Bridge Ratings just released some data relevant to Pandora fans that purports to show evidence that the popular online music service is showing signs of weakness. The study examined Pandora users of various stripes and segmented them by their tenure with the service, as follows:
The study concludes that "over time...the satisfaction level is affected by 'fatigue or boredom'" and that "the longer consumers use Pandora, fatigue and/or tedium sets in as users become accustomed to the programming 'style' of Pandora."
It's tempting to read into this data that Pandora's appeal has some weak spots. Reality intrudes here, as Clear Channel, CBS and other webcasters can attest, since Pandora is cleaning their clocks right now online. But does this data show evidence of future bumps in the road? Well, maybe--but maybe not.
First of all, we don't really have a benchmark for these numbers. How would overall satisfaction look for any channel or service over a three year period? Do you think it might show declines as the "halo" effects wear off? Of course it would. The only way these numbers have meaning is if we can see a comparison of "satisfaction over time" between other webcast services and Pandora--you might find that Pandora's "eroding" satisfaction levels are actually better than you think. Or not--you can't tell from this data.
But the more sinister bias of this conclusion is that the longer you expose a body of listeners to Pandora, the more "bored" or dissatisfied they are likely to become. Unfortunately, that assumes facts not in evidence with this study. What we do see is that longtime listeners report greater levels of dissatisfaction with the service than do new listeners. Some of that, as noted, is certainly the halo effect of a new service. But it might also be that the people who have been listening to Pandora for three years or more are simply different people than Pandora's newest fans. Early adopters are also early rejecters (I know--I am one!) As Pandora becomes more and more mainstream, it's attracting more middle-of-the-bell-curve listeners, who will accept or reject the service using different criteria than Pandora's early fans. That isn't speculation--we see this very clearly in our tracking of Satellite radio satisfaction over time, and I have little reason to doubt that Pandora's user base is undergoing a similar shift.
There's nothing wrong with asking several hundred Pandora listeners how satisfied they are with the service, and with asking them how long they've been Pandora listeners. Those are perfectly legitimate data points to report. Where studies like this get into trouble, however, is in reporting some kind of longitudinal effect with a one-time data snapshot. The assumption made in this study's conclusion, which I've seen parroted all over the Interwebs, is that the longer you listen to Pandora, the more likely you are to migrate from "highly satisfied" to lower levels of satisfaction. But the only way to really prove that is to ask the same people the same question over time.
Remember--you could just as easily read this data to show that Pandora has recently uncovered new users that love its service even more than its early adopters did.