Is Tiered Cable Internet Access a Scam in the U.S.?
March 4, 2016 3:24 AM Subscribe
At least here in the U.S., cable companies like Comcast and Time Warner sell net access in tiered bandwidth packages. I.e., the more you pay per month, the more bandwidth you are supposed to get.
Is this a scam? Or, is there a technical basis for this?
What additional costs does a cable provider take on to provide more bandwidth to a single existing customer? The network infrastructure does not change. They don't install new cable lines.
What actually changes and does it cost anything?
Looking at the Time Warner cable plans, I'm taking that you're asking whether it's a scam that companies are charging different rates for different speeds?
I'm not in the US, but I work for an internet service provider, and for us we definitely pay the wholesaler more when customers are on faster speeds. For example, if we had a customer on 50 mbps but accidentally provisioned them at 100 mbps with the wholesaler, we'd be losing money.
posted by kinddieserzeit at 3:45 AM on March 4, 2016 [2 favorites]
I'm not in the US, but I work for an internet service provider, and for us we definitely pay the wholesaler more when customers are on faster speeds. For example, if we had a customer on 50 mbps but accidentally provisioned them at 100 mbps with the wholesaler, we'd be losing money.
posted by kinddieserzeit at 3:45 AM on March 4, 2016 [2 favorites]
Response by poster: Same question applies.
My provider, Time Warner, uses both terms in its advertising.
Every year, TW wants to boost my fees significantly. I call and complain, threaten to leave, moan and groan, and eventually they shuffle me around to someone who offers me a deal for another year at just about the same price I'm currently paying. Presumably, they are not selling service to me at a loss.
posted by justcorbly at 3:48 AM on March 4, 2016
My provider, Time Warner, uses both terms in its advertising.
Every year, TW wants to boost my fees significantly. I call and complain, threaten to leave, moan and groan, and eventually they shuffle me around to someone who offers me a deal for another year at just about the same price I'm currently paying. Presumably, they are not selling service to me at a loss.
posted by justcorbly at 3:48 AM on March 4, 2016
If by scam you mean "not justified by marginal cost" the answer is probably. At some point there is a requirement for incremental capital investment to support higher capacities but mostly it is pure profit to the provider.
posted by JPD at 3:54 AM on March 4, 2016 [2 favorites]
posted by JPD at 3:54 AM on March 4, 2016 [2 favorites]
Response by poster: @kinddieserzeit: Do you know what extra costs your wholesaler does to provide the additional speed? How do you know those are legitimate increases?
posted by justcorbly at 3:54 AM on March 4, 2016
posted by justcorbly at 3:54 AM on March 4, 2016
Well, they're not making a loss but they're not making the profit margin that they would like to make. They probably find that when they put their prices up, enough people go with it that it makes up for the few that complain.
In Australia providers tend to let people stay on "grandfathered" plans when the plans and prices do change.
I'd suggest that you keep doing what you doing. Each time that they try to make you pay more, request that they keep you on your current plan. Unless the new one is better, of course.
posted by kinddieserzeit at 3:56 AM on March 4, 2016
In Australia providers tend to let people stay on "grandfathered" plans when the plans and prices do change.
I'd suggest that you keep doing what you doing. Each time that they try to make you pay more, request that they keep you on your current plan. Unless the new one is better, of course.
posted by kinddieserzeit at 3:56 AM on March 4, 2016
Response by poster: @jpd: I'm want to know what a cable company actually does when I either buy increased speed or pay more to maintain the current level of service. What is it and is it a cost to them?
posted by justcorbly at 3:57 AM on March 4, 2016
posted by justcorbly at 3:57 AM on March 4, 2016
No you are right. Any data provider essentially had zero marginal costs to serve a customer once the network is built out. However the average costs are quite high and fixed, so as long as mr>mc they have an incentive to keep you on the network. Tv requires content so mc is non-zero but for broadband it is pretty low.
