A Note to Cable Companies Regarding Bandwidth Caps

Time Warner has taken a PR beating in its efforts to do the right thing.  I would like to offer them a suggestion.

You are not going to win the PR battle if you make usage caps an issue related to cost.  There is far too large a contingency on the internet who will do their best to shout down anyone  and every corporation that challenge the dogma of “free internet” or “unlimited bandwidth”.   You will have folks like Saul Hansell at the NY Times who doesn’t seem to realize how often he contradicts himself writing articles that try to support the notion that all bandwidth is equal, and cheap.

Saul ignores the fact that any person or company can buy a server from any hosting company to get direct access to unlimited pay as you go bandwith, at very low commercial rates. He cites the fact that an ISP would pay $1k or more per 100mbs, without questioning the value. Then goes on to suggest that its acceptable to share the same amount of bandwidth delivered to the home 500 ways.  Which of course is another way of saying that ” you can have as much bandwidth as you want, as long as you don’t use it”

Saul says that if the consumer does use it,  the ISP just “splits the nodes” as many times as needed till consumers get the bandwidth they need.  Of course this doesn’t take into account the differences between upstream (which is more limited for technical reasons) and downstream, but let’s set that aside as not being an issue.  The natural maximum amount of noding is to the point where  there is a 1 to 1 ratio between node and home.  At that point the consumer has a link to the network that is not much different than that of the ISP that Saul said was paying $1k per month per  100mbs.     Should the consumer pay $1k per month for 100mbs of switched bandwidth  ?

The differenceof course is the quality of service required.  The ISP needs the full 100mbs available to it 100pct of the time.  Without the certainty of availability, it has no way to reliably offer service to its customers. The presumption is that consumers don’t  require any specific level of quality of service for their internet needs.

That is where the cable companies are getting their PR approach wrong.  They know what happens to their customer service lines if TV service is interrupted for even the shortest periods of time.  Look at the big to do made about the transition to Digital TV. Politicians were counting lost votes from consumers who couldn’t get their convertors for their analog TVs.  People get upset when they can’t watch TV.

Netizens are just as passionate about their internet.  Not only do they want access, but they want it fast.  Unfortunately for 99pct of the vocal minority of netizens, those who post all over the net and start websites about the importance of unlimited bandwidth, they think bigger numbers mean faster access.  It doesn’t.  They just don’t know it.  Firefox with too many tabs open on an Airbook will slow how fast a website comes up or your gmail loads more than whether or not you have a 2mbs, 10mbs or 100mbs connection.  The size of your Outlook database on a PC will impact how quickly your new email downloads and is available for you to reply to than the speed of your internet connection.

It won’t be easy, but the battle ground has to switch from amount of bandwidth and cost, to performance .

Consumers understand that there is a huge difference between quality and quantity.  Consumers understand that realworld speed comes at a premium price. We see it with our toll tags. We see it in carpool lanes. We see it with Fedex and UPS.  Amazon gives us delivery options with different costs. We as consumers make judgements every day about whether or not we want to pay for better performance.

The cable companies need to stop making this about gigabytes per month, and make it all about performance. You need to start making it clear that usage caps SERVE AND PROTECT the performance of your customers.  That you want to offer your customers the fastest access to your favorite websites at the lowest possible cost. That your network is finetuned to optimize the performance for 99pct of your customers. That you want to offer streaming video that doesn’t buffer. That you want to allow for fast downloads.  That can’t happen when a few people take advantage of the network.  Let people know you have special offerings for those who use corporate level quantities of bandwidth and that usage caps are in place to identify those people.

I like usage caps because there is nothing that pisses me off more than the network slowing down when Im trying to get work done. I’m a consumer that places a premium on quality over quantity.  I’m one of the 99pct of your consumers that you are putting 2nd to the 1pct. The vocal minority that probably don’t turn their lights off and then wonder why the electric bill is so high.

Focus your message to us. Those who care about quality of service.

50 thoughts on “A Note to Cable Companies Regarding Bandwidth Caps

  1. Mark, I think you are a little out of touch buddy. Most people can’t pay 1K a month for internet access. Though the majority don’t “need” a huge amount of speed and bandwidth, the fact is that the United States is behind the eightball (big suprise) on internet speed. Other countries have multiple times the speed we get for less money. Please explain how that is possible. $50 a month is a lot of money for just access. They are not providing content and are not providing ANYTHING when the consumer isn’t using the internet. You would think for that $50, we could get at least the speed someone in Singapore can get. Again, like the car companies, this is another example of US companies being too big to care that they are not up to speed with foreign competitors. If there wasn’t a monopoly on phone and cable systems, and/or you could upload to satellite instead of just download, these companies would be put under by foreign companies with a quickness. Just like Honda and Toyota make far superior cars than Ford or GM. Its a freakin’ car! They are on the street every day. Take one apart and see how they did it better! I’m tired of the excuses put forth by large US companies. Between corporate laziness, the governments idiotic constraints on research, and monopolies, the United States is going to become a producer of NOTHING!

    Comment by Mike Libidinsky -

  2. Mark,

    This post ignores the larger reasoning behind caps. It’s about greed, the caps and tiered offerings are NOT about quality, but control.

    If Mark Downloader normally downloads things like movie streams from netflix, streams from youtube, broadcast.com and whatnot, he will go over an in theory cap of 10 GB a month @ $40 within less than a week and be charged a dollar for each GB thereafter. Now, if Mark Downloader got the same content from his ISP’s streaming video site (like say Time Warner, Comcast or whatnot) offers or will offer it, that content downloaded from the ISP’s streaming video does NOT count against the cap because it’s written in the subscriber agreement.

    It’s about money, always has been. Want to know the funny thing? 1 GB of bandwidth costs on average 4 cents on the ISP side.

    Mark Downloader can download 100 GB a month worth of torrents, downloads, streaming video, and it’d only cost his ISP $4 against the $50 Mark Downloader pays the ISP.

    Tiered pricing? What a joke, you aren’t going to get any better service than you do now if we revert to tiered pricing. It’s just another way for ISP’s to make more money on the backs of those uneducated about the costs and understandings of how ISP’s and networks function.

    Do a little googling, look up how much better Japan’s internet is than ours, in speeds and costs. We need to model ourselves after that. We’re the world’s supposedly best nation, we seem to be falling short more and more each and every passing day.

    Comment by Joel Hicks -

  3. Mark,

    One of the issues for me is that I am promised 20mb/s. That is what my contract tells me that I am paying for. So is my neighbor, and the neighbor on the other side. So shouldn’t we all be able to use all 20mb/s all the time? You are telling me that it is actually OUR FAULT that we are using what we are paying for?

    Now, I don’t at all have a problem with limiting. What I have a problem with is the companies blaming paying customers for problems in the network.

    You are a business man. From a business sense does it make any sense to blame your customers for taking exactly what you promised them? Sounds like a bad business plan to me.

    Comment by Brandon Hansen -

  4. I have had Time Warner Roadrunner for MANY years my bill is way to HIGH and their customer service is now overseas.
    I can’t wait till we don’t pay a penny and it all comes from satellite 2-3 years from now Time Warner will be toast.
    We will get all this media for free mark my words 🙂
    throw your cell phone in there also

    Comment by John Sullivan -

  5. Just like electric utilities – ISPs could implement rate ratchets.

    You use more – you pay more – AND you pay a higher rate.

    Comment by Jeff @ another71.com -

  6. Well Mark I disagree wholeheartedly on your argument that this is about “quality bandwidth”
    Fiber does interfere with the load on the network.

    I’m a big fan of your outlook and perspective on not only technology but business as well. But this process of holding the thumb down on the consumer for rights to a “fountain” of bandwidth that services like fiber based technologies provide is wrong.

    This blog entry contradicts your view of bandwidth and the american family from March 8th 2006. Yes we will consume, but the point of the provider is not quality of the bandwidth availability but the quantity. As you stated directly at the end, it will take around 100 mb/s which Europe has coming in the pipeline, why not us? Go ahead and charge per MB, watch the community switch providers to fiber based technology.

    Bandwidth to the home, how much is enough
    from Blog Maverick by Mark Cuban

    Its the trillion dollar question. Telcos and cable companies are retooling and/or rebuilding their networks to squeeze as much bandwidth as possible out of their plants. Satellite providers are looking at every alternative means of increasing banwidth to, and adding a return path from the home.

    None have a choice. There is no question that households will consume an ever increasing amount of bandwidth. The question is how much. What will peak simultaneous bandwidth consumption to the home be in the next 5 and 10 years ?

    Television, High Speed Data, Telephony, sharing home movies and pictures, remote backup and applications that havent been thought of will consume bandwidth. A lot of it. But lets break it down to get some minimum numbers. Because if a provider cant hit minimums, they will have problems.

    First of all, lets start with Television since thats a battleground that all providers are fighting on. The very near term future of television is high definition. A high definition stream is going to require a minimum of 8mbs per second on average. Thats a minimum. Now some people are trying to say that they can do high definition in far less bandwidth, they cant.

    Even with the best compression, lower bit rates fail the smell test. More importantly, because its easy to see the difference in picture quality as bit rates are reduced, viewers will complain about reduced picture quality, and picture quality will become a competitive element. That competition will keep bitrates at 8mbs, if not higher for sports and movies. So lets work with 8mbs.

    Today its difficult for people to imagine a High Def TV in every room, but within 10 years HDTVs will be ubiquitous. More importantly, over the next 5 years, the homes with a HDTV in every bedroom and the family room, with HD PVR with Terabyte drives centrally housed, or connected to each PVR will be the most important homes in the neighborhood.

    Why ? Because if the household can afford to be the first in the block to be an all HD household, they will be a household able to buy everything that the provider sells. They will also be the more technically sophisticated household, so they will be more likely to buy all the options for high speed data, in home wireless data and eventually media, online backup of PCs to a central location and anything else the provider can think of. These are the “whale” customers. The most profitable customers that always pay their bills and never churn off.

    So lets look at our customer, The Whales, and their 3 kids and see what services they use.

    First, each of the 3 kids has an LCD HDTV that operates both as a HDTV and a PC monitor. THeir PC is of course connected to the net and is their stereo. It is not their TV PVR because of the hassles of cable card or lack of satellite PC connectivity for programming. Instead they have a provider installed HD PVR that shares a multi terabyte drive with their PC.

    The kids are collectors. They save every bit of music and internet content that catches their fancy on their hard drives. They used to use ITunes, but instead they use a freeware desktop that front ends Itunes and all the different broadband environments that have been created and presents it as a unified front. It of course strips out any and all commercials by identifying the tracking information that is part of the internet url or embedded in the content itself.

    When it comes to TV content, they use the same front end to programatically control the provided PVR. With the front end, they dont use season passes any more. They save networks. Everything on MTV. It gets saved to the PVR. Everything from HDNet and HDNet Movies, CBS, NBC, HBO, Showtime, ABC, TNT, ESPNs, they all just automatically get saved. They understand the concept of mutliple terabytes and at 8mbs a stream, they know they can save content to their hearts content and if they need more storage, they can delete something or just add more terabytes. Its cheap. So their PVRs have basically become network spiders pulling in content 24x7x365.

