Companies Trying To Restrict Usage Isn't Evidence Of Any Broadband Crunch

from the anecdotes,-not-data dept

This is fascinating. A few weeks back, we noted that a press report in the UK was repeating Nemertes' questionable research on how there was apparently some big bandwidth crunch on the way that would harm the internet. However, someone from Nemertes quickly showed up to claim that the original reporter took their (old) report out of context, and Nemertes wasn't saying what was claimed. However, Zubin Madon, alerts us to the fact that Nemertes' Johna Till Johnson (oddly, without any clear disclaimer that she works for Nemertes) has published a piece at ComputerWorld claiming that, indeed, the internet is running out of bandwidth and that "the Internet sky really is falling." Not only that, but that it's happening even faster than their original estimates?

The proof? Well, there isn't any. She takes two data points (that have nothing to do with actual bandwidth) and extrapolates that we're running out of bandwidth. First, she points out that YouTube was discontinuing servicing certain geographies due to "lack of access capacity." First of all, I'd be interested in some more details on this, because I don't recall seeing the news, and a quick look around isn't turning up much on a story that I would have imagined would have generated a ton of press. Even if it's true, it doesn't seem to support Johnson/Nemertes' point. If YouTube really were pulling out of certain regions due to a lack of access, that would just mean the company is focusing on regions where there is more bandwidth, not that bandwidth is somehow running out. It just means Google is focusing on markets where there's a larger market.

The second data point is Time Warner Cable's weak attempt to try to force metered broadband. However, as the actual research has shown, Time Warner's actions had nothing to do with a bandwidth crunch, and everything to do with simply trying to abuse a market monopoly position to squeeze more money out of customers, even as its own costs were decreasing.

So it's again difficult to take Nemertes' research seriously when these examples are the best it can roll out in support of its position -- especially when Nemertes seemed to step away from its own supposed position just days before Johnson's column.


Reader Comments (rss)

(Flattened / Threaded)

  1.  
    identicon
    Anonymous Coward, May 13th, 2009 @ 4:26am

    The monopolies that offer internet access have done a great job selling the Malthusian threat of running out of internet. I think that your average person hears that and just believes it.

    Meanwhile, Time Warner isn't willing to upgrade their network, and even attempt to enact legislation to prevent competition.

     

    reply to this | link to this | view in thread ]

  2.  
    identicon
    Matt T., May 13th, 2009 @ 4:30am

    Re:

    If the threat is really Malthusian, wouldn't everyone left after the bandwidth war get a T1 connection for $10 a month?

     

    reply to this | link to this | view in thread ]

  3.  
    identicon
    BrooksT, May 13th, 2009 @ 4:48am

    Lack of access capacity?

    Sure sounds like she heard someone say something about *excess* capacity and misunderstood it. The point may be the same and equally valid/invalid, but I'm suspicious of someone who seems to be touching on common economic terms, but getting them wrong. It's like reading an article about from an aerospace analyst who talks about "commercial hairliners" -- regardless of the validity of the actual claim, getting that wrong raises doubts about credibility.

     

    reply to this | link to this | view in thread ]

  4.  
    identicon
    Anonymous Coward, May 13th, 2009 @ 5:44am

    Re: Re:

    Yes.

     

    reply to this | link to this | view in thread ]

  5.  
    identicon
    Steven, May 13th, 2009 @ 6:31am

    Pulling out.

    I vaguely remember reading about poorer countries not being good for youtube. The argument basically went: streaming a ton of video is expensive, Google supports it through advertising, people in poorer countries don't buy enough to pull in the advertising revenue, and these poorer countries tend to actually stream more video than others on average (per capita).

    If I remember correctly there was speculation that video streaming sites, such as youtube, would block those countries (I think they were specifically talking about African countries), but I think it was just speculation.

    None of this had to do with a bandwidth problem.

     

    reply to this | link to this | view in thread ]

  6.  
    identicon
    johnos, May 13th, 2009 @ 6:39am

    Youtube source

    I'm guessing she is referring this story in the New York Times from April 26.
    The key quote:
    Web companies that rely on advertising are enjoying some of their most vibrant growth in developing countries. But those are also the same places where it can be the most expensive to operate, since Web companies often need more servers to make content available to parts of the world with limited bandwidth. And in those countries, online display advertising is least likely to translate into results.
    The issue is not lack of bandwidth. Its the expense of bandwidth coupled with low ad revenues. No matter how you slice it, insufficient bandwidth is not the problem.