posted by JPD at 4:00 AM on March 4, 2016 [1 favorite]
posted by JPD at 4:00 AM on March 4, 2016 [1 favorite]
Someone is just turning a nob somewhere. But eventually if enough people take more data they will need to upgrade equipment.
posted by JPD at 4:01 AM on March 4, 2016
posted by JPD at 4:01 AM on March 4, 2016
Response by poster: @jpb: I don't buy TV. Used to, though. After I cancelled the TV side, my price for broadband-only increased to more than the previous broadband+TV fee within the year.
posted by justcorbly at 4:04 AM on March 4, 2016
posted by justcorbly at 4:04 AM on March 4, 2016
Again direct costs are close to zero. Indirectly there is some capital charge that needs to be paid for. Depending on who you ask that capital charge is either meaningful or a rounding error. I suspect that depends on the characteristics of the network.
So from a pure marginal cost perspective your mental model can be zero - but thats not economically correct because you need to fund and earn a return on a fixed cost network that will require some unknown level of future investment ro sustain future demand growth. What that number really is i suspect the provider doesnt even know.
posted by JPD at 4:13 AM on March 4, 2016 [1 favorite]
So from a pure marginal cost perspective your mental model can be zero - but thats not economically correct because you need to fund and earn a return on a fixed cost network that will require some unknown level of future investment ro sustain future demand growth. What that number really is i suspect the provider doesnt even know.
posted by JPD at 4:13 AM on March 4, 2016 [1 favorite]
Mod note: OP, Ask Metafilter is not well suited to a conversational style. You can ask any question you like — but at that point it's basically a one-way street, folks can provide answers and you can decide what might be helpful to you. So please limit any follow-ups to the odd clarification or correction; and please resist the temptation to discuss answers or ask followup questions, thank you.
posted by goodnewsfortheinsane (staff) at 4:26 AM on March 4, 2016 [1 favorite]
posted by goodnewsfortheinsane (staff) at 4:26 AM on March 4, 2016 [1 favorite]
I call and complain, threaten to leave, moan and groan, and eventually they shuffle me around to someone who offers me a deal for another year at just about the same price I'm currently paying. Presumably, they are not selling service to me at a loss.
They've already paid for their infrastructure and their upstream bandwidth; provided the fraction of that that you actually use is less than what you're paying them to make available, which for the vast majority of their customers it will be, they're making money off you.
I'm want to know what a cable company actually does when I either buy increased speed or pay more to maintain the current level of service. What is it and is it a cost to them?
They're just varying a rate throttling setting that affects your connection. It costs them nothing (well, one minute of employee time) to make that specific change.
ISPs make money on the difference between what they charge and what they pay for the aggregate bandwidth that flows through their routers; upgrades to routers and backhaul provisioning are also a cost for them. Customers with a higher data allowance and/or higher maximum speeds put more pressure on all their cost items, so they charge them more.
I've long believed that the way data gets charged for is fundamentally broken. It seems to me that if the underlying concept was that every entity had to pay a download fee per megabyte entering its boundary, and got paid an upload rebate per megabyte leaving its boundary, and the download fee was a little bit more than the upload rebate, then we could have a commercially sustainable Internet without any need for advertising whatsoever.
posted by flabdablet at 4:30 AM on March 4, 2016 [4 favorites]
They've already paid for their infrastructure and their upstream bandwidth; provided the fraction of that that you actually use is less than what you're paying them to make available, which for the vast majority of their customers it will be, they're making money off you.
I'm want to know what a cable company actually does when I either buy increased speed or pay more to maintain the current level of service. What is it and is it a cost to them?
They're just varying a rate throttling setting that affects your connection. It costs them nothing (well, one minute of employee time) to make that specific change.
ISPs make money on the difference between what they charge and what they pay for the aggregate bandwidth that flows through their routers; upgrades to routers and backhaul provisioning are also a cost for them. Customers with a higher data allowance and/or higher maximum speeds put more pressure on all their cost items, so they charge them more.