    Of course they cant watch it all, but so what. When something they want to watch in realtime is on, like an NBA game, they watch it. If they arent at home, they use the front end to re route it to their personal IP address that they bought their name-url for. At the mall during the game ? Just program the front end to send it to markcuban.pda. Never miss a minute, just watch it on my phone/PDA whatever at the mall. Of course, if all their buddies want to watch it, they have added their personal urls to a buddy list and its multicast to them all. Their own personal version of slingbox.

    But wait, there is more. Because they have collected everything on disk, they can use the front end to progam their own TV networks and share it from their goowy.com pages. A little MTV, a little ESPN, a little HDNet and boom, their own tV network. Of course, they can use redswoosh.net to insert contextual commercials and even make a little money from it if anybody watchs. For fun, they program in some home videos and pictures from the party they went to last week, set to music of course

    None of this is far fetched. In fact, its likely. But back to the original question. What is the max amount of simultaneous bandwidth being consumed during a day ?

    3 Tuners bringing in 3 networks in bedroom 1 , one being watched, two being saved. Thats 24 mbs. Same thing going on in bedrooms 2 and 3. Thats another 48mbs. thats 72mbs per sec and thats just the kids rooms.

    Of course mom is watching a day and date release on HDNet Movies in the living room while saving Desperate Housewives for the hubby to watch later. Thats another 16mbs. We are up to 88mbs and going strong !

    Dad is working on the collage of movies and pictures that he wants to give to the grandparents as a gift. So he is uploading and downloading digital pics and HDV files to and from the family storage site on box.net . He hates that the kids use so much bandwidth, but thank goodness he is able to finally buy the 100mbs package.

    When he is done with the collage, or at least after he makes some progress on it, he is going to plug in the portable hard drive he got from the satellite company to the USB 2.0 port on his 80” plasma. Ever since they bought Netflix, he has subscribed to their movie service. They send him a hard drive full of hundreds of movies that Netflix customizes to his tastes, he loves murder mysteries. He watches as many as he wants to/can and when he is done, he sends it back and gets another disk with the genre of his choice. His only complaint is that they wont split genres on a disk, so he cant order half chick flicks for the wife, which of course creates problems from time to time.

    But it works itself out. There is enough bandwidth, and enough TVs to go around and every Sunday they look forward to the family tradition of going to the movies together.

    So there you have it. My over simplified vision of the bandwidth and technology consumption of a family of the future. Not necessarily your typical family, but one of the millions of upper income families that every provider will do whatever it takes to make happy. Which they just might….

    If they offer more than 100mbs of bandwidth.

    Comment by The Avid Techist -

  7. From MC> You miss one point. The content providers that you think will pay for access, are getting paid by the same ISPs for the content they provide on the TV side

    No… the last time I checked Time Warner (or any other MSO) wasn’t paying Apple, YouTube or Netflix for content.

    CBB isn’t about making networks faster. It’s about changing the fundamental architecture of the Internet. And their sad story about how their bandwidth costs are too much is crap. Ars Technica reported as much after reviewing costs stated in the TWC 2008 annual reoprt. And the NYTimes had a similar article regarding Comcast’s bandwidth expenses.

    You need to be a leader on this Mark! I know you’re chomping at the bit to pay ISPs big dollars so HDNet can stream content on the fast lane but some things are more important!

    Comment by Joe -

  8. Comment by monster -

  9. Another interesting study:
    Of course, if there was actual competition instead of a sanctioned duopoly, we wouldn’t be talking about any of this, would we?

    Comment by Chris K -

  10. Mark,

    My concern is that at least some of the motivation for these caps is an attempt to gain a competitive advantage for the cable companies own PPV services against the net based alternatives: Netflix, Hulu, Apple…

    I’d be willing to pay for tiered service with guaranteed performance — but I don’t want my cable provider to be able to provide PPV movies to me at a lower price than a web based service solely because they control my pipe and can arbitrarily price PPV one way and data another.

    To put it more plainly, if Time Warner wants to provide PPV they should be forced to provide it as a service that counts against any bandwidth cap they enforce.

    Comment by John -

  11. I watch almost all of my movies via Netflix and TV shows via Hulu. In my neighborhood, we have FIOS and I have never experienced any down time or slow down in the last 2 years. Bandwidth caps are not only a bad idea but like many here have said, a step in the wrong direction. I cannot believe someone like Cuban can support caps when he himself supports HD content. I thought for sure he would be on the side of the consumer here and not big business. Streaming media is the future and anyone that does not see that has blinders on.

    Comment by Scot -

  12. I thought this piece was interesting: http://arstechnica.com/tech-policy/news/2009/04/twc-without-data-caps-internet-upgrades-now-in-doubt.ars

    Comment by Geoffrey -

  13. My issues with this solution:
    1. I do not know my current bandwith usage, and am unaware of any meter I can check to monitor my usage the way I can my electricity, water, and cell phone.

    2. Knowing that the end of a billing cycle is coming and I am low on bandwith, would encourage me to hold-off on key OS and anti-virus updates. This is of course rectified by off-peak incentives ie. free or discounted nights and weekends.

    3. Human nature is not to limit usage until supply is almost gone. People spend freely on payday, and are broke the day before the next payday. The same will happen here, people will still use high amounts of bandwith at the beginning of their billing cycle and less at the end. So this will really only increase quality for the last half of the billing cycle.

    4. As others have stated, I am weary of any company’s motives for policy changes. My paranoia runs to attempts by cable companies to force me to pay and use their crappy VOD systems without them upgrading them or making them more user-friendly by making competing systems like Hulu too expensive for consumers.

