Notes on cable – Part 4 – Billing for Internet

There are different approaches to billing users for Internet usage.  This post looks at the ways in which ISPs might raise prices and implement price tiering for its customers.

A value-based model

The value of the Internet to a consumer should be roughly proportional to the amount of bandwidth consumer.  Charging for bandwidth is a very straightforward approach to implement price tiering.

Cost-based model

The ISP’s costs mainly relate to capacity during peak hours (roughly 8-12PM).  As technology improves, the cost of capacity comes down.  As well, there are some minor economies of scale.  See the white paper “The Future of the Internet – Innovation & Investment in IP Interconnection – 2014” for data on infrastructure costs.

A logical method would be to bill users based on time-of-day usage, similar to the way electricity is billed (e.g. higher rates for peak hours).  In markets where the incumbents are forced to open up their networks to competing ISPs, the independent ISPs may shift to this model if their strategy is to compete on price.

However, the reality is somewhat complicated.  1GB of Netflix usage may have dramatically lower costs to the ISP than 1GB of other Internet usage.  The cost depends on the path that the data takes.  If the ISP is able to charge connection fees to Netflix, then the cost of Netflix bandwidth is negative.  ISPs that account for a large chunk of Netflix’s revenue will be able to extort higher fees from Netflix.

Don’t charge for bandwidth / the ‘overselling’ model

A very small number of users will account for a significant portion of bandwidth consumed.  One approach is for the ISP to simply absorb these costs.  Advantages are:

  1. Low customer support costs.  Billing users for usage will increase the number of customer service inquiries because some people will complain about the extra charges.  If the ISP is known to reduce charges to increase customer satisfaction, then some consumers may simply call support to reduce their monthly bill.
  2. Sometimes the metering systems do not work correctly.  Some customers may (rightly or wrongly) claim that the metering systems do not work correctly.  This will raise support costs.
  3. Consumer friendly.  Having to worry about excessive bandwidth usage is annoying.
  4. Some users will accidentally use massive amounts of bandwidth if their computer has been compromised by hackers.  Not charging for bandwidth reduces complaints about billing.
  5. Allows the ISP to advertise “unlimited” bandwidth.  In reality, the ISP may reserve the right to throttle heavy users (which effectively reduces their bandwidth/capacity usage).

Throttling during peak hours

Teksavvy (the ISP that I use) is experimenting with a program that offers consumers unlimited bandwidth in exchange for lower download speeds during peak hours.  I see this as a cost-based approach with the benefits of not charging for bandwidth.

Overage charges

Some ISPs will gouge users for exceeding their data caps.  The cost of bandwidth may jump several times once the data cap is exceeded.

Personally, I think that this is a terrible billing strategy.  It:

  1. Generates the most customer service costs.  Rogers for example currently uses this model.  Many customers who accidentally exceed their caps will call customer service to try to get their bill reduced (knowing that Rogers will do this).
  2. Is consumer unfriendly and value destroying.  Consumers need to carefully monitor and curtail their Internet usage to avoid exceeding caps.
  3. Encourages consumers to switch to competitors.  This practice makes consumers angry.
  4. Makes the marketing message to the consumer very complicated.  Not marketing-friendly.

On the “upside”, the ISP might make some short-term profits at the expense of its users.  In the long run, I do not believe that these billing practices are sustainable in competitive markets.  This strikes me as a clear sign that an ISP is poorly managed.  As a former Rogers Internet customer, I can say that this practice made me switch my Internet to Teksavvy.


A sensible retentions strategy is to offer consumers plans that are enticing compared to competitors’ plans.  When I switched from Rogers to Teksavvy several years ago, there was no such strategy.  I wanted low speeds with lots of data/bandwidth.  I would have been willing to pay Rogers slightly more money than Teksavvy because:

  1. I was unfamiliar with Teksavvy.
  2. Switching ISPs is a pain.  The consumer loses Internet for a day.  If there is a problem with the new Internet, it may be difficult to troubleshoot because the customer cannot go online to do research on Google.
  3. Teksavvy charges upfront setup fees and makes its customers buy their own modem.

At the time, Rogers did not offer a plan comparable to what Teksavvy was offering.  (Nowadays it does.)  I do not know what Rogers’ current retentions execution is like.

Overall business strategy

The pricing structures of ISPs are often linked to an ISP’s overall strategy.  Cable companies that are trying to defend large monthly bills for video services may try to engage in practices that hurt competing services:

  1. Throttling BitTorrent to reduce illegal downloading.  This practice started a technological arms race with the BitTorrent ecosystem.  BitTorrent software clients began implementing encryption to get around throttling.  Later, the ISPs found a way to detect BitTorrent traffic patterns.  Unfortunately, that technology would also throttle legitimate Internet activity such as World of Warcraft.  I believe most ISPs no longer throttle BitTorrent because the false positives created too many customer service issues.
  2. Slowing down Netflix, either by underinvesting in their infrastructure connections with Netflix servers or by explicitly throttling Netflix.  Later on, this practice allows the ISP to extort Netflix for connection fees.  The problem with this approach is that the ISP has to intentionally slow down its network to outside connections.  Slow Internet leads to a lower quality service for the consumer.
  3. Charging high prices for bandwidth while offering unlimited bandwidth for in-house video streaming services.

Cable companies with very low penetration (e.g. Charter in the US, European countries where DSL has higher market share) may try to offer high-quality Internet that is very consumer friendly because they are trying to grow market share.  Charter for example has fast Netflix and does not currently charge egregious overage fees.

The future

I think it definitely makes sense for some form of price tiering to occur.  Probably the best way to do this is to charge a flat rate per GB or to throttle bandwidth during peak hours.

Ideally, an ISP would only offer 2 or 3 Internet options to the consumer.  I believe that fewer choices would make marketing easier to understand and more effective.  Fewer choices would also reduce the amount of time spent explaining pricing plans to the consumer and simplify billing systems (especially if the ISP supports grandfathered plans and plans specifically for retentions).

In practice, ISPs will likely offer a wide range of Internet plans because that’s just what they do.  On top of that, consumer choices will be more complicated thanks to double and triple play bundles of Internet + voice + video.  And on top of that, ISPs may offer a wide range of promotional offers to try to entice customers to switch providers.


Competition will definitely shape how ISPs charge for usage.  In countries where incumbents are forced to open their networks to competitors, I think that we will see ISPs compete on price.  Therefore, the landscape will be dominated by cost-based pricing.  ISPs will bill for peak-usage bandwidth or they will throttle users during peak hours (unless the user pays some type of premium for unlimited bandwidth during peak hours).

In  less competitive markets, we will likely see ISPs charge based on the value of the data (Internet, voice, and video) that the customer consumes.  I think that Internet TV will greatly increase the value of the data flowing over the pipes, whether it is a coaxial cable or phone line.  The consumer will have access (legally or illegally) to a broad universe of content at a low price.  This gives the cable companies pricing power to increase ARPU (average revenue per user) over time. If history repeats itself, eventually regulators will step in to cap prices if too many consumers complain about high prices.

Because voice has to compete with cellular networks, competition may continue to drive revenues down.I do think that management matters.  No ISP is so wonderful that any idiot can run the business.  Bad management is the reason why Charter went bankrupt.  To spot good management, one thing to consider is whether or not their pricing strategies make sense given their market.  Some companies such as Rogers have been increasing their prices to gouge their customers.  However, I believe that Rogers’ pricing strategy is very short-sighted and unsustainable in the long term.*Disclosure:  No position in RCI (or LBTYA).  I own LBRDA (which owns Charter).LinksThe Canadian media landscape

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s