Why Marco Arment Built A Podcast App

Developer Marco Arment isn’t exactly a household name, but that could depend on your household. As a frequent podcaster and writer, he’s taken on a somewhat polarizing role as a commentator on the Apple developer ecosystem. As a developer, he was Tumblr’s first employee and went on to create the still popular Instapaper (now owned by Betaworks) and iPad periodical The Magazine (now owned and operated by Glenn Fleishman).

After Tumblr’s sale to Yahoo, Arment went on to shed his other holdings and become an ironically-self-labeled ‘analyst’. His posts on Marco.org are frequently upvoted on Hacker News, where he is then praised and/or vilified based on the current developer meta mentality.

But, over the past few months he’s been working on an app called Overcast, and it launched late last week. The app is a podcast player — which could sound like an odd niche to tackle — until you realize that read later apps, iPad magazines and (at some point) web logs were all niches too.

I’ve been testing Overcast for a couple of weeks (and enjoying it) and I thought it would be interesting to reach out to Marco to ask the obvious questions and a few more.

What follows is a lightly edited transcript of our chat. It gets a bit into the weeds on the developer side of things, but my feeling is that if you’re reading this you won’t mind that terribly. Enjoy.

Matthew: The easiest question to start with is why a podcast app?

Marco: The short version is just that I love podcasts. I’ve always loved podcasts. Ever since iTunes added them in 2005, I’ve been listening.

I’ve actually been listening to podcasts since 2005, and it was always harder back then. It got a lot easier over time, and smartphones of course made it easier, so now there’s this great market. You have the combination of smartphones, and Bluetooth audio in cars, which helps a lot.

I wanted to support the podcast medium, and from a selfish perspective, I don’t use a whole lot of apps on my phone. I’m not one of those people who has all the SpringBoard pages full. I have two pages of apps, and one of them has folders.

I don’t have a lot of apps and I don’t use a lot regularly. My most commonly launched app is probably either Tweetbot or Mail. I’m not using tons of stuff. My podcast client is one of my most frequently used apps, so I was sitting there using other podcast clients all day, every day, and I’m like, “I would love to do it this way instead,” or, “I would love to add this missing feature.”

It was just that developer itch that was — not only is this a market that I love, and a medium that I love, but I’m sitting here using someone else’s app every day that I don’t like very much and I want to do it my way, so that’s basically the reason.

Matthew: It’s like a machinist. He’s using a tool, and he doesn’t really like it, but he can build himself a new tool.

Marco: Right.

Matthew: Obviously there’s no shortage of podcast apps, right?

Marco: No, not even close.

Matthew: It seems like a market that’s really, really deep with stuff, so what were the things that you saw that really flipped that switch for you, from, “I like podcasts. I listen to them. I open this app all the time. I really can’t find these things and that’s what I want. That’s why I want to build it”?

Marco: The biggest two were…I wrote about this…you’ll see on my site, but the biggest two were, by far, when I had ideas for what are now Smart Speed and Voice Boost. I first had the idea for Smart Speed — “What if we could do this dynamic shifting of the speed based on silence, and speed up the silence more than the rest of it,” because I wanted to pick up extra speed, but I didn’t want to distort the sound too badly.

Just as a listener, I hear…when you listen to sped-up podcasts, it’s a waste. You hear all the silences that are longer than they have to be, and then someone has to say their words super fast, and it distorts them, so I said, “Let me try to even this out a little bit.” I wanted to play with that, and play with different time-stretching algorithms also, because at the time I first started doing this in the fall of 2012 — I don’t think Apple shipped any fancy speed algorithms on the iPhone.

They had the one you could use that had the preset notches, the preset speeds. It was the old iPod algorithm. It was even called iPod Time in the API. It’s very old. It’s fine. It doesn’t use a lot of CPU power, it saves the battery, but it doesn’t sound very good and only has those preset notches.

I started playing with different libraries, different even third-party commercial libraries to time shift and time bend, and I ended up using some of the Apple’s newer APIs using some new settings they had launched.

In fact, if you look, Castro did an update a couple weeks ago — a couple months ago at this point — where they improved the voice quality of their speed-up algorithm, and what that was was choosing one of these newer APIs.

There’s a low-level API you can set to say, “All right, I want the time stretching algorithm to be this,” and it’s a trade-off between CPU time and quality, so they changed that option. At my level that I’m working on, I have the same option that I can take, so I made the same choice there.

