There’s been a lot of back and forth today about some comments AT&T CEO Randall Stephenson (yes, him again) made recently during a Q&A session. When an annoyed customer asked why it takes so long for AT&T to roll out new Android releases, Stephenson said the following:

Google determines what platform gets the newest releases and when. A lot of times, that’s a negotiated arrangement and that’s something we work at hard. We know that’s important to our customers. That’s kind of an ambiguous answer because I can’t give you a direct answer in this setting.

That’s the CEO of the nation’s second-largest carrier placing the blame solely on Google for the poor Android update timeliness. Obviously, Google is not going to be happy about that. So they gave the following response to 9to5 Google:

Mr. Stephenson’s carefully worded quote caught our attention and frankly we don’t understand what he is referring to. Google does not have any agreements in place that require a negotiation before a handset launches.  Google has always made the latest release of Android available as open source at source.android.com as soon as the first device based on it has launched. This way, we know the software runs error-free on hardware that has been accepted and approved by manufacturers, operators and regulatory agencies such as the FCC. We then release it to the world.

So what’s going on here?

If you actually watch the video (embedded in the 9to5 Google post), my sense is that Stephenson is talking specifically about one of two things.

Either he means Google’s “flagship” handset launches. Those absolutely do require Google working with the OEM/carrier beforehand to get both the device and the new OS ready to go. The last one of these was the Galaxy Nexus which launched exclusively with Verizon, for example.

Or he could mean the Android Compatibility Program. That is, the certification a device must go through in order to be able to get the Google Play software license (in order to come with Google apps pre-installed). See more here.

It doesn’t seem like either of those are exactly the answers the audience member was looking for — he probably just wanted an easy answer as to why only a handful of devices have access to Ice Cream Sandwich months after launch — but that’s the one he got.

Google’s response can also be read two ways: that they really don’t understand what Stephenson could have meant. Or that they’re just being coy — playing dumb — to deflect something that is actually a real issue.

Stephenson’s comments out of context are a little hard to follow and perhaps poorly worded, but come on, they’re not that hard to follow when you think about it.

While it’s intriguing to think that the CEO of AT&T doesn’t understand how his own phones get updated, this is spin trying to make it seem as if the company that just got thrown under the bus is actually the one driving the bus.

When comScore released their latest U.S. smartphone market share numbers earlier today, I was a bit confused. According to comScore, Google (Android) usage surpassed 51% last quarter. Apple (iPhone) meanwhile, was at 30.7%. Those numbers alone aren’t necessarily surprising, comScore measures overall market usage, not just new sales and Android devices (as a whole) had been outselling iPhones for much of the last couple years.

But something happened last quarter. On the nation’s two largest carriers, Verizon and AT&T, the iPhone actually outsold all Android devices — combined. The nation’s third-largest carrier, Sprint, did not give a number for total smartphones sold last quarter. But they did disclose that they sold 1.5 million iPhones, which was higher than expected. Given the numbers, it sure seems like the iPhone is the majority of their smartphone sales as well — maybe by a lot — but it’s hard to know for sure. Yet, according to comScore’s numbers, Android market share rose 3.7% versus 1.1% for the iPhone.

This leaves two distinct possibilities.

First, that T-Mobile and regional carriers (which don’t offer the iPhone) more than made of the difference between Android and iPhone sales at the big guys. Eric Slivka of MacRumors notes there are around 70 million wireless subscribers in the U.S. without access to the iPhone, so it may be possible. That assumes that basically all of those people chose Android devices, but that’s also possible given the shrinking market share of Microsoft and RIM.

Second, that comScore’s method of measuring smartphone market share is flawed.

It certainly could be the case that the first scenario is correct, but it just doesn’t feel right. I’ll concede that the people without access to the iPhone could have offset the iPhone dominating Verizon, AT&T, and Sprint, but enough for Android to see nearly 4x the growth rate of the iPhone? That seems suspect.

Then consider the numbers from NPD. As a rival to comScore, NPD has their own methods for gathering smartphone market share. In their most recent report, they had Android controlling 48% of the U.S. smartphone market versus 43% for iPhone in Q4 of last year. In the same span, comScore had Android at 47.3% and the iPhone at 29.6%. 

Forget the actual numbers, focus on the differences in the numbers. It’s pretty clear that the methods these firms are using to measure smartphone usage aren’t an exact science.

Digging deeper, you’ll find that the way comScore gets its numbers is through a service they call MobiLens. How is MobiLens calculated?

MobiLens data is derived from an intelligent online survey of a nationally representative sample of mobile subscribers age 13 and older. Data on mobile phone usage refers to a respondent’s primary mobile phone and does not include data related to a respondent’s secondary device.

A survey. 

So on one hand, we have actual, verified and legally reported public data from the three largest U.S. carriers. On the other hand, we have a survey. 

I’m not denying that Android still has the larger overall market share in the U.S. I’m just disputing comScore’s stats that it’s still growing faster then the iPhone. 

Regardless, one thing is very clear: when the iPhone is available on a carrier, it dominates. This is backed up by cold hard sales numbers, not surveys. If Android is still “winning” in some segments of the market in the U.S., it’s only because Apple is allowing it to. 

Update: comScore has written to clarify things a bit. It turns out their numbers do show iPhone subscriber growth outpacing Android on the “Big 3” carriers (13% vs. 11% from December to March). But the overall growth Android saw came mainly from other carriers (T-Mobile and regionals) where Android is dominating.

Fair enough. This reinforces the last point: that Android is dominating the areas where the iPhone isn’t competing. Yet.