Why is this page text-only?
 

« Rural Broadband Won't Be Solved By Profit-Maximizing Means | Main | We Should Count All Bandwidth Equally - Up And Down »

December 9, 2009 11:17 AM

Bandwidth Hogs Are Real, They're Just Misunderstood

The term "bandwidth hogs" has been much debated in broadband policy circles over the last couple of years. Most recently, Benoit Felten, someone who I respect tremendously as a clear-thinking industry expert, has raised the question of whether or not bandwidth hogs are actually a myth.

I wanted to take a moment today to make the case that bandwidth hogs do exist, and they in fact represent the heart of the evolutionary transitional period that broadband providers face.

It used to be that no one used their broadband connections much. Broadband providers counted on this minimal usage so they could oversubscribe their networks. They also offered unlimited service to gain another competitive edge over dialup Internet access. And this all worked fine so long as no one was consuming that much bandwidth.

But over the last few years the demands on these networks have risen dramatically. This increase stems from the dual factors of power users having more ways to use more bandwidth than ever--like P2P video sharing, Hulu and YouTube, hosted applications, and the use of broadband to enable telecommuting--and more people finding reasons and ways to become power users.

This growth in demand is straining traditional broadband business models, forcing operators to invest in increasing capacity without necessarily adding any new revenue to offset this cost.

This reality is what's leading operators to explore options like metered bandwidth, so they can recapture their investments and drive additional profits from their heaviest users. This is also what's leading operators to use things like traffic shaping and application blocking.

While I disagree with the notion that bandwidth hogs don't exist or aren't a problem, Felton's more specific question of whether or not it's a specific small group of people who are stressing the network is still worth more detailed consideration.

But regardless of this, there's no denying that there are a small group of people who are using broadband networks way more than the average person, and that the heaviest users are straining the capacity and affecting the performance of broadband networks.

That brings us to the real question at the heart of this matter: is it the users' fault?

I've long recoiled at the label "bandwidth hogs" as it implies that users are being overly greedy, that they're doing something they shouldn't when really what they are are power users, or people who are taking full advantage of what the Internet has to offer today.

Put another way, I think so-called "bandwidth hogs" actually represent the leading edge of the next generation of broadband usage and demand. They're bringing to light how in the Internet of the future, users won't just be consumers of data, they'll be producers, and they'll be using applications that cause their computers to act like servers.

To support this evolution, we need broadband networks that can support these trends in usage, that will encourage rather than dissuade this growing demand. Because it shouldn't be a bad thing when people want to use more of a service.

At the same time, I am respectful of the fact that we can't expect private operators to continually invest in upgrading their network capacity without having some way of recouping those costs. Right now, the only incentive to invest is competition. But without a competitor forcing their hand, or a way to drive additional revenue off of offering additional capacity, then they have no reason in invest in expanding and upgrading their networks.

This raises an interesting conundrum that I think can be highlighted by using the analogy of a buffet. Who's fault is it that the crab legs keep running out on the broadband buffet? Is it the hungry customers for eating too much, or the buffet owners for not refilling the bin fast enough?

It may be easy to blame the buffet owners in this scenario, but what if their customers' appetites went up tenfold. What are they supposed to do? Supplying this increase in demand without realizing a subsequent increase in revenue would likely mean putting the buffet out of business. And yet if a customer's paying for an all-you-can-eat buffet then they should be allowed to eat all that they can.

Resolving this tension is a key to unlocking a better tomorrow for American broadband. My hope is that we can move past demonizing power users as bandwidth hogs and network operators as bandwidth misers to instead focus on how we can spur the deployment of networks with enough capacity to supply the demand that's there today and that will continue to grow on into the future.

Because if we don't get this resolved, then we're going to end up with barren broadband buffets and hungry bandwidth hogs at a time when we should be living in a land of plenty with Americans free to consume all the bandwidth the desire.