I've long believed that the way data gets charged for is fundamentally broken. It seems to me that if the underlying concept was that every entity had to pay a download fee per megabyte entering its boundary, and got paid an upload rebate per megabyte leaving its boundary, and the download fee was a little bit more than the upload rebate, then we could have a commercially sustainable Internet without any need for advertising whatsoever.
posted by flabdablet at 4:30 AM on March 4, 2016 [4 favorites]
yeah but the upstream bandwidth is mostly the same fixed cost kind of characteristic.
posted by JPD at 6:05 AM on March 4, 2016
posted by JPD at 6:05 AM on March 4, 2016
Every year, TW wants to boost my fees significantly. I call and complain, threaten to leave, moan and groan, and eventually they shuffle me around to someone who offers me a deal for another year at just about the same price I'm currently paying. Presumably, they are not selling service to me at a loss.
This is just common price segmentation, not a scam. The prices they charge particular customers have very little to do with direct costs. Since various customers are willing and able to pay different prices for the same service, providers would be foolish to charge everyone the same amount. So, just as with used cars, there's a high sticker price that can be negotiated down. They have to make the negotiation process a PITA, so that only people who really, really aren't willing to pay the higher price will make the effort to get the lower prices.
posted by jon1270 at 6:06 AM on March 4, 2016 [2 favorites]
This is just common price segmentation, not a scam. The prices they charge particular customers have very little to do with direct costs. Since various customers are willing and able to pay different prices for the same service, providers would be foolish to charge everyone the same amount. So, just as with used cars, there's a high sticker price that can be negotiated down. They have to make the negotiation process a PITA, so that only people who really, really aren't willing to pay the higher price will make the effort to get the lower prices.
posted by jon1270 at 6:06 AM on March 4, 2016 [2 favorites]
You're asking about two different things:
Speed - the provider has to invest in the infrastructure needed to provide higher speeds. The average speed of home broadband has slowly increased over the years as newer technologies are rolled out: DOCSIS 2, DOCSIS 3, fiber, etc. Speed is how the providers have traditionally tiered their service to the customer. Yes, there is a capital investment here but that cost is spread out over all the customers. I would note that speed being the structure of the price tiering is somewhat artificial also. The max speed of a typical cable connection is set by the provider, not typically limited by the technology itself. DOCSIS3.1 is capable of 10Gb/s but nobody gets that to their house.
Data caps - Data caps are a profit grab. as the marginal cost to the provider of say 300GB (Comcast's current cap) used per month by a customer compared to 400GB is near zero.
One could argue that the rise of Netflix-like services strains the peering arrangements of the broadband providers but I would argue that's just the bandwidth providers strong arming the content providers for a share of their profits. Netflix even has their Open Connect Program to alleviate the peering issues their subscribers causes ISPs. Comcast and Verizon boycotted the Open Connect Program during the fight with Netflix a couple years ago.
TL,DR: capitalism
posted by LoveHam at 6:09 AM on March 4, 2016 [1 favorite]
Speed - the provider has to invest in the infrastructure needed to provide higher speeds. The average speed of home broadband has slowly increased over the years as newer technologies are rolled out: DOCSIS 2, DOCSIS 3, fiber, etc. Speed is how the providers have traditionally tiered their service to the customer. Yes, there is a capital investment here but that cost is spread out over all the customers. I would note that speed being the structure of the price tiering is somewhat artificial also. The max speed of a typical cable connection is set by the provider, not typically limited by the technology itself. DOCSIS3.1 is capable of 10Gb/s but nobody gets that to their house.
Data caps - Data caps are a profit grab. as the marginal cost to the provider of say 300GB (Comcast's current cap) used per month by a customer compared to 400GB is near zero.
One could argue that the rise of Netflix-like services strains the peering arrangements of the broadband providers but I would argue that's just the bandwidth providers strong arming the content providers for a share of their profits. Netflix even has their Open Connect Program to alleviate the peering issues their subscribers causes ISPs. Comcast and Verizon boycotted the Open Connect Program during the fight with Netflix a couple years ago.
