So is the decline in satisfaction shown by Bridge Ratings for folks who have listened to Pandora for a long time a sign of "fatigue or boredom" as my good friend Dave Van Dyke suggests?
1. Peak satisfaction naturally declines for most products after using them for a while. That, after all, is why we replace things that aren't broken. This is hardly unique to Pandora.
2. All Pandora "users" are not created equal, just as all your station's listeners are unequal. How do we know that the longer-term users are as "heavy" as the newer ones?
3. We assume that this pattern is not shown in our stations, but in general it is. Don't you think fans are more passionate about a station that's new than one which isn't? Of course they are. In format change situations we always – ALWAYS – assume that a successful station will settle 20% below the ratings at its peak. And that's the successful ones. We also know about stations who switch formats three years after a booming opening act because the bottom falls out.
4. The odd figure isn't the 50% satisfaction after three years, it's the 90% satisfaction in less than six months. That's insane and patently unsustainable. Indeed, 50% of your audience being characterized as rabid fans is far better than most products will ever do. Tell me, would 50% of your cume rate themselves "highly satisfied" with your station? Speaking as someone who sees these numbers, my educated guess would be "no."
5. "Satisfaction" is an attitudinal behavior that is not the same as measuring usage. What are the real world statistics on usage among short-term and long-term users? Do long-time fans listen less than short-time ones? Pandora knows these answers, but surveys such as this do not.
6. Don't discount the effect of Pandora's recent cap on heavy usage designed to motivate users to pay for the service. That is likely to strike long-time users as more distasteful than short-time ones and reflect in their scores of "satisfaction."
7. As much as extreme satisfaction is down over time, note that extreme DISsatisfaction in not particularly high.
None of these points mean that the conclusion is incorrect, of course. I am simply arguing that you should not assume it to be true.
I think when we see numbers like these we too often view what we hope to view rather than what might actually be in the numbers (or beyond their reach).
Just as we relish the foibles of the fabulous and wealthy, so do we relish the imperfections of radio's most highly publicized substitutes.
Throwing stones doesn't make our glass houses any more secure.