and you don't do a lot of like fancy hacks and things

00:20:51   that might break in future OS updates, that also will help you just make sure that you

00:20:56   can keep that going for a long time. Anyway we are sponsored this week as mentioned by

00:21:01   Linode. Linode gives you fast powerful web hosting for your projects. You can set up

00:21:06   in just seconds and leave running for years if you want to. Linode plans start at just

00:21:11   five dollars a month. They get to a server with one gig of RAM in the Linode cloud and

00:21:16   you can get a lot more after that if you want to. Five bucks a month for a gig of RAM. To

00:21:21   give you some idea, my first VPS was fifty dollars a month for a hundred and twenty-eight

00:21:26   megs of RAM and that wasn't that long ago so this is pretty great. Anyway Linode has

00:21:32   over four hundred thousand customers who are all serviced by their friendly twenty-four

00:21:36   seven support team. You can email them, call them or even chat over IRC in the Linode community.

00:21:41   If you need help Linode is there for you in whatever way you want. They have comprehensive

00:21:45   guides and tons of support documentation to teach you everything you need to know about

00:21:49   setting up and managing your own Linux virtual server. Their control panel is beautifully

00:21:53   designed with a focus on ease and simplicity. You can deploy, boot, resize and clone servers

00:21:58   in just a few clicks. They have two-factor authentication to keep you safe and you can

00:22:03   also create snapshot images of your disk volumes for backup and replication. Linode is the

00:22:07   full package for your server infrastructure needs. They have the power you need as well

00:22:12   as the infrastructure and assistance that you want. So check it out today. You can get

00:22:15   one gig of RAM for just five bucks a month. You can go all the way up with lots of things

00:22:19   in between to sixteen gigs of RAM for just sixty bucks a month and of course there's

00:22:23   even higher ones after that if you want to. Their two gig plan now includes thirty gigs

00:22:27   of storage all for just ten dollars a month. There's tons of great options at Linode at

00:22:31   pretty much every price point. These servers are great and they're very powerful for the

00:22:34   money. As a listener of this show if you sign up at linode.com/radar you'll not only be

00:22:39   supporting us but you'll also get twenty dollars towards any Linode plan and with a seven day

00:22:43   money back guarantee there's nothing to lose. So go to linode.com/radar. To learn more sign

00:22:50   up and take advantage of that twenty dollar credit or use promo code radar 2017 at checkout.

00:22:55   Thank you so much to Linode for supporting this show and for keeping our apps running

00:22:58   for years into the future without much effort from us.

00:23:01   Yup, years and years and years. So the last thing that I think I would, it seemed like

00:23:07   I would be good to bookend this discussion with would also just be to say that this approach

00:23:13   of thinking about an application as an annuity that you kind of build and then can kind of

00:23:17   leave on its own for the most part or do lightly maintained potentially is probably not for

00:23:23   everybody. And I was thinking about it a little bit. There's two groups that I feel like this

00:23:27   approach really wouldn't work for. And that is people who are perfectionists and people

00:23:34   who kind of have the like a growth mindset in the sense of you kind of have that vision

00:23:41   of the venture started kind of company that's trying to be worth a billion dollars and has

00:23:49   that kind of runaway hockey stick magic fanciness. I don't know. Like those two groups of people

00:23:55   I feel like would not do well with this kind of approach because by its nature it is optimizing

00:24:01   for minimizing input and rather than trying to maximize output which works well for me.

00:24:10   I love that I can sort of, once you get a business to this, you put in years, it takes

00:24:18   years to get there, but if you put in years of work with this kind of mentality you can

00:24:22   get to a point where your core business requires relatively little input and gives you a steady,

00:24:30   reliable, slightly decaying over time but can be bolstered periodically by putting inputs

00:24:37   into the system rather than something that is growing month after month, year after year.