    Comment by Joe -

  14. I agree with what David Langlands says, and wanted to just add a secondary point to his.

    The bandwidth caps that you are discussing are not a solution to the poor performance problem (but a small part of it). The real solution is dynamically limiting network connections per subscriber.

    The problem is that the bandwidth throughput (I’ll call throughput to differentiate from the bandwidth used in the cap discussions) that is given to each customer is oversold, so that 200mbps pipe to the homes in the node is distributed in 12/16/20mbps packages to 200 subscribers for example (numbers are not real, but the point is valid). And thus the network congestion problem occurs when the number of users downloading data reaches a threshold. David explains well what the providers need to do is limit the connection throughput throughput for each customer dynamically during peak times. This would give everyone proportionally less throughput based on their performance tier they are subscribed to.

    This solution can work in the other direction in the form of bursting bandwdith, which Comcast already has this in place. Dynamic throughput allows users to have move download throughput for the first 10MB of a file in the case of Comcast’s Powerboost(tm).

    The bandwidth caps that providers are trying to use, while a cap can help limit some users from saturating the pipe 24/7, doesn’t really address the problem in a logical way. Even with bandwidth caps, network saturation can still occur, until enough subscribers stop using the network because they’ve gone over or are close to their capped usage.

    Node throughput – 300mbps
    # Subscribers per tier: 10

    Tier 1 (T1) – 20mpbs down
    Tier 2 (T2) – 16mbps down
    Tier 3 (T3) – 12mbps down

    Throughput needed during peak times (no limit):
    T1 – 200mbps
    T2 – 160mbps
    T3 – 120mbps
    All Tiers – 480mbps > 300mbps (network congestion, packets are lost, internet video streams become choppy)

    Throughput when limited to accomodate all customers:
    All Tiers must be < 300 mbps
    Thus, a limiting solution could be to cut all connection throughput throughput in half
    Tier 1 (T1) – 10mpbs down (limited)
    Tier 2 (T2) – 8mbps down (limited)
    Tier 3 (T3) – 6mbps down (limited)
    Total – 100+80+60 = 240mpbs < 300

    Obviously the limited connection could be a bit higher, but assuming network overhead this seems to be a logical limit.

    There are issues with this method of limitation, as driving the limited throughput low enough will prevent certain types of content to be inaccessible.

    The limitations stem from the necessary overselling technique that the companies use. However, these limitations can be overcome with innovative thinking all the way up the stack from the Hardware Network to the Content Delivery.

    Please just don’t use bandwidth caps in their current form as a solution to the network saturation problems.

    From MC> The points are well taken and correct. Cap do not eliminate congestion. That in fact, normal usage in an overbooked environment will lead to not only congestion, but more latency. HOWEVER, in an environment where some users not only have the desire to fully utilize every mbs available to them, but in some cases “resell or trade ” it for P2P applications, caps are a needed step. I do agree with you when you suggest that there are network management tools that need to be used as well. But I don’t think that universal throttling will make many people happy. The approach you mention in reality is just a dynamic cap that is set based on utlization. The arguments against it will be slightly different, but none the less the same as the complaints against caps today.

    Comment by Steve Tranby -

  15. Mr. Cuban, usage caps will do nothing to keep your network from slowing down at peak hours.

    Consider 100 people with 10 Mbps connections on a network with a total capacity of 500 Mbps (it’s called overselling bandwidth and it happens all the time, usually by a factor of more than 2).

    If 50 of those 100 people use their connection at once the network will be at full capacity. If 100 of those people use their connection at once everyone will slow down. It will slow down even if everyone is far under their data cap simply because they’re all using it at once.

    You said in reply to another comment:
    From MC> There were caps. They were called 56k modems. There were per hour charges from many ISPs. If you think caps stifle innovation, you might want to start over and do some homework. Anyone who needs bandwidth to develop anything can get hosting services, cheap. With far better opportunity than trying to do something from their home ISP.

    The problem with your retort is that 56Kbps modems and per hour charges are all about access and bandwidth. In that 1 hour you could download as much as your 56Kbps modem could pull which was more than your 48Kbps modem could pull in the same hour (you did upgrade right?). Now the modems are 1 – 30 Mbps instead of 56Kbps and instead of paying by the hour we pay by the month. It’s only a data cap in so far as it was physically impossible to transfer more than X amount of data in 1 hour over a 56Kbps modem, but you could still get a busy signal when you tried to dial in and everyone was using the network!

    Data caps aren’t the answer. Managing bandwidth (access) is what will give you your improved quality of service. That or just don’t go online in the morning, during lunch, or after dinner.

    Comment by Brion Swanson -

  16. I’m one of those people who can easily use my maximum bandwidth (not usage or cap, but bandwidth) for 36-48 hours straight. When I encounter usage caps, it is usually within hours of getting online. I will never again subscribe to a service that has a usage cap. In that same vein, if my ISP wants to charge me more because I use a disproportionate amount of bandwidth, I will either pay it or I will find an alternative to being online. Right now, I pay a premium to get a higher bandwidth. I pay $80/month while most people pay $50/month. I pay it without complaint because it makes those 96 hour stretches into 48 hour stretches.

    But I’m not the only person. My cable company provides this service because they perceived a market for it and they continue to provide it because they have apparently achieved a stable market for it. So, I share the cost of my premium with everyone else who wants the same premium. I certainly don’t WANT to pay more for my internet. However, it is entirely hypocritical of me to demand that others share in the cost of my egregious bandwidth usage. I think pay-as-you-go is possible but ridiculous, but that is a different story.

    And, alas, I am no billionaire. Maybe a thousand-aire.

    Maybe if the ISP provided more tiers. Maybe we should just wait until wireless becomes ubiquitous before we gripe too much about what we have now.

