Blog Post

Why Tiered Broadband Is the Enemy of Innovation

It should come as no surprise: Incumbents are beginning to act like incumbents. But while the cable companies are the first ones to jump on the tiered broadband bandwagon, they won’t be the last. Their argument for limiting bandwidth and data transfers based on price sounds like a good idea, especially as a way to get bargain hunters to buy. In the long run, however, tiered broadband is a terrible idea that will bring the innovation inspired by flat-rate broadband to a screeching halt.

Flat-rate broadband – however cheap or expensive (depending on your point of view) it might be – inspired the formation of Skype, YouTube, Facebook, Apple’s iTunes and MySpace, amongst others. It allowed us to freely experiment, to embrace both the applications and the ideas they represented, such as VoIP, online video, digital downloads and social networking. [digg=]

The emergence of these applications has, in turn, spurred demand for broadband in the U.S., much like the illegal version of Napster jump-started the demand for cable and DSL broadband in the late 1990s. And they’ve helped lift the number of broadband subscriptions to U.S. cable and DSL companies to 69 million by the end of 2007, subscriptions that have brought in enough cash to pay for the cable companies’ foray into voice and to help with their digital transition. Yet now these guys want to slaughter the golden goose. Why?


The answer is in my living room. Thanks to a fast connection from Covad, I now get my video fix over the broadband pipe. Apple’s iTunes, Jaman,,, CBS and scores of other services make it possible from me to watch shows either on my laptop screen or, in some cases, on my big-screen TV via Apple TV.

I used to pay Comcast about $150 a month, but now I pay them zilch, instead forking over a mere $30 a month to Covad. Oops! In the future, the emergence of much higher-speed DOCSIS 3.0 and fiber-based broadband will make it even easier to download or stream videos, which scares the bejesus out of the phone companies. And that is one of the reasons they are introducing tiered broadband.

But consider the bandwidth caps. I asked some of my telecom sources to help me put into perspective the new tiered-pricing structure with which Time Warner Cable is experimenting. TWC’s lowest price tier – 768 kbps at $29.95 a month for 5 Gbytes and $1 per GB – may seem reasonable, but it isn’t.

If you assume that we’re pulling down data at a steady 20 kilobits per second for every second of the month, the total monthly transfer comes to about 6.8 gigabytes. At a higher speed of 768 kbps, that jumps to over 250 gigabytes, and at 1 megabits per second, the monthly download will hit 324 gigabytes. At first blush, those look like awfully generous numbers. After all, who uses their connections consistently?


However, if you take into account our average behavior online, data transfers start to add up really fast. Stacey crunched the numbers yesterday and came up with an interesting conclusion: If you bought the monthly 15 mbps/40 GB transfer option for about $56 a month, you’d get about 40 hours of standard definition video along with enough bandwidth for your normal browsing and surfing habits. That’s just over 75 minutes of SD Internet video every day – two or three shows at best – which means you might need to continue buying the “video connection” in order to watch more television. Sure you can slice and dice the data transfers with other online activities, but this is all about video.

From that perspective, you would think that Comcast’s proposal for 250 GB a month is pretty reasonable. Actually it’s not, especially if you factor in how quickly we’re moving towards HD downloads. With HD, each roughly 2-hour long movie is going to consume about 8 GB, while live sports events, etc., when watched in higher quality can take up some 13 GB. Remember we share our Internet connections with multiple people in a household. So Before you know it, that 250 GB isn’t enough.

Cable companies are trying to convince Wall Street that they need to upgrade their networks to DOCSIS 3.0 in order to compete with telecom operators, especially those with fiber connections. The idea of metered broadband makes the big spending on these networks more palatable for Wall Street.

As for consumers, the cable companies have evoked the P2P bogeyman. I spoke with Time Warner spokesperson Alex Dudley, who claimed that some 5 percent of its user base abuses its network through the use of P2P, causing problems for the remaining subscribers. “Video is the most bandwidth-intensive use right now, and it is not people that go to iTunes but instead it is P2P which sucks bandwidth in the system,” he said. There are some questions about that claim.

My biggest fear is that as these companies try and protect their video revenues, they are
doing more harm than good, and putting roadblocks in the way of interesting services
that make broadband worth having. When I asked Dudley if his company was putting innovation at risk by limiting flat-rate broadband — if they might be throwing the baby out with the bathwater — he noted that many of these startups and services are built on their infrastructure.

“You need to understand that the networks are going to be managed and we need to make profit,” he said. “We are trying to find a balance here, and it is too soon to say that we are throwing baby out of the bathwater.”

