What Does a Gigabyte Cost, Revisited

Earlier this year, I wrote a report and a blog post that attempted to estimate the cost for Canadian ISPs to transport a gigabyte of data. While there is nothing magical about a gigabyte, since overage charges typically levy by the gigabyte, it is useful to link actual costs with consumer costs that can reach as much as $10 per gigabyte. In fact, this data is likely to become increasingly important as the wholesale usage based billing battle has shifted toward a cost-based approach.

Unfortunately, there is little reliable information on actual costs with ISPs typically claiming that such information is a trade secret that should be kept confidential. My report noted that there are two elements to the actual costs. One is the Internet-facing data costs, which arise once a user’s traffic travels onto the public Internet. This cost is very low, estimated in the report at about one cent per GB and falling. This is consistent with public transit arrangement pricing and is likely even cheaper for large ISPs that use peering arrangements to cover off most actual costs.

The more difficult calculation involves the internal ISP network leading to the public Internet. As CRTC Commissioner Candice Molnar noted during the usage based billing hearing, “we all, I think, can hopefully agree that there is no marginal cost to using the network when you are not causing augmentation.” While there are no marginal costs, there is a capital cost of building the network and ongoing maintenance and augmentation costs when congestion arises due to traffic growth. My report used Bell’s data in the deferral account case (one of the only ones to put information on the public record) to estimate that seven cents per gigabyte (for a total of eight cents) was a best guess among a range of possibilities.

Bell’s response has been to target the study with claims that it is inaccurate. At the Canadian Telecom Summit, Bell’s Mirko Bibic specifically focused on the study, saying he was the “happiest person in the country because we can finally get the facts on the table and deal with some of this nonsense that is out there. One particular piece of nonsense is this study that Michael Geist did for Netflix on costs.” During the hearing, Bibic told the Wire Report my report was “very, very flawed.”

Yet Bell’s recent comments about actual costs suggests that it relies on similar ranges to provide a best guess at cost per gigabyte. Over the span of one week, it effectively provided three different numbers. On July 13th, Bibic told the Wire Report that 19.5 cents per gigabyte, the original number in its AVP proposal, was close to its actual costs. Bibic claimed “they bear a fairly close relationship to the nineteen-and-a-half cents. The nineteen-and-a-half cents [per gigabyte] is justifiable based on the actual costs of a gigabyte on the network.” A week later, Bibic told the CRTC that actual wholesale costs were $39.15 per Mbps, which works out to 12.4 cents per gigabyte at full capacity. A day or two later, it submitted that 17.5 cents per gigabyte reflected actual costs plus markup. The 17.5 cent number appears to be based on fairly low wholesale usage, however (roughly 63% usage).  Increasing estimated usage would result in a significantly lower estimate on the cost per gigabyte, likely closer to 12.4 cent estimate.

Bell’s inconsistency aside, the company submitted a response in April on my report after Telus conveniently asked about it during the interrogatory phase. It claims that my estimate is based on assuming maximum usage every month (900 GB), which obviously results in a lower cost per gigabyte. It notes that average users do not come close to using that maximum capacity, particularly since it is a shared network. My report included a pricing scenario based on theoretical maximum output (F) and it sets the lowest possible price per GB for each equipment configuration [$0.30 for a pure OC3 configuration; $0.08 for a pure OC12 configuration, and $0.02 for a pure OC48 configuration). Bell seems to assume that this is the OC12 F scenario that is the source of the eight cents per gigabyte figure.

Yet that is not how the report arrived at the figure. It identifies a range of pricing options based on different equipment configurations and usage patterns which range from as high as from $0.34 to as low as 0.019/GB (see Table 5 at page 34). The report notes how the pricing data relied on for the study was created for rural areas and for a complete buildout of the network (as opposed to simple network maintenance through provisioning for increased flows on an existing network). This suggests that actual prices will be at the lower end of the spectrum presented in table 5 (the costs for an urban network are likely far lower than a network that Bell itself says would not be built based solely on market considerations).

The report never suggested that the cost should be based on the assumption of maximum usage. Instead, it concluded:

So, in the long run, the low end of the range would again seem to be the better estimate as demand for peak bandwidth increases. Consequently, while there is a range of plausible values, a comparison of Tables 5 and 7 suggests a figure of $0.07/GB may be the best single estimate for the costs of the internal network.

In other words, there are a number of possibilities, but the seven cents/gigabyte figure appears to be the best guess based on Bell’s own numbers. Bell’s claims now cover a range as well, reflecting the reality that per gigabyte pricing is subject to many variables that result in a spectrum of possibilities (the most important variable is capacity/usage since the higher the assumed capacity and usage, the lower the cost per gigabyte – fixed costs to build divided by larger or smaller amounts of data depending on use). Arriving at a common metric for identifying per gigabyte pricing will be a crucial aspect to resolving the wholesale usage based billing issue since the issue is subject to considerable variation. It is hoped Bell will step up with public data for all to scrutinize, or at the least that the Commission treat Bell’s proposed figures on cost with the skepticism it deserves.


  1. I think you meant OC3 instead of OC4, right?

  2. This is all very confusing. Bell says its cost is 17.5 cents at 63% fill. You say its costs are 1.9 to 34 cents under a range of fill and pipe-size scenarios, and have suggested 7 or 8 cents is the best single number to go with within that range. Is that about right?

  3. We’re still being gouged
    I find it sadly hilarious that Bell is quibbling over a few cents on an estimate while charging customers at least 10x more than any of these figures.


  4. Top Hat Connoisseur
    Nice. A markup of 992¢ on an 8¢ purchase. I wish that was my business model.

    When are these guys gong to get into banking?

  5. Bill MacEachern says:

    As if…
    Bell wouldn’t fudge the numbers to support whatever self-serving claim they make.

  6. Ryan- They are already into banking. They are laughing and running all the way to the bank with our money

  7. High mark-ups…
    … High mark-ups are perfectly acceptable in low volumes.

    Unfortunately, Bell controls the volumes with caps, and network augmentation. Where oh where does this leave us?

  8. Why does a liter of water that costs 0.003 cents when it comes out of your tap cost $1.50 at the corner gas station. It’s the same stuff! Let’s start a petition and march on Ottawa – after all water is more essential to life than a gigabyte!

  9. Karl Stevens says:

    @Cynic – if the water companies were limiting the amount of water that you could get from your tap so that you would be forced to buy bottled water from them, you might have a point.

    If that’s the best argument you have, why don’t you just admit that you’re wrong?

  10. @Karl Stevens
    Wrong about what? If you were deprived of water you would be dead in a couple of days, no?

    If you were deprived of Gigabytes I suspect you would probably last a bit longer … depending on how severe your internet addiction is.

    Why do we get all bent out of shape over the allegedly unjust cost of a Gigabyte when we accept the same injustice for water which is even more important to our survival?

    I find this whole debate about the cost of a gigabyte to be disturbing in that so many people assume that you build it once and forget about it. The telcos primarily use DSL technology in their access network and it has been through 7 generations of upgrades in the past 12 years costing billions of dollars. If they suspended their capital investment programs and just sucked the cash out of their existing network it would make sense to price service purely on a cost plus 15% basis but that would leave us all very angry in a few months when the exponential traffic growth uses up all the fixed capacity of the networks.

  11. Devil's Advocate says:

    Yeah, it’s just too bad “gigabytes” don’t come to us through a “tap”, in order for your analogy to even begin to work.

    You should have found Karl’s version much more fitting as a parallel scenario, if you had cared to absorb it.


  12. @Cynic
    I have been working in networking for well over 30 years. I am well aware of the various factors involved.

    You can’t rely on a physical analogy to justify a position on a virtual concept. Attempting to do so shows your lack of technical understanding in the issues.

    The analysis above does a reasonably decent job of presenting the complicated issues involved around determining the “cost” of transferring a gigabyte of data. They are not that far off (his are slightly higher) what I have seen in various businesses and industries, even to the observation of “the higher the volume/speed, the lower the unit cost”.

    Capital investment, ROI, and profitability all enter into the picture. It has been pretty well established that UBB is the wrong mechanism to apply to the issues. In Canada, in the US, in Australia, in Japan, and in Europe. The “cost and amount of gigabytes transferred” bears no relationship to any of the above points.
    You might consider that in Japan, the “cost per gigabyte transferred” is higher than we have over here, yet they manage to supply customers with higher speeds at cheaper prices, and no UBB. Yes, they are different geographies and markets, but the point is that UBB is effectively irrelevant in the market.

    So the question is why are the ISP’s even considering a UBB billing scheme? The answer is simply “because they can”. It’s “free money” as far as they are concerned. Once established, it will be hard to remove. Especially since they can use such a scheme to constrain or eliminate competition.