    And I agree that this has NOTHING to do with net neutrality.

    Comment by Jason -

  17. Mark, your entire QOS argument falls apart when you consider that usage caps do not prevent people from saturating their pipe. If a user is willing to pay $150/mo, they can suck down GB after GB at their maximum downstream bandwidth, 24x7x365. Get enough people doing that on the same node and you’ve got congestion.

    Let’s say everybody’s staying under their cap, even. For those first 40 GB of the month, or however much it is, there will inevitably be times when people are downloading using their maximum bandwidth.

    There is no correlation between amount of bandwidth in use at any given time and how much a user consumes. You can argue that people who consume more are more likely to be saturating their pipe at any given time, but you can’t draw a statistical correlation.

    Comment by Scott Cranfill -

  18. 1. The objection with usage caps that I have is that ISPs are really trying to get more money out of people through overage fees. I would be much happier if there were no overage fees: have a cap and after you use up the cap, you move to the slow lane. In the slow lane, you get 500kbs or so… something enough to check email or listen to audio but not to stream video.

    2. If I am going to pay for bandwidth, I don’t want to be paying to download ads. There needs to be a mechanism where the advertiser pays for this bandwidth.

    3. If the last mile bandwidth is truly the problem, then have free/cheaper periods (i.e. free nights and weekends on my cell phone). Then I can schedule downloads (microsoft updates/podcast subscriptions) for this low usage period.

    Comment by Alex -

  19. Agreed. Even the highest levels of low-latency transaction-processing on private networks owned by F100 companies only worked if “most users, doing nothing, most of the time” was the creed that defined transaction processing. The tiny minority burning bandwidth shouldn’t be able to impact everyone else.

    Comment by Jay Krackeler -

  20. This post is so off base and ignores so many core issues I don’t even know where to start.

    But I’ll just say this — if the equivalent data caps were in place in 1998, no one would have ever heard of Broadcast.com or Mark Cuban, and you might be worth a cool million right now, tops. So glad you’ve cashed out and now don’t care a lick about innovation or advancement of technology. It’s good that’s working for you — but leave the rest of us alone, please.

    From MC> There were caps. They were called 56k modems. There were per hour charges from many ISPs. If you think caps stifle innovation, you might want to start over and do some homework. Anyone who needs bandwidth to develop anything can get hosting services, cheap. With far better opportunity than trying to do something from their home ISP.

    Comment by Brock -

  21. There are slippery slopes no matter which way you go.

    Good business will always follow the path of most revenue. Therefor, “Comment by David — April 21, 2009 @ 5:30 pm” has very serious and legitimate concerns.

    An old cliché, necessity is the mother of invention, applies as well. As soon as networks start capping bandwidth they have much less of a reason to invent ways to improve bandwidth.

    History has proven again, and again, and again that what makes the most coin isn’t always what is best in the long run.

    By the way, I completely agree with your points on quality over quantity. If we compare today’s Internet with electricity it would be like having your lights flickering on regular basis.

    Comment by Anthony Topper -

  22. Love to read your views. Agree mostly, disagree with some of the analysis in that a key input is left out.

    Where I am the dominant ISP’s are also in the content delivery business, I think if you put in the “glass steagal act” for Internet providers, for example if you have the pipe you can have no interest in the content, you would see the industry argument change.

    It is protecting a model that limits choice to protect their content distribution channel, to promote services they own, to limit advertising revenue to services they provide.

    Comment by Ross -

  23. People have to update Windows. If I want to download updates, it could be several hundred megs and if I have more then one computer, it could be several gigs. Putting caps on bandwidth will deter people from keeping their computers patched and updated and the internet will be run down with virus infected PCs and bot-nets and anything else you can think of. Bandwidth limitations of any kind are not only anti-consumer but they are dangerous.

    Comment by Bill -

  24. Mark,

    as a former Tech VP of a cable MSO, I appreciate your sympathy for the cable companies when it comes to providing a reliable quality of service to all customers. Personally however, I feel that most of the congestion problems that many consumers experience today are the result of poor engineering or short-sighted cable plant decisions made in the late nineties/early 2000’s.

    The oft-cited problem with abusers who think it’s okay to peg their connection at 8mbps 24x7x365 and only pay $40/mo is largely solved by economics. The cost for the Internet bandwidth of that full 8mbps to a large MSO (~$20/mo?) is less than the net margin on that user’s triple-play connection. DOCSIS 3.0 and the law of large numbers allow those abusers to draw from a larger pool and greatly reduces the chance they can saturate the link for more typical users on the node. There truly do need to be caps for the significant outliers, but today’s technology makes this only relevant for the 15+ mbps abusers.

    Or, why don’t we just rate-shape the abuser’s traffic? Packeteer’s been around since 1996. Practically every network vendor has similar capabilities baked-into their products for the past 4-5 years.

    Please let the cable companies shoulder their share of the blame for opting for substandard technologies and outdated thinking. We typically pull a dozen or more strands to the node. Only a single strand is needed for video, it’s trivial to boost the uplink capacity per node to 1Gbps, or even 10Gbps, over commodity Ethernet switching on an additional 2 strands. Switch to a micro or pico-node strategy for the RF channels and you can certainly dedicate 10 or more 6MHz (38 x 10 = 380mpbs) downstream channels per 50 homes. Cull out the < 2% abusers, and you could profitably offer shared 100mbps per user to your entire plant without conflict at less than $50/month.

    This was the dream I believe you foresaw at the time Broadcast.com was purchased. Unfortunately, it never happened, and honestly it isn’t the BitTorrent crowd that prevented it from taking place. It was old telco thinking and lack of innovation on the part of a handful of firmly entrenched monopolies.

