Database Reference
In-Depth Information
three years who are active in this community. It is great to see that sort of
overlap. Anytime you run into data scientists, you can talk shop and you can hear
about what is going on in their respective fields. This is great because we all over-
lap in many ways, so that is also a great way to keep up with all that is going on.
Gutierrez:
Tell me about the social and music industry data that Next Big
Sound looks at.
Hu:
Next Big Sound provides analytics and insights for the music industry by
tracking different data signals to help record labels, artists, and band managers
make better decisions. To that end, the data that we focus on is combining
social media data with sales data, radio airplay data, events data, and any other
proprietary data we can get to provide context and cross-sectional insights to
the music industry. We keep track of social media on at least thirty different
sources—Spotify, Facebook, YouTube, Twitter, you name it—and combine that
with sales data from the record labels. The data is relatively granular. We
are talking about album and track sales for each particular artist on a daily
basis, broken down a lot of the time by different geographical regions and
demographics.
Gutierrez:
What do you seek to answer with the collected data?
Hu:
The original vision of Next Big Sound was: How do we find the next big
sound? Our core belief is that you can do this better with data, specifically
social data. So we take the various data sets and analyze a couple of different
things. Number one, what is the impact of social media on sales? Or what is
the impact of radio on concerts? What is the impact of events on stream-
ing? And then, what is the capability for forecasting album sales or future
engagements in social media? So when an artist plays a concert or makes a TV
appearance, what is the impact on their social media and sales? How much can
we measure that? Then how can we find up-and-coming artists?
What is interesting is that you can get indications of who is becoming popular,
who is going to be the next Justin Bieber, Katy Perry, Kanye West from the
earliest seedings of a couple hundred people listening to one of their tracks
on SoundCloud. You can see the growth of that artist. For example, Gotye
had a big song, “Somebody That I Used to Know” which came out in 2011.
Before the song exploded, he was relatively unknown outside of Australia.
Yet, you could see right when that song came out, if you were looking at
SoundCloud, that you were looking at the fastest-growing artist. Before he
was signed to a major label, before anybody knew about him, he was number-
one for months just based on that metric. So if you were tracking the data,
you would know that he was going to become big, even though he was not
signed until a few months after that. So one of the tools that we built was
essentially a reverse lookup on these metrics. Who has the most SoundCloud
plays in a given week, in a given month, in a given year? And that is based on
all the data that we have. We can actually build a database that allows us to
make that type of grid.
