If you look very closely at the Feb. 2011 data in the chart at right you will see slight uptick in the green line that represents Apple's (aapl) share of the U.S. smartphone market.
It's a subtle change. You might have to enlarge the chart (by clicking) to see it. It reflects, according to the ComScore data released Friday, an increase in the iPhone's market share from 24.7% in January to 25.2% in February -- a month, according to ComScore, in which the Verizon iPhone was the best-selling smartphone in America.
I point this out because one of the great unsolved mysteries of the smartphone wars is the effect of the arrival in February of the long-awaited iPhone for Verizon (vz). Judging by ComScore's data, it didn't do much to slow the advance of Google's (goog) Android, whose U.S. market share rose that month from 31.2% to 33%.
And I tried to
would argue that it puts the lie to claims in the press that Verizon sold 1 million iPhones in the first weekend of sales.
Proving this is a little tricky, because Apple doesn't break down iPhone sales by country. Moreover, what ComScore is reporting is market shares, not actual unit sold, in a market that itself is growing, according to Gartner, at a rate of more than 70% per year.
But here we go:
[Update: given the known flaws in what follows, maybe you should skip the next two graphs]
Using Gartner's numbers, 296.6 million smartphones were shipped to users worldwide in 2010 (up from 172 million in 2009), and 46.6 million of them were iPhones. According to ComScore, Apple's share of the U.S. market was 25%, give or take, all year. That suggests that the U.S. smartphone market in 2010 was around 186 million -- or 15.5 million per month -- and that Apple was selling iPhones last year at a rate of about 4 million per month.
If it were true that the Verizon iPhone bumped Apple's sales from 4 million units in January to more than 5 million in February, then Apple's share of the market would have jumped 6 or 7 percentage points, not the measly 0.5% ComScore observed.
[Reader Dan in Philadelphia, who seems to have a firmer grasp on the numbers than I do, suggests that the bump caused by 1 million extra iPhone sales would be on the order of 1.4%. See his comments below. Reader KenC in Gardiner, Maine, (or is it Shanghai, China?) points out that ComScore's February numbers are actually an average of surveys taken over three months, not a snapshot of February's sales. And BrianL in Houston, using some mathematical legerdemain I don't quite follow, reaches this conclusion: "Looking at iOS additions by month, Nov to Dec showed 425,000 additions, Dec to January 450,000, and Jan to February 1,262,000. Looks to me that Verizon has made a difference."]
OK, so my math was
pretty shaky full of holes. But it jibed with some first-hand observations:
I've seen the kind of crowds that gather outside Apple Stores on those weekends that the company sells 1 million units of anything. Even if, as Verizon reports, 60% of its initial sales were pre-ordered online, you would think if 400,000 people showed up to buy iPhones between Feb. 10 and Feb. 13, someone would have spotted them.
Besides, Verizon has been cagey -- and Apple silent -- about the size of the Verizon iPhone bump. If 1 million iPhones had been purchased first weekend, don't you think -- given Steve Jobs' penchant for nice round numbers -- that Apple would have issued a press release?
We may finally get some answers on April 20, when Apple reports its earnings for the second quarter of 2011. Tune in here for our analysis.
Update from ComScore senior analyst Mark Donovan: "I wanted to point out that the data we report in our mobile benchmark releases is based on a 3mo moving average, so events that happen in a single month—like the launch of iPhone on Verizon—will tend to be a bit muted in this view. The release noted that the Verizon iPhone was the most acquired handset in February. While we don’t release model-level data publicly, our data show that more than 1 million Verizon subscribers owned an iPhone by the end of February." (emphasis ours)
Also on Fortune.com:
- Why are the Verizon lines so short?
- Verizon halts iPhone orders. An estimated 100,000 units pre-sold
- Survey: 44% of Verizon Android users likely to switch to iPhone on Day One
Follow Philip Elmer-DeWitt on Twitter @philiped.