    BTW.. In your “bottled water” analogy, a better analogy would be the water company looking at what you pay for water in the bottle, and deciding to increase the price of the water from the tap to match. Why? “Because it is what consumers are willing to pay.” Cost, ROI, and capital investment, would have nothing to do with such a decision.
    Just as cost, ROI, and capital investment, have nothing to do with UBB.

  13. @oldguy
    You got me beat by a year. 29 years ago I started gluing semiconductor lasers to strands of glass fiber to see if there was any practical use for some new fangled idea call fiber-optics – it turned out to be a fairly good idea once we figured out how to stabilize the glass so it wouldn’t go dark after a few years.

    I respectfully disagree with your assertion that cost, ROI and capital investment have nothing to do with UBB. Personally I don’t like the UBB idea either, but it is a legitimate method of managing the supply/demand equation. AT&T announced just a few hours ago that they will manage the supply/demand issue on their wirless broadband networks by throttling the access speed of “unlimited” users when they reach the top 5% of heaviest users in a billing cycle – check it out here:

    Personally I would like to see a tiered service model with flat monthly fee for a given speed/capacity and you buy a plan that meets your needs. Light users pay less and heavy users pay more – just like gasoline, water, electricity or beer. What could be more fair than that?

  14. Karl Stevens says:

    @cynic – wrong about your entire post. Everything. You are making claims that are entirely unsupportable.

    To wit: your post contained the most asinie analogy I have ever heard. The fact that you *continue* to attempt to use it, when I have proven that it has no bearing whatsoever shows that not only do you know you are wrong, but you are intellectually dishonest as well.

  15. @Karl Stevens
    Don’t know who you are and don’t really care but if your only contribution to this discussion is to suggest you are intellectually superior and anyone who doesn’t agree with you is wrong then I would suggest you find another playground, this one is for grown ups. Oh, by the way genious you mispelled “asinine”.

  16. @Cynic
    I have you beat by more than that. I have been working with this stuff a little longer than ethernet, never mind the internet. I say well over 30 years, because modern networking was in it’s infancy around that time, older networking isn’t relevant. Go check on when thick ethernet first came out and add around 10 years.

    The problem with UBB, is it rewards heavy peak time burst users, and penalizes steady trickle users. When you provision a network, you provision for peak time capacity, not overall “usage”. In fact, “usage” is only relevant within the peak capacity times as a measure of the “use” of that capacity.
    UBB has nothing to do with peak capacity planning or provisioning. Once you provision for peak times, any other “usage” is irrelevant. It takes the same amount of maint to keep a network running at 1% of capacity as it does to keep it running at 99% of capacity.

    I know what AT&T is doing, as are Comcast and others. But that doesn’t change the fact that UBB won’t “solve” anything. It isn’t the “high usage” customers that are driving bigger capacity requirements, it is the narrow time frames when every “low usage” customer is hitting the network at the same time. And each of those “low usage” customers are needing more and more.
    This isn’t anything new, every business and network for the last 20 years has seen this behavior. You plan for it and provision for it. Identifying and even curtailing “high volume users” doesn’t change the stats curve one bit. The peak capacity “usage” stats don’t budge. Outside of peak times, high or low “usage” is effectively no additional cost, it’s lost in the rounding errors.

    This is why I say “usage” (as in UBB) is irrelevant in cost, ROI, or capital investment planning. You plan for the peaks, and that has nothing to do with individual monthly transfer stats. Once you provision for those peaks, individual monthly transfer stats are irrelevant.

    Throttling during peak times is a valid temporary measure, until you have increased peak capability. UBB doesn’t fit into any of the models, except as a “because we can” money grab.

    The water pipe analogy isn’t a bad one, but you have to be very careful. It’s not like you have to treat each gigabyte for usability, nor do you have a limited size reservoir. What you are delivering “through” those pipes is zero cost and unlimited.

  17. random data guy says:

    On costs
    These days, OC3/12/48 are going out of use for data transport and relegated mainly to telephony (eventually they will be gone there as well). Major carrier networks use far cheaper and more capable (as in capacity) technology based on long-haul optical ethernet. All the usual suspects in canada are using it already (and have been for a little while). 10GE (10 Gbps) is now common, and everybody is looking at 40GE (already available), 100GE. With this, the cost of transport becomes more and more irrelevant. The initial investment to install the fiber plant, and the access equipment (DSLAMs in the case of Bell) is the major factor in the cost, as well as the network operation; and then the question comes as to what amortization period to use. For equipment that could be 5 years; for the fiber plant that could be 10.

