SPOT - Spotify

[KJP]

 

Where might I find them? Thanks.

 

Here:  https://www.reddit.com/r/SecurityAnalysis/wiki/letters

 

[Broeb22]

https://www.dropbox.com/sh/imqj7tk724xpw7n/AAB1u8DA3Um08fL_jC2IDKuPa?dl=0&preview=Coho+Capital+2020+Q2+Letter.pdf

 

Long-form write-up on Spotify...Coho per usual has some very thorough long-term thoughts to provide.

 

I think this one of the most thoughtful pieces written on Spotify. Thanks for sharing. Would love to read other stuff from Coho. Where might I find them? Thanks.

 

Coho does great work. They are very much so in the growth world but their writing style reminds me a lot of Nicholas Sleep. At my prior job, we had a PDF version of Sleep's letters. I didn't know how rare they were but I have since been unable to find his letters either.

 

[thefatbaboon]

Thanks for the Coho writeup on Spotify it was an interesting read.

 

In the concluding prognosis of a future where Spotify doubles ARPU and has 600m paying subscribers ...he doesn't say what Apple, Amazon and Google will be doing. Does he expect them to be charging 50% of what Spotify charges?  If the tech giants who reach every internet connected human alive are happy taking tomorrow what they take today for unlimited music streaming and charging half what Spotify does how many of the projected 600m consumers are going to stick with Spotify?

 

I had  another question too but more from the label side of things...like if Spotify tries to change the labels' share from 80% to 50% or whatever we are imagining here...again what if the tech giants simply continue to pay the labels their share as they do currently?  Why would the labels not withhold their content from Spotify and trust that all of Spotify users would be able to find one of: Apple, Google or Amazon?

 

Also we never get any discussion of CRB and how the regulations and compulsory licenses work in the USA.

[griezeman23]

Does he expect them to be charging 50% of what Spotify charges?  If the tech giants who reach every internet connected human alive are happy taking tomorrow what they take today for unlimited music streaming and charging half what Spotify does how many of the projected 600m consumers are going to stick with Spotify?

 

I am guessing the labels are going to pressure all the distributors to raise prices once they have reached a certain level of penetration. Likely the days of $3 / month for 3 months or free for 6 months days will end.

 

But the more important sticking point for Spotify (and other distributors) to the labels since the start is that the labels were overcharging customers in the pre-Internet days. They were only attracting the high-end customers. While it was more profitable per unit, the total $ amount of profits captured was far lower as it excluded a whole class of customers who chose to listen to music for free (via ads). So, while assuming ARPU is doubled and subscribers go to 600mn may seem unreasonable, the bigger thing that Spotify is able to do is aggregate audio listeners... Meaning that the big money to be made (on monetization) is not on the paid subscriber side but rather the ad-supported side.

 

This is the whole reason for the push into podcasts, because there is a huge chance to aggregate the podcast space akin to the early internet days (h/t Stratechery). Plus, with more data, the better targeting gets, prices rise and the flywheel begins to churn. After this last call, I am becoming increasingly confident that Spotify is going to look to somehow enter the audiobook space as well.

 

if Spotify tries to change the labels' share from 80% to 50% or whatever we are imagining here...again what if the tech giants simply continue to pay the labels their share as they do currently?  Why would the labels not withhold their content from Spotify and trust that all of Spotify users would be able to find one of: Apple, Google or Amazon?

 

Well, Spotify is roughly 20% of Warner music's revenues (30% b/w Apple and Spotify per their K, and Spotify's subscriber base is 2x that of Apple Music) and roughly 30% of UMG's revenues (based on Dec 2019 investor presentation and 50% share). So, I'd think for financial reasons it would be hard to argue that they pull their music plus the most active music listeners (users spend 2x as much time on average listening to Spotify than they do at Apple Music) are on Spotify so there would likely be HUGE backlash against the labels if this happened.

 

As Spotify scales and their market share continues to grow (I am guessing Apple Music has gone nowhere fast since they disclosed 60mn subscribers in 2Q19 and anecdotally, all my friends are switching to Spotify because of the social integration), then they'll be able to extract a better seat at the negotiating table. Personally, I don't bake in any upside to GMs for label negotiations but I'd say it is more likely that Spotify continues to gain power over the labels rather than the other way around.

 

While Spotify has doubled, the stock still trades 2x lower than Netflix does and the pathway to GMs in the 30s seems clearer today than a year ago.

[thefatbaboon]

I really want to understand the bullish spotify argument better. Please bear with me a bit...

 

I asked what the giant DMPs would do in a world of doubling ARPU for Spotify and your answer is that the Labels push through increases for all DMPs.  OK so in this environment where everyone's ARPU doubles isn't Spot still kicking 70% revenues to the labels?  I mean if they aren't it kind of begs my original question, ie what do the other DMPs do if the labels accede margin to them. The issue I have is can Spotify's costs for music ever be meaningfully less than Apple, Amazon and Google and if these DMPs want to can't they always pass on music to their customers at cost? 

 

I also don't see the comparison to Netflix. 

 

Distribution: Netflix had and continues to have 100s of relatively small local distributors:  all the worlds cable, copper and cellular companies.  On the other hand every Spotify relationship is mediated by Apple and Google, who have competing music products that they can toggle and bundle more or less aggressively. 

Content: music is regulated, and truly exclusive music is nearly impossible in most major markets. So if a service wants to provide Beatles they can so long as they pay.

