Has streaming ruined the charts?

Charts like the Billboard Hot 100 are meant to be a measure of what's popular in music. A popular song should be one that's widely known by the general public. In the old days, determining what was popular was easy enough. Single sales and radio play were counted to find the biggest song of the moment. Overall those were accurate measures. Of course, radio play can be controversial. After all, record labels in the form of payola do pay-to-play to get songs onto the radio. But still, radio has a strong incentive to mainly play songs people want to hear. If they play too many songs people don't like, they'll lose listeners. Radio stations use surveys like callout scores and other methods to determine what their listeners like. So, while radio is somewhat imperfect, it's still a pretty good indicator of what people actually like.

But now streaming is having a big impact on the charts and in many ways it's a poor indicator of what's popular. Here's one example of why that is. Someone goes to YouTube to leave a comment on how much they hate Taylor Swift's new song. But by doing so, they've contributed to streams and to the song's chart success. Thousands of people who want to express their disdain for an artist or song by leaving a negative comment are inadvertently making the song more popular. Taylor Swift's Look What You Made Me Do has one million thumbs down and six million thumbs up. Every one of those million thumbs down helped increase its chart position.

If the thumbs up/thumbs down ratio is an accurate representation of how many people disliked Look What You Made Me Do, it would mean that about 84 million of the song's 600 million streams were by people who didn't like it. People won't buy a song they dislike. But for many reasons they will stream songs they dislike. As an example, a new release that gets mentioned on lots of websites, forums, and blogs will inevitably include an embedded video. Many people will click play out of curiosity and contribute to the song's chart success even if they never listen to it again. In that case, streaming reflects how much hype and attention the song got. Not how much people like it.

Curated playlists are another problem. All streaming services have top 40 playlists and hot songs playlists. Many people run these playlists in the background and may not always skip songs they don't care for. Think of someone who's studying or doing housework with music they aren't fully paying attention to running in the background. Even songs they dislike will play and count toward the charts for no other reason than no one bothered to skip them.

Songs that are popular on iTunes but weak on streaming may struggle to enter the top 20 on the Hot 100. And yet a lot more is required to buy rather than stream a song. In Digital Spy, Tom Eames complained about the effect streaming was having on the UK charts. He pointed out that Drake's One Dance was #1 on the overall chart for 15 weeks. But on a sales only chart, it was #1 for only four weeks. In one of those weeks where the song was #1 overall, it was only #14 on the sales only chart. Did One Dance hold onto the #1 for 11 extra weeks because people were deliberately streaming it (either searching for it or playing it from personal playlists) or because it was in so many popular curated playlists?

The daily Spotify chart often has little in common with the iTunes chart. iTunes, Google Play, and Amazon typically have a lot of overlap in their biggest selling songs. Thunder by Imagine Dragons is #1 on all three. But it's at #39 on the Spotify United States Top 50. The #1 on Spotify is Rockstar by Post Malone. It isn't in the Amazon top 10. It's #4 on iTunes and #2 on Google Play. That's fairly close. But the #2 on Spotify Bank Account by 21 Savage is way off compared to sales. It's #55 on Google Play and #57 on iTunes. It isn't in the Amazon top 20. Thanks to streaming, it's #14 on the Billboard Hot 100, just 6 places behind Thunder. This is the case for several songs in the Spotify Top 10. Many don't even rank in the top 25, and in some cases the top 50, on sales. To quote Tom Eames,

Many Spotify users simply play Spotify's curated playlists or the 'UK Top 50' playlist of the most popular songs around right now. Over and over. This means it takes yonks for songs to finally go away or to allow others to have a bloody turn. And it depends on Spotify's algorithm.

In other words, it's not listeners determining what's popular. It's the software they're listening on making that determination. The more types of popular curated playlists a song is placed in, the higher it will rank on the charts even if few people in the general public have ever heard it. These few songs that are played over and over make it harder for new songs to break through. The BBC's Mark Savage points out that this has consequences for musicians and music fans.

In the first six months of 2016, there were 86 new entries in the UK singles chart. Ten years ago, that figure was 230. And while artists like Beyonce, Rihanna and Calvin Harris continue to rack up the hits, new artists are being pushed out of the race. 

There are also concerns about chart manipulation with streaming. Republic Records was accused of manipulating the charts by putting a 3 minute 38 second loop of the chorus of Post Malone's Rockstar on their YouTube page. The loop had 42 million views, which all counted toward the charts, even though none of those listeners heard the full song.

Tom Eames recommends taking streaming out of overall charts to more accurately reflect what truly is popular. But this doesn't make sense because increasingly people are consuming music through streaming services. If you remove tens of millions of streamers from the charts in favor of a dwindling number of purchasers, the charts still won't be accurate.

So, what can be done to make the charts more accurate and give more songs and artists a chance to shine? 

• Exclude curated playlists. That way only songs people actually choose to listen to through searches or from their own playlists count. 
• Weight streamed songs in a way that aligns more with digital downloads. 
• Eliminate radio services like Pandora, which don't involve much in the way of choice. 
• Take thumbs up versus thumbs down and average watch time into account on YouTube to better reflect how much listeners like or dislike a song.

Billboard is frequently changing their formula. Right now, no one (other than the few artists who have long running number 1 songs) seems to be happy with the current formula. 

I helped contribute to the success of Taylor Swift's Look What You Made Me Do when I clicked on it to find how many thumbs up versus thumbs down it had.