     

    reply to this | link to this | view in thread ]

  7.  
    identicon
    Mike Raphone, May 13th, 2009 @ 7:31am

    Bandwidth CAPS

    The Broadband ISP are proving that you do not necessarily have to go to a Delicatessen to get Bologna!

     

    reply to this | link to this | view in thread ]

  8.  
    identicon
    Anonymous Coward, May 13th, 2009 @ 7:32am

    You're confusing two issues

    Johna's article about The Sky was fundamentally about the problems with the TCP/IP protocol suite, which is senescent and showing serious cracks. The IETF has fiddled for 15 years and come up with IPv6, which makes it worse -- the real problem is route proliferation in BGP, which is invisible to end users. So she (and I, and some others) are promoting the Pouzin Society's alternative approach, PNA, which is the first novel thinking about network protocols in decades.

    The capacity problem is a different one. As others have noted, bandwidth isn't really free, we just charge as if it were, and that isn't profitable for high-bandwidth low-revenue applications in high-cost areas. (YouTube in Africa, for instance.) The Internet business model is not mature yet. Whinin that you want more, free, forever, doesn't make it so.

     

    reply to this | link to this | view in thread ]

  9.  
    identicon
    Anonymous Coward, May 13th, 2009 @ 7:48am

    Re: You're confusing two issues

    that isn't profitable for high-bandwidth low-revenue applications in high-cost areas.

    You don't sell ice in Alaska, you don't sell ads in Africa. I don't see how either has anything to do with selling ice in Egypt or Ads in Japan.

     

    reply to this | link to this | view in thread ]

  10.  
    identicon
    Chronno S. Trigger, May 13th, 2009 @ 8:06am

    Re: Youtube source

    From that little clip it wouldn't be the expense of bandwidth at all. The bandwidth limitation is between their country and the Google main servers so they have to spend all that money to get servers to their country where the bandwidth seems good enough for Youtube. And due to the fact that people there wouldn't buy enough to recuperate the servers threw ad revenue it becomes a problem.

    Why one would want to cut off those countries entirely and completely eliminate a potential revenue stream is another question entirely. Just because one doesn't have servers in their country doesn't mean that no one will deal with the international download speeds (and potentially buy something from those ads). It's not like Google has to pay for bandwidth over there if they don't have servers over there. And it wasn't Google who said it, it was Joost, and Joost didn't close it's borders, Veoh did.

     

    reply to this | link to this | view in thread ]

  11.  
    identicon
    Anonymous Coward, May 13th, 2009 @ 8:23am

    There isn't an Internet bandwidth issue but there is a to the curb bandwidth issue, at least for cable.

    Do you know what percentage of bandwidth 5% of a companies customers use up? It is over 80%. That means that 5% of a companies customers consume over 80% of their product. If a restaurant worked that way, you could go into McDonalds and order everything on the menu and pay the same as someone who only ordered a Happy Meal.

    Is it the same? Kind of. I am a cable customer and my connection is slower because some of my neighbors consume large amounts of bandwidth, so yes, their usage affects me. My usage isn't as high, so why shouldn't they pay more? Why should my rates go up because they have to improve their network when my usage isn't the one who causes the need for improved service?

     

    reply to this | link to this | view in thread ]

  12.  
    identicon
    Brooks, May 13th, 2009 @ 9:08am

    Re:

    Bandwidth and happy meals are "kind of" the same? I guess if you start with that, your argument makes sense, but if you accept that intangible time share services are different than tangible consumables, it breaks down.

    Almost all bandwidth is measured and billed at 95th %ile usage, not bits transferred. How do you know your peak usage isn't as high as your neighbors? How do you know that they don't already pay more (I pay $70/month for cable internet; my cable company's lower speed plans start at $30/month).

    And, as has already been said, nothing about increased rates has anything to do with network improvements; last mile ISP's have plenty of upstream bandwidth, and metering is strictly about using their monopoly power to squeeze more profits out of customers.