TL,DR: capitalism
posted by LoveHam at 6:09 AM on March 4, 2016 [1 favorite]
I'd challenge the assumption that price segmentation has to reflect cost of goods or services or else it's a scam. Pricing a product or service is an attempt to find the absolute highest amount the customer will pay. But if I have something that costs me little to obtain, I can still charge as much as people are willing to pay.
The weasel words around "speeds up to XX Mbps" are worth examining for scamminess, however. That is where I'd look for a failure to deliver on an implied promise, were I looking for one.
posted by chesty_a_arthur at 6:44 AM on March 4, 2016 [5 favorites]
The weasel words around "speeds up to XX Mbps" are worth examining for scamminess, however. That is where I'd look for a failure to deliver on an implied promise, were I looking for one.
posted by chesty_a_arthur at 6:44 AM on March 4, 2016 [5 favorites]
The equipment and lines that they own (starting down the street from your place) have to be good enough to provide the max offered speed on the basis that you, or a subsequent tenant at the same place, will want the max account. That is true for your particular line, and for your neighbor's particular line, and each other particular line.
However, they probably don't have the capacity to serve all of the customers at full speed; that would be unnecessary for them to invest in, as long as they have way of keeping some customers at lower speed. And they do-- the prices. They can use the prices to manage the bandwidth load from customers, while simultaneously pacing their bandwidth increases (from year to year) based on capacity and competition (which, of course, doesn't always exist).
Here's an analogy: You've probably heard of the CAFE standards for automobile fleet fuel efficiency, which sets a lower limit on the fuel efficiency of cars sold by an automaker. Does that mean they can't sell SUVs that have lower fuel efficiency than that number? Hell no, this is America. The fuel efficiency is averaged over the entire fleet of cars sold by the automaker, so they just have to sell more fuel-efficient cars. But geez, what if gas gets cheap and people buy Hummers again? You make Hummers expensive so fewer people will buy them.
It's estimated that carmakers add an average of $10K to the cost of SUVs, because adding to that price discourages enough sales to limit the effect of SUVs on the fleet average.
Comcast can't give everyone gigabit tomorrow, but they can probably support giving gigabit to 1% of potential customers (both current and new), and 300Mbps to 10%, and 100Mbps to 30%, and 20Mbps to the remainder. Then they figure out which combination of service tiers and prices will provide maximum revenue (best market price) while avoiding overselling their total capacity.
TL;DR: they use tiers and prices to maximize revenue while keeping total usage within their capacity. Any given customer can get the full speed, but they likely cannot support all customers at full speed.
posted by Sunburnt at 9:59 AM on March 4, 2016
However, they probably don't have the capacity to serve all of the customers at full speed; that would be unnecessary for them to invest in, as long as they have way of keeping some customers at lower speed. And they do-- the prices. They can use the prices to manage the bandwidth load from customers, while simultaneously pacing their bandwidth increases (from year to year) based on capacity and competition (which, of course, doesn't always exist).
Here's an analogy: You've probably heard of the CAFE standards for automobile fleet fuel efficiency, which sets a lower limit on the fuel efficiency of cars sold by an automaker. Does that mean they can't sell SUVs that have lower fuel efficiency than that number? Hell no, this is America. The fuel efficiency is averaged over the entire fleet of cars sold by the automaker, so they just have to sell more fuel-efficient cars. But geez, what if gas gets cheap and people buy Hummers again? You make Hummers expensive so fewer people will buy them.
It's estimated that carmakers add an average of $10K to the cost of SUVs, because adding to that price discourages enough sales to limit the effect of SUVs on the fleet average.
Comcast can't give everyone gigabit tomorrow, but they can probably support giving gigabit to 1% of potential customers (both current and new), and 300Mbps to 10%, and 100Mbps to 30%, and 20Mbps to the remainder. Then they figure out which combination of service tiers and prices will provide maximum revenue (best market price) while avoiding overselling their total capacity.
TL;DR: they use tiers and prices to maximize revenue while keeping total usage within their capacity. Any given customer can get the full speed, but they likely cannot support all customers at full speed.
posted by Sunburnt at 9:59 AM on March 4, 2016
they likely cannot support all customers at full speed
...though there's a chance they might be able to, if all customers were paying what they charge for the full speed service tier.
posted by flabdablet at 10:35 AM on March 4, 2016
...though there's a chance they might be able to, if all customers were paying what they charge for the full speed service tier.
posted by flabdablet at 10:35 AM on March 4, 2016
This is a standard marketing strategy related to market segmentation and price discrimination.
In a simplified model there is one pair of supply and demand curves that dictates one optimal price that provides the most income and profit. But in reality there are multiple price points that will be attractive to different segments of customers. It is not necessarily a scam. It's just a way of attracting the most customers at the highest prices for each. These techniques are most effective in markets with the least competition, because competition should drive all prices to lowest possible level.
For example airlines have found that they can charge extra for an aisle seat even though it costs them no more to fly the airplane. Store coupons are another example. There are lots of customers that are willing to pay full price for a product. But by issuing coupons, a seller can lure in a percentage of customers that are more price sensitive without lowering the overall price. Amazon will show different prices for different viewers depending on demographic information they compile.
So what the cable company is doing is segmenting the market into those willing to pay a higher price for full service and simultaneously able to lure in some more customers not willing to pay the higher price. If the cable company faced real competition, the price would likely be driven to the lower level for everyone.
posted by JackFlash at 10:54 AM on March 4, 2016
In a simplified model there is one pair of supply and demand curves that dictates one optimal price that provides the most income and profit. But in reality there are multiple price points that will be attractive to different segments of customers. It is not necessarily a scam. It's just a way of attracting the most customers at the highest prices for each. These techniques are most effective in markets with the least competition, because competition should drive all prices to lowest possible level.
For example airlines have found that they can charge extra for an aisle seat even though it costs them no more to fly the airplane. Store coupons are another example. There are lots of customers that are willing to pay full price for a product. But by issuing coupons, a seller can lure in a percentage of customers that are more price sensitive without lowering the overall price. Amazon will show different prices for different viewers depending on demographic information they compile.
So what the cable company is doing is segmenting the market into those willing to pay a higher price for full service and simultaneously able to lure in some more customers not willing to pay the higher price. If the cable company faced real competition, the price would likely be driven to the lower level for everyone.
posted by JackFlash at 10:54 AM on March 4, 2016
So long as the company is not misrepresenting the products or the prices they charge, it is not a scam when a company sells a more desirable product at a higher price than a less desirable product — even if the company's costs for the two products are exactly the same.
posted by DevilsAdvocate at 11:01 AM on March 4, 2016
posted by DevilsAdvocate at 11:01 AM on March 4, 2016
even if the company's costs for the two products are exactly the same
...which, in the case of ISPs, they're not; the lower-tier customers each represent a smaller fraction of the total bandwidth resource that the ISP needs to supply and maintain.
posted by flabdablet at 11:15 AM on March 4, 2016
...which, in the case of ISPs, they're not; the lower-tier customers each represent a smaller fraction of the total bandwidth resource that the ISP needs to supply and maintain.
posted by flabdablet at 11:15 AM on March 4, 2016
I work for an ISP, I do not speak for the ISP. Where the ISP I work for owns the infrastructure (fiber), we have two-tiered bandwidth: 100 megabit for $40+taxes etc, and gigabit for some little extra. Where they have to lease the infrastructure (ie: copper pairs from AT&T), they deliver the fastest they can over that infrastructure, and tiers relate to how many pairs they have to buy.
Bandwidth to the outside world is a negligible part of the cost of delivering the service to the customer, and for the most part one can just ask Akamai and/or Netflix to put another cache machine in the data center long before one needs to upgrade external pipes.
