Music Business

Are Streaming Music Services Providing Useless Artist Data?

BaddataWhile many streaming services may talk up the fact that they provide artists with data and analytics regarding their listeners, closer inspection reveals that, in addition to not being all that useful, a lot of this information isn't always accurate, says Chris Castle.

 

_______________________________

Op-ed by Chris Castle of Music Technology Policy

Pandora and Spotify (among others) have made a big deal out of providing “data” and “analytics” about streaming uses to artists–and particularly managers–about how the artist is performing on their respective services.  The “artist data” meme is also offered up as a value add to counter complaints of low royalties.  There is a real question of how useful this “artist data” is and a recent CNBC article calls into question just how accurate it really is in the first place.

Of course the most valuable piece of “artist data” that services could at least help the artist acquire–the fan’s email address–they won’t touch.  Obviously, I’m not suggesting that the service hand over the fan’s email address to the artist without the fan’s consent, so let’s not go down that rabbit hole, a favorite of the services trying to avoid this issue.

What I am suggesting is that the service provide the fan with an opportunity to sign up for the artist’s own email list.  This could be as simple as a link that would take the fan outside of the service momentarily to the artist’s email list sign up page.  That way I don’t believe there are any privacy law issues for the service as there would be if the service just handed over the email address.

I have raised this with senior executives at Apple and Spotify and it went nowhere.  The Spotify person rejected the idea outright because it would take the Spotify user (aka the artist’s fan) outside of Spotify.  Strange, because the fan would be offered a choice.  You know–the fan of the artist who most likely was driven to the service by the artist they are streaming.  (This would produce another interesting metric based on the number of email list sign ups by service, but I digress.)

Aside from whether the type of information being provided is even useful to artists, there’s another question of whether the “artist data” is even accurate in the first place.  And how would you even know.

CNBC did a little fact checking on streaming data provided by iHeart Radio analyzing the recent Grammy nominees.  This isn’t exactly the same as the “artist data” being hawked by streaming services, but it is perhaps a good proxy (since it’s hard for artists to see each others artist data results).

Iheart-radio-logoiHeartRadio (owned by Clear Channel) gave CNBC the “artist data” for the most popular tracks on Clear Channel’s massive streaming operations.  But CNBC discovered by using simple logic–aka sequential thought–that Clear Channel’s “artist data” was wrong.  Because CNBC concerned itself only with massive hits, data checking was relatively simple (which of course makes the Clear Channel screw ups look even more idiotic).

Keep in mind as you read this that if you’re an artist using the “artist data” for its recommended purpose–discovering nuances about the service’s listening audience–it will almost certainly be more difficult if you’re not Ed Sheeran or Taylor Swift.

In a blog post based on the original (aka wrong) data, iHeartRadio said Ed Sheeran took home the honor of “most-thumbed up” track of 2015 with “Thinking Out Loud.” Taylor Swift’s three big songs relegated her to second, third, and eighth place. (Update: the original iHeartRadio blog post was taken down. The link above is a cached version.)

42f5ce85More impressive in the original data was that Drake’s “Hotline Bling” was the most-thumbed up track in 26 states in 2015. That would be amazing because the song only came out in July and didn’t really catch on until the colorful video was released in October. Here is exactly what the blog post said about Drake:

Although the track didn’t premiere until mid-2015, with the unforgettable music video coming out just weeks ago, the last-minute addition of Drake’s “Hotline Bling” surprised and delighted the ears of listeners across the country, leading it to become the No. 1 most thumbed song of 2015 across more than 20 states!

But this didn’t make sense to us. Sure Drake’s song was popular, but how could it be the top song in half of all states even though it was only hot at the very end of the year?

More weirdly: how could Drake’s song be No. 1 in half of the states but not appear anywhere in the top 10 songs nationally.

CNBC pointed this out to Clear Channel, who admitted their mistake and “corrected” their bad data.  Yet that “corrected” data was FUBAR also according to CNBC:

The new data included the actual number of “thumbs up” and “thumbs down” by state. Of note is that Ed Sheeran’s “Thinking Out Loud” is the “most thumbed up” track in only one state — Hawaii. We find it is extremely unlikely that his track was actually the top track in the nation overall. IHeartRadio has not yet responded with further clarifications about the rest of its data set. We can’t check that against its original report, but we’re hoping to find out soon.

Of course, the new data set only included raw “thumbed” numbers for the top track in each state, so it’s possible that “Thinking Out Loud” had enough second-ranked “thumbs” to overcome Taylor Swift’s “Style,” but that seems unlikely.

So the thumbs didn’t have it after all?  Clear Channel miscounted?  Hard to say because we are entirely dependent on Clear Channel to provide the data backing up their work product.  But we know there was some kind of screw up because CNBC had to provide a link to a cached copy of the original Clear Channel blog post with the bad data.  If there’s no agenda here, why would Clear Channel try to hide their mistake from artists?

But good work by CNBC–except for one thing.  They continued the streaming service meme that somehow talent buyers and concert promoters care about how many thumbs you get when deciding to book a gig.

[C]oncert planners and promoters depend on data like this to create touring routes.

No, no, a thousand times no.  Clearly CNBC have never met a talent buyer. This is just simply not true.

Nonetheless, this CNBC post is a great example of how unreliable this “artist data” can be and just how hard it is to verify.  So hard that it may well be simply misleading to ask artists to take a lower royalty rate for what is most likely some species of snake oil.

They can keep their thumbs.  I’ll just settle for asking the fan to sign up on the artist’s site, thank you very much.

Share on:

2 Comments

  1. Really interesting article, it can even be said that in most cases they’re using the artists image/draw to market their product yet the artist receives no formal affiliation for this either.

  2. “Email capture” is brought up all the time as the gold standard for quid-pro-quo with labels and artists doing deals with streaming services. Services like bandsintown, songkick, and topspin have offered drop-in solutions for this but without backing from the major players (spotify, pandora, youtube) there’s not enough traction to make it work. End users need either brand recognition (“oh, I know songkick so I’ll trust them with my email”) or some kind of carrot (a free download or discounted tickets) to make the leap to signing up for an email list. Streaming services can help reduce the friction but the email aggregation needs to be centralized or at least standardized.

Comments are closed.