     

    reply to this | link to this | view in thread ]

  13.  
    identicon
    E, May 13th, 2009 @ 9:13am

    Re: You're confusing two issues

    It may not be the entire article, but the second paragraph is all about a lack of bandwidth, and has back-to-back misleading statements (you could even call them fabrications). In fact, you even make the correct point about Youtube limiting access to "certain geographies" due to it being bad business, and not "due to [condescending 'ahem'] lack of access capacity." So no, that's not the primary subject of the article, but she still throws all that BS out there at one point.

    P.S. The problem is with IPv4, not the entire TCP/IP protocol suite. Leave my bread 'n' butter alone. All we need is a government-backed move to IPv6 and better routing algorithms (yeah, I know, easier said than done, but it's still the most reasonable solution).

     

    reply to this | link to this | view in thread ]

  14.  
    identicon
    E, May 13th, 2009 @ 9:32am

    Re: Re: Youtube source

    Why one would want to cut off those countries entirely and completely eliminate a potential revenue stream is another question entirely. Just because one doesn't have servers in their country doesn't mean that no one will deal with the international download speeds (and potentially buy something from those ads). It's not like Google has to pay for bandwidth over there if they don't have servers over there.

    A swing and a miss! It costs Youtube money every time someone streams a video, no matter where the viewer or servers are located, which is money that they need to more than make up for in advertising sales to be profitable. If they're making too little money in a certain large area from ad sales and it doesn't seem to be a trend likely to change, it makes plenty of sense to close up shop in those areas.

    I'm not saying that just because it makes sense it's necessarily true, I'm just saying that it would make sense.

     

    reply to this | link to this | view in thread ]

  15.  
    identicon
    Chronno S. Trigger, May 13th, 2009 @ 9:48am

    Re:

    First, hamburgers != electrons. Hamburgers cost money to get the parts and man hours to put together and move. Electrons travel easily down copper wires and require little money to move. And electricity is considered a fixed cost so it douse not add into the price calculation (translation: those 1s and 0s equal $0.00 to produce). I would have to pay for the parts to make a hamburger myself but I don't pay one cent to transfer a file from one PC to another.

    Second, you're looking at the wrong numbers. When they say that 5% of the people use 80% of the bandwidth they mean 80% of the bandwidth used. The used bandwidth could be only 50% of their max, so the 5% of users are only using 40% of the max. These numbers are probably wrong, it's probably more like 20% or 10%. The used vs. max numbers are the important ones and they will probably never be released.

    And if your ISP douse truly have problems (FiOS doesn't nor douse Comcast) and they chose to do these stupid moves instead of just upgrading their network (it's not that hard or expensive) then that would seem to be their problem with not staying up with technology (and they're making yours).

     

    reply to this | link to this | view in thread ]

  16.  
    identicon
    Chronno S. Trigger, May 13th, 2009 @ 9:52am

    Re: Re: Re: Youtube source

    "It costs Youtube money every time someone streams a video"

    Where? My company doesn't pay for every single file that is downloaded from our website. How is Google different?

     

    reply to this | link to this | view in thread ]

  17.  
    identicon
    E, May 13th, 2009 @ 10:20am

    Re: Re: Re: Re: Youtube source

    Fair enough, we can go for semantics.
    It costs Google a large amount of money to stream a large amount of video. They may have ample amounts of bandwidth and server time to the point that a single video is not of great importance, but streaming millions of Youtube videos is certainly a sizable cost to them.

     

    reply to this | link to this | view in thread ]

  18.  
    identicon
    Chronno S. Trigger, May 13th, 2009 @ 10:31am

    Re: Re: Re: Re: Re: Youtube source

    So they pay for bandwidth witch is a fixed monthly cost. They would be paying that price for that connection weather they streamed one million videos or two. I can't see Google paying per bit (especially since we as a company don't).

    Hay Mike, douse it cost more for Techdirt to put up say 10 articles a day instead of say 5? Or to have 700,000 readers instead of 100,000? Am I costing you more money just by posting this response?

     

    reply to this | link to this | view in thread ]

  19.  
    identicon
    E, May 13th, 2009 @ 10:59am

    Re: Re: Re: Re: Re: Re: Youtube source

    Ok, let me get this straight: you honestly don't believe that Google would save money on bandwidth costs by shutting out millions of users? (ignoring, for the moment, whether or not this is a good business strategy)

     

    reply to this | link to this | view in thread ]

  20.  
    identicon
    Anonymous Coward, May 13th, 2009 @ 11:17am

    Chronno, so why does my service slow down when the kids get out of school?

