Monday, March 29, 2010

How Many iPads Have Been Ordered?

Two attempts that received widespread attention in the tech blogosphere attempted to decode Apple order numbers. Working off the assumption that these numbers were assigned consecutively to orders from the company’s online store, bloggers used a series of orders to attempt to construct a sales rate.

Andrew Erlichson, chief executive of the photo-sharing site Phanfare, based his estimate on just two sales. Erlichson said his own company assigns order numbers consecutively, which has the advantage of simplicity but may reduce the security of the system.

Another effort came from a group with considerably more experience decoding Apple sales. Members of a private message board about the company, led by Apple investors and watchers Daniel Tello and Victor Castroll, have gathered 500 order numbers and used them to estimate sales, in the iPad’s first week, of 190,000. These are based on several assumptions and exclude in-person orders and corporate sales, so may be too low.

However, Tello does have a strong track record, beating several analysts in estimates of Apple earnings and contributing to a successful effort to count iPhone sales in 2008. Having access to a group of Apple investors and customers “is the most valuable tool in my toolbox,” said Tello, a former bank data analyst who now manages his investments full-time. He added, “In a way I think analyzing a company like Apple shouldn’t be too hard. The most important part, in my opinion, is to be able to create your own assessment of the direction things are going, and somehow avoid all the punditry and biased opinions found all over the tech media coverage of this company.”

Castroll primarily follows Apple as an investor because clues are available to create such an assessment. “If you want to know Apple sales, you can look at order numbers or you can see people walking into a store,” he said. Other companies, by contrast, are “a black box.”

Other industry watchers use different techniques to measure sales. Flurry, which helps makers of mobile applications track users, can use its tools to recognize which phones people are using, through unique signatures for each device. Flurry’s technique led it to produce numbers that were disappointingly low for a new Google phone, the Nexus One. “We’re pleased with our sales volumes and with how well the Nexus One has been received by our customers,” a Google spokeswoman said, while declining to comment on specific Nexus One estimates. “The Nexus One is one of a fast-growing number of Android handsets. … Our partners are shipping more than 60,000 Android handsets each day compared with 30,000 just three months ago.”

Peter Farago, vice president of marketing for Flurry, said that users of applications which incorporate his company’s software probably cover 80% of advanced devices. He called it “a number we’re willing to live with.” He said he isn’t sure whether Flurry will estimate iPad sales, once the device starts being used by the market, because Google, Apple and other rivals in the mobile device industry are partners. “I know it would be a pretty strong scoop,” Farago said. “But when we release news, we don’t want to injure ourselves.”

Yet early sales numbers are far from definitive. Studio executives can tell their technology counterparts all about the dangers of paying such figures too much heed. A year ago, “Watchmen” debuted to an impressive opening-weekend box-office take of $55.2 million, according to It didn’t make that much in the rest of its domestic run, finishing with a gross of $107.5 million. Eight months later, its studio, Warner Bros., released “The Blind Side” and raked in a more modest opening-weekend total of $34.1 million. That film’s box-office total is $252.7 million and counting — more than twice the total for “Watchmen.”

Gene Munster, senior research analyst at Piper Jaffray, recalled taking phone calls from investors who considered the iPhone “dead on arrival” soon after its launch when early numbers from AT&T Wireless missed early sales forecasts. Munster pointed out that word of mouth and strong sustained sales rates rewrote that first impression.

Munster’s own forecast was among those that overstated early iPhone sales, predicting 355,000 purchases in the first two days of availability, compared to an announced total of 270,000 from Apple a month after the iPhone launch. “We did our work and our work was wrong,” Munster said. Learning from that experience, he attempted to predict early sales for the iPhone 3G S when it was released last year — and undershot, predicting a half million sales in the first weekend. It took three days for the handset to sell one million units.

No comments: