Should You Have a Data Cap?

data recovery

data recovery (Photo credit: Sean MacEntee)

Over the last few years most of the cable companies and some telcos have implemented data caps on high-speed Internet access. They always claimed that caps were necessary to help protect their networks from congestion. They claimed that heavy users would clog the networks and make data speeds slow for everybody else. But as someone who sees hundreds of networks, this claim holds no technical validity, except in some isolated instances and in some parts of some networks.

Michael Powell, the head of the National Cable and Telecommunications Association admitted publicly last week that caps are not about congestion, but rather are about ‘pricing fairness”. In the telecom world there is a general rule of thumb that the most active 15% of your users will use 85% of any resource, be that minutes, data, etc. And it’s a pretty good rule of thumb. If cable companies had come along and lowered prices for the 85% who are not heavy users and then made up the difference on higher rates for the 15%, then his argument would resonate with the public. But nobody saw any rate reductions and it’s hard to see data caps as anything more than a way to make even more money from data service.

One has to just note that the US has some of the highest-priced Internet services in the world to poke holes in NCTA’s announcement. If you compare US rates to the Far East or Europe it is easy to see that our rates are way out of line on a cost per megabit of service available to customers.

To make it worse, cable companies are starting to raise data rates. And this follows a ten-year period where the underlying cost of raw data has gotten cheaper every year. When a cable company set a monthly rate of $40 or $50 a decade ago, during that decade the cost of buying wholesale access to the Internet has probably dropped by 90%. It’s my opinion that cable companies know that in another decade that they are going to mostly become ISPs since cable and telephone are both dying products. They are starting to creep the rates up now to hedge against the day when that is their only product.

But even assuming that our rates are too high and that profits are really high, should any ISP consider any sort of cap or limitation on how customers use their data. I think the answer is yes, and it is not for any of the reasons that the cable companies have given.

Using my metric, 15% of the users on a network create most of the data usage. But absent any rules on how the network can be used, a small number of them could be using most of the usage for that group. For example, customers who operate servers and operate ecommerce site or other very busy sites like a pornography server can use huge amounts of data on the network. Much of that data is sent in the upload direction and doesn’t cost as much for a carrier as downloaded data, but a few such sites actually can clog a part of the network if they are busy continuously. The way around this problem is a prohibition against using servers on a basic residential data product. But if you are going to have this kind of policy you also need to have some way to measure how much data each customer is using.

On the download side of the equation, there are always a few customers who abuse any system. There are internet hoarders just as there are hoarders of anything else, and so you might want to set a cap that discourages continuous downloading. Comcast has implemented data caps of around 300 Gb in a lot of markets lately. If a customer downloads movies at a very high quality rate, they can use around 2 Gb per hour. If they watch non-HD movies it’s about half of that. And so a 300 Gb data cap would limit people to watching 150 hours of HD programming or 300 hours of normal programming per month. That works out to a limit of 5 hours per day or HD programming or 10 hours per day of normal programming. That may seem like a lot, but if each person in the family is watching their own programming, that is a really small limit.

I have advised my clients to institute a fairer cap, but to still have one. For instance, a cap set at 1 Tb (1,000 Gb) allows for over three times the usage than the Comcast cap. Anybody going over a 1 Tb cap is likely a data hoarder because that requires somebody to be downloading video more or less continuously every day of the week. Every network has a different configuration and so this is not a hard and fast limit. But I suggest some limit on data, at a very high rate that will only affect a truly small handful of people. The Comcast rate is set to make any family who actually uses their bandwidth to pay more. My suggestion is to set a cap that stops bad abuse, while giving people what they have paid for.

3 thoughts on “Should You Have a Data Cap?

  1. Dear Doug:
    So the interesting sidelight to the data cap regimen is the perceived migration of customers from Cable TV to various streaming video platforms on the Internet. Folks have called them ‘cable-cutters’, among other terms.
    My guess-timate is that as more customers migrate from Cable TV to streaming video, their aggregate amount of data, either on a daily or monthly basis, will increase precipitously.
    Is this basically where the industries want their telecom customers — using telephone service via VOIP, and streaming video?
    Just a thought…

  2. Part 2 (Cut off prematurely…)
    Thus, as more customers migrate to the Internet-based platforms and utilize higher and higher amounts of data, they too will “trip” the data caps. Is this what carriers want?

  3. I think that is exactly what the big cable companies want. They see the caps as strictly a revenue opportunity and the hell with the customer. They say otherwise, but that is exactly what they had in mind when they created the somewhat low caps.

Leave a Reply