By John Gruber
Journey across Bhutan with Gray Langur. October 16-29, limited availability.
I try my best to avoid writing sensational headlines for my pieces here.1 I slipped, though, with my piece on the differences between the Android and iOS app markets two weeks ago. “Where Are the Android Killer Apps?” was too broad. It implies there are no third-party “killer apps” for Android, period. That’s not true, and it distracted from the point I was trying to make.
A better headline for that piece would have been, “Where Are the iOS-Style Killer Apps for Android?” That gets to the heart of my argument.
In my piece, I wrote:
At this point, I’m guessing, Android fans are ready to exclaim that the fact that Android supports things like home screen replacements (or other system-level tools, such as touchscreen keyboard replacements) — and that iOS does not — is precisely why they prefer Android, and/or consider iOS to be an unacceptable toy, or what have you. But, again, that’s not the argument I’m making. I’m talking about third-party developer exclusives — and the only ones Android has are ones that Apple doesn’t want.
Many Android supporters took that as an attempt on my part to dismiss Android’s strengths as somehow not counting. But I wasn’t arguing about the merits of iOS vs. Android overall. I was (and am now, again) trying to make a specific point about third-party developer support for these two platforms.
But OTHER THAN widgets, navigation, voice search, inter-app communication, and tethering, what have the Romans ever done for us?
That’s a good summary of many Android fans’ favorite things about Android. But none of that pertains to the point I’m trying to make about third-party developer support. For the sake of argument, let’s concede that an Android phone loaded with only Google apps may well be a lot more appealing than an iPhone loaded only with Apple apps.
On the Hacker News thread for my piece, the highest-rated comment concludes:
If everything Apple does is exactly what you want, Android probably doesn’t have any “killer features” worth talking about, but as soon as you want to do something that diverges from the Cupertino Grand Unified Vision, Android starts to become a lot more attractive.
Sort of true, sort of not true. Most of the apps Android users claim they couldn’t live without (if they were forced to switch to iOS) come from Google. Most of the apps iOS users claim they couldn’t live without (if they were forced to switch to Android) are from the App Store. (Feel free to replace “couldn’t live without” with the un-hyperbolic “would miss the most”.)
One fundamental difference between the two is that Android allows third-party software to change the system itself in certain ways. You can add components that show up in standard menus across the system. There’s a standard “do something with this URL” menu, for example. Third parties can add items to that menu, very much like the way you can add system-wide Services in Mac OS X. Third parties can also replace the Android home screen, and replace the touchscreen keyboard. There are lots of Swype fans out there.
So lots of Android fans argue, with merit, that most of their favorite apps/additions/add-ons/whatever for Android would not be allowed or even possible in the iOS App Store. This gets to the point of my article, though. Most apps for the iPhone could be ported to Android. They’d lack a bit of polish, in many cases. It’s highly doubtful something like Rage HD is even possible on any Android device today. Media apps like Netflix and MLB At Bat are problematic for Android because Android devices don’t have a standard DRM system. But most of my personal favorite iOS apps, including games, could, technically, exist today on Android in a more or less equivalent form. But they don’t.
iOS’s best apps could exist for Android but don’t. Android’s best apps couldn’t exist for iPhone.
In theory, then, Android could be beating iOS in both regards. Android could be the platform with exclusive apps like Reeder, Twitterrific, Things, Simplenote, Instagram, Calvetica, PCalc, and Pastebot — in addition to the exclusives it already has like Swype and home screen replacements that the iPhone can’t have. What I find interesting is that Android just doesn’t have apps like this. There is no Iconfactory or Tapbots for Android.
A lot of the Android-vs.-iOS horse race coverage focuses on unit sales and market share, but never even mentions this disparity in developer support, except to simply list that the iTunes App Store has X total apps and Android Market has Y total apps.
I’m not saying (here, with this argument, at least) that iOS is better than Android. I’m just saying that the iOS App Store is thriving in a way that the Android Market is not. And, that the Android Market would be thriving in this way if handset market share were all that mattered.
Tim Bray, via Twitter, suggests an optimistic answer:
@gruber Simplest explanation: Android numbers didn’t get interesting to devs till last summer. Re-check in a year.
That could be. The iPhone had developer appeal from day one. In fact, it had developer appeal before it was even released. Android didn’t have that appeal out of the gate. But what Android now has are raw numbers — millions and millions of users. According to Andy Rubin, 300,000 phone activations per day. As the size of the overall Android user base grows, it will inevitably attract more developer attention, so the argument goes.
Third-party developer support for Android is likely to grow — it almost seems like a certainty given the growth in Android handset sales — but I don’t think it’s going to grow in an iOS-like way.
Note, too, that the U.S. market looks different than the worldwide market. Here’s a recent report from Comscore that claims Android is very close to catching the iPhone in total U.S. smartphone users. Note that those numbers are U.S.-only, and only for phones — thus excluding the iPad and iPod Touch. Compare that to these worldwide numbers for overall web usage by mobile OS, which shows iOS with more than double the share of Android.
But even considering only the U.S. smartphone market, keep in mind that the iPhone is exclusive to a single carrier — the one with the worst network quality and worst customer service. These market share numbers are true today, and we shouldn’t start counting our Verizon iPhone chickens before they’re hatched, but surely, a Verizon iPhone in early 2011 would change these numbers significantly.
