Perhaps the question analysts should be asking is how fast could it build them?
When estimating unit sales in a highly competitive market — like computers or cell phones — the traditional method is to sample some part of the sales channel and use that as a proxy for the broader market.
In a post entitled “Predicting iPhone sales for dummies,” Asymco’s Horace Dediu makes a strong case that in markets where demand outstrips supply — as it often does for Apple’s AAPL new devices — a better way to estimate sales is to determine the constraints on production and try to measure those.
Dediu was writing about the iPhone, which Apple says it still can’t build fast enough to meet demand. Using ramp-up rates for the previous three years, Dediu estimates that Apple will sell twice as many iPhone 4s as it sold iPhone 3GSs.
But the logic he uses for the iPhone applies in spades for the iPad — both the original model, which was in short supply most of last year, and the iPad 2, for which customers have been queuing up in record numbers.
So how many iPads did Apple sell in the quarter that ended Saturday (which includes two weeks of iPad 2 sales in the U.S. and two days in 25 other countries)? We put the question to a few dozen analysts — both professionals and amateurs like Dediu. Their answers range from a low of 5 million (from Deutsche Bank’s Chris Whitmore) to a high of 8.8 million (from Gambelli’s Hendi Susanto). The average estimate is 6.23 million, the median is 6 million.
Dediu’s number — 7.3 million — is, as usual, on the high end. But his estimates also tend to be pretty close to the mark.
Below: The numbers we’ve gathered so far. Unit sales are in millions.