    Everytime carriers change transport technology, their per-bit transport costs drop dramatically (if OC3 is “X”, GE ~ X/10; 10GE ~ X/100 etc); and they haven’t been passing those savings to the subscribers for the last 6 or 7 years. What Bell and Rogers are doing with their ridiculous caps is just plain outrageous.

    The internet access itself, as you’ve identified, costs relatively nothing. Major IPSs with tier-2 status (Bell, Rogers & friends) have very very little IP transit costs (other than their equipment costs); even for Tier 3’s (typical “wholesaler ISP” case) the costs are next to nothing (e.g. $1.5/Mbps/Mth for a 10GE pipe to a Tier 1)

    While I haven’t been in the industry 30 years, I was there when DSL deployments took off working on engineering, design and costing (back when we did use DS3 and OC3 links to the tier ones). Then I moved to wireless (cellular/3G and up). It is no secret that the average cost for 3G internet access services for North American operators is in the range of 1c per served MB. That is approximately 100 times more than in wireline (specifically: dsl – cable is even cheaper).

    Hence the comparison of ATT/Wireless (same as T-Mobile) with Bell Canada is incorrect. In wireless, the caps are in the range of 5GB (from 3 to 10GB). And the caps are shifting to “hard” in the sense that you can’t have overage (ATT new plan, T-Mobile, and others here). This is done to drive down average consumption and avoid operating at a net loss, as well as to remove the risk of bill-shock to subs.

    If Bell were to do the equivalent of ATT Wireless or T-Mo, their cap should be in the 500GB to 1TB range; Rogers should probably be in the 1.5TB range. That’s why the 25GB “CRTC Cap” or the current ~60GB caps from Bell and Rogers are a complete joke. It’s like going backwards on time, to an era were GE / 10GE / 40GE / 100GE transport did not exist and DSLAM ports had a cost of 500 bucks.

    Essentially, were it not for the wholesalers and the absence of UBB, broadband access in Canada would be a total rip-off.

  18. User Loser says:

    Competition who needs competition?
    Saw a thing on Here and Now a pbs program. I think everybody should see it. It is pathetic that the United States and Canada are going down the road they are.

  19. Lol
    If there is no additional costs after base then charging per gig will result in ever increasing profits the more gigs that is run through existing connections. More gigs less the cost basically. Meaning increasing price lowers the use which increases the rate. Counter to free market functionality. These assholes did this in the long distance telecom days and made a fortune at the expense of Canadian globalization.

  20. Un-Trusted Computing says:

    Punitive overages
    Basically this is about protecting higher margin legacy businesses nothing more. Why subscribe to FibeTV or Cable when you can get Netflix or another over the top video provider? Punitive overages will make these services more expensive than traditional TV services which is why the overages are priced the way they are.

  21. Canadian Consumer says:

    Canadian internet innovation is getting destroyed due to Bells greed. At what point does More’s law figure in to the equation? How will the consumer benefit from the fact that our speeds and hardware increase every 18 months?

    It seems once they put this toll on the net, it sends CANADA into the dark ages. CRTC needs to be disbanded and Bells and Rogers internet monopoly needs to torn into sunders.

  22. Asking the wrong questions…
    It seems that we are asking the wrong questions. Maybe instead of asking “How much is a gig worth?” we should be concentrating on “What is the cost to have someone online?”

    Really, transporting a gig of information costs less the more the network is used so it sounds like they are putting a fixed cost on a variable. Why not steer the discussion to the right source of the revenue, the cost of the connection period? That way we can get Bell and all the other providers off the per gig band wagon and into the real cost, the connection period.

  23. blatanville says:

    @Un-Trusted Computing
    Ding! Ding! Ding!

    All discussion of the “price per gigabyte” (or other metric) become irrelevant when those who control the channels also control the content (BELL, Rogers, Shaw). It is strongly in their interests to throttle your access to other content providers (Netflix, CBC streaming service, Spotify, etc.) who don’t have their own delivery channels.
    This argument isn’t about the channels, it’s about the content! Divorce the content providers from the supply channels (at the corporate level), and you can get entirely away from the UBB discussion.

    Is it sacrilegious to suggest a national Crown Corporation to handle the channels, leaving the content in the hands of the private corporations?

  24. Asking the wrong questions…
    Right. The cost of provisioning the network for a certain capacity is a fixed one. Ongoing maint and support is the same regardless of what capability you provision. Nor does it matter if the network is “used” at 1% of capacity or 99% of capacity.

