In Part I I focused on what I feel is a major issue with Pandora (P) -- cost, based on data usage, assuming the user is using Pandora as their primary source of audio entertainment in a mobile environment.
In Part II, I will focus on problems with Pandora's advertising model, and how it contrasts with SiriusXM's (SIRI) subscriber-based model.
Pandora derives much (actually a majority) of its revenue (83 million expected for Q4 2011) from advertising. This model is, for now, similar to terrestrial radio, in that the user is subject to interjections between a certain number of songs of typically 15 to 30 seconds, and forced to listen to an advertisement. This is part of the "fee" the user pays, in the form of time.
SiriusXM derives almost all of its revenue from subscription fees. Its two points of worry are retaining current customers, and bringing new customers in. SiriusXM does not need to worry about advertising, where its users are listening, or if they are even listening at all. As long as they are paying for the service, it's all the same to SiriusXM.
In order to be successful as a company Pandora must grow advertising revenue to offset cost of content, either by increasing the number of ads (detrimental to the user base) or by increasing the cost of the ads through making them more valuable to the advertisers.
I have witnessed firsthand f the increasing number of ads that come through the Pandora stream, and I think Pandora may already be at point break for ad quantity. Thus, I feel their only source of increasing revenue beyond increasing the user base, is to increase ad "quality" or rather, value, so they may demand a higher premium for their advertisements.
The problem with this? In order to drive the cost up on advertising, they must prove that the user base is attentive, and also that ads are targeted towards the correct demographic. While I have no doubts Pandora can and will target appropriately based upon the service being delivered over IP, I have my doubts that they can prove the user base is attentive without said user base being in a vehicle.
When you are in a vehicle, you're concentrating on driving, and perhaps listening to the radio. You're not working around the house, you're not doing homework, you're not in the other room with a bit of background music playing. Advertisers want the attention of your ears, and the vehicle is the place to virtually ensure this. The problem? The vehicle is where Pandora is either absent, or is present, but with issues that prevent it from being viable as a radio replacement.
That "issue" is mobile internet infrastructure, which is not reliable enough to make Pandora a viable radio replacement for the automobile. SiriusXM is the only viable radio replacement, and has done an impressive job building their own infrastructure, for near spotless reception anywhere and everywhere in the US.
I do not see this changing anytime soon (less than 3 years) and I have a great concern for the short term viability of Pandora. I think, for all intents and purposes, they are just a bit too far ahead of the infrastructure curve. This is not a bad place to be, in most situations, but waiting for technology and infrastructure to catch up to your company can be perilous in the short term.
And in this wait period, the barrier of entry is relatively low for competitors to crop up, or for established juggernauts to develop their own internet radio systems. This leads me into Part III (upcoming), Pandora's third pitfall: low cost to entry.