Magnatune discontinues Shoutcast Radio StationsI decided to run a test to see if Internet radio was worth the cost. We put up all our shoutcast radio stations for 2 weeks (they were previously down) to see if it had any impact on sales or web site hits.
In the two weeks time, we saw no increase whatsoever in web site hits or sales, and the radio stations were often hitting the maximum number of listeners 200mbits of bandwidth could handle (2000 listeners).
It appears to me that most people who are listening to Internet radio run it all the time (the average listener time is 3h 20m, so that most who stay around are actually listening about 8h at a time, since 50% leave right away) and are not often motivated to purchase what they hear. Rather, they listen to it as background music.
Another thing I noticed is that the 200mbits of bandwidth aren't enough to support all the listeners who want to hear our radio stations.
I suspect that what's happening is that people new to Magnatune aren't able to get a "slot" in our radio feeds, because the slots are all filled up by people who listen all the time. That's probably why shoutcast radio used to be a good sales generator for us: we were new and everyone finding us was hearing us for the first time, and many people converted into buyers. As time proceeds, more and more of the radio station bandwidth is dedicated to full-time-listeners who don't buy very often.
Since the cost of running Internet radio are very high -- $5000 per month in bandwidth for just 2000 listeners, and either not many new people are able to get on, or else the existing ones are not motivated to buy, I've decided to permanently discontinue our shoutcast feeds.
That being said, the ratio of listeners at our web site, to buyers, is still quite good -- about 40 listeners to 1 buyer, and therefore we'll absolutely continue to do that as that's a business model that can survive. It looks to me like web site visitors are much more involved with the "Magnatune experience" and consciously looking for music that they may buy, rather than radio listeners, who are looking for a background music experience.
We don't see a plausible business model for Internet radio at this time, at least not one where one has to actually pay for the bandwidth used.
That being said, if a) bandwidth prices plummet, b) we get free bandwidth (gigabit+) through peering arrangements or donations or c) p2p radio technology takes off, I can definitely see restarting our radio feeds.