    As you put it, what is the “cost to have a customer online?” That’s not an easy question, technology costs are always dropping, volumes of scale come into play. Looking back, I would say the “cost per customer” is less (perhaps considerably less) than it was 5 years ago, even if you factor in increased peak usage per customer.

    The problem comes when the ISP builds their pricing plans per customer. They start with an assumption that each customer will only use x% of capacity, and prorate the “price” based on that “usage”. Fair enough, but they miss the fact that “usage” must be factored into peak time periods. If you have 1000 customers, each using 0.1% of capacity during peak periods, your network is at capacity. If 5 of those customers continue to use the network at 0.1% of capacity outside of those peak periods, it is hardly noticeable.
    If the average “usage” per customer jumps from 0.1% to 0.2%, the ISP has a problem. It needs more capacity. But the “usage” outside peak times is still hardly noticeable. This is the situation we have today. The “jump” in peak usage has been dramatic over the last 2 years, possibly 3-4 times what the ISP built their provisioning plans with.

    If the ISP throttled or blocked every one of the customers with high “monthly usage” (the ones that continue to use outside peak times), they wouldn’t put a dent in the network peak usage stats.
    Throttling each of those 1000 customers to 0.1% during peak times allows you to “fair share” the available capacity, but there is another problem. The ISP has “sold” each of those customers on a “speed” that is 100 times what they planned the customer would actually “use”. Throttling would upset a lot of those customers.

    There is a way to spread out the “peak time usage” issue. Put a premium on network usage during peak capacity times. Even though this would directly address the problem, it would upset a lot of people.

    UBB concepts come from someone that sees “usage” as physical consumption, not from someone with a background in network provisioning. They hear the word “usage”, and jump to false conclusions.

  25. wrong direction.
    There should be no usage based billing. if it costs $60 per month to deliver 5 meg service to a persons house 24 hours a day 7 days a week then that is what we should be billed. EVERYONE. The internet is a community. People meeting, doing business, playing games… you don’t pay $20 to enter an amusment park then pay another $70 because you stayed for 7 hours… The internet is NOT a commodity…

  26. I don’t like the idea of UBB at all but we do need to have a funding model that allows the network providers to continually upgrade the network. When I first got a 1.5 Mb/s severvice from Telus back in 1999 it had a 2 GB per month cap and cost $43 per month. Today I pay $37 per month for a 15 Mb/s service with a 250 GB cap. The service more than meets my needs and I feel the price is very fair when I compare it to the the $40 a month I pay for a 1.5 Mb/s service at my place in Arizona.

    Let’s not forget that the CRTC proceeding isn’t about retail UBB its only looking at wholesale UBB for Bell. As “oldguy” has correctly pointed out the network is engineered for peak usage and we know that peak keeps growing exponentially so figuring out the cost of a GB today is a waste of pencil lead. Bell has already figured out a funding model that allows its retail customers to fund ongoing network capacity upgrades now they just need a mechanism that allows their wholesale clients to contribute as well. As much as I hate Bell, it’s not fair to make Bell’s retail customers subsidize service for the retail customers of their resellers.

  27. @Cynic
    That explains a lot. Very few Canadians actually have access to Telus’s service, which sounds awesome btw.

    In Ontario, for example, we have the choice between rogers and bell. 60$+/month for 5-7Mbps and 60GB monthly cap. See the problem? and they want to add (or are already adding) UBB on top of this.

    The point of the CRTC hearings is to allow for maximum competition. If resellers can’t price like Telus does, or are stuck at exactly what Bell wants them to be via twisted UBB models, what’s the point of having resellers at all? Bell should be charging the base rate ONLY, end of story, for the use of the last mile, and UBB should not even enter the calculations at this stage.

    Bottom line is that for real competition to exist here in Ontario and in other provinces, Bell must be forced to lease their lines for the absolute least possible – i.e. what it costs them to operate it, no ifs and or buts. UBB is not a model that accurately depicts operational costs, and therefore is completely unfair to be imposed on resellers.

  28. @Cynic
    …”figuring out the cost of a GB today is a waste of pencil lead.”

    Right. And as such, UBB is irrelevant in ROI and capital investment planning. It is simply a “source of revenue” unrelated to any measurable costs. It penalizes a subset of customers unfairly.

    Comparing Canada to the US isn’t appropriate, both are failing at about the same rate in the quality, speeds, and prices of internet usage.
    Although Telus has published data caps, they have stated they have never enforced them. Nor have I ever heard of anyone running afoul of them.