Dudley was, however, quick to point out that TWC’s experiment in Texas was just that – a test. If consumers don’t want it, the company is going to back away from it. “I think this is a trial and we are going to learn from this trial,” he said. If the results of our poll are any indication, they would be wise to back away from it — and soon.

123 Responses to “Why Tiered Broadband Is the Enemy of Innovation”

  1. As has been pointed out in a few posts above, let’s not confuse innovation with free-riding. Many of the “innovators” Om points to took advantage of a distribution network they didn’t have to pay for. That’s not their fault but it is hardly an example of innovation.

    As soon as the operators of those “free” distribution networks started expressing some discomfort with things, the “innovators” rounded up the “lawyers” and headed off to the epicenter of anti-innovation, Washington DC. But that’s a story for another day.

    Nevertheless, I agree with Om that “metered” (not “tiered,” which we’ve had for years) broadband packages are ill-advised. And not because they will stifle innovation (er, free-riding), but rather because consumers don’t like them. Any business that foists products on its customers that they don’t want is bound to fail.

    A better approach, for all sides of the debate, would be to open up the pipes to any and all traffic. What will inevitably happen, in fact what is already happening, is that some traffic (like long-form, high-definition video) will perform poorly. Ditto VOIP. Ditto gaming. The broadband provider can then step in with quality of service (QOS) packages that enhance performance for select traffic flows…at a price.

    Everybody is happy. Consumers can get better QOS if they want. Innovators don’t have to pay for network access. No one is capped. And broadband operators develop new revenue streams that help pay for broadband upgrades.

  2. ronald

    I don’t know, this setup (semi competition market, and heavily regulated) reminds me of our health care system and the mess we have created.

  3. Martin Rijssel

    BTW don’t you think deep packet inspection is not an *innovation*?

    I hate it for privacy reasons, but technically it’s a bigger innovation than YouTube.

    It was created to manage the cable networks better — that was the initial market.

    But it sells well to the Chinese government…

  4. Martin Rijssel

    you live in the lala land of people thinking all goods should be free. Why should innovation only work with plentiful resources and NOT under some constraints?

    Some commenters above talk about Facebook. How would they be affected by a metered internet?

    I agree with the statement above that current low ISP prices are due to the oversupply since the late 90s. In a more bandwidth-constrained market we must understand that broadcasting (over cable or DSL) is orders of magnitude more efficient than sending each video individually to each user. P2P reduces the cost of the websites at the expense of the ISPs, but doesn’t use bandwidth more efficiently.

    Now that’s a place for innovation — in a realistic environment.

    Give up your communism. Russia went down because actual cost was widely ignored, not because of the strength of the USA. Let the market win.

  5. Hello Om
    Maybe this is what they should do !
    I believe that they can’t put up with their customers with unlimited service.
    Nit tot only the U.S BUT ALSO in Canada and Australia the service have become limited !

    I still think it’s really bad

  6. broadband_fan


    Why do you presume that metered broadband will quash innovation? Quite the contrary, I think if we value a resource more efficiently, users will become more innovative in its use. Perhaps forcing protocls and applications will to work more efficiently.

    I think smarter broadband pricing is coming whether we like it or not. As it stands today, a small number of users are consuming a disproportionate amount of capacity for a given price level. This needs to be reconciled somehow. In the end, an efficient price scheme allow everyone to pay for the capacity they consume.

  7. @ John Thacker…. Now that you bring it up…. You might be right. I should have called this “metered broadband” instead of tiered broadband. Now i need to figure out how to correct that…. :-D

  8. @ vijay gill about the comments in RSS…. Since we use for hosting our sites, we are limited by what we can do the feeds etc. But good point and I will put that forward to the WP team and see if they can help us out.

    @ vijay gill … regarding your other comments, while I understand that building networks isn’t cheap and costs a lot of money, but remind me, isn’t that cost of doing business if you sell broadband. I think you might have an interesting information on how this can be done :-)

    Randomly musing: I wonder why people at or similar carriers in other countries don’t complain about the costs?

  9. John Thacker

    I don’t think your point is really about “tiered broadband.” “Tiered broadband” is simply paying different prices for different levels of speed, which is a separate issue (though some people do complain about that as well). What you seem to be complaining about is “metered” broadband versus unmetered– prices per GB. At least you do use “metered” elsewhere in the piece

    It’s certainly true that metered local phone calls kept countries outside of North America back when Internet access was mostly via modems. At the same time, the very expense of narrowband connections contributed to the rapid adoption of broadband in those countries, in comparison to the US where many people have decided that they would rather stick to dial-up, even in areas where broadband is available. In practice, people greater prefer unmetered access, so I’m have strong doubts that this will be effective.