Anyway, what made me want to do it was a combination of just wanting to do it my own way and then also coming up with these couple of features that I thought were pretty good, and that no one else was doing. I also knew the voice boost and the higher speed algorithm, you can do that with AVPlayer, which is the API all these other things are using. Every other app that I know about is using AVPlayer.

You can’t do smart speed with it though, and I knew that if I did it, that would give me a competitive advantage for a while. It’s probably not going to last forever, but certainly a while, because if I’m willing to go through the technical hurdles to write my own audio engine, and use Core Audio down to the raw levels, and bypass the easy but limiting frameworks that everyone else was using, I knew that would be unique, at least temporarily. We’ll see how long it lasts.

Matthew: Like all good features, if you execute them well, then they’re just going to pop up everywhere. Just ask anyone who creates a category-defining or element-defining piece of design, or code, or whatever.

Marco: Exactly. The difference here is that smart speed, in the best case scenario, if every other podcast app drops everything they’re working on now and starts trying to do this, it’s probably a six-month project, so at least there’s that going for me.

Matthew: You’ve got six months of feature uniqueness going on.

Marco: Right, which I have to spend catching up with all their features.

Matthew: It’s interesting because you’re launching a podcast app at a time where you have been “pre-Sherlocked” by Apple. It seems like this is one of the first times that I’ve seen that somebody’s launching an app that has been ‘pre-killed’, in theory. Podcast is built into iOS as of iOS 8.

Marco: If you look at, whether it’s built-in or not, people are making a big deal of this because it’s a change that is happening, but if it was always built-in, which for the first x years of iPhones it was, they didn’t split it out into its own app until [iOS 5].

For all the time before that, there were podcast apps, and they were fine, and if you look now, there are tons of apps that have been always built-in. Notes. Weather. Look at all these things, and they’re massive app categories in the store.

Matthew: What do you mean, Marco? Nobody makes notes apps.

Marco: And no one makes weather apps.

Marco: You have these massive categories in the store that are technically Sherlocked in the most complete, obvious way, which is that every iPhone has always come with a free app, often on the first screen, to do that, often with capabilities that third-party apps couldn’t do. Different integrations; different features; integration with Siri, stuff that third-parties can’t do.

This is not a new situation, and the fact is you can look all around the market and you can see it’s really no big deal.

Matthew: It seems like there’s definitely a case to be made that a podcast app on the home screen will expose many people to the idea of podcasts that have never been exposed to them before. It was on the iPhone, but you would be forgiven for not knowing about them because it was a section of the iTunes app.

Then it was a freely downloadable but not pre-installed app from Apple, but now it’s been elevated to first-party status.

Apple is saying, “We are as interested in podcasts as we are in weather or your calendar,” so it seems like there’s a case to be made that they are elevating the conversation around podcasts.

Marco: I hope so. The fact is I’m not sure they are, but I hope they are. It could just be this is easier for them. It could just be that…and maybe they were losing too much control by not having their own podcast app out be there. Maybe it had to be there for some other reason. Some kind of integration with iTunes or iCloud where it was causing issues with sync to have it be a separate thing.

Who knows? It could be any of those reasons. Either way, again, I just don’t think it’s going to affect third-party clients at all. I saw with Instapaper when they added Reading Lists and everyone’s like, “Oh, this is going to kill Instapaper.”

At first it was — “Well, this isn’t really a full competitor yet because it doesn’t have sync or it doesn’t have text view or it doesn’t have offline support.” Over the next two releases, they added all those things. Then obviously it was a direct competitor, and it had all these advantages of things like system-wide integration before iOS 8.

Everything could send to reading list. In fact, you couldn’t even disable that!

They had this thing that completely Sherlocked me in the most clear way and had all sorts of advantages. It was built in. It could do all these special things I couldn’t do. I had the same thought at first which was like, “Oh, god, I’m going to die.”

Then next I made this post saying, “Oh, well, people are going to see Reading List, get into the idea of saving things for later, and then search for a better app and find me. So it should actually help me, right?”

Looking back on it, I’m pretty sure it did nothing either way. It had no impact at all.

Matthew: The people who were going to download Instapaper downloaded it and the people who weren’t, didn’t.

Marco: Exactly. I really don’t…looking at the stats over those years, and even just remembering what it was like to be there, I can’t see any correlation of any effect at all.

Matthew: It will be interesting to see what happens with the podcast thing. Apple’s always been a front-runner on podcasts, but you get the sense that, early on, there was one dude there who said, “no, this is super important, we should do this,” and built podcasts into iTunes.

