Something very fishy is going on in the world of music this week.
Before diving in, let’s lay down the groundwork of this piece. The story we are about to detail for you is potentially still developing, surfacing just after the release of BTS’ newest single, “Butter,” on May 21, 2021. As such, what we are covering here today is a current issue at the time of this writing. Updates will be forthcoming as we learn more, so watch this space for additional information as it is unearthed.
Here’s what we know so far.
It’s no secret that ARMY runs like a well-oiled machine when it comes to supporting BTS, whether it be through voting for them at award shows (like our recent BBMA win for Top Social Artist, where ARMY won by 100 million votes) or streaming newly-released BTS content. We’ve worked hard to perfect our strategies for streaming, creating streaming guides and step-by-step articles on the “do’s and don’ts” of streaming BTS content on a variety of apps. So, it’s no surprise that as soon as “Butter” dropped, ARMYs were at attention and ready to let the streaming commence.
Interestingly, the group announced via the official BTS Twitter account that they would be hosting their own streaming parties every day for a week, from May 21 to May 27 (the last one is today, for anyone who hasn’t already joined in, starting at 9:00 PM EST). Though it wasn’t an unwelcome addition — ARMYs, after all, are perfectly willing to listen to official BTS-curated playlists as well as fan-made ones — it was a little unusual. BTS had never openly endorsed such events for previous music launches. Why would they hold their own streaming parties this time?
After just a short time, a possible answer surfaced.
Discerning ARMYs began to notice that there was something amiss with their Spotify streams. Numbers of listens for the song were taking an abnormally long time to update, leaving the number of total streams for “Butter” at below 1,000 for an extended period of time. Unsurprisingly, ARMYs immediately grew suspicious. This suspicion was only exacerbated by the discovery of some other discrepancies between the Spotify charts and playlists.
“Butter” debuted on the chart with over 11 million global streams, according to Spotify’s official numbers for the day, which is staggering on its own. As Forbes explains in their article, it broke the record previously held by Ed Sheeran and Justin Bieber for their hit “I Don’t Care,” which was released in 2019, making it the biggest debut in Spotify history. However, it somehow did not net the boys the top spot of the day, which instead went to Olivia Rodrigo’s “Good 4 U,” putting BTS in second place.
In and of itself, this wouldn’t necessarily be impossible. There are other popular artists out there who have a lot of fans; this is something we have to admit. Defeat must sometimes be gracefully accepted. However, in some aspects, the math doesn’t quite add up.
We have multiple sources that were live-tweeting or otherwise posting throughout the day, indicating that the numbers were higher than the 11 million which Spotify officially reported. For instance, this article from Naver, which the official Big Hit account shared, states that “Butter” debuted with almost 21 million streams. Here’s the direct quote in the original Korean article:
And here it is translated via Google:
So where did this “official” 11 million number come from?
View and stream “filtering” has long been considered an issue. Platforms are known to adjust views or streams, ostensibly to cut out streams from bots. According to Venture Music, a music marketing company based out of Nashville, TN, “bots are programmed to control individual Spotify accounts to play the same song(s) or playlist(s) repeatedly on a loop.” They also explain that labels are sometimes paying for botting services to push their songs up the charts.
On the surface, eliminating these bots is an admirable goal — we would naturally want to have accurate numbers reported and would not want people to cheat. However, ARMYs have suspected for a long time that our views and streams might be targeted unfairly, resulting in excessive fluctuations in stream numbers. As an example, one Twitter user snapped two screenshots of the official music video for “Butter” on YouTube. You can see in the screenshots that the lower number of views comes from a later time based on the number of likes being higher.
Might some of the deleted views or streams be from bots? Sure. The numbers are particularly large, however, and it’s made ARMY wary. This sort of situation is specifically why the community created the previously-mentioned streaming and viewing guides so that we could avoid looking like bots and having the numbers of views/streams adjusted. These guides were created after much observation and trial-and-error; a simple Google search for how to properly stream music videos or songs brings up multiple blogs, videos, and the occasional Reddit post from fans.
The Spotify Terms and Conditions of Use does include some of their policies on botting in Section 9, User Guidelines (yes, we did read through this for you; click here to skip straight to Section 9). The section states at the top that “The following is not permitted for any reason whatsoever,” and the following points apply directly to botting:
“8. artificially increasing play counts, follow counts or otherwise manipulating the Service by (i) using any bot, script or other automated processes, (ii) providing or accepting any form of compensation (financial or otherwise), or (iii) any other means;
“12. "crawling" the Spotify Service or otherwise using any automated means (including bots, scrapers, and spiders) to view, access or collect information from Spotify or the Spotify Service;
“13. selling a user account or playlist, or otherwise accepting or offering to accept any compensation, financial or otherwise, to influence the name of an account or playlist or the content included on an account or playlist; or
“14. artificially promoting Content by automated means or otherwise.”
