Estimating iPad mini Cannibalization

Today there was a survey out from Cowan and Co that got written about in almost every blog. It is about the effect of iPad mini on iPad sales. Most quickly jumped to the obvious conclusion (in articles with catch headlines too) that iPad mini will go on to add significantly to Apple’s profit with “inconsequential” cannibalization of iPad.

Here is my representation of the survey results using the data from Cowan and Co and let me tell about the problems with this survey and hence the conclusions.

First this is a stated preference study measuring the attitude of the customer and not the actual behavior. It is well established in marketing research literature that stated preference surveys overestimate behavior at point of purchase.

Second the survey question was specific to iPad mini, asking them specifically if they would buy iPad mini in the next 18 months. They did not ask them about other tablets in their consideration set nor did they ask them if they planned to purchase Kindle Fire, Nexus or nook. That is too narrow, anchors them on single choice and ignores other possibilities. Had they asked, “Which tablet will you buy?” and reported percentage distribution of different tablets it would have been much better.

Third, they slice and dice the 24 samples who reported switching from another device to report cannibalization and conversion from other tablets. The number 24 is too small to make any meaningful estimate, especially when 8 report switching from iPad and 3 report switching from Fire. If these were accurate estimate then it also point to even smaller impact on Kindle and other tablets.

Fourth, a more significant problem with the question  “What device will iPad mini replace?” is it is just plain wrong as respondents were not primed to compare price and value of each. One right way is to do a (choice based) conjoint analysis to find the respective share of different tablets – iPad, iPad mini, Fire, Nexus etc.

Fifth, let us take the 6.1% new buyers number at face value. You can interpret this as 6.1% of all those who would buy a tablet would choose iPad mini. That is, iPad mini is not bringing in many new buyers into the market. Had they asked what tablets would they buy this number would likely pale in comparison to others. So it is overreaching to say, “its low-price will bring in new customers”.

Netting it out, there is not enough validity in the data to make bold predictions about iPad mini. There are indeed many uncertainties and those are not considered let alone quantified by this study. What you have is someone’s wishful thinking supported with non-scientific sampling and analysis.

For the record. here is my statistical analysis on iPad mini numbers and incremental profit it would drive.

One thought on “Estimating iPad mini Cannibalization

Comments are closed.