    Incidentally, when I was at Cornell a few years back the University switched from a flat-rate price of around $50/month for campus connections to a plan similar to this– $20/month with 2GB of transfer free, pay per amount transferred after that, break even with the old price being around 20 or 30 GB per month. The only amount metered was from outside Cornell’s network; intranetwork transfers were free. Students responded by creating locally hosted P2P services for popular large downloads. Indeed; most students and users saw their monthly costs go down, as did the University’s bandwidth costs. The ability of students to form locally hosted P2P networks and the division between Cornell’s network and the outside made it fairly different from a typical consumer experience, though.

    Bryan said:
    I’m for Net Neutrality. The United States is known for innovation. … We need to take examples from South Korea and apply it to ourselves

    Oddly enough, South Korea doesn’t have Net Neutrality, neither legally nor in practices. The Big ISPs in South Korea all run their own websites and services, and have been repeatedly caught blatantly favoring their own traffic over that of competitors.

  10. We’ve always had implicitly tiered rates, and it hasn’t quashed innovation. There was a time, not that long ago, when many more people were on dial-up than on broadband, and those who were on broadband paid more for it. That’s a tiered rate structure in the market. And it didn’t quash innovation. If you force a single rate structure, then it will either be (a) too expensive for most people, or (b) too cheap to make it worth the providers’ while to provide it. If you allow tiering, then people who really want fast speeds will pay for it, and it will be worthwhile for providers to do the innovation they need to do to create faster speeds – not to mention create incentives for competitors like Covad to jump in.

  11. Why don’t the telcos and cable companies take a look at the CDNs?

    Netflix, iTunes, Jaman,, and CBS typically pay a CDN to deliver their content. Shouldn’t those CDNs share some of the fees with the ISPs over whose networks the content is delivered to the end user?

    Rather than punishing the end user for using the service they were sold, they should look at they guys who are dumping the traffic onto their networks in the first place…

  12. @bostonwolf: If they really are doing this to inhibit competition from Netflix, that is just one more bit of justification for forcing a split between the business of communication provider and the business of application (in this case, televison) provider. (Not that I need any more justification, but some people might.)

  13. Allen

    Current prices for core internet access are largely a result of oversupply – remember the tech crash of the late ninety’s – early naughty’s? Billions were invested (and lost). You cant compare this with consumer access – while they are both fruits, apples are not oranges.

    Up until the last few years existing copper lines and coax were largely sufficient to support consumer access – basic web browsing and email really aren’t all that demanding.

    Then the fundamental nature of the traffic changed – the design rules no longer apply. Huston we have a problem.

    It’s hardly surprising that these companies, responsible to their shareholders to *gasp* make money are weary of throwing it away in expensive upgrades for dubious incremental revenue.

    It seems to me that a user pays approach is a rational alternative to say deep packet inspection and P2P rate limiting, that just might allow someone to build a business case to upgrade the access to cope with what has been a fundamental shift in traffic behaviour.

  14. Dimitrios Matsoulis

    Om I agree with you. Tiered 3G has led nowhere in Europe and the people that use it is those that have the bill paid by their employer. If that was introduced in landlines it would definitely make the Internet’s progress slower.

  15. vijay,

    why is it that you think broadband subscribers should be pleased to contribute to the obscene and excessive 40+% operating margins earned by the telco/cable duopoly? Both cable and telco have high fixed, low variable cost structures, so what’s with this charging a high varible price for service? That’s monopolistic abuse, plain and simple. Don’t give me all that bs about matching pricing to capital requirement. Cable cos are notoriously negligent at upgrading their plant – that’s not the subscriber’s fault, and the fact that cablecos now face large bills for their “deferred maintenance” is a risk their shareholders should appopriately bear with new sources of financing, not their customers via price increases. There is no free lunch – not even for cable cos.

  16. @Vijay — sort of my point. If carriers are introducing tiered services, that means they don’t fear any competition. Or they are milking customers until the other half of the duopoly arrives. So then the argument goes, time for government to step in.

  17. vijay gill

    @paul: What competition? Can you point me at some companies? Clearwire doesn’t count.

    @andrew: the core costs are on the order of $10-$20 per mbps depending on how much handwaving you can do with your books. The tough costs are in getting to the provisioned core, the last mile, and the capital needed to build out there

  18. I reside outside of Philadelphia and have already been affected today by the illegal Comcast throttle. I’ve been downloading a single torrent and have watched it go from 100kbs to almost zero in a pattern. This is beyond annoying; their TV programming is awful, now they’re ruling over my download habits which I pay too much for…

    Please FCC, remove thy thumb from thy =asses= and freaking do something about this. Gas prices are being ignored, don’t let it spread to our way of communication…

  19. You mention “In the future, the emergence of much higher-speed DOCSIS 3.0 and fiber-based broadband will make it even easier to download or stream videos, which scares the bejesus out of the phone companies. And that is one of the reasons they are introducing tiered broadband.”

