This morning, I woke up to hear an American Public Media report on how data traffic on U.S. mobile phone networks is doubling every year. Steve Henn of Marketplace was interviewing Chetan Sharma about how data traffic on cell phones will hit an Exabyte by the end of this year. (Henn mentioned that an Exabyte is more than 18,000 times the amount of information contained in all the books ever written, a figure that will wake up pretty much anybody.) Sharma commented that it's not just phones hooking into the mobile phone network, but anything that can be broadband enabled.

This seemingly innocent but dramatic bit of information set the tone for the day. By the time I reached my desk, I had received an email from someone clarifying an earlier email about wireless data growth. (Of course, I had not yet seen the earlier email, but the words "clarification" or "correction" tend to jump out at journalists.)

In the second email, Sanford Bernstein senior analyst Craig Moffett explained that his earlier note about AT&T's mobile network traffic growth of 3,000 percent had omitted the time frame. That growth occurred over the past three years. AT&T's corresponding revenue growth was 166 percent over the same time frame, he added.

All of this led me to track down the original message, which then led to a review of some of Bernstein's latest research.  Much of it is centered around consumers' attitudes toward usage-based pricing, but it also hinges on a central question, which is whether revenue will grow faster or slower than capital spending. Now, it's quite clear from quarterly financial reports that operators are raking in a ton of money from wireless services. However, you have to factor in how much operators are spending, too. Moffett & Company say wireless capital spending at AT&T over the past nine months is up 55 percent, while overall wireless revenue grew by 10.5 percent. Of course, one assumes you will get reimbursed later for your network investments, and that's a risk you take, but there's quite a challenge in here too.

After reviewing the Bernstein research, I decided that I haven't appreciated the usage-based pricing conundrum nearly as much as I should. Of course, I was reminded just a week ago what a conundrum this is by my colleague Andrew Berg's recent story about data pricing. I was reminded then and again today how consumers need to be given some kind of clear way to understand their data usage. (A shout out to T-Mobile USA, which tells customers their service may suffer if they go over 5GB of data in a month – it doesn't shut off service or charge extra, but customers may see a difference in their service quality.)

Back to the Bernstein study, it appears people are very much interested in keeping their "unlimited" data plans even if it means they're overpaying. That's despite the fact that several years ago, I recall a lot of folks were complaining that nothing was really "unlimited." At some point, data hogs get tagged and, more or less, penalized. But that's another story. Or is it? 

As Moffett points out, carriers are faced with the challenge of increasing adoption of mobile data but limiting consumption (and I don't think 4G is going to save the day.) Even with usage-based pricing, it's not that easy. Bernstein's study found that customers don't like or even really understand usage-based pricing. A lot of industry folks have pointed out that what we all need are some clear-cut parameters for consumers to understand their consumption and the price for that consumption. Like 10 YouTube videos a week equals XYZ in service costs. But with so many variables and consumers' usage varying as much as their individual personalities, how can you come up with a meaningful chart?  

With the dawning of the Exabyte, it's a fine time to come up with some answers. That just might require an Exabyte of brain cells to accomplish.