It will be interesting to see if they start to take it more seriously now that it’s a built-in thing or if, as you said, it’s just to facilitate something else — or just easier for them, architecturally speaking.

Marco: They are still bound and held back by the Apple strategy tax with what they do in the podcast app. Because, first of all, whatever is in the podcast app has to be locked and tied to the iTunes podcast store.

Which means it is tied to the iTunes store. Which — that does not sound like something that would make your job easier. That’s something where I would imagine it’s hard for them to get things done, to get changes made, because they are tied to the iTunes store ecosystem and they are completely reliant on it, and they can’t do anything outside of it because obviously they’re in that company.

They also can’t do geeky, specialty features because that’s not generally what Apple does. That would not work in their aesthetic, or it wouldn’t pass the committee, or whatever.

Matthew: Or it doesn’t fit their ’80 percent’ philosophy, necessarily.

Marco: Exactly. Everyone else in the market is not bound by that. We can do whatever we want to, within reason. But for the most part, we can add whatever features we want to. You can have really geeky things, like per-feed settings and priority playlists, the stuff that I don’t think Apple would ever add.

We can do things like that. That will always be able to set us apart from them. There’s nothing stopping them from doing this if they wanted to. They could do it, but they won’t do it.

Matthew: I don’t think the 80 percent philosophy’s going anywhere. That’s an opportunity, obviously, for everybody else. Because 20 percent is a very healthy business when you’re talking about hundreds and millions of downloads.

Marco: If you look at the landscape now, the Apple podcasts app is already way more popular than any other podcast app on iOS, or in the world, for that matter. The Apple podcast app is even more popular than Android podcasts, despite the market share difference. The Apple podcast app is by far the biggest podcast app in the world, and it always has been.

There’s nowhere for us to go but up, right? It’s not like Apple’s going to steal our market share.

Matthew: They already have the market share, right?

Marco: We’re stealing it from them.

Marco: That’s the direction that this is going. I also have always gotten the feeling that Apple didn’t really know what to do with podcasts. They invented podcasts, basically.

They didn’t technically invent enclosures and feeds and everything, but they were the ones to popularize it and get the whole model that we have now.

They still have the biggest and most important podcast directory in the world that they maintain. They do features. They do the new and noteworthy section that I was looking through the other night. They have tons of categories in this directory. This isn’t just in the US. They have different ones for different countries. It’s just like, man, they have a big staff working on this.

Here they are and they’re putting a decent amount of resources into keeping the directory running. They are in the most, the best position in the entire medium. They basically control the medium. Yet, they do so little with it.

I was always curious. Why doesn’t Apple have the best podcast app? Why hasn’t Apple done anything with the format by promoting new standards? Like, ‘this is going to be the standard for paged feeds’ or ‘this is going to be the standard for inline links in the podcast’, stuff like that.

Some of the stuff they have the basics for, and then they just drop the ball. It just always seems like that working on podcasts is never important enough to Apple compared to their other stuff. It’s never a high enough priority. They just let it sit and wither.

Unlike almost everything else they’ve made, there’s no lock-in. There’s no DRM. There’s nothing proprietary about the iTunes podcast store at all. Everyone else can do all of it. It’s this great opportunity for people like me to, like, the fact that podcasts publish their RSS feeds and they have to be publicly accessible to be useful to be most people.

All this stuff and anyone can just walk into this market. It’s not totally locked down. You don’t have to go deal with anybody.

Matthew: That is interesting, and it is relatively unique to other Apple products. I know that they’ve been making some effort to do that a little bit, with expanding store linking and stuff like that. But for the most part it’s the only thing that Apple maintains that’s like that. You can argue WebKit, sort of, but they keep all the best bits for themselves anyway now.

Marco: The App Store is not this open. Not even close. Certainly not the iTunes Store, the music store. That’s not even close.

Matthew: They don’t have full feeds for everything. They don’t want you recreating a store. But they don’t care if you recreate a podcast directory.

Marco: Exactly.

Matthew: It’s an interesting dichotomy there — I don’t know. Maybe it stems from the fact that podcasting was essentially an Internet radio-like paradigm of some sort, or like a mutation of radio for the Internet. That openness of radio. It just like grew that way, organically.

It seems like if the podcast store was created today, it would be created with an entirely different architecture. Mostly likely much more closed off. Because everybody’s locking down their silos these days. It’s like Twitter and the Instagram friend finder — Twitter asked themselves ‘why are we giving you that information?’