Fred Wilson, principal of Union Square Ventures and author of the A VC weblog, wrote a piece last week recommending that mobile developers focus first on Android:
So, when thinking about where to invest your precious mobile development resources, I’d say Android first and iPhone second. And think hard about HTML5. You may want to hedge your bets by having a kick ass HTML5 experience. I learned in the comments to this HTML5 post last week that there is an awesome open source library called Phone Gap that lets you port HTML5 apps to Android, iOS, Blackberry, Palm, and Symbian. Seems like developing in HTML5 and then porting to the mobile OS platforms is an interesting option as well.
One thing I am sure of is that developing solely for iOS, which is a very common thing I see out there, is not the right strategy unless you only want to serve 25% of the market.
He bases this advice on market share trends, citing the same U.S. smartphone market share numbers from Comscore I linked to above:
Apple is stuck at about 25% of the smartphone market. RIM is losing share and Google is gaining share. If we have two more quarters like this past quarter, Google will have 37% market share, RIM will be at 29%, and Apple will be at 26%.
Of course there is no certainty that the next two quarters will play out the same way the past quarter went. Many people replied that getting the iPhone on Verizon will be a boost to Apple’s numbers. I suspect that will help. Apple could get into the mid 30s with the help of Verizon.
Again, I think it’s unwise to focus solely on U.S. market share, and even more unwise to focus on “smartphone” market share, ignoring non-phones like the iPod Touch and iPad (and, now, Android devices like the Galaxy Tab). But even with just U.S. smartphones, I think he’s underestimating the impact a Verizon iPhone would have. Even if Wilson is right that the iPhone would gain only 10 percent share — from around 25 percent to “the mid 30s” — would not a large chunk of that gain come specifically at Android’s expense? Verizon is the number one carrier for Android phones in the U.S. — and the U.S. is Android’s strongest market.
The bigger flaw in Wilson’s reasoning is that he seems to think that all smartphone users are equivalent. Based on those Comscore numbers, iOS and Android each account for roughly 25 percent of the U.S. smartphone market. But from a developer’s perspective, are those audiences equivalent? Wilson is arguing that they are. I don’t think that’s true at all. If Instagram were an Android exclusive rather than an iOS exclusive, would it have picked up 100,000 users in its first week and a million in its first month? No way, I say.
I can see the argument that betting on Android may be a worthwhile risk. An argument along the lines of, “The iOS App Store is crowded, and Android has a dearth of high-quality apps. A great app that might get lost in the iOS App Store would have a better chance of standing out on Android.” But I don’t see how betting on Android — following Wilson’s advice and making it your primary target for mobile development — is not a gamble. Even with its market share gains and strong momentum in handset sales, I don’t see any evidence of corresponding gains in sales momentum for Android apps.
But perhaps that’s a key word: sales.
In a tweet, Wilson wrote to me:
I believe mobile economics will trend toward web economics. Not paid, maybe freemium, often free.
In that context, his advice that mobile developers focus first on Android makes sense. If all you want are eyeballs, go to where the most eyeballs are.
But what if you don’t merely want eyeballs, but rather want to sell your app directly to users? In that case, you should go to where the sales are, and that means iOS.
From the perspective of developers who wish to sell their apps, there seems to be something wrong with Android Market. Note, for example, that Swype, the popular keyboard replacement for Android, is not sold through the Android Market. There’s a “beta” version, but the beta is invitation-only. The company is monetizing the software not by selling it to users, but by working out licensing deals with handset makers to have Swype pre-installed. (And it is, in fact, pre-installed on many (most?) new Android handsets sold today in the U.S.) This matches what I heard from an iPhone game developer — that rather than sell their upcoming Android ports, they’re trying to work out deals with carriers to have the games pre-installed on phones. Angry Birds, which has sold — sold — millions of copies for iOS, is only available for Android as a free app with on-screen ads.
One problem with the web as a model for mobile economics is that the web economy doesn’t seem to work for many product categories. The web economy works great for Google, not so much for anything that is intended to be sold.
I don’t know how the purely ad-based Angry Birds is faring on Android. Perhaps Rovio (Angry Birds’s developer) will make even more money from ads on Android than they do from sales on iOS. We shall see. Sounds good so far. But what if you don’t want ads everywhere?
The economy for Android apps may well trend toward resembling the economy of the web. I think that’s the mindset with which Android was designed. But I don’t think the iOS App Store resembles the web economy at all. Isn’t that the whole point, the reason Apple created the App Store in the first place? Apple’s pre-App Store “just write mobile web apps for Mobile Safari” developer story for the iPhone in 2007 was unsatisfying, both for users and developers.
Where I think Wilson is wrong is in his expectation that there will be a single collective “mobile economy”. The economics of music on iTunes do not resemble at all the economics of “digital music” in general — so why expect the iTunes app economy to grow to resemble that for “mobile apps” in general?
The differences between the iOS App Store and Android Market are a microcosm of the differences between Apple and Google. Apple is a retailer, a purveyor of well-crafted goods that people will line up to purchase. Google is an advertising company that builds popular services that command large audiences.
There’s a difference in culture — from the platform creators, from the developers writing software for the two platforms, and from what the users of these devices expect. For iOS, it’s about emotional appeal — art, design, the ineffable.
Sensational headlines are one of the worst aspects of web-based news media. Pre-web, there wasn’t much temptation for publications to juice their headlines, with the notable exception of tabloids. In print, all publications care about is the circulation for entire issues. With the web, particularly with CPM advertising, publications are motivated to drive traffic to each individual article. Total page views, rather than issue sales, are the metric, and one easy way to juice page views is to use sensational headlines. I think Hacker News and Techmeme have succeeded largely because they seem designed to work around the problem of sensational headlines. Digg, for example, tends to promote articles that sound interesting. Hacker News tends to promote articles that are interesting. ↩︎