     

    reply to this | link to this | view in thread ]

  21.  
    identicon
    Josh, May 13th, 2009 @ 11:43am

    Re:

    This is due to the ineffecient way that cable companies provide you with bandwith. Instead of giving you a dedicated link, such as DSL does, they give you a shared link. You share this link internally with your TV signal, even if you don't use TV it's on the same wire, and with anyone else in your house, and you also share this link with everyone else in your area. So yes. When the kids get home from school you will see a slowdown, you will also see a slowdown if you live in an area that is full of business people, because they all get on the internet or turn on the TV when they get home from work as well.

    So to say that it's the heavy users that are causing you to experience a slowdown is a fallacy. I consider myself a heavy user and that's only because I play online video games after my children go to bed. But my wife does video editing and is always uploading and downloading videos to her clients and we never get anywhere close to the maximum amount of up/down speeds that we are offered. The closest is when the kids are watching netflix on the xbox, my wife is working and watching a show on netflix and I am playing a game. We hit about 85%-90% usage of our personal allotted bandwith.

    So yes. I can see that you will notice a slowdown when the kids get home if everyone is trying to do what they normally do with what they have payed for.

     

    reply to this | link to this | view in thread ]

  22.  
    identicon
    Anonymous Coward, May 13th, 2009 @ 11:45am

    So what is wrong with charging those that do cause the slowdown?

     

    reply to this | link to this | view in thread ]

  23.  
    identicon
    Josh, May 13th, 2009 @ 12:28pm

    Re:

    Because I am already paying for what I am using. Why do you need to charge me more?

    For example, in my area the local cable provider (Mediacom) offers three levels of service. The basic service is $29 a month and allows for a download speed of up to 3MB/sec, the mid tier is $49 and allows for a download speed of up to 8MB/sec, and the top level offers up to 20MB/sec download speed for ~$70 a month. So if I am paying for the mid-level tier and I am not using more than that amount, which I can't due to the limitations that the cable company puts on the line, and lets say that all of my neighbors have the same plan. Now lets say that we all get online at the same time and we all expect to see the webpages load fast. Except because of how cable interent works none of us get to see that fast speed we are expecting. Yet none of us are going over our purchased limit. So why are we seeing slower speeds? And why should the person that maybe uses 95% of the speed they purchased have to pay more for that speed than the person that purchased the same package yet only used 5% of what's available?

    If we follow this through to your comment then we should be charging our respective cable providers for the slowdown since they have not provided an adacute infrustructure to handle all of the purchased needs of their users.

    This should never fall onto the backs of the users. If I purchase a printer and it says that it can print 55,000 pages on one cartridge but I only get 5,000 pages out of it before it slows down am I required to pay more money to get the rest of my pages? No. The company either fixes the printer or replaces it for me. The same holds true, granted in a slight stretch, to the internet. If I pay for a specific service, 2/4/6/8/50 MB/sec then I should see those speeds regardless of how many people are using the service. If I'm not, the fault for that does not lie with me. It lies with the company that provided me the service. They either oversold what they had available or they lied about what they could provide. Yet nothing is ever done about this.

     

    reply to this | link to this | view in thread ]

  24.  
    identicon
    Chronno S. Trigger, May 13th, 2009 @ 1:14pm

    Re: Re: Re: Re: Re: Re: Re: Youtube source

    "Ok, let me get this straight: you honestly don't believe that Google would save money on bandwidth costs by shutting out millions of users?"

    if they only do that, no, it will not drop the price of a fixed connection. I pay $70 for my internet connection weather I use 20M or 2,000G. My company pays $200-$300 for one of our connections weather we upload 20M or 2,000G. I would assume that Google is the same only to a factor of a few thousand.

    The question is: Do you honestly believe that a company with the leverage of Google pays per bit? That is the only way they would save money.

     

    reply to this | link to this | view in thread ]

  25.  
    identicon
    Chronno S. Trigger, May 13th, 2009 @ 1:24pm

    Re:

    "So what is wrong with charging those that do cause the slowdown?"

