There has been a lot of hubbub this week around Eric Schmidt’s Android comments. Eric, the chairman of Google, said at LeWeb that “whether you like [Android] or not, and again I like it a great deal, you will want to develop for that platform, and perhaps even first.” He set a six-month time horizon for this Android-first change. Android has a lot going for it, that is for sure. Its smartphone “market share” is over 50% and clearly many customers prefer Android over other platforms.

Marco Arment, the creator of Instapaper (an iOS, web and Kindle app), asked why anything would change in the next six months? Developers aren’t making money on Android right now (link). Gene Muster from analyst Piper Jaffray backs up Marco’s assessment. He claims that Apple has sent $3.5 billion to developers while Google has paid out $240 million (link). That’s a huge difference. The folks at Shifty Jelly, an iOS and Android developer, have had the opposite effect, though. Their Android sales far outweigh their iOS sales (link).

“Market Share”?

I put “market share” in quotes before. Why? Because market share numbers to developers is completely different than how most consumers view the popularly reported figures.

As others have pointed out before me, we can’t really compare Android market share to iPhone market share. That is a platform versus a device. As a developer we generally write for a platform. To compare apples to androids, we would need to count iPhone, iPad and iPod touch market share, combined called iOS market share, to Android market share. And given Apple’s commanding lead on the iPad and the number of iPod touch devices sold, that number favors Apple. According to comScore, Apple’s US market share as of August 2011 is 43% to 34% (source). Worldwide as of October 2011, according to Net Applications, Apple is favored 61% to 19% (source).

A Touch of Accounting

This comparison comes into even sharper contrast, favoring Apple, when you realize that Android (and all other mobile device) sales are counted completely different than Apple sales. Apple’s sales are counted as the actual sales to a customer. I walk into a store, buy an iPad and Apple adds a 1 in the sales column. Android’s (and other mobile device’s) sales are counted as sales to channel.

The difference is huge. A sale to channel is a sale to Verizon or Best Buy, not a sale to a customer. Samsung, HTC, Motorola, RIM, and Nokia all get to count the sale into the channel as revenue and as a sale toward market share [1]. But it doesn’t mean that device was sold to the public. RIM just showed why the two — iOS market share and other operating system market share — aren’t equivalent as it is taking a write-down (meaning the channel sent back devices) of $485 million and possibly as many as 2.4 million devices (link). RIM, when figuring BlackBerry/QNX operating market share, got to count those 2.4 million devices toward their share.

I only care about devices sold to customers.

If Market Share Really Mattered…

Even if you think I am full of it and don’t believe the numbers from above, then why are many developers focusing on iOS over other platforms with much greater penetration? John Gruber points out, for instance, that Windows and the web have much larger installed bases than Android and iOS combined. If market share matters so much, why do developers ignore the web and Windows, choosing to develop exclusive for iOS, Android, RIM or a combination of these “low market share” devices (link)?

The real answer, which John misses, is they don’t and you can’t, at least when it comes to the web [correction: I mean Internet so I might not actually disagree with John]. The hottest applications of 2011 are all web-connected applications. Instagram was just named Apple’s app of 2011 (link). The year before, I believe, it was Flipboard. Neither app works without the web. Both require a web connection to share images and gather stories with your network. Just because the front-end isn’t in a browser doesn’t make them less of a web application. Neither of these apps could have existed before the web, even on a desktop computer.

If you want to create a highly successful mobile app today, I believe, you better figure out how the device and web work together.

I’m No Idealogue

Look, I’m not an Apple fanboy nor am I an Android one. My allegiance is to the platform(s) that pay the bills. For that matter I write apps that work with numbers and have made a living at it (sometimes just barely) on multiple platforms for 14 years now (link to powerOne). I do put a lot of stock in the numbers and so far they are telling me that for an app like powerOne, one that relies on customers paying me $4.99 per copy, and for a team like we have, which is too small to focus on many platforms, that iOS — iPhone, iPad and iPod touch — is still the primary game in town.

—

[1] Amazon Kindle Fire sales are most likely counted like Apple sales since they are sold directly to customers. I’m not certain how Barnes and Noble accounts for Nook sales.