The funny thing about the music industry, as he says, is that it is fundamentally unconcerned with musicians and audiences. In the old days, it did this by making money regardless of the level of success a band experienced, as documented in its glory by "The Problem With Music". Now, it does so through analytics.
The article, on Shazam, treats data as a verifiable way to determine what music is going to be popular. It is premised on two assumptions that are quite debatable:
1. Musicians will continue to produce music in a range of genres, thus providing a constant stream of product to be sold. The quality of this music is unimportant, only whether people will buy it.
2. The important thing about audience reaction is that it exists. Why an audience reacts to a song is inconsequential.
In other words, Shazam's decisive technological victory is to take songs that have already been written, recorded and distributed, and use the fact that a growing audience has already liked the song, to determine that yet more future people will like it. Presumably, the song popular with a growing audience will continue to grow quite apart from Shazam or any other analytics. "Royals," the article's example, seems to have done just fine without concerted record industry action. Synthesizing this information isn't any kind of innovation at all: one would have to demonstrate that analytics make songs more popular than they would have been otherwise.* If you were looking for evidence that the music industry thinks of music as widgets and fans as lines on a balance sheet, you could hardly do better.
*Analytics make sense in some contexts. For example sports, where there are objective (if subjectively-defined) measures of success, and correlations can be found for the regular production of these measures. Very little is similarly objective in the arts. Certainly not amongst the whole population. If the purpose of analytics in music is to determine which sub-demographics might appreciate a certain type of music, then you're just re-creating the distribution models for independent music, or for niche markets like jazz and classical.