Measuring online video growth

There is a very interesting discussion over at the Nyquist Capital blog ^[1] about the growth rate of online video (read the comments!).  Essentially, one measure of the video market shows a decrease of 9% in the last year, which seems to contradict claims by Cisco, AT&T and others that video is going to grow so fast that the internet will be under severe stress.  In a response on his blog ^[2], Ike Elliott cites a Comscore study showing 66% increase in the number of videos viewed, and he is quite right that these two results are wildly inconsistent. Indeed, the measurement of internet trends has always been a very nebulous proposition, the error bars are a mile wide.  Methodology is everything, but it gets short schrift in the press.

However, while counting the videos viewed on the internet is interesting, I would like to point out that this is not necessarily the metric that will determine whether online video growth leads to explosive bandwidth growth.  The most important factor right now is not a count of videos viewed, rather it is the video bits transferred.  There is a huge difference between watching a 5 minute low resolution video on Youtube and watching a full screen episode of [insert your favorite sitcom here] – it is several orders of magnitude.

The greatest growth in video traffic will not come from growth in clicks, it will come from the resolution of the videos watched and their length.  YouTube-like video will not cause an explosion, it is everyday viewing of the big stuff that will add the bits if it happens. Folks seeking to measure the growth of traffic from online video have to look past the videos viewed and take into account the changing parameters of those videos.  The explosion, if it happens, will come from more videos being viewed at higher average bitrates and for longer times – a triple effect.