Link To Article

The amount of online “chatter” about an upcoming album release
directly correlates to higher physical album sales, according to two
researchers with New York University’s Stern Business School.
Professor Vasant Dhar and former student Elaine Chang observed the
trends of 108 albums released during the first two months of 2007 to
see how different outside elements affected (or predicted) sales once
the albums became available, and found that all of them had some
effect or another. But certain elements of online chatter—namely blogs
and social networks—seemed to be fairly accurate predictors of future
success.

The researchers followed the Amazon sales ranks for each of the 108
albums over a period of eight weeks (they said that Nielsen SoundScan
stats would have been ideal, but they are costly and proprietary), as
well as articles, blog postings, and MySpace friend counts about them.
The blogosphere appeared to be most strongly correlated to better
album sales—if 40 or more legitimate (written by normal people and not
by marketers) blog posts were made before an album’s release, sales
ended up being three times the average.

That trend doesn’t just apply to music from the Big Four, either.
Albums from independent labels enjoyed the same level of success. But
if an album was from a Big Four record label, sales increased five-
fold after 40 legit blog posts. If blog posts crossed 250, album sales
turned out to be six times the average, regardless of label.

The number of MySpace friends a particular band has also correlates to
better album sales. “The number of friends a band has is displayed on
its MySpace page is like a public badge of popularity,” wrote Dhar and
Chang, while observing the change in the number of MySpace friends
from week to week. The bigger the increase in MySpace friends, the
better an album’s sales turned out to be. The change, however, was not
as big as that related to blog posts, which the researchers believe is
because adding someone as a friend on MySpace is a relatively passive
process compared to putting the effort into composing a blog post.

Despite all of this new data, a good review in Rolling Stone still
can’t be beat. “Although we found that user-generated content is a
good predictor of music album sales, our analysis showed that
traditional factors cannot be ignored,” the researchers wrote. Music
from major labels traditionally sold 12 times as much as that from
independent labels, and the more mainstream media reviews an album
got, the higher the sales.

As to whether all this online chatter actually causes or merely
predicts online sales, the researchers can’t say. “It is not possible
to make such a conclusion based on this study,” they wrote, nothing
that it was probably a mixture of both. The quality of the artist and
expectations about the album causes people to talk about the album
more before release, but new buyers could be swayed as a result of the
increased chatter. But Dhar and Chang warn that if there is any
causality involved, it has to be totally organic in order for the
effect to work. “[I]f blog posts start becoming manipulated because
people think they have an impact on sales, that the predictive power
might disappear because the underlying reasons for it disappear,” they
wrote.