dslreports logo
 story category
2-5% Of Your Bill Actually Goes To Bandwidth
Netflix Is Not The Enemy, And The Sky Is Not Falling

Somewhere between 2-5% of your monthly broadband bill actually goes to bandwidth, long-time broadband industry analyst Dave Burstein reminds readers. While there's obviously plenty of additional costs beyond that -- such as support, lobbying, labor and marketing, there's also abundant new revenue streams (advertising via webmail, BVAS, selling your clickstream data, DNS Redirection revenue, charging to get around spam filters, targeted behavioral advertising). Burstein notes that the fact that bandwidth margins are about 90% is important to remember as ISPs try to convince consumers that Netflix bandwidth demand is unmanageable without low caps and high overages, or content company subsidies:

quote:
2 cents to 5 cents per gigabyte. The actual bandwidth cost to a large carrier like Time Warner or AT&T, depending on how you do the accounting. $1/month/customer. The industry standard figure for the cost of bandwidth. Fortunately, Moore's Law has been bringing down the cost per bit of bandwidth at 25-40% per year, allowing the industry to thrive as video drives usage. When fiber is in place, the main cost for additional bandwidth is upgrading routers, switches, wave division multiplexers and the like. They've become much cheaper at a predictable rate. Result: the cost per customer of bandwidth has been about $1/month since 2004 or so. Since broadband prices are $20-$50, that's 2-5% of the price charged.
Burstein's thoughts were apparently triggered by this Washington Post article, one of several lately that paints Netflix as a sort of scary network monster -- one that requires carriers to either get additional subsidies from content companies to handle traffic, or shift to new costly usage-based plans for consumers, lest the entire Internet come crashing to a halt under the immense traffic load. The problem, again, is that these arguments aren't factual. Burstein notes that bandwidth usage growth per customer is actually going down, a fact that ultimately deflated the carrier myth known as the Exaflood, a fear-mongering construct of lobbyists, used to bully and scare politicians into awful government telecom policy.

Assuming such nuanced accounting were even possible from outside an ISP, Burstein's politely avoids going into what percentage of your bill is pure profit, how much goes to exorbitant executive compensation, how much goes into lobbying to erode consumer protections or anti-competitive behavior -- or winds up being used to ban, sue or threaten community broadband efforts. He also doesn't mention how free perks like newsgroups continue to disappear, or the fact that carrier advertising ambitions will open up entirely new revenue windows, creating savings consumers will never see, while becoming more invasive and obnoxious. We're also not addressing ISP savings from layoffs and employment perk reductions.

Burstein does remind readers that while costs drop and revenues rise, actual re-investment back into the network continues to shrink, with companies like AT&T and Verizon cutting their U-Verse and FiOS projects considerably. In short, the broadband sky is not falling, and again -- there's absolutely zero real financial reason (outside of TV revenue protection and greed) to shift the U.S. from a flat-rate to a per byte overage pricing model.

Most recommended from 141 comments



gatorkram
Need for Speed
Premium Member
join:2002-07-22
Winterville, NC
kudos:3
·Suddenlink

gatorkram

Premium Member

Greed

This is the kind of greed, in every industry in this country, that is bringing it all down around us.

And then people sit and wonder, why the US keeps falling futher and further behind in everything you can measure.

Education, Healthcare, Technology....

kapil
The Kapil
join:2000-04-26
Chicago, IL

kapil

Member

To be fair...

...there is no actual way, for a large ISP anyway, to break out the cost of bandwidth from other operating costs. I mean, its not like an end user customer buying a DS1 or DS3 etc. where you're paying a fixed amount for a fixed number of Mbps.

With peering, the bandwidth cost for a network operator is essentially the cost maintaining its own network....the corner of the internet it runs....which is not the same thing as Joe Entrepreneur buying a T1 from, say, AT&T.

There is very little cash changing hands for bandwidth amongst the large backbone providers. Every so often, if the traffic exchange is extremely lop-sided, a peering dispute will break out but providers settle those quickly as it tends to piss off paying customers on both sides.

So, how do you determine the actual cost of bandwidth? Take all of your operating cost and divide it by the number of customers you have? I don't know. I suspect the providers don't either. No one does...and the telcos are using this nebulous idea of bandwidth cost and the lack of understanding amongst the unwashed masses to their advantage....look at their balance sheets....all large end-user ISPs, all of whom by now are also large backbone operators, are very profitable. They are not hurting for money...they just want more of it.

If Comcast can buy NBC...it is clearly not dying because netflix customers are bleeding it dry with excess bandwidth usage.

AT&T and Verizon aren't content with charging you outrageous amounts for a lousy DSL connection and then raping their enterprise customers by charging them obscene fees for connectivity....they want those enterprise customers to pay for providing you and me that lousy DSL connection while they also bill us for it. This isn't "my side" of the story...that is the entire story. There is no other way to explain what they are trying to pull.
BlueC
join:2009-11-26
Minneapolis, MN
kudos:1
·Xfinity
·Integra Telecom

BlueC

Member

Very True

It's true, in the core, bandwidth is more likely very cheap for large providers.

The big cost comes from the last mile delivery. There's a lot of equipment to maintain and customers to support. It still doesn't give providers a reason to force content providers to pay in to cover the costs (especially when a lot of larger ISPs have settlement-free peering in the core), the residential ISPs have a last mile to maintain and they will have an obligation to do so regardless of what content is available.

There has to be some sort of ROI on all the cabling/fiber that is installed in order to make future upgrades. That's obviously a big cost, but it will always be present.

How about ..