Searching on the Spotify website in their support section only turns up brief posts about how to sort and filter playlists; no results are available for “botting,” and other variations of “bot” turn up only general “getting started” pages.
One person asked on the Spotify forum how streaming actually works, and the responding moderator stated that the information could be found on the Spotify for Artists page...which does not really answer the question. Following the provided link leads to the home page, not an actual post about how streaming works, and any posts in the Spotify for Artists page about promotions or playlists are geared towards the artists promoting their work, not a fan who wants to make sure their streams are properly counted (which is who the inquirer was if their question is read properly).
With vague or nonexistent information from Spotify, people who wish to support their favorite artists through streaming are left in the dark about why some streams count and others don’t; we can only do our best to interpret what we see, adjust our actions accordingly, and create our own guidelines based on the observations of others, such as the generally-accepted rule that “if the song plays for 30 seconds it counts for streams.”
Streams and views being filtered is old news to ARMYs; it’s frustrating and needs to be addressed on an industry level. However, this was not the only discrepancy that ARMYs found.
Olivia Rodrigo’s “Good 4 U” earned the top slot for number of streams on the Butter’s release day and was the number one song on the Spotify “New Music Fridays” playlist in a number of countries, with “Butter” appearing second. However, “Butter” appeared higher on the song charts for those same countries, with “Good 4 U” appearing below or not at all. Numerous Twitter threads have been made on the subject, and screenshots of the Spotify charts are provided to show the differences both by the original user and then by others who added to the thread as replies; here’s one example so you can look through and see for yourselves how many ARMYs reported this discrepancy globally.
Why would this discrepancy exist? Unfortunately, there’s no easy answer; the pathway is much more convoluted than it looks at first glance.
First, we have to look at how Spotify playlists are created. In this case, the Spotify for Artists website can help us. They break down exactly what kind of playlists you can find on the app: personalized playlists, editorial playlists, and listener playlists.
Listener playlists are the easiest to understand, in that they are somewhat self-explanatory: they are playlists specifically created by users of the app — any playlist you may have made for yourself or your friends would be a listener playlist, where you manually add to it and change it based on what you’re vibing with. Personalized playlists are a variation of this; they are created algorithmically based on what you play, like, save, share, and skip, as well as habits from similar listeners. Yes, they are tracking what you’re listening to, and this tracking allows them to suggest new music for you in playlists such as “Discover Weekly” or “Release Radar.” So, in theory, someone who likes K-pop but hasn’t (somehow) heard of BTS might see “Butter” pop up on their discovery playlist based on the fact that they’ve listened to a lot of other K-pop.
Most playlists on Spotify, however, are editorial playlists. According to the Spotify website, editorial playlists are “curated by our playlist editors. Our editors are genre, lifestyle, and culture experts with diverse backgrounds. Many playlists have several editors in different locations to find music from all around the world.”
Here’s where things start to get murky. These playlists — of which “New Music Fridays” is one — have people in charge of them. Not math. Not users’ input. People. People chosen by Spotify.
As we’ve seen time and time again, people have biases. That’s unavoidable. We, as authors, and you, as readers, have a bias towards BTS. We have always been transparent about the fact that we are writing with this bias and have dedicated ourselves to covering the issues we discuss as objectively as possible, with the aid of thorough research (as we have done here for this very article). As soon as a human factor is placed into the process, those biases start to affect the outcomes, especially if they don’t actively try to counteract them.
In an interview with Forbes last year, Jeremy Erlich (Spotify’s Co-Head of Music Strategy) states that “editorial decisions are based purely on the quality of the song and its fit in the playlist.” He also, when asked about whether or not the editorial curators are “gatekeepers,” says “I don’t believe on the editorial side that we’re gatekeepers. We’re connectors. We’re the ones who should know what song goes in the right place so that they can have a path to the top.”
Digest that for a moment.
By his own admission, Erlich says that songs are being placed. By people. People who have thoughts and opinions. Is this really the most objective method? It seems like it might not be, considering the fact that a more objective method, an algorithmic method, exists. That being said, even an algorithm might be subject to some slight bias via the person programming it. Erlich also is stating that they are the ones who should know what song goes in the right place. It seems a little self-aggrandizing, does it not? This sort of sentiment feels very much like the industry award shows, which have recently come under fire for their vague criteria and lack of diversity; we have commented on this in the past.