Del.icio.us Digg Yahoo! My Web Seed Newsvine reddit Technorati

TrackBack

TrackBack URL for this entry:
http://www.app-rising.com/cgi-bin/mt/mt-tb.cgi/1497

Comments (4)

"My hope is that we can move past demonizing power users as bandwidth hogs and network operators as bandwidth misers to instead focus on how we can spur the deployment of networks with enough capacity to supply the demand that's there today and that will continue to grow on into the future."

Indeed. This and the so-called net neutrality debate is based on a false premise of bandwidth poverty -- a contrivance of the incumbent telco/cable duopoly -- which is a prescription for a race to the bottom when it comes to deploying advanced telecommunications infrastructure in the U.S. and elsewhere.

Posted by Fred Pilot on December 9, 2009 9:34 PM

Good point -- one man's hog is another man's power user.

Benoit hopes to get data from ISPs to test his hypothesis that heavy users do not matter. That being said, would we/he trust the data? Do we trust ISPs?

ATT is spreading the word that 3% of their 3G users account for 40% of traffic. (See, for example,
http://www.wired.com/epicenter/2009/12/iphone-caps/). I guess they will use that as a PR excuse to cap traffic or raise prices.

Do you believe them? How was the data gathered? Where? What time of day? What difference did it make to the service the other 97% received? Did they just make the figure up?

Posted by Larry Press on December 10, 2009 9:36 AM

"This growth in demand is straining traditional broadband business models, forcing operators to invest in increasing capacity without necessarily adding any new revenue to offset this cost. This reality is what's leading operators to explore options like metered bandwidth..."

Geoff, this kind of fact-free "third-way" "analysis" is why you are such a useful tool for industry, and why no one of importance takes you seriously.

Have you bothered to actually learn and understand about the realities of network congestion, and ISP's cost structure? Here are a few facts that completely undermine your basic premise:

1) Despite years of crowing about congestion, and ample opportunity to show any evidence of the degree and nature of this congestion, no ISP has. Telcos like Verizon and Qwest have flatly said it is not a problem on their networks.

2) Even if we assume your congestion premise is correct, the notion that managing such congestion is forcing investment that cannot be recovered in the normal fashion of running a business is asinine. DOCSIS 3.0 upgrades cost $8 per customer, that's it. Again, there is no evidence that the telcos have a congestion problem. And the wireless congestion problem is vastly overstated; even still, wireless revenues and profits are growing rapidly, even as relative investment levels decline.

3) Every single ISP, be they wireline or mobile wireless have profits that are up, while cap-ex spending as a percentage of revenues are down. And their cap-ex plans over the next 5-years show a future decline.

4) "Metering" (as you call it) is purely driven by Wall St. myopic profit hunger. The ISPs are considering true metering, they are considering overage charging. The ISPs have no desire to charge fairly, or else the bottom 50 percent of users that use just 5 percent of bandwidth would see a huge drop in their bill.

Bottom line: ISPs are investing somewhat, and revenues are increasing along with profits. This is the nature of the ISP business. Your shrieking about congestion and the inability to recover investments are plainly not supported by the facts.

Please, stop trying so hard to offer up this third-way garbage. It is killing America.

Posted by Jane Crews on December 10, 2009 9:40 AM

DOCSIS 3.0 upgrades cost $8 per customer, that's it.



So in a community of 10,000 a DOCSIS 2.0 provider can upgrade to 3.0 for $80,000??? that's the cheapest DOCSIS 3.0 system I've ever heard of. Just because your community equated to $8/customer doesn't mean they all do. Do some more research.



the wireless congestion problem is vastly overstated



Sorry, but I've actually run an ISP, monitored it's traffic and received problem calls when some user on the shared traffic network decided to bit-torrent the latest 4 GB iso file of their favorite linux distribution.... OK it was me. I've also seen at home on a largely over-subscribed neighborhood POP when my gamer neighbors held a lan party and my connection suffered.


Sure I'll agree that as a whole, most providers do not see their internet connection suffer. But at the premise-based POPs there is definite signs of congestion during certain events.

Posted by Bryan F. on December 10, 2009 11:44 AM

Post a comment

(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)