    Bell has stated that the “last mile” doesn’t have a congestion problem (DSL and Cable are different, as are wireless and 3/4G). “3rd party” ISPs have been trying to bypass Bell’s “network” and co-locate directly into the Bell network, they are even willing to pay extra for it. Bell has resisted this.

    …”Bell has already figured out a funding model that allows its retail customers to fund ongoing network capacity upgrades”

    And this model is arbitrarily penalizing a subset of their retail customers unfairly. Because these customers then move to the 3rd party ISPs, Bell is trying to force that same unrealistic model onto them. They have decided on a “business model” that has no relationship to costs or reality or planning, and are attempting to force that model on everybody.
    If you look at the places where Bell has real competition, they carefully avoid penalizing those customers.
    In essence, UBB is simply a “because we can” funding model. And they are attempting to artificially force this same model on the 3rd party ISPs.

    I appreciate the issues involved, but the funding mechanisms Bell (and others) have chosen have no connection to reality. CRTC mandated wholesale prices are supposed to be “cost plus”, not a passing along to 3rd parties of cost disconnected Bell management “funding models”.

  29. fulanomengano says:

    @Cynic Your bottled tap water analogy is ridculous
    On the same gas station that YOU are willing to pay $1.50 for a bottle of water, you can drink from their bathroom tap for free. I have the option to do that, and I exercise it, so I won’t be screw by them. The fact that you (and millions of other people) are willing to pay $1.50 for that bottle of water shows how much do you like to get screwed.

  30. reality check says:

    @Cynic re: funding
    You’d have me more convinced if it weren’t for the profit margins under which these companies operate…

    Aside from that, if UBB is about charging their customers usage fees for a unit that has little actual cost (as this article explains), even if it costs them $0.34 per gigabyte at the highest end, does that justify a $10 per gigabyte fee? Does it even justify a $2/gb fee?

    If it was about funding upgrades, how would a punitive measure, intended to have people use their internet less, increase that funding? If they actually “got their way” they’d be no further ahead.

    Lets assume for a second that a 60gig per month cap makes any kind of actual logical sense, and that somehow, if a user goes over that 60gigs in a month, it actually costs them anywhere close to say $2.50 per gigabyte over what the user pays for their plan (lets forget any sub-$1 amount which is probably the reality). So lets say all their users only use 2gigs a day to stay under their cap. At 15mbps, at full tilt that takes ~18 minutes to reach. If all the users decided to go full tilt for 18 minutes at the exact same time (ie: peak usage), still nobody would break their 60 gig caps, but the ISP would have the same “congestion” problem they do now. UBB doesn’t solve the problem since they would make $0 additional per customer since everyone stayed in their cap. That’s one reason why it’s not a good, nor fair, solution. In this scenario they’d have to, lower the caps even more (to the point where nobody uses the service anyway), charge per gigabyte ON TOP of the monthly plan cost, or only charge usage (which at their supposed costs would be ridiculous). Bottom line UBB doesn’t solve anything.

    It’s an anti-competitive cash grab, plain and simple. Meant to stifle smaller resellers (their only real competition, since incumbents don’t really compete with each other), stifle their competition for content (ie: netflix), and generally just increase their profits.

  31. @Reality Check

    You have convinced me that UBB isn’t the right mechanism for managing congestion and funding capacity growth – actually I was already there … but what are you suggesting as a solution?

    Are you suggesting that the ISPs simply stop investing in network capacity upgrades? Are you suggesting the government expropriate the networks and operate them using tax dollars? Are we to divert funding for health care to the internet instead?

    I have worked for both small and large companies and profit is not a dirty word. In fact profit is absolutley essential – most publicly traded companies direct 50-60% of their profits back into news products, R&D, and new technology to deliver their products/services. The remainder goes back to the shareholders in the way of dividends. No profits means no new internet stuff for us to play with.

    If the ISPs were not making any profits then we would be debating the cost of a minute of dial-up service instead of broadband.

  32. Why is this even up for debate?
    The fact that Bell has come up with three different numbers on their cost per Gb, how Rogers recently changed their pricing tiers AND how Shaw admitted to “monetizing the Internet” a few months ago, all of this just PROVES that the incumbents’ pricing is arbitrary, has nothing to do with infrastructure maintenance and upgrades and has EVERYTHING to do with gouging the customer base.

    I find it absolutely mind boggling that, in spite of all the backlash and information that’s come to light, companies like Rogers and Shaw STILL have the audacity to claim that they are listening to their customers.