00:24:43   It's like, "Oh, we just need to be doubling every six months," or whatever. This is not

00:24:47   an approach or a mindset for an application that will lead to that kind of a result. And

00:24:54   I think in our next episode, like I said at the beginning of this one, if you're looking

00:24:57   to sell an application or have a big exit or something, you might structure your application

00:25:02   and your approach differently than this. But it can work for some people or it wouldn't

00:25:07   work for others. And I think also if you're a perfectionist, if you're someone who just

00:25:09   wants to keep tinkering and polishing over and over and over again and keep working on

00:25:15   something until it's almost like the asymptote that you're approaching is perfection rather

00:25:22   than looking at it the other way around of hitting good enough and then just letting

00:25:27   it slide into the future, that's probably not for you either.

00:25:32   - Yeah, I mean, in many ways it's kind of like the culmination of everything we've been

00:25:39   trying to teach the audience for this entire podcast, this entire run of this show so far,

00:25:45   is like when you're an indie, if you want to optimize your limited resources, you have

00:25:52   no other staff or you have very little other staff, you don't have a huge amount of cash

00:25:57   with VC money to build up stuff at a large scale very quickly up front. And so in order

00:26:04   to optimize your limited resources and especially mostly your limited time, it does help to

00:26:10   have these qualities to really not be a perfectionist and to be willing to ship something that is

00:26:17   good enough and then to decide afterwards, you know what, now that's kind of done or

00:26:24   at least seems to be done for a while and now I can move on and do something else or

00:26:29   I can move on to feature expansion to expand into new markets instead of doing things like

00:26:36   perfectly sending down every single individual feature to be totally perfect. In a lot of

00:26:42   ways the whole point of the show is to tell people if you want to run a business the way

00:26:48   we do and especially the way you do, then this is how you do it. It is possible to make

00:26:55   a living in the app store but if you do it in these other ways that might be more high

00:27:01   profile it's harder and it's less likely to succeed. Whereas if you do it this way

00:27:06   where you're really just building up a whole bunch of asymptotes of gradual income from

00:27:13   a larger quantity of apps that are not necessarily all perfect but you can get them all to be

00:27:19   good enough and to be long term businesses, that is actually a way more likely approach

00:27:24   to succeed in the app store than hoping that you are going to be the new tank game that

00:27:31   makes a million dollars in a day.

00:27:34   Yeah, and I think the biggest adjective that comes to mind in this approach is tolerance

00:27:41   and it's being okay with letting things go and being like this is fine, I think this

00:27:49   is good enough and being honest with yourself about that and having that as your goal. I

00:27:55   know that's not the approach that everyone feels comfortable with or will take and I

00:28:00   think especially the Apple developer community, at least early on and probably to some degree,

00:28:07   has that bit of the mentality that they took from Apple and the way that they build things,

00:28:16   which is not this at all. This is in no way the approach that Apple builds most of their

00:28:21   products that they take the other extreme of the highly engineered, highly polished

00:28:28   approach to something that's a premium product and that works and that model, the premium

00:28:34   product market works clearly. It works for Apple, it works for Tiffany's, you can think

00:28:38   of any number of car manufacturers that take that approach and that works but I think it

00:28:45   struggles when you're a small company, when you're one person, to be viable. If your goal

00:28:53   is to win the lottery or to have that premium approach, so many things have to go right

00:29:00   for you that are outside of your control but if your goal is more modest and your goal

00:29:05   is to say, it's much less scary to say my goal is to make an app that makes $50 a day.

00:29:17   That is much less scary than trying to make an app that makes tens of thousands of dollars

00:29:23   a day and could potentially be purchased for millions of dollars. Those are big scary things.

00:29:29   Making it much smaller, much more modest, I think is a much more approachable and reasonable

00:29:33   thing for small developers to take and I can speak from experience that over time, it works.

00:29:40   Thanks for listening everybody. Tune in next week when we talk about selling your apps

00:29:44   instead of keeping them forever. Thanks and we'll talk to you next week. Bye.

00:29:48   [BLANK_AUDIO]