There are also examples of maxed out peers of cable providers offering to buy the additional switching gear in order to provide the cable network the bandwidth necessary for their customers to get full speed access to the wider Internet. (ie: Here's Level 3 calling out Verizon)
Now you'll notice there that one of my examples directly competes with cable TV. So if the question you're asking is "do the cable companies charge proportionally more for bandwidth than it costs them in order to punish activity which would compete with their cable TV business?", the answer is a very solid "hell yes."
posted by straw at 12:06 PM on March 4, 2016 [1 favorite]
Bandwidth to the outside world is a negligible part of the cost of delivering the service to the customer, and for the most part one can just ask Akamai and/or Netflix to put another cache machine in the data center long before one needs to upgrade external pipes.
There are also examples of maxed out peers of cable providers offering to buy the additional switching gear in order to provide the cable network the bandwidth necessary for their customers to get full speed access to the wider Internet. (ie: Here's Level 3 calling out Verizon)
Now you'll notice there that one of my examples directly competes with cable TV. So if the question you're asking is "do the cable companies charge proportionally more for bandwidth than it costs them in order to punish activity which would compete with their cable TV business?", the answer is a very solid "hell yes."
posted by straw at 12:06 PM on March 4, 2016 [1 favorite]
Bandwidth is a scarce resource, and there are fixed and marginal costs to the ISP to provide the service to you. The amount of bandwidth and/or speed you consume has a direct impact on either 1) what is available to sell to other customers or 2) Actual costs incurred by the ISP.
Now, the $/mb you pay to the ISP may not increase/decrease in the same proportion as the increase/decrease of the costs to the ISP, but that's the point of capitalism.
posted by blue_beetle at 1:49 PM on March 4, 2016
Now, the $/mb you pay to the ISP may not increase/decrease in the same proportion as the increase/decrease of the costs to the ISP, but that's the point of capitalism.
posted by blue_beetle at 1:49 PM on March 4, 2016
The equipment and lines that they own (starting down the street from your place) have to be good enough to provide the max offered speed on the basis that you, or a subsequent tenant at the same place, will want the max account. That is true for your particular line, and for your neighbor's particular line, and each other particular line.
Eh. I'm pretty sure the SOP is to act as if all the physical plant is in tip-top shape and can provide whatever service level is being advertised, and only to upgrade the last mile after it's been shown to be inadequate (I've run into this with SBC, Verizon, and Comcast, in three states and the District of Columbia). But that's neither here nor there.
Anyway, as to the original question: no, tiered access is not a scam. The reason essentially boils down to the difficult nature of network upgrades. Let's say (hypothetically) that your ISP has a 1 gigabyte circuit to your neighborhood. Whatever service they sell to you and your neighbors has to add up to no more than 1 gigabyte, or there will be trouble. (Note: networks aren't 100% efficient, so they're probably targeting something like 80% sold … but they're probably also overselling that 80%, assuming that not everybody is going to be online at exactly the same time, so for the sake of easier math and ignoring some estimation error, I'm just going to stick with round multiples of 10). So, they can sell 100 megabit service to ten of you, or they can sell 10 megabit service to 100 of you. They make more money, and recover more of their infrastructure costs, in the latter scenario, so they're going to price the 100 megabit service pretty high to compensate.
Say they've sold the 100 megabit service to a number of people, though, and somebody else wants to subscribe, but the circuit that feeds your neighborhood is (more or less) saturated. Then they have to upgrade that circuit in one way or another. Maybe they've got dark fiber they can just connect; maybe they've got a fiber link that isn't actually running at its rated capacity and they can upgrade the equipment without running new fiber; maybe the fiber's fine but the rack in the data center is full, and they need to add physical capacity; maybe an upgrade for your network requires a new fiber run. Those upgrades can be stupidly expensive, and they don't want to do them if they don't have to. Big ISPs have whole departments doing capacity planning to make sure the utilization is in a sweet spot between "not too full" and "not too empty."
So they have an incentive to keep the subscriber numbers low enough that the aggregate peak usage is still below saturation, which means they're going to price the higher tiers of service higher, since those users are more likely to saturate the link and slow things down for everybody. The economics still hold true if the ISP has a 10 gigabit link instead of a 1 gigabit link: getting too close to saturation is bad, and upgrading around it can be really expensive. Capacity planning is a real factor in last mile pricing.