When it comes to art, who is really qualified to say that one thing is higher quality than another? One person may prefer different things compared to every other person around them. It’s the same award show situation all over again, just in a different setting.
As Rootnote reports, label services company Systemic did a study of the “New Music Fridays” playlist as well as “Singled Out” and found that for the week that they examined, 75% of the songs on the playlists were from major labels or imprints of major labels. This means that there was a larger marketing budget to back the release of the song or album. Ostensibly, this shouldn’t matter, according to Erlich’s statements, but it does seem like an interesting coincidence. Does this imply that indie music, for example, would not be as good as music from major labels? Do they categorically not “fit in” with the major label music? It exudes an aura of the odds stacked against indie artists. Based on the Western reception of BTS and other global acts at award shows thus far (including but not limited to K-pop), it would not be a huge stretch to assume that the same logic could apply to non-Western acts in the case of Spotify playlist curation.
In the same interview with Forbes, Erlich states that “we’re always cognizant that the artists do all the hard work. They’re the blood that comes through our veins and we’re the conduit.” This is an interesting statement to make, considering how Spotify came under scrutiny in November 2020 with the announcement of its new “Discovery Mode,” which Rolling Stone calls “payola” and Music Business Worldwide calls “pay for influence.” A brief explanation is that artists and/or labels can agree to lower royalty payments in exchange for Spotify “prioritizing” their music in playlists.
Not only does this sound directly contradictory to Erlich’s statement that the curators at Spotify are not gatekeeping, but it also could be damaging to the artists themselves, who are the supposed lifeblood of the streaming service. Artists already receive less in royalties from Spotify than they do from other streaming services, such as Apple Music. According to Wall Street Journal, Apple Music pays artists a penny per stream, which doesn’t seem like a lot. Compare it to Spotify’s rate, which, according to Business Insider, is between $.003 and $.005, and it seems like a veritable gold mine. This is apparently quite a bit lower than it was in past years; Business Insider states that “in 2014, they paid $.00521 on average, but two years later, the average rate dropped to $.00437. By 2017, the average pay rate had been reduced again to around $.00397.”
Now, with the new “Discovery Mode,” artists can choose to accept even less money in order to get more exposure? By some logic, because Spotify is so big, you’re more likely to get a larger number of streams on their site than you are on some others. However, this hasn’t stopped Apple Music from offering higher royalties, and the Apple corporation is among the biggest in the world.
Is it possible that Rodrigo and her label, Interscope Records, opted into the “Discovery Mode” scheme? Very much so. Something to note is that Interscope is part of the larger umbrella corporation Universal Music Group and, interestingly, is the former employer of Jeremy Erlich.
While this connection exists and could present a conflict of interest, it could easily be a coincidence, and to state that something shady is happening is pure speculation. People change careers and employers all the time. What we have for concrete information right now is that a) stream numbers for “Butter” were likely heavily filtered, b) most Spotify playlists are curated by a person or multiple people who are making decisions about what goes where, and c) artists now have the option to accept decreased royalties for boosted content.
Now that we have all this to factor in, it seems like there is a plausible explanation for “Butter” being placed where it was on playlists. We don’t know at this time whether or not HYBE might be opting in to the “Discovery Mode” plan (as of February 2021, it was launched in beta, according to Rolling Stone). We don’t know for certain if Interscope has opted into this, either. Much of what we have here is information that may or may not be connected, and we may only safely theorize about what biases may play a part and what may be occurring behind Spotify’s closed doors.
Regardless of what may or may not have played a part in this particular case, the “pay for influence” scheme does seem like it would be damaging for artists, particularly those who haven’t yet gotten their “big break” in the industry and may not be able to afford sacrificing the royalties they deserve for their work. The “Discovery Mode” announcement and launch time is also questionable, considering the amount of money coming in via streaming services and the amount of revenue lost for artists who were not able to tour as they normally would in the last year due to the COVID-19 pandemic.
As of right now, the situation has an odd feel. The story, however, may continue to unfold over the next weeks. And even if things quiet down, for now, this is likely not the last time we will see this issue resurface. Keep an eye out here. If and when more information becomes available, we’ll keep you updated. Regardless of the circumstances around it, “Butter” still broke records — records previously held by Western artists. As Forbes reported, even with its likely-filtered stream counts, “Butter” had the biggest song debut in Spotify history.
Against the odds, on the record, we still triumphed.
And we will again.
DISCLAIMER: We do not own any audio & visual content in this video except for the editing. ALL RIGHTS BELONG TO THE RIGHTFUL OWNERS. No copyright infringement intended.
Written by: Anna Moon
Edited by: Vera
Checked by: Clare