    This is a vary good question. The answer is: the people who cause the slow down are the 95% that only use 20% of the connection just all at once. The other 5% have it spread out over the entire month. The 5% don't cause the slow down, the kids who play on youtube, myspace, facebook, XBL, all at the same time do.

    Plus, I haven't noticed this for at least 5 years. Even in my back water home town our cable company fixed this. So if your ISP is still having the problem than they should upgrade their network just like the rest did.

     

    reply to this | link to this | view in thread ]

  26.  
    identicon
    Brett Glass, May 13th, 2009 @ 3:10pm

    The author confuses cause and effect

    It's ironic. The author says that ISPs' desire to regulate bandwidth is not evidence of a bandwidth crunch. But the author confuses cause and effect. Bandwidth restrictions would never be evidence that there was a crunch; it would be a consequence of it. Furthermore, the author presents no evidence that there is not a bandwidth crunch. Rather, he seems just to be ranting at ISPs because he harbors a grudge against them.

     

    reply to this | link to this | view in thread ]

  27.  
    identicon
    Brett Glass, May 13th, 2009 @ 3:10pm

    The author confuses cause and effect

    It's ironic. The author says that ISPs' desire to regulate bandwidth is not evidence of a bandwidth crunch. But the author confuses cause and effect. Bandwidth restrictions would never be evidence that there was a crunch; it would be a consequence of it.

    Furthermore, the author presents no evidence that there is not a bandwidth crunch. Rather, he seems just to be ranting at ISPs because he harbors a grudge against them.

     

    reply to this | link to this | view in thread ]

  28.  
    identicon
    Anonymous Coward, May 13th, 2009 @ 5:13pm

    It just had to happen. We're using so much internets these days, that they were bound to run out. We need to conserve our internets, maybe build a big internet battery that will efficiently manage our stockpiles of internets. It is one of those cases where demand has truly surpassed supply. The factories are not producing enough internets to meet the demand, and it's possible that the mass amount of layoffs due to the economic turndown will mean we will have to import our internets from China or otherwise outsource the manufacture of internets.

     

    reply to this | link to this | view in thread ]

  29.  
    icon
    Mike (profile), May 13th, 2009 @ 5:24pm

    Re: The author confuses cause and effect

    It's ironic. The author says that ISPs' desire to regulate bandwidth is not evidence of a bandwidth crunch. But the author confuses cause and effect. Bandwidth restrictions would never be evidence that there was a crunch; it would be a consequence of it. Furthermore, the author presents no evidence that there is not a bandwidth crunch. Rather, he seems just to be ranting at ISPs because he harbors a grudge against them.

    Hi Brett. I'm sorry we didn't get to meet the past few days, though I know you were here at the Tech Policy Summit as well.

    That said, you are wrong in your statements about me and what I said.

    Nemertes was the one who claimed that the metered bandwidth plans are evidence of a crunch. I, in fact, pointed to evidence that it was not. To claim I have not is simple fallacy. I did right here in this article, and we've done so multiple times in the past.

    Outright lying is beneath you.

     

    reply to this | link to this | view in thread ]

  30.  
    identicon
    Anonymous Coward, May 13th, 2009 @ 6:37pm

    The internet is not a big truck

    its a series of tubes, tangled up tubes

    I sent my staff an internet last thursday and they got it on tuesday

     

    reply to this | link to this | view in thread ]

  31.  
    identicon
    Anonymous Coward, May 14th, 2009 @ 10:50am

    Chronno, my cable provider is Cablevision. You know, the one that wins all the JD Power awards for service? The one who offers "up to 15 down and 2 up" The same one when I run speed tests on their own site, never have been over 8? That cable provider that needs help in upgrading their service?

     

    reply to this | link to this | view in thread ]


Add Your Comment

Have a Techdirt Account? Sign in now. Want one? Register here
Get Techdirt’s Daily Email
Save me a cookie
  • Note: A CRLF will be replaced by a break tag (<br>), all other allowable HTML will remain intact
  • Allowed HTML Tags: <b> <i> <a> <em> <br> <strong> <blockquote> <hr> <tt>
Follow Techdirt
A word from our sponsors...
Essential Reading
Techdirt Reading List
Techdirt Insider Chat
A word from our sponsors...
Recent Stories
A word from our sponsors...

Close

Email This