How Many "Old" iPads Were Sold Last Quarter?

Note: I refer to the iPad 3 numerous times during the post. I am well aware that Apple refers to it as the New iPad, but for the sake of clarity (and since I will be discussing “new” and “old” iPads in a different context) I have ignored the Apple-approved naming convention.

Update: Daniel Tello, author of the wonderful deagol’s AAPL model, indicated to me that I had incorrectly calculated the cumulative and 3-period average ASP for the iPad. I had calculated a straight average instead of a weighted average, an embarrassing mistake considering I have taught the difference many times to introductory accounting students. The data has been updated.

Matt Richman observed a significant decrease in iPad average selling price (6% quarter-over quarter, 8% year-over-year; tip-of-the-hat to John Gruber) and concluded that the iPad 2 (and its $399.00 entry point) must be selling well. I thought I would put this to the test.

The following two charts give some background on the trend in iPad selling price thus far. The first shows the disparity between the growth in unit and dollar sales Apple brings in from the iPad, while the second calculates the quotient of these (the average selling price, or ASP) and shows it over time.

iPad Unit and Dollar Sales per Quarter

iPad ASP per Quarter

We know that the iPad had an ASP of $558.57 in the quarter just ended. In order to estimate the relative mix of “new” iPads (iPad 2’s before the iPad 3 launch, iPad 3’s after the launch) and “old” iPads (iPad 2’s after the iPad 3 launch) we need to know the ASP of each. This presents an interesting problem as neither can be determined for certain.

On the “new” iPad front, we can choose a number of figures from a historical context. The average ASP since the iPad launch is $632.31; the average ASP over the prior three quarters is $621.47. We could also estimate that the ASP of “new” iPads remained constant with the prior quarter. Alternatively, we could extrapolate from the prior quarter’s quarter-over-quarter decrease (yielding an ASP of $569.59). I feel, though, that this last choice is highly unlikely. There was significant press surrounding the need for users to purchase higher capacity iPad 3’s to make up for the larger retina images, and the ASP actually increased by 8.23% quarter-over-quarter during the launch quarter for the iPad 2. I will use all of these figures, save the last one, in estimating the “new” iPad ASP. I will also punch in my own estimate of $600 (a quarter-over-quarter increase of 1%).

The “old” iPad is an equal challenge. There are now only two price points for the iPad 2: $399 for the wifi-only iPad and $519 for the wifi + 3G version. I inputted six possible ASPs for my analysis: $399 (only Wifi models sold), $529 (only 3G models sold), $450 (my own personal estimate), and three other price points that were calculated so as to correspond to the average (since launch), average (trailing three quarters) and prior period ASPs used for the “new” iPad.

Putting these together and applying my proprietary formula (actually, the data is freely available here), we get estimates ranging from 2.1 million to 8.4 million, or 17.8% to 71.4% of all iPads sold during the quarter. My personal estimates yielded very similar results to the numbers given using the prior quarter’s ASPs (3.3 and 3.4 million, respectively), so I feel very comfortable forecasting the iPad 2 sold between 2.5 and 4 million units since the iPad 3 launch. The results of this analysis are shown in the following chart:

iPad 2 Unit Sales Estimates

My estimate is actually at the low end of the range, but the number is still tremendous. The quarter ended just two weeks after the iPad 3 launched, indicating an iPad 2 sales rate of 1.5 to 2 million units per week even after the iPad 3 release. This may have been helped by the iPad continuing to be the newest available iPad for most countries around the world, though I have a feeling that customers’ desire to wait for the newer version to be released in their country would have balanced this effect. In any case, this is still an order of magnitude faster than Android tablets have ever sold, combined (typically between 50,000 and 200,000 per week), let alone when a brand-new successor had just been announced and released.

You can get the data used in this post here. You can also follow me on Twitter or subscribe to my RSS feed.