    The problem truly is that now these substandard cable plants are being leveraged to deliver “on-demand” and PPV products and as a result many MSO’s are unable or unwilling to free up the additional spectrum for Internet bandwidth. Unfortunately, too few of my former peers “get it” and they’ve only recently begun to adopt cheaper technologies to get higher bandwidth to the node. It’s only costly because their lack of foresight made it so, and it truly isn’t that expensive to fix it.

    We all benefit from greater access to information. Low caps are a cop-out for poor engineering and ignorance of better alternatives.

    Comment by David Langlands -

  25. Bandwith capping is such a backwards move that I honestly can’t see how you support it. Why would I want to support this act of preserving out of date technology. These companies should take cues from places like Japan or Sweden, whos connections are far faster than ours. Your rich anyway if it’s so important to you why don’t you just get a dedicated line ran to your house?
    The only way this makes any sense is from a business stand-point. I can only guess that is all you care about since you make it a point to stress your feelings constantly about this issue.
    I want to know that in the future just because I feel like it, I can download multiple distros of linux OS, blu-ray movies released through vendors, stream HD content. I don’t want to think to myself about how much I am going to first pay to purchase the content, and then add on top of that my bandwidth fees.
    What this really seems like is that your comparing bandwidth to taxes.

    Comment by James -

  26. If you want to talk PR, then the truth is that “framing” the usage cap issue differently isn’t going to help. Thanks to the AIGs of the world, corporate greed is the current topic of conversation, and people simply don’t trust companies. People are losing their jobs and their houses…they’re pissed off. They are going to be skeptical when a company tries to “improve” service.

    Besides the trust issue, I’d be willing to bet people are generally happy with their internet service. I’ve had Time-Warner broadband for a long time. It’s one of the few services I’m happy with. Why would I want to risk losing that? I want changes to things that aren’t working, not to things I’m happy with.

    Maybe when the economy improves and trust is restored in corporate America, this issue can be addressed again. But until then, people aren’t going to take it lying down. And making logical arguments like you are isn’t going to change anyone’s mind. These are emotional times, logic won’t work.

    Comment by Kevin -

  27. Sorry Mark. High speed internet companies have had since 1996, when they took over $200 Billion from the American taxpayers, to improve their connections to handle this traffic. What have they done? Nothing, except take our money. As a result, the US is now 16th in broadband internet technology and falling. FiOS? UVerse? Now Broadband Caps? Give me a fucking break. Time Warner will be doing the right thing when the internet is as fast as was promised back in 1996. Until then, it’s bullshit.

    Visit http://www.teletruth.org for how they’re screwing us.

    Comment by Eric Ogunbase -

  28. Hi Mark,

    I don’t think that bandwidth caps alone will resolve the problem of users desiring speed over quantity. Like all traffic jams, it’s not only an issue of bandwidth but a matter of WHEN the traffic takes place.

    Congestion charging is one method of alleviating the problem during peak times, alternatively users should be requesting that their IP packets have higher priority (QOS) than the others in the node. During heavy usage, their speed should be unaffected.

    It’s not a difficult concept to grasp. Take a flight for instance. There are people willing to fly in first class (and pay for that experience) while others desire coach.

    Comment by Mike -

  29. Mark,

    They have a much bigger problem than what you lay out. And it’s really a basic one, one of trust. No one believes that they are out to best serve the 80, 90, or 99%. Everyone believes this is to earn more money, and to prevent loss by people moving away from cable. And since they are a near monopoly in most locations – DSL and FiOS aside – they can afford to keep doing this.

    Last year, similar moves were talked about as targeting P2P users, allegedly swamping the networks. No one believes it now, and almost everyone acknowledges it is a hedge against downloadable and streaming content.

    As to your “one kid downloading 5 movies”, well, that’s often rate limited by the modem, and I think even in DOCSIS 1.1 you could target single users to reduce their rate.

    If I can get ala carte programming, I will fully support ala carte internet. Until then, this just looks like a money grab.

    Comment by Dave -

  30. I don’t have a problem with tiered pricing as long as I get what I pay for. I’m a web developer and day trader by day and a gamer by night. I use a ton of bandwidth everyday. I have a computer that can handle it. Now if I’m paying for ‘unlimited’ bandwidth and I can’t get get over 1.2mbps…there’s a problem. Just the problem I’m dealing with now.

    So go ahead, cap me – but god help you if you don’t live up to your end.

    Comment by DOUBLEZER00 -

  31. Usage caps are irrelevant as related to performance.

    Take a lesson from electric companies and their business users.

    It’s all about PEAK usage.

    If broadband ISP’s really care about their customers (and not just a bogus profit stream), they will maintain quality of service by limiting ALL customers to an equal lower bandwidth on a maxed out node (for the duration of that stress). If a node is limited to 38MBS then just throttle back the max when that threshold it hit. Worst case, 500 nodes are limited to 76KBS if they are all downloading at the same time.

    Comment by Internet guy -

  32. However much I disagree with you about the business reasons behind the caps, your PR spin on the situation is far superior to anything that’s come out of Time Warner.

    Some committee in TWC thought the “divide and conquer” strategy made sense, trying to turn the vast middle against anonymous heavy users. That didn’t work, for two reasons (1) the extreme was loud in protest and (2) people started to fear that even they might hit the cap as the internet grew. My retired Mom now watches Netflix streaming, and she’s not exactly on the leading edge. If nothing else, people worry that their kids will put them in that category.

    But your story would be “we’re all in this together.” Done well, it would avoid that kind of vilification of the heaviest internet users. Instead, it would be selling the idea of better service. And the message (at least the marketing pitch) is about better enabling the use of online services, not stunting them.