    Were that the case, why is there still Retail UBB?

  33. Stop with the water analogy.
    Everything I wanted to say was said by people in these comments. I saw people going against the water analogy but I didn’t see anyone mentioning the real problem with it. How can you compare a finite resource like water to data. It’s not like we’ll run out of bytes anytime soon.

  34. Mikael Andersson says:

    By Bell’s own numbers
    Mmm, alright, let’s for a moment accept Bell’s highest numbers at face value: a cost of $0.195 / GB. This cost is sufficient to cover wholesale transport fees and network augmentation. Let’s highball that Bell’s “customer service” and “tech support” (in quotes for a reason) costs them $5 / month / subscriber (an average of 10 minutes of support per month @ a staffing cost of $30/hour), and that their allowable bitcap is 60GB / month before overage charges. This means $11.70 in usage-based costs, $5 for staffing overhead, for a total of $16.70 in costs per subscriber (assuming everyone hits their bitcap). Let’s not take modem rental costs into account for either provider or subscriber, though it’s pretty much a cash-grab for the provider ($5 extra profit / month after the 4-to-6-month mark).

    Let’s further assume that Bell is currently charging $50 / month to an average subscriber for 60GB of usage (this fee takes the dry-loop cost into account, covered on the provider’s side by the “tech support” costs above – it will be charged to the subscriber one way or another, either on your phone bill or your Internet bill).

    That’s a total of $33.30 / month profit per subscriber, or a 300% markup. Where is this money going? Marketing? Lobbyists? Paying execs? What about all the tax money earmarked for rural expansion, how does that play into the augmentation costs?

    If UBB pricing was reasonable, we could talk. I much prefer fixed pricing with a high cap, as it subsidizes rather than penalizes early adopters which drives innovation, but if I could get my Internet access at a fixed cost of $20 / month (including dry-loop fee) and get charged in the vicinity of $0.25 / GB for every bit I transfer, that’s a more than healthy profit margin for the provider and not too different from my current bill (an effective “bitcap” of 120GB / month to reach the $50 / month subscriber cost).

    But this model won’t fly for Bell – their profits would plummet, as by their own numbers, the majority of their subscribers happily pays $50 / month for less than 25GB!! / month. Every one of these subs would be paying Bell less than half what they’re paying now under the “reasonable” UBB model above.

    No, Bell wants to have the cake – the 300% profit margin on basic subscribers – and eat it too, in the form of 500% to 10000% profit margins on bits transferred in excess of their bitcap. How can this be even remotely allowable by a duopoly providing what’s in essence – if not by classification – an essential service? All of Bell’s analogies with utility companies – hydro, gas, water etc – are put in a very different perspective if we imagined that utilities were raking in dough at that rate.

  35. Fair competition rules is all we need for the underdogs.
    Do you for Bell or Rogers? Why else would you argue that you should pay more when there are competitors out there willing to offer way more for a lot much cheaper? Bell and Rogers couldn’t possibly compete in a real, open market, they need their buddies at CRTC to cover their asses so they can keep their oligopoly.

    Re: your first analogy..
    You can go ahead and buy the most expensive bottled water out there and no one will give a rats ass about it. We just don’t want your bottled water company to use the government to kill off the cheaper alternative.

  36. @Cynic
    …”but what are you suggesting as a solution?”

    I already suggested a solution, place a premium on usage during the times when the network is at peak capacity. It directly addresses the peak times problem. But that would be a solution nobody would be happy with.

    The alternative is to properly plan for capacity upgrades out of normal income, and build in a fudge factor. It may be a higher initial cost to plan for 10 times what you think you will need in 2 years from now, but if your plans are off by a factor of 3-4 you will still be ok. You will need to revise the timeframes and costs for the next upgrade.

    You certainly can’t say these companies aren’t making money. They have the income and margins to justify the upgrades. It’s simply that they are overly focused on the shareholder returns and ignoring customer requirements.
    When the “plan” made 2-5 years ago doesn’t match current reality, do you take your lumps (shareholder returns) and redo the plan(s)? Or do you find someone else to blame it on?
    Profit isn’t a 4 letter word. On the other hand, CYA should be.