    But, this whole post is about cable companies. Have any phone companies introduced tiered bandwith caps? It seems that only CableCos have, not PhoneCos.

  20. Ronin8317

    There is a huge danger with data limit. It will be a ‘backdoor’ way to get around net neutrality. ISP will tell Google and Apple to pay them money to stop data downloaded from being counted toward the cap. The fee will lock out anyone else coming into competition. Sites like Facebook and YouTube will never get off the ground.

    Getting a higher return on investment is fine. Controlling what user can see on the internet is not.

  21. Right now we don’t have effective competition. We have artificial competition between regulated monopolies (Phone & Cable companies). Lets take away the regulation safety blanket from these guys and allow competition to regulate. This is the only really effective way to regulate this activity. If you misbehave and you have competition you will lose customers. Right now they don’t have to worry much because they know the other guy is just as much of schmuck as they are.

  22. bryan lor

    I hate to see this being implemented throughout the nation. I’m for Net Neutrality. The United States is known for innovation. By having tiered services, that innovation will no longer exist. We need to upgrade the infrastructure, not squeeze every last drop of money out of the consumers. Oil companies are already trying to squeeze as much as they can from consumers and soon, those cable companies will become the new “oil companies” of next generation. If you look at South Korea’s infrastructure, they have one of the fastest residential broadband services worldwide, but that’s because their digital information infrastructure is superior to the United States. We need to take examples from South Korea and apply it to ourselves. I hope to pay $50 one day…for speeds as fast as T1, like in S. Korea, but with how things are looking, not sure if it’ll ever happen.

  23. Tiered pricing and complaints about how costly it is to build out infrastructure seems to give credence to the opinion that last-mile connectivity should be funded by the government, since it is too important to be left to the whims of Wall Street. These schemes just smell like someone trying to squeeze every last drop out before competition arrives.

  24. To some extent I can understand why the cable companies are doing this.

    Netflix just unveiled a box to deliver video over your broadband connection. Essentially they are piggybacking on your provider’s bandwidth for nothing to provide a service that directly competes with the cable company’s video on demand service.

    I would probably try to protect my investment as well, unless Netflix was willing to cut me a slice of their box income.

  25. Andrew

    An interesting observation is that core bandwidth (i.e. bandwidth that is guaranteed all the way to the backbone exchanges) only runs about $10-20 per Mbps-month for any decent sized network operation, and less if you are really big. Or to put it another way, it is costing the telcos on the order of $0.03-0.05 per GB for core, non-oversubscribed bandwidth. The rest is overhead and margin, and the installation charges are designed to cover the overhead; once a circuit is up the overhead is negligible, so the bandwidth billing becomes gravy.

  26. vijay gill

    @brian: because like I said, look at their free cash flow. if they just spent that on builds and not on buybacks, their stock would tank.

    @don jones: precisely, some one does need to disrupt them. Since it appears to be quite an easy job, I am eagerly awaiting the legions of business plans and associated funding.

  27. brian

    If it is so costly to upgrade their networks to support greater bandwidth demands (which btw is still a fraction of Asian nations like Japan and Korea today), why do they insist on using their cash flow to buy back stock?!

  28. vijay gill

    Also, your platform doesn’t show comments in RSS readers for some reason, unlike techcrunch. And the thumbs up thumbs down links appear to be broken

  29. vijay gill

    The fundamental problem is that upgrading the infrastructure to support very high demand _all the time_ is a non trivial exercise. A node split esp. on a full plant is non-trivial exercise in expense, not to mention all that backhaul across expensive gear and up to the big I Internet. An 10GigE linecard or port, plus the router allocation, is not that cheap either. Might want to ping juniper or cisco for what that stuff costs. Assuming you can discount 50 points off list, it’s still real capital. A good example would be to look at the capital spending of public cable companies like Comcast, vs. their free cash flow. That exercise would be illuminating – I did a back of the envelope and the return from free cash flow if someone bought the company, would be over 30 years. Not very tempting. As a result, as usage spikes, and buildouts continue, someone somewhere is going to get the squeeze or their stock will go to zero very quickly. Tiered broadband is one way to push out usage before upgrades, or alternatively, get more money to finance those upgrades. It is entirely possible that the first mile is a natural monopoly, and while there are going to be a lot of folks shouting about how it doesn’t cost that much to build out broadband, to them I have only one thing to say – if it was that easy, get a business plan together, fund it, and profit!