Prediction is hard. Especially about the future.1

Ten years ago today, still in the fairly early years of this blog, I shared some numbers in a very short post, the entire substance of which was:

Last year, 37% of us had an MP3 player.  This year, it’s 57%.  Next year, it will be higher still – if those who say they are going to buy, do, it will go up to 73%.

Who wants to bet on the accuracy of channel predictions five years out?

The last sentence is probably (though I really can’t remember) the clue to why I chose to share the numbers. At the time, there was still a widespread assumption that digital channels would only ever be for a small minority, indicators such as that one were sending a message to anybody who chose to hear that it might not be like that.

But that’s not the reason for looking back those ten years. Instead, let’s think about what longer term predictions might have looked like from the perspective of 2006. Given the rate of growth, 73% would not have felt like an upper limit. This was a market clearly heading rapidly to saturation – so the question would have been what counted as saturation. Not everybody likes listening to music, so it wouldn’t have been 100%, but perhaps it would be pretty close.2 Whatever that level might be, it would have been pretty clear that it would have been reached within five years and completely stable by ten – following the classic innovation diffusion curve.

But of course if you had made that particular prediction, you would have been spectacularly wrong. A few months later the iphone was unveiled, making it much more apparent that the future was about pocket computers which happened to play music (and at a pinch could even make phone calls) than it was about dedicated music players.

Ten years on, you can still buy an MP3 player if you really want one. You might well be consuming music predominantly in MP3 form. But you’re probably not using an MP3 player to do it. And ten years on, Mary Meeker wonders whether we have passsed peak iphone.

Peak iphone

It’s easy to think that the future has run out of surprises, but there is little sign that it has. People have been predicting better predictions over the thousands of years since the prediction business got underway. Simple extrapolation is rarely the smartest move.


Header image by Giandomenico Ricci

  1. Often attributed to Niels Bohr, but of uncertain provenance.
  2. Not just listening to music, of course, but podcasts were still a long way from mainstream.


  1. Indeed, the future is so uncertain and especially with new technologies replacing or making older tech obsolete it becomes harder and harder to predict the future with any accuracy, especially the further into the future you try to predict.

Comments are closed.