Posts Tagged ‘Big Data’


May 14, 2018 Comments off

One of the big pitches we have heard for years from digital services is how they can provide artists with data resources to connect with fans.  That is–everything except a meaningful way to connect with the fans that the artist isn’t already driving to the service in the first place.

Of course, the most laughable part of this pitch is that somehow knowing you’ve been streamed in Shoreditch, Tyler, Yellowknife and Brooklyn is going to be meaningful to a talent buyer, even if that talent buyer books in those towns.  Yet we frequently see journalists dutifully spout this received wisdom as if it meant something other than trying to gin up a reason to pay artists and songwriters a still lower royalty to offset the cost of rent at World Trade Center.

Alan Graham’s recent post in Music Business Worldwide has put his finger right on another problem that defies the conventional wisdom and Spotify narrative–big data ain’t all it’s cracked up to be and may be going the way of Cambridge Analytica.

Big data was a solution pitched and sold to the music industry as a panacea to fan engagement problems. While big data seems very attractive, using personal data and profiling fans may in fact turn out to be, like oil and plastics, already outdated and toxic….

In a 2014 New Yorker article, Spotify was keenly aware of the power of such data:

All this, Ek explained, will help Spotify to better program the “moments” of a user’s day. “We’re not in the music space—we’re in the moment space,” he told me. The idea is to use song analytics and user data to help both human and A.I. curators select the right songs for certain activities or moods, and build playlists for those moments. Playlists can be customized according to an individual user’s “taste profile.” You just broke up with your boyfriend, you’re in a bad mood, and Justin Timberlake’s “Cry Me a River,” from the “Better Off Without You” playlist, starts. Are you playing the music, or is the music playing you?

…In fact, when you agree to use Spotify, third-parties who install the Spotify widgets on their sites may also send data as to which page on what site you are visiting.

With this knowledge, just how comfortable are you knowing that Facebook is now contextualizing your private chat messages to suggest music on Spotify?


Sharing is Caring: What is Google’s Position on Data Sharing with Artists?

February 25, 2013 2 comments

While it remains to be seen exactly what the contours of a deal might be, Zoë Keating‘s advocacy of data sharing with artists by online music retailers is getting some traction.  Beats is making positive noises in that direction and I would expect others to follow shortly.  This is something of a privacy law challenge, but it could be something as simple as a “sign up here” button for the artist’s email list next to the “buy here” buttons to buy the artist’s downloads or CDs.

I’m willing to be educated otherwise, but it seems that an email opt in would be unlikely to present a greater privacy issue at all for the retailer.  This is because the fan would be in control of the opt-in decision, and any privacy rules applicable to the email list would live at the artist site and would (or should) already be in place regarding the existing artist email list.

However–this is a good moment to bring up a subject we have banged the drum about at MTP for years now:  Nondisplay uses of music.  Meaning uses that are made by a Big Data company like Google of information about music use, consumption, sales, distribution (or whatever) that are (1) not reported to the artist, (2) are covered in the cracks of the Big Data user privacy policy, and (3) that are sold, resold, and sliced and diced by the Big Data company like there is no tomorrow, now and forever amen.  And of course–none of the revenue derived from nondisplay uses goes to anyone other than the Big Data company–in this case, Google.

It is very, very unlikely that a Big Data company like Google is interested in a music platform just because they really want to be in the music business and make their cut of subscription, download or webcasting revenue.  No, the real cash cow for Big Data is Big Data.  That’s not a tautology–it was confirmed as recently as last week by the Financial Times (with the hysterically funny title, “Google Looks to Beat Music Rivals“):

Google is in talks with big music labels to launch a streaming service to compete with companies such as Spotify and Deezer, as it looks to expand into one of the fastest growing areas of the music market.

The discussions reflect the technology company’s ambition to extend its influence into new business areas and diversify away from advertising, which accounts for 95 per cent of its revenues.  [Good thinkin’, Dob….]

[But, wait, not so fast…there’s more….] Advertising executives also speculated that by scrutinising consumers’ listening habits, Google could build a valuable database for advertisers.

“It will be another piece of the puzzle for understanding consumers,” said Christophe Cauvy, European head of digital at advertising agency JWT. “This will be very interesting for brands where purchases are emotionally or status driven.”

So…think about that.  Where’s the real value going to lie…so to speak…for Google?  Making a tiny vig off of music, or collecting a bunch of information about fans that can be used in other Google products?  Hmmm?  As Ben Sisario identified it in the New York Times, which of the Two Googles are we talking to?  Aaron or Roy?

The Financial Times is even fuzzy on this issue as their article starts like they think that Google intends to get further into the music business because they want to make money off of streaming.  Given what we know about artist royalties from streaming, that seems highly unlikely.

What seems more likely is that Google Play gets consumers to put a Google entertainment center in their home and then Google monitors them all the live long day to serve advertising to the fans.  Maybe not while the fans are listening to a no-advertising subscription service, but when the fan leaves that environment and uses Google for something else.

Of course, Google will add this information from non-display uses to the data that it has already collected from serving ads to pirate sites offering the identical music, movies and books.

Do you think for one second that Google (or any other Big Data company) would share that information with the artists whose music gave it value?  Much less share it for free?

Hello, Roy.  I thought that was you.

%d bloggers like this: