Why AT&T is right about usage-based pricing for 3G data

It’s not that often that I come out in support of AT&T or any cellular carrier, given their predatory business practices and continually honed techniques to trap users into paying more than is appropriate. But AT&T is right to charge smartphone users based on bandwidth consumed, and I hope the other carriers follow suit.

Simply put, unlimited wireless broadband is a bad idea. The capacity of the airwaves is not unlimited, and as mobile devices get more and more desktop PC-like capabilities, such as video streaming and VoIP telephony, bandwidth usage can only soar. Any AT&T user in a major city knows what that leads to: poor connectivity and slow speeds.

With wireless bandwidth a limited resource, it makes sense to pay on a usage basis, as we do for water, electricity, and gas. Surprisingly, AT&T’s new pricing looks reasonable: $15 per month for 200MB of cellular data, with $15 per additional 200MB, or $25 per month for 2GB of cellular data, with $10 per additional 1GB. Of course, if you use only 1K of that additional bandwidth before the month expires, too bad — you still pay the extra $15 or $10. That should avoid the nasty surprise of overage charges, and we can all be happy that AT&T’s pricing is not based on actual usage but instead on baskets of usage (the same way voice minutes are typically assessed).

No one should be surprised by this change, given AT&T’s tiered, usage-based pricing scheme for the iPad in late April. Such a pricing scheme is not unique to AT&T; other carriers abroad use it as well. And of course, most broadband providers used tiered pricing on their landline services, though they typically price based on speed rather than bits. It pretty much ends up having the same effect: Those who use more pay more.

Some people are objecting to the end of unlimited 3G data bandwidth, even though we’ve never really had it; the carriers have all had fine print in their agreements letting them cancel your account for undefined “excessive” usage. But it is a reasonable response to wireless’s bandwidth limitations.

If I were an AT&T competitor, I would introduce similar pricing, though I’d probably raise the first tier’s cap to 250MB. I’d definitely offer a limited form of rollover, so the overage is charged only if the average usage of, say, the previous two months exceeded the cap. Alternately, maybe the overage fee is refunded of the next month’s usage is less by at least as much as the current month’s is over the limit. But the principle is sound.

However, if anyone thinks that AT&T’s usage-based pricing will fix the poor coverage that people in San Francisco, New York, and other cities typically experience on their iPhones and BlackBerrys, they’d be wrong. AT&T claims its new pricing covers 98 percent of all users’ current data usage patterns. That means that at best, 2 percent of customers’ usage will decline — the worst 2 percent, to be sure — as they end up paying more. That reduction won’t free up enough bandwidth for the rest of us.

AT&T still has to beef up its 3G networks — both the in-air spectrum and the backhaul to the Internet — to handle its current customer base, never mind all the new users now buying iPads as if they were candy, all the new owners of iPhones that will result after next week’s widely predicted announcement of a next-generation iPhone model, and all the Android OS devices projected to be sold as more Android devices ship this summer and fall. We may even see demand driven by the forthcoming Windows Phone 7 around Christmas. Mobile usage is going nowhere but up, as is the sophistication of the services available to them.

AT&T and other carriers need to figure out something better for international 3G data roaming than the exorbitant fees now the norm. So, AT&T, please follow up this new tiered, usage-based plan with with a reasonable international roaming scheme for others to copy. Your customers may actually begin to start liking you again.

This article, “Why AT&T is right about usage-based pricing for 3G data,” was originally published at InfoWorld.com.