[Brad Adrian]

Both Sven and Michael make some very important and valid points, not the least of which is how convoluted statistical analysis can be.

Perhaps one of the most important points is to remember that when it comes to measuring loyalty, the mobile phone market has a very complex bag of influencers. Because things like contracts, device availability and corporate requirements can all impact how likely a customer is to continue patronizing a particular vendor, looking at churn alone is a very dicey way to try to understand loyalty.

And, it is of no little consequence that this data was collected prior to Verizon's iPhone announcement and the availability of that device through Verizon. Those factors alone can have a huge impact on how people would respond to this survey. For example, two months ago I might have stated a very small likelihood to switch carriers because I'm really attached to my iPhone; today, though, I'm much more likely to switch to Verizon if it means I can still have an iPhone.

Food for thought...

[Said the Market Researcher...!]