  37. telecom industry analyst
    I am trying (key word is trying) to find out how much investment Canadian service providers (fixed-line, long-haul, cable, mobile) are making in Carrier-grade Ethernet and Optical transport equipment. Any information, suggestions, pointers would be immensely appreciated…if you like you can email me directly at

  38. twinkletoes says:

    once again .. MG silly nonsense…
    ISP’s (globally) must invest into network expansion like no other time in history to keep up with demand. The CRTC knows this, ISP’s are painfully aware and Geist know this too. It makes for great print tho to take some obscure reference to pricing and extrapolate a cost of bandwidth that seems outrageous.

    The facts are that the capital required for ISP’s to meet the demand for bandwidth (doubling every year) can’t be supported by today’s subscription fee models. This isn’t unique to Canada. UBB is .. so far .. the only model that works.

    If we truly consider internet as essential as a utility we should be petitioning our governments to ease things like the taxes buried in our service fees. Like paying utilities $10 a hydro pole monthly to hang cables. Any idea how many hydro poles there are in canada? All in your $/month internet fees.

    Geist’s assertions are thinly veiled incitements that only profit the Geist franchise and not the average Joe.

  39. @twinkletoes
    Do you have any references to the costs involved? To the investments that have been done? Raw data of any kind?

    It looks like Sam would like to get his hands on this data. I would be interested in seeing it as well. So far, it looks like this is closely guarded by the ISP’s. Understandable. The only data we do have access to, is their stockholder reports. Which would indicate they are not suffering in any way for investment cash.

    As far as models are concerned, how do other ISP’s around the world manage to avoid UBB, and still charge less that our incumbents?

  40. @twinkletoes
    This is one of the more blatent troll posts that I’ve seen here, or you are working in some capacity for Rogers, Bell or one of their compatriots. Posts above have made trash out of the UBB model as the solution to peak demand issues. How about ditching the corporate line and discussing this in an open fashion rather then simply repeating the lines coming from the big ISPs.

  41. What Really Matters
    What really matters is actually the variable cost per GB at full capacity. Since this discussion is really about overage on the cap imposed by the ISP we need to assume that the ISP isnt losing money (or even breaking even) as their fixed fee per month covers all fixed costs at whatever realistic % of capacity their network is running at.

    Bottom line is that whatever the number is 7 cents, 8 cents or 19 cents this is pure profit to the ISP until they need to upgrade their infrastructure

  42. ISP costs from an ISP
    I’m the net admin for a small ISP. The cost for fibre is $60/Mbps/month plus tax. A client of ours pays $45/month taxes included. That covers our costs to deliver what amounts to on average, a 3Mbps download speed, to the client.

    The network has a certain bandwidth capacity, that is, a certain amount of bps. If a client downloads 1PB or 0B in a month, the cost to operate the network is exactly the same.

    I say this to put the lie to the idea of general UBB.

    However, if everyone on the network tried to download 1PB at the same time, it would be SLOW. Assuming about 1000 clients all trying to download 1PB when the fibre is 30Mbps means that they each get 3kbps. That’s when they complain.

    Most of the time, internet traffic is bursty. Just because all 1000 clients may be sitting at their computers at the same time, reading various web sites doesn’t mean that they are all downloading websites at the same time. Usually they are not. And that is what allows ISPs to squeeze 1000 clients through a 30Mbps fibre.

    What we tell people is, between 16:00 and 22:30 ET, the internet is at its peak and it’s going to be slow. Traffic jams and congestion can and do occur on the internet as well. If you want to download a movie, do it between midnight and 4AM when no one is on.

    So the use of UBB as a means of discouraging high-bandwidth transfers during peak times is something that may be examined.

    An example I often use is to describe the internet as a network of pipes which deliver water. The water is always there, you just need to open the tap. How fast the water comes through depends on how big the pipe is and how many people are drawing water from it at the same time. How much does 1L of water cost? It costs the network nothing. When you pay your water bill, you’re paying the water provider, the people who treat the water for you so it can be drunk. The price of connecting to the water network is fixed, regardless of how much water you use.

    So the issue is not how much does a Gigabyte cost, but rather what are your expectations on how fast it’s going to be delivered.

  43. A gigabyte costs whatever people are willing to pay for it.

    How much less than that amount that it costs the provider of the bandwidth to obtain is entirely superfluous unless one wishes to waste their time coveting other people’s profits.

  44. High quality replica Rolex watches says:

    Watch online stores
    What has perhaps fallen by the wayside is the sense of a connected community that is built into the systems.

  45. 7 cents is fine
    I’d be okay with paying that. But I’m paying 2 dollars. Why the big difference?

    If I went over 14 gigs today, I owe 28 bucks, when it only costs them a dollar. Something’s rotten at the heart of mainframe.