There is an article in the BBC feed this morning on American use of podcasts.[Link] It claims that only 0.07 of Americans who use the internet have ever downloaded a podcast and only 0.01 download on any “typical” day.
Lots of flaws here, both in terms of the reportage and the folks analyzing the situation. The latter largely follows from the way the use of podcasts develops and how they try to analyze it.
Most of these trend analysts aren’t taught any real maths, they just get taught how to use some standardized techniques for forecasting without understanding the underlying models. And they end up working in commercial organizations where the maths literacy is downhill from them.
As I have blotted before, this type of thing spreads by a diffusion mechanism, which doesn’t necessarily mean a Drunkard’s Walk. In fact, I expect that this is diffusing ala either a logistic model, akin to how diseases are spread by contact between someone with the disease and someone without it, or a broadcast model, akin to seeing something on a TV program or reading in a newspaper and then trying it. Or its possibly a mixture of the two.
There are some complications on all this since the process is probably stochastic and not stationary – definitely not stationary. There may be competing processes, not all of which may be obvious – as we’ll explore.
I should comment at this point that this is how both Firefox and Linux are spreading. I probably should also admit that I am one of those few who do use podcasts. I can attest that they’re few because I have no acquaintances who subscribe to podcasts.
There are a couple of microdynamics matters here, I think. First is that the whole podcast thing is still pretty much in its amateur infancy stage allthough there has been some considerable growth recently in the professional sector. Partly because of this podcasts are pretty much divided between amateur special interest podcasts, most of which are pretty bad and very audience specific, and organizational outreach podcasts. The latter include folks like the BBC, the scientific journals, universities and Williamsburg, to name some that I subscribe to.
The software environment of podcasts are very much amateurish. Almost all of the software is open source, and the clearing sites for podcasts that maintain databases of podcasts are all ad hoc and open. So if you’re going to subscribe to a podcast you’ve got to deal with software that isn’t Megahard and may be kind of brutal in assuming you know a bit from a byte and a URL from an IP address.
Once you’ve got the podcast subscribed to, you have to receive it, which means it comes in over the internet to your computer – although I understand the phone companies are sorta working on cellular reception. Then once you get it, you have to either listen to it on your computer, or download it onto some device like a CD that you can listen to on your automobile wireless/CD player (if its MP3 capable,) on onto your MP3 player. And as I understand how iPods work, if you’ve got one of them you’re either left with more open source and socially discouraged software or you’re out of luck. Luckily I found out about this before I bought an MP3 player and so I don’t have this problem.
The whole podcast on an MP3 player is a niche thing since MP3 players were primarily intended for folks to listen to specially bought music on. Indeed, the whole MP3 community is pretty well divided into two groups of people – those who only want to listen to popular music and those who want to cart around all their tunes. Interestingly, there is a pretty good age differential between the two groups and there is a big difference in MP3 players between them.
If all you’re going to listen to is pop music and you’re going to dump the music in a couple of weeks, then a little (small storage capacity) player is ok and the price is low, ~ $100 unless you just have to have an iPod. If you’re going to cart around 100 albums of whatever, then you need a big player and that runs upwards of ~$250.
The other factor is that these players are fixed size buckets and once they’re full, they’re full until you throw something away. Its a bit of a bookcase metaphor. And podcasts are big files compared to songs. They may have lower sound fidelity (bits per second) but they’re 10-60 minutes each as compared to 3 minutes for a song.
So the bottom line, which you all thought I’d never get to:
- Find a podcast you want to try;
- Get the software to download the podcast;
- Move the podcast onto your MP3 player;
- Listen to it; and
- Throw it away.
Lot of touching here. Lot of places to get discouraged, especially if you are a baby boomer and not as computer literate as you know you should be.
Which is part of why the numbers are small. There’s also the factor that most of the podcasts are pretty bad, even though they don’t have much in the way of commercials (as compared to TV,) and almost all of them have something annoying about them. So what the numbers aren’t getting is the folks intimidated by the process or turned off by not finding what they want.
And the fluctuations? That part is easy. Given all of this process, its pretty easy to see that what you do is set up a routine. I have the software set to download to this desktop all during the week whenever the software wants to and can get bandwidth, and then I transfer to the MP3 player on Sunday to listen to through the week.