How Net Discrimination Works (Or Doesn't Work…)

3 Mar

Dailymotion

I'm increasingly puzzled by the net discrimination stance that a number of European Incumbents have started to push forward in the last few months despite mounting evidence that the "alternative business model for the internet" that they are pushing forward simply cannot work. I summarized my views a few months ago in a piece entitled The Slow Suicide of Net Discrimination. Now I'm looking for evidence that backs or disproves my thesis.

A few weeks ago, one of the players who has been vocal about establishing a Net Discrimination regime for the internet, France Telecom, gave us a wonderful example of this and a fun case study for me. Two things happened on the same week and the confluence of these two events made my cortex tingle:

  • AT Kearney released a study sponsored by European incumbents (including France Telecom) that recommended that a traffic tax be put in place for content and application owners accessing the networks of telecom operators.

AT Kearney released a study sponsored by Telefonica, Deutsche Telekom, Telecom Italia and France Telecom entitled A Viable Future Model for the Internet and subtitled "Investment, innovation andf more efficient use of the Internet for the benefit of all sectors of the value chain". I will probably post a full critical review of the study in a few weeks, but suffice it to say at this stage that the fact that it is sponsored only by telcos may cast doubt on the "all sectors of the value chain" assertion.

One of the key conclusion of this study is that internet traffic passing through telco networks should be taxed for the "model to be viable". The recommended tax is as follows:

  • 0,05 €/GB of fixed termination traffic
  • 3,03 €/GB of mobile termination traffic

Since the model is supposed to be viable and since France Telecom, that very same week, purchased Dailymotion, I thought it would make sense to apply these rates to an estimate of Dailymotion's traffic to see how "viable" the model was. 

First, the hard data:

  • Dailymotion shows roughly 1 billion videos per month (Source Dailymotion 2009)
  • The average video on Youtube was 10MB in 2007 (Source). I couldn't find anything more recent or specific to Dailymotion, but in all likelihood using this figure actually underestimates the size.
  • I couldn't find a source that stated the exact proportion or number of monthly videos served over mobile networks so I assumed 10%. (If anyone has any clear source for this, please share it and I'll run the model again). All the press releases about the 

Based on these numbers, Dailymotion would have to pay roughly €3.5m per month for distribution of its content. That's €41,76m per year.

Dailymotion's 2010 turnover was €18m. In other words, Dailymotion not only wouldn't be able to pay the tax if it was instituted, it would instantly go bankrupt. That's, apparently, what our European Incumbents define as a "viable future model for the Internet".

Additionally, this back of the napkin calculation raises an interesting question about France Telecom's intentions in buying Dailymotion. I can see one of two explanations: 

  • France Telecom doesn't believe that the model it's lobbying for is actually viable
  • France Telecom is trying to kill Dailymotion

Take your pick…

2 Responses to “How Net Discrimination Works (Or Doesn't Work…)”

  1. Josh Holbrook August 26, 2011 at 3:48 pm #

    Benoit -
    As always a very interesting piece. However, I think you jump to the conclusion that the tax model doesn’t work a little too quickly. You may ultimately be right, but I don’t think the analysis above is proof of the policy’s failing. It seems equally feasible that Dailymotion has an unsustainable business model. It’s not Oranges responsibility to ensure Dailymotion makes money. It’s in Orange’s best interest that Dailymotion survive (even pre-acquisition), but not their responsibility to make sure their (hypothetical) tax is low enough to allow content providers to survive.
    I hope all is well.
    Josh

  2. Benoît FELTEN August 26, 2011 at 5:01 pm #

    If the tax model is deemed to be “viable” it shouldn’t endanger companies that are currently making profit. It’s an ecosystem issue. I used a single example (and maybe an extreme one), but there are probably hundreds that could be used. I used this one because it seemed paradoxical to me that Orange would buy out a business (which suggests they think it’s profitable) and at the exact same time issue policy recommendations that would make that business no longer profitable.
    You say it’s “not their responsibility to make sure their (hypothetical) tax is low enough to allow content providers to survive.” Well then fine, but let’s not call the model viable. If the (hypothetical) tax destroys the online content industry in one fell swoop, would you call it “viable” ?
    Remember, one of the core assumptions of the ATK piece was that OSPs could afford to pay. Admittedly, I only desmontrated that one of them couldn’t pay, and an in-depth analysis (which I’d love to do if time allowed) should replicate what I’ve done for dozens or hundreds of businesses representative of the OSP space. My point is, if you’re saying the model is viable, you should be able to prove that it doesn’t kill the industry that it’s supposed to “grow in a healthy way”.

Leave a Reply