Reality Check (Special Edition): Why AT&T changed their data pricing

Editor’s Note: Welcome to our weekly Reality Check column. We’ve gathered a group of visionaries and veterans in the mobile industry to give their insights into the marketplace.
I am taking a brief respite from my start-up focus to weigh in on the AT&T announcement this week that they are ending unlimited plans. This just as video is introduced on the iPhone 4.0 and as the iPad starts to take off.
Many others have entered into emotional rants (the Motley Fool is one example) – here, we’ll save our rants for the FCC. Let’s look at the economics of providing data that AT&T is facing:
1. AT&T has real network costs to carry data. More spectrum means more electronics, presumably on more towers.
2. More towers and more electronics drives more bandwidth from the tower to AT&T’s local IP infrastructure.
3. There are some increased costs to the IP network from additional traffic. These costs are marginal.
4. iPhones and Blackberries carry pretty hefty subsidies. Yes, that was the reason for requiring a data plan with all of these phones in the first place, but $99 has become the new entry price for the iPhone with $199 being the standard price (Blackberry phones have long been free for individuals who are willing to switch from Verizon to AT&T).
Let’s begin with some very basic assumptions. First, let’s assume that at 2 gigabytes of traffic at $25 makes AT&T money – at 2 GB, let’s assume they make a marginal 10% operating income before depreciation and amortization. That would imply a cost per MB of roughly 1.25 cents. Yes, if this logic holds (costs do not change even though buckets on pricing plans do), the $15 for 200 megabyte plan carries a cost of $2.50 and yields an 83% OIBDA. The $15 pricing plan has become, in the words of a telecom veteran, the “casual caller” plan of the smartphone generation. An innocent price point with a whopping margin.
So, using this logic, 1.25 cents per MB is AT&T’s cost of the current network without high speed upgrades. We know that Clearwire is experiencing an average of 7 GB of usage and is at least breaking even at $40 per month. That implies a cost structure of .6 cents per MB. A 50% better cost structure might be aggressive for 4G, so let’s taper that back to 25% improvement to .94 cents per MB. After all, AT&T owns the access pipe to the tower for many of their covered pops, right?
At .94 cents per MB, 2 GB costs AT&T $18.75 and would yield a 25% OIBDA margin (yes, the $15 plan margin goes to 88%). And the additional GB for $10 looks about right at this cost structure as well (assuming there is some breakage on the 1 GB usage). Just to bookend this, 5 GB (the current cap on an wireless modem) would cost AT&T $46.88 which implies a 22% OIBDA margin.
Bottom line: AT&T saw a negative OIBDA in their future if they didn’t take quick action. That 3rd GB of usage, which is easily driven by the Pandora One app on the iPad or in my car on a summer vacation, was going to come soon. The same engineers who underestimated the growth of data from iPhone applications are now predicting a bandwidth explosion with the new iPhone and iPad. There’s no two ways about it, and, despite the outcries, these are real costs.
How will Verizon respond? Well, at the beginning of the year, Verizon introduced a mandatory $10 pricing plan for 25 MB of data (yes, at the costs above, this would yield a 97% OIBDA. The first reflection of a full quarter of this mandate on pricing will be seen in 2Q earnings). They have unlimited data at $30 now and, with AT&T’s pricing move, likely could take this to $40 and the higher end of the market would move en masse to Android devices, provided that applications developers create Android versions of their Apple products (I talked to two developers yesterday who are working on video iPhone applications – they anticipate a slowdown in iPhone app development as a result of yesterday’s announcement).
Verizon will likely double bandwidth to 50 MB for $10 which is still a very healthy margin. They should triple or quadruple their applications focus and get two or three of the larger developers to work on differentiating 4G apps (perhaps with some local storage linkage – you can cache a lot of Pandora or last.fm in a 32 GB SD card).
AT&T needed to make this move. The economics shown above are close to reality (although, with the access costs removed to reflect inter-company transactions, they could have squeezed in that 3rd GB and still made some money). However, it’s not likely that Verizon will follow, except to roll out 4G faster. T-Mobile USA already played their card with user-specific rate adaptation, and Sprint Nextel’s marketing message would be significantly tainted if they came off of unlimited right now.
Rethink Possible. Yep, they did it. And it’s the right move.
Welcome your comments.
Jim Patterson is CEO & co-founder of Mobile Symmetry, a start-up created for carriers to solve the problems of an increasingly mobile-only society. He was most recently President – Wholesale Services for Sprint and has a career that spans over eighteen years in telecom and technology. He welcomes your comments at jim@mobilesymmetry.com.