Calling the Broadband Bluff

Publish date:
Updated on

Here’s another way to frame the debate over usage-based billing: After years of broadband providers marketing super-fast, unlimited Internet access at a fixed monthly price, customers are calling their bluff.

Broadband networks were originally designed with oversubscription ratios (the total theoretical peak capacity available to all subscribers vs. the actual amount of bandwidth available) of 40:1 or even higher. That worked because not everyone was online at the same time, so the average utilization of the network was far below the maximum possible peak. Meanwhile, people were mainly using broadband to check e-mail and look at Web pages: that generates bursts of traffic, not sustained flows.

Now, first with the rise of peer-to-peer networking and then the explosion of online video, the standard broadband usage profile is destroying the assumptions the networks were built on.

Video is now estimated to be 27% of all Internet traffic (per Sandvine’s 2009 survey of 20 ISPs), surpassing peer-to-peer — formerly the top bête noire of network providers. Online video viewing is increasing in the double digits year-to-year. Expectations are that the bandwidth usage will only increase; Cisco forecasts a fivefold increase in global Internet traffic from 2008-2013. The rate of wireline Internet growth is 40%-50% per year, according to the University of Minnesota’s MINTS project — somewhat slower than in past years, but still considerable (thanks to Karl Bode at BroadbandReports for the link). It sure isn’t leveling off.

The model for the unlimited-access, eat-all-you-want network worked when people were running up to the salad bar and taking a broccoli floret or a carrot stick every half an hour. Now they’re noshing for hours at a time and the crowds are only going to get thicker: time to build a bigger salad bar.

Is it possible there will not be a significant extra cost associated with accommodating additional demand, as some (including Bode) argue? Pointing to recent quarterly filings is irrelevant, as we’re talking about longer-horizon trends. If you expect incremental, single-digit-per-year growth in Internet usage, I’d agree the costs are controllable. But with 5X growth in five years, you’re going to have to bottlenecks unless you add downstream and/or upstream capacity.

It will be a marketing challenge, clearly, to introduce the idea that if you use more of a broadband service you should pay more — after the industry has successfully spent years pushing the flat-rate deal.

Time Warner Cable’s attempt to do this blew up in its face. The New York Times, for example, noted that under TWC’s proposed cap-and-overage pricing plan, customers would have to pay $150 per month for unlimited usage — more than three times what they pay for unlimited use today. The episode was a “a bit of a debacle,” TWC CEO Glenn Britt acknowledged.

There’s no way to sugarcoat the real issue: that the old oversubscription models have been shattered, and that an all-you-can-eat plan is only viable as long as you don’t eat too much.

So far, in the U.S. anyway, the suggestion that consumption-based pricing is the fairest way to offer broadband has backfired, as TWC’s lead balloon was attacked by parties ranging from U.S. Sen. Chuck Schumer to consumer activists and tech bloggers.

A lesson from Canada, where Rogers, Cogeco and Vidéotron have introduced consumption-based billing, as summed up by Multichannel News contributor Leslie Ellis: “A slow, customer-focused approach to additional usage billing works to gently unclog the network.”

Rogers, for example, spent two years implementing the plan, which included extensive customer outreach, online tools and notifications. Cogeco, according to BroadbandReports, has had issues with its own usage-based billing plan, including inaccurate bandwidth meters.

But Rogers has demonstrated it’s possible to introduce the concept of metered billing, if it’s done correctly. BendBroadband in Oregon employs overage billing for usage beyond 100 GB per month. The question for the other U.S. ISPs is when and how they’ll figure out how to do this.