It seems like it might be much different if it came about today.

Matthew: One cool thing I’ve noticed is that you put links to all of your independent podcast app competitors into Overcast, which I thought was an interesting decision.

Marco: I was surprised Apple allowed it.

Matthew: Oh, really?

Marco: That was the one thing I was concerned about, about rejection. I was lucky. This actually got approved on its first submission, which is a first for me.They have rules against showing unrelated apps, or showing apps that you didn’t make.

Matthew: I remember that. It was aimed at apps recreating the App Store.

Marco: It was aimed at those app specials — “hot apps” — kind of apps. That was the one thing I was little concerned…I’m like, “I wonder if they’re going to allow me to do this?” But they did. Who knows? I might be removed in the future.

Matthew: It seems like maybe they just agreed with the philosophy behind it?

Marco: Maybe. Honestly, I don’t think the App Store reviewers have the leeway to make exceptions to the rules based on their own philosophies.

Matthew: You might be surprised.

Marco: The reason I did that was — it was nothing weird or sinister — it’s just that this is a small market full of independent developers that I knew to some degree. I felt bad going in at a free price point, so this was a way to alleviate that, but it wasn’t that much thought put into it. It was more like, “You know what? This is a cool thing I want to do, and let me see what happens.”

I was hoping it wouldn’t be perceived as an asshole move, and fortunately it hasn’t been, because there was a little bit of a risk there. But most people have been very good about seeing it as just what it is, just something that to me is a relatively small but fun, nice thing to do, so that’s what I hope it is.

These are friends of mine, some of them.

Matthew: It’s not such a big market that you have this distance from all these people.

Marco: No, it really isn’t. It’s really a very small market. You look at the top podcasts in the world, something like “This American Life,” which is usually the biggest podcast in the world. They might get a million downloads of a popular episode. Compare that to traffic numbers that you get on blogs, and big news sites, and everything.

The whole podcast medium is a pretty small community relatively, so it doesn’t make sense to do crazy poison-the-water kind of things in this small, well-meaning community.

Matthew: One last thing. I saw you mentioned something. I can’t remember if you wrote something about it or what, but you mentioned that it was much easier for you to do design work now that Apple’s made some changes to their frameworks that they offer for developers looking to design their own apps?

Marco: Yeah. It’s less about the frameworks. The frameworks helped a little bit in…ever since iOS 6, they’ve been slowly adding things. Maybe even five.

They’ve been adding things that make it easier to style the built-in elements, because back in forever ago, in the iOS 3 and 4 days, everyone would try to do crazy, ridiculous hacks to try to customize but not totally replace UI kit controls, so you’d have crazy hacks to try to dig in, and dive into the subclasses of something, and try to replace the texture with your own texture, and all sorts of crazy text.

Matthew: Without having to create something from scratch?

Marco: Exactly. Over the last few releases, Apple has added a lot of customization so that now you can customize the stock stuff pretty effectively without hacks and without rewriting everything, so that helps, but what also helped a lot.

What probably helped more, is just the change in fashion that happened with iOS 7, because if I was making this app for iOS 5 or 6, and I released it looking like the way it does, a lot of people would say it looks terrible because it didn’t fit in. It was not the fashion of the highly textured style of app. It didn’t have that.

I’ve never been good at making those kind of apps because I can’t draw textures, I’m not a graphic artist, so Instapaper never really had that, and that was mostly because I couldn’t do it. Now it’s nice because that’s out of style, and what’s in style is exactly what I happen to be able to do by pure coincidence, which is text, lines, and white space.

Now I can turn in something that looks like the wireframe I would’ve handed a designer for iOS 5 and that’s the final app.

Matthew: I know there are a lot of reasons why they made those changes, but making at least decent design more accessible to more developers is a great side effect.

I don’t think they got together in a room and said, “What do we need to do? We need to make it easier for people to make decent-looking apps and that’s the primarily goal for this release,” but it probably came up at some point as an effect of what changes they were going to make to the system, which had a lot more to do with flexibility on different screen sizes and all of that stuff.

It seems like it’s a good side effect and it should make it more accessible, which will make, in turn, help Apple because of expanded App Store offerings for their customers that look at least decent.

Marco: Exactly.

Matthew: I think that’s about all I had, I’ll let you go. Thank you very much for talking.

Marco: I have 387 emails to read.

Matthew: I’ll let you get to it.