I wish we didn’t have to have this discussion over at TechCrunch, but it really is happening.  Internet radio and its cousins are facing asymmetric pricing compared to terrestrial radio.  Many call it stupid of labels to cripple internet radio.  After all, doesn’t internet do a much better job of promoting music purchases?  Listeners can actually identify the song and artist playing, and the interface even provides convenient links to purchase channels via iTunes or Amazon.   Surely that must be a great sales driver that can not be replicated by the non-interactive interfaces of terrestrial radio!  Well it probably is, but internet radio is also deeply disruptive to the fame-creation process, an important source of value to record labels.

Funny signals

Many Web 2.0ers look forward an era where fans can easily signal their tastes, generating preference data which could then be used to form robust recommendations.  More data is a good thing right?  However this signaling has no place under the current business model.  Today, labels invest money in production and promotion to make signed artists famous, and any outside signaling pathways would be more of an interference than an asset to the fame-creation process.   Part of this promotion engine is terrestrial radio, where the playlist is influenced by the content industry.  Internet radio is a radical departure from this because its playlists are generated algorithmically.

In internet radio, the audience signals preferences to the industry, while in terrestrial radio, the industry signals preferences to the audience.  Both are promotional channels to create fame, but its “tell us what you like” vs “tell you what to like”

This is critical to labels because much value is derived from the fame-creation process.  In the old paradigm, before the internet made preference signaling easy, it was difficult to gather early-stage information on which artists are likely to grow famous.  To deal with this uncertainty, the industry developed a model where talent scouts selected artists, and then fed these artist through a powerful production/promotion engine which virtually guarantees fame.  That way, labels don’t have to worry about betting on the wrong horse: just juice them up to secure the win.  Because the artists are signed pre-fame, the initial contract is far more favorable to the label than to the artist.

The degree to which the initial contract favors the label is directly proportional to the uncertainty around the budding artist’s potential fame, an uncertainty that is asymmetric given the label’s ability to influence the outcome

Essentially, the fame-creation process then was actually a form of arbitrage.  But the introduction of early-stage signaling changes all that:

If an artists arrives at a label having already been made famous by audience-to-industry signaling pathways, the label loses bargaining power when signing the deal.  Additionally, the introduction of externally-generated fame competes against the internally-generated fame in which labels are already deeply vested.

The three pillars

You could certainly build an internet radio/listening service that fits within the old promotion paradigm.  It is neither the internet nor the radio that embodies the threat, it is the user-generated content (preference data) and its democratizing effect.  Specifically, democratization knocks out one of the three pillars of value in the recording industry: marketing.  To learn more about the three pillars, I suggest you read the classic Ian Rogers letter to Guy Hands.  It is pretty much required-reading for anyone looking to transform the music industry.