Back in early March, we covered
news that AT&T was instituting broadband meters after years of providing unlimited service. For standard DSL customers, 'unlimited' now means 150GB per month, while U-verse subscribers have a higher cap of 250GB/s. AT&T
claims that these caps are necessary due to the fact that a small number of customers consume a huge amount of bandwidth.
"AT&T has experienced a dramatic increase in the amount of data that is sent and received over its wireline broadband networks. This dramatic increase is driven primarily by a small fraction of our customers. In fact, the top 2% of customers use about 20% of the total capacity on our network. A single high traffic user can utilize the same amount of data capacity as 19 typical households. Lopsided usage patterns can cause congestion at certain points in the network, which can slow Internet speeds and interfere with other customers' access to and use of the network."
AT&T provides a web utility for viewing your additional usage; a screenshot of the author's usage for the month of April is shown below:
AT&T intends to pursue a sort of three-strikes policy.
In the following months, we will send you additional notices each month when your usage exceeds 65%, 90% and 100% of your monthly usage allowance. If you exceed your monthly allowance a second time, AT&T will send you a notice advising you that the next time you exceed your allowance - the third time - you will be billed $10 for each 50 GB of data over your allowance.
This Isn't About Network Traffic
We're skeptical that AT&T's new bandwidth caps have anything to do with traffic shaping or maintaining QOS. Bandwidth caps are a poor way to deal with network congestion—they're unpopular with customers, and they don't address the fact that bandwidth loads occur at certain peak hours. There's no proof that capping monthly bandwidth leads to any sort of change in a person or household's usage at peak.
If AT&T actually wanted to save bandwidth, they'd be throttling certain kinds of traffic and using packet shaping. Granted, this can backfire with customers as well--Comcast
took a lot of heat for it once upon a time—but at least throttling/packet shaping attempts to directly treat the issue.
The gap between U-Verse and regular customers is also frustrating, given that there are those of us who live in places where U-Verse isn't available (or those who don't want it). The 100GB difference is a none-too-subtle shove in the direction of signing up for AT&T's own rich multimedia services. Such caps also impact the attractiveness of services like GoogleTV—bandwidth hungry families are less likely to buy such products if they come with a $20-$30 surcharge from their Internet provider.
At least one of us at Hot Hardware will be leaving AT&T as a result of this change, though we'll be surprised if consumers in general raise a ruckus.