    Now, can you please, please turn your PR talents to the Dampier situation? The cable business will be here next week, and the last thing we need is Tony Parker getting D-Wade-level special treatment from the refs.

    Comment by doncruse -

  33. Can someone provide a real breakdown of the cost of operating a network? How much does it really cost a provider to deliver bits? Do the networks really get overrun by bittorrent users? If so, where is the bottle neck? Can it be alleviated by upgrading the consumers modems or does it require laying new fiber? Or, is the talk of bandwidth caps all about preventing the competition from using “their” network? That is, making sure people use the networks phone service instead of Skype or Vonage or using their TV service instead of iTunes or bittorent.

    Comment by Anthony -

  34. Maybe some of the other commenters could explain: I don’t understand where the content-neutrality issue comes in to the question of terms for consumer bandwidth consumption. I agree content-neutrality is a concern since there’s not much transparency and competition is limited due to the natural monopoly characteristics of infrastructure investment. But the way to keep things neutral is to make any contract between the cable companies and content providers nonexclusive, right? So long as the packages offered to consumers are bandwidth-based (and thus content-neutral) there’s no issue with neutrality. Are some net-neutrality advocates conflating the political issue with the congestion issue?

    Comment by Michael F. Martin -

  35. Hi Mark
    I read your posts about broadband bandwidth, and I keep missing something in your opinion. Yes, you have a valid argument about use and costs, but that is the distraction here. There are not that many users who are throwing the system out of balance.

    It is about content, and access to content, and messages delivered. It is about whoever controls the access directing or forcing the user to access certain information and blocking certain information. That is the bigger game, and one that is played out here in Italy, in China, and in the US. I haven’t seen you address this issue properly.

    The reason people get up in arms about net neutrality is that finally there is a place that information can move freely. If it just a usage cap,that is what it OK, but that quickly turns into a control on usage, and content. That is heartbreaking, and simply wrong.

    I would really like to hear your opinion about this.


    From MC> Its not about content. its about quality
    In a typical cable ISP, users share 38mbs on the node. That means everyone in the 200 to 500 or more homes share a meager 38mbs. So if at any given time the people living in those homes or apartments are consuming more than 38mbs, then EVERYONE using the net in those homes get diminished performance. Your youtube buffers, your websites time out. Your performance basically sucks. That is the issue. Its not about what kindof content, or who’s content. Traffic on your ISP is not much different than traffic on a highway. Your car may be capable of going 100mph. The highway can support 6 lines of cars going 100mph side by side. But as more and more people drive the highway, congestion happens and EVERYONE slows down. Your ISP’s network is no different. Now some would argue that its relatively inexpensive to add more capacity or more “lanes”. Its a fair argument. But it doesnt take into account basic economics. The more capacity people have , the more they will use and people’s ability to find ways to use bandwidth expands faster than the ISPs ability to expand and pay for the expansion. All it takes is one kid in the neighborhood trying to download 5 HD movies at once and 38mbs is toast. One or two people trying to fill up terabytes of drives with free movies and tv shows and the entire neighborhood suffers. Thats what usage caps try to stop

    Comment by danielmcvicar -

  36. Nice post Mark but you’ve failed to look at the big picture. It’s not about making networks faster, more efficient or punishing bittorrent users playing “hogs of the road”.

    CBB will allow operators to have a tiered architecture as they’ll be able to choose which traffic counts against this bandwidth counter. Content providers will pay operators to have their bits delivered in the toll-free lane which would fundamentally change the inherent equality that makes the Internet such an efficient disruptor. They’ll be able to strike a deal with Netflix so their bits won’t count but leave Apple’s iTunes service alone, thereby controlling which IP service provider will receive more users.

    It’s obvious this is what they plan to do. Operators might also pitch different packages to customers the same way channels are offered for televison; $50 / month gets you 250GB and all-you-can-eat Netflix, Amazon and YouTube traffic; $75 / month gets you 50GB extra plus unlimited iTunes streaming, etc.

    Of course this wouldn’t be a problem if ISPs in the country weren’t quasi monopolies. Whether it be through spectrum licenses or franchises obtained from municipalities, these business are allowed to operate in a monopolistic way as dictated by we the people. The people have every right to make demands that ISPs operate in the public interest because there is no true competition.

    Telcos are no longer required to share DSL lines which was an incentive to get them to upgrade their networks. It’s time this investment started to pay off without ISPs asking for more.

    Consumption based billing, non-network neutrality, etc. are the result of one thing; ISPs in this country have failed miserably at rolling out their own IP services.

    From MC> You miss one point. The content providers that you think will pay for access, are getting paid by the same ISPs for the content they provide on the TV side

    Comment by Joe -

  37. Fine, forget my contrived example. My point was about why it is easy for a billionaire to say they don’t like usage caps. If I was a billionaire I would feel the same way.

    Comment by David -

  38. Also, I’m pretty sure (having worked for an ISP for almost three years) that it’s not a technical reason that your upload is capped. Usually it’s a nice profit center – selling that bandwidth on the corporate side.

    That said, I agree that the purpose of limiting is not about protecting the QOS for other customers, but protecting the voip, vod, etc from those pesky upstart interweb companies.

    Your provider is merely doing anything and everything they can to avoid becoming a commodity provider.

    Comment by Chris K -

  39. David,

    Why should everybody else pay for the fact that you can’t keep track of your kid’s Internet usage? I got grounded for a month after running up the phone bill using my dial-up modem in junior high. What’s cheaper for us as a society? To expect parents to be parents or the cable company to be parents?

    Comment by Michael F. Martin -

  40. Mark,

    It is easy for you to say you like usage caps because you aren’t affected if you go over the cap. A $500 overage fee because your kid went crazy for a week doesn’t materially change your life. It would be a disaster to a lot of people.