Also, all those links about price discrimination and market segmentation are also true. What may not be obvious is that a customer at a higher tier will also have a higher expectation of service (both the actual service, and the customer service when they call with a problem). Somebody who's connecting at the bargain basement rate may be happy just to have internet at all and they won't want to rock the boat; somebody who's paying $100/mo may call and scream and post to Twitter when they're unhappy, since they want perceived value. So say one of your neighbors is a gamer and he calls regularly to complain that he's got 21ms ping and he keeps getting killed by people with 16ms ping. He's not going to switch to DSL because then he can't play at all, but he's paying for the top tier of service and (by his metrics) he's not getting it. So he's going to keep calling, and he will be a very expensive customer to have. The network doesn't want you to be that guy.
posted by fedward at 4:02 PM on March 4, 2016 [1 favorite]
Eh. I'm pretty sure the SOP is to act as if all the physical plant is in tip-top shape and can provide whatever service level is being advertised, and only to upgrade the last mile after it's been shown to be inadequate (I've run into this with SBC, Verizon, and Comcast, in three states and the District of Columbia). But that's neither here nor there.
Anyway, as to the original question: no, tiered access is not a scam. The reason essentially boils down to the difficult nature of network upgrades. Let's say (hypothetically) that your ISP has a 1 gigabyte circuit to your neighborhood. Whatever service they sell to you and your neighbors has to add up to no more than 1 gigabyte, or there will be trouble. (Note: networks aren't 100% efficient, so they're probably targeting something like 80% sold … but they're probably also overselling that 80%, assuming that not everybody is going to be online at exactly the same time, so for the sake of easier math and ignoring some estimation error, I'm just going to stick with round multiples of 10). So, they can sell 100 megabit service to ten of you, or they can sell 10 megabit service to 100 of you. They make more money, and recover more of their infrastructure costs, in the latter scenario, so they're going to price the 100 megabit service pretty high to compensate.
Say they've sold the 100 megabit service to a number of people, though, and somebody else wants to subscribe, but the circuit that feeds your neighborhood is (more or less) saturated. Then they have to upgrade that circuit in one way or another. Maybe they've got dark fiber they can just connect; maybe they've got a fiber link that isn't actually running at its rated capacity and they can upgrade the equipment without running new fiber; maybe the fiber's fine but the rack in the data center is full, and they need to add physical capacity; maybe an upgrade for your network requires a new fiber run. Those upgrades can be stupidly expensive, and they don't want to do them if they don't have to. Big ISPs have whole departments doing capacity planning to make sure the utilization is in a sweet spot between "not too full" and "not too empty."
So they have an incentive to keep the subscriber numbers low enough that the aggregate peak usage is still below saturation, which means they're going to price the higher tiers of service higher, since those users are more likely to saturate the link and slow things down for everybody. The economics still hold true if the ISP has a 10 gigabit link instead of a 1 gigabit link: getting too close to saturation is bad, and upgrading around it can be really expensive. Capacity planning is a real factor in last mile pricing.
Also, all those links about price discrimination and market segmentation are also true. What may not be obvious is that a customer at a higher tier will also have a higher expectation of service (both the actual service, and the customer service when they call with a problem). Somebody who's connecting at the bargain basement rate may be happy just to have internet at all and they won't want to rock the boat; somebody who's paying $100/mo may call and scream and post to Twitter when they're unhappy, since they want perceived value. So say one of your neighbors is a gamer and he calls regularly to complain that he's got 21ms ping and he keeps getting killed by people with 16ms ping. He's not going to switch to DSL because then he can't play at all, but he's paying for the top tier of service and (by his metrics) he's not getting it. So he's going to keep calling, and he will be a very expensive customer to have. The network doesn't want you to be that guy.
posted by fedward at 4:02 PM on March 4, 2016 [1 favorite]
« Older What is the actual scientific issue with ketones... | Time and Energy to Paint/Craft with a Dayjob Newer »
This thread is closed to new comments.
posted by kuanes at 3:42 AM on March 4, 2016 [2 favorites]