FORTUNE — Can you spot what’s wrong with this picture?
Apple (AAPL) on Monday announced first weekend sales of the iPhone 5S and 5C. They “topped” 9 million. Wall Street was expecting 5 to 6 million. Big surprise, egg on faces. Estimates revised. Price points raised.
Then the second-guessing began. How solid, some analysts asked, were those 9 million sales?
- Jeffries Peter Misek estimated that 2.5 million of the iPhone 5Cs Apple “sold” weren’t really sold, but were “sitting on shelves,” somewhere in Europe or Asia.
- Piper Jaffray’s Gene Munster told Bloomberg TV that “sell in” for the iPhone 5C could be even higher — as high as 3.5 million — which would mean that only 5.5 million iPhones actually “sold through” to customers last weekend.
- Given that the iPhone 5’s launch-weekend number last year was 5 million, 5.5 million is not, as Misek puts it, “a huge amount of upside.”
Coming to Apple’s defense was Goldman Sach’s Bill Shope, who saw a “significant upside” in those 9 million iPhones. He made several points:
- Apple this year used the same sales recognition rules it has always used. From an accounting perspective, this year’s 9 million is directly comparable to last year’s 5 million.
- There were differences — two new phones, not one; 11 countries (including China), not 9; the availability of preorders — but Misek, Munster and the rest knew all about those differences in advance, and they still offered their clients estimates that were off by 3 or 4 million units.
- There was certainly unsold product on hand at the end of the weekend (“sell in to channel,” in the jargon of the trade). But most of those phones would have been sent to Apple’s own stores, where Apple’s sales recognition rules say they aren’t counted as sold until they walk out the door. Supplies to Best Buy and other non-Apple stores (which are counted as sold) are probably just as tight this year as they were last year. That makes the so-called channel fill issue is a wash.
Arguing from yet another side, The Core‘s Matt Lew points out that Apple’s 9 million figure does not include any of the iPhones — tens of millions of them, he estimates — that were ordered online but have not yet been delivered. Lew agrees that 9 million distorts the picture, but from the other direction, making sales look lower than they are.
Here’s what’s strikes me as strange about all this: No other smartphone manufacturer’s sales figures are subjected to this kind of scrutiny.
Take, for example, Samsung. You won’t see analysts questioning Samsung’s unit sales numbers for the Galaxy S3 or S4. Nobody writes notes to clients asking what percentage of those sales were sell-in or sell-through.
Samsung doesn’t get this kind of scrutiny because it doesn’t tell anybody — not analysts, not investors, not the SEC — how many smartphones it sells.
And that’s what’s wrong with this picture.