    Comment by David -

    • just cause I have money doesnt make me stupid and want to waste it.

      I have switched ISPs and wireless data providers for cost and quality of service reasons

      Comment by markcuban -

  41. Another problem is that a lot of data the user cannot control. What about all the traffic from hackers going into the home computer trying to take it over. Ever watch a log of incoming traffic? That stuff is coming in all day long, constantly. Yes the size of the packets are small, but the overall daily usage is not insignificant. What about when MS pushes out a patch? Those things aren’t small. What about an online gamer who wants to download a demo, there goes 1-5GB right there. Mark if your internet is slowing down at night it isn’t because of usage, it is because your provider is refusing to spend a little of its huge amount of profits to deliver the quality that you want. Don’t blame your neighbors for using the pipe they are paying for, blame your provider. Time Warner’s recent news was not about preserving the network for the majority, it was all about a money grab. More and more people are going to the internet to stream tv shows and get their entertainment which means less people subscribing to cable and paying for PPV and VOD. This means less money for TW.

    TW is not paying $1k a month for a 100MB pipe. The cost for a data center connection (as you know or should know) is not in the transfer of data, it is all in the services to make sure your data can be reached, security, 24×7 availability, power backup, etc. The actual cost of transferring a GB of data across a pipe is super cheap, but like cell phone companies and SMS they feel they can rape the customer.

    Comment by Michael -

  42. There are so many problems with the oped by Hansell that you linked to, it’s hard to know where to start.

    But I think you’re basically right that they’ve got the wrong marketing. Another angle for them to pursue would be to analogize bandwidth to property rights. Americans love property rights, and completely get the need to enforce them against trespassers. Maybe cable companies that build the last mile should start selling it off to consumers as “their bandwidth.” That would align the consumers whose bandwidth is being hogged quite nicely with the cable companies.

    Comment by Michael F. Martin -

  43. Except that Time Warner already *does* sell their service as performance.

    I pay for a 10mbps link into their service, and guess what, if I try to saturate my link and get massive downloads from say microsoft.com coming my way.. the link caps out at right around 10mbps.

    The problem here is not PR messages not focusing on performance and instead focusing on GB per $$, the problem is about PR messages that are trying to draw attention away from a promise of performance that cannot be delivered with the current infrastructure they have.

    Comment by jason -

  44. Exactly!

    I have been with Time Warner (road runner) for nearly 10 years. I paid around $50/month for internet. I get about 600 KBps download and about 60 KBps upload. I pay a little less now because I’m taking advantage of a special (switched to digital phone service through them) and am getting a discount for a few month (or 1 year, forgot).

    For many people, these specs look abysmal. Let me tell you, though: This is consistent, day in, day out, middle of the day, middle of the night 600KBps download and 60KBps upload. If they are completely down, it’s because of an issue, but they fix it in a very short time. These outages are very rare, maybe 1 or 2 per year, and some years have no problems at all.

    My wife uses the internet for work (at-home translation) and she cannot have disruption or slowness. I am a programmer. I remote into client servers with my connection. I cannot have disruption or slowness.

    Time Warner cable (Los Angeles West Valley) has been completely stellar in delivering consistent, reliable internet access. My wife enjoys the occasional youtube video, and that’s always working. I like to watch movies on hulu, and that’s always working. I connect to two, three, four remote servers (rdp and ssh) for 12 hour coding sessions, and never get performance issues or problems.

    For me, Time Warner delivers great value.

    However, if I ever hit a bandwidth cap, and I get disconnected in the middle of work, or my wife has to give up an $800 translation job because of no internet access, I’ll be mad as hell and will be looking for alternatives faster than you can Google “internet access provider in zip code 91325”. So far, though, everything’s been peachy.

    Disclosure: Nobody in my immediate or extended family works for Time Warner or any of its affiliates, nor do we receive any sort of preferential treatment, discount, or other compensation for speaking so highly of them. I’m just a very satisfied customer.

    Comment by Chris Mahan -

  45. Why don’t we just go back to pay-per-minute, like we did in the good old days of AOL, Mark? Sorry, nothing is going to convince me that this mid-stream horse change is anything but cable companies trying to kill bittorrent by pricing it to death.

    Comment by researchrants -

  46. Usegae caps do nothing but make it where companies like Time Warner dont have to upgrade their network, expand it but put more users on the same pipe. It gives them a tool to leverage us as the consumer to slow our useage, slow development, slow streaming, so they wont have to spend their profits so fast on upgrading an already aging technology in twisted copper and coax cable.

    Comment by Jason S -

  47. This is really only an issue for cable companies serving bandwidth over coax and regional telcos serving bandwidth over copper. Telcos using fiber shouldn’t even have to cap their customers (and would avoid the PR headaches cable companies are suffering through now).

    Comment by Deven Nongbri -

  48. I think they should have approached it as tiered pricing for minimal internet users. Instead of saying they are introducing bandwidth caps, they should have pushed out a lower transfer amount for lesser internet users. Opting into the tier.

    I agree with Ricky too, if they would give people a record of their actually real time usage, then they would know what tier they want and move to it easily.

    Comment by Trae -

  49. Also, if you’re going to introduce tiered pricing, I need to have a tool that I can download and use – on any ISP – to know what my usage is. If I pay $50/mo for ‘unlimited’ right now, and you want to switch me to $30/mo for 10GB, I’m going to resist it every which way. But if you can offer me a tool that proves to me that I’m only using 5GB anyways, then that extra $20/mo starts to look MUCH more tempting, knowing that it won’t interfere with my normal usage.

    Comment by Ricky Cadden -

Comments are closed.