 

Spotify just seems like an app joint venture between the labels and the phone duopoly.  They get enough of a margin "to taste" but nothing more, step out of line and they'll just get squeezed back into line.

[griezeman23]

So, on your inability to see the bull case for Spotify: I think it stems from two facts: 1. It is a commodity competing against big tech; 2. It is only about music and the labels will always control content. Please correct me if I am wrong.

 

1) My reasoning on why Spotify is not a commodity music player:

 

Today, music listening is still relatively akin to broadcast TV. We know the largest artists make a disproportionate amount of all streams / money in the industry. I think one of the key things Coho picked up on was that last year, the top 90% of all streams were from 16k artists. Now, 32k artists account for the top 90% of streams. I imagine that music will become more and more niche, as discovery becomes an even bigger part of folks lives. This is not to say that there won't be big winners in music, there always will, but I imagine that their importance or share of the music industry revenues and mind share will decline. Who's driving this? Spotify. Look at mature streaming markets: https://twitter.com/compound100x/status/1294530406297948160/photo/1. https://www.digitalmusicnews.com/2017/08/14/spotify-youtube-sweden-norway-streaming/ (last two charts).

 

Now, if these markets were widely distributed in terms of DMPs, then I'd think this would point to your argument that Spotify is a commodity. But that is not the case. I think this is because Spotify has more data on listening than anyone. On average, Spotify listeners listen to 2x the music that Apple Music listeners do, and Spotify is more than twice as large (per last Apple Music update). More data means a better recommendation algorithm. And that is what is driving users to listen to more (and new) music. I think the algorithm then means that Spotify is not a commodity.

 

But the key thing is to continue making music and listening social in order to create network effects. Which they are doing with the ability to make "live" playlists with friends, as well as follow each other's playlists. They also have hugely influential playlists like Rap Caviar that dictates what's popular in today's music. https://www.vulture.com/2017/09/spotify-rapcaviar-most-influential-playlist-in-music.html. This is showing up in continually improving engagement (https://www.edisonresearch.com/the-infinite-dial-2020/). Amazon Music is seemingly the only other DMP making meaningful gains outside of Spotify.

 

So, I don't think Spotify is a commodity. Is the overwhelming in my favor? No. This is why this a heavily debated stock and on an EV / EBITDA per user basis, trades nowhere near NFLX.

 

2) My reasoning on why Spotify is not just beholden to the label:

 

So, if Spotify is driving what people want to listen to and knows what type of music each person wants to listen to, I think this erodes the power of the labels. They'll be able to help these labels figure out how popular an artist may be. I think this is a key part of the two-sided marketplace.

 

Compared to other DMPs (I am assuming this means Apple Music, etc.), right now, it seems unlikely that Spotify should have a better take rate than the others. But, if they own say, 70%+ of listeners (at about 50% and anecdotally more folks I know are switching to Spotify away from other DMPs), when the business is mature and they are "dictating" what folks are listening to relative to peers because of better (and more) data, then I do think it is possible they improve the take rate. To do this, they need to have network effects and take a larger share of listeners (and this is where podcasts help).

 

Content: music is regulated, and truly exclusive music is nearly impossible in most major markets. So if a service wants to provide Beatles they can so long as they pay.

 

On content: agree on music. This is why podcasts are an important piece of the pie. I don't think the opportunity is as large as some may think but I do think it is a meaningful and important business for Spotify. I think the importance of podcasts are both the ability to grow the business but hold power over the labels.

 

I am not sure how the contracts with the labels are written, but I would be suprised if the labels can take 70% of the subscription fee, when say, a 50% of a user's time is spent listening to podcasts. Even if this is the case, I think this may have been how Spotify signed up UMG for the two-sided marketplace: i.e. ignore podcasts from listening and keep the pay rate the same. I believe this is the power of diversifying away from music. I believe this is the point of the two-sided marketplace is that it is a hedge against an inability to improve their take rate.

 

So, the story is no longer just about music. Music is the most important piece, but I believe that podcasts and other forms of audio (I imagine they'll get to audio books at some point) will become a large portion of the business (30% of the value). 

[NewbieD]

They have a job posting for Head of Audiobooks up 8)

[griezeman23]

They have a job posting for Head of Audiobooks up 8)

 

LOL had no idea! Just posted last week... https://www.linkedin.com/jobs/search/?currentJobId=1980737916&keywords=spotify%20audiobooks

[Castanza]

https://www.businessinsider.com/spotify-report-joe-rogan-transphobia-fight-employees2020-9

[Liberty]

Interview with Daniel Ek:

 

https://www.theobservereffect.org/daniel.html

[chesko182]

Matthew Ball's long-form essay a history and the future of Audio, worth a read. Some very interesting comparisons to other mediums like TV, video games and even texting.

 

https://www.matthewball.vc/all/audiotech

[Xaston]

I love his writing, excited to read this

[aglittell]

It's that time of the year. Spotify released their "Wrapped Up" review today - showing users all sorts of statistics on their listening activity for the year. This has become a big social event, at least on my Instagram feed for the last two years (my feed may be biased as it is filled with college kids). People will post their listening stats to share with the rest of their followers. As a shareholder, I love it. It's free marketing and it instills a sense of social FOMO in the non-Spotify subs. The most common stat shared on my feed has been listening activity. I just scrolled past a classmate of mine who had spent over 123k minutes listening on the app. To put that into perspective - that is 2050 hours, 85 days, or 23% of his year. He is an outlier in terms of listening activity but even if the average user is someone like myself who had 46k minutes streamed, the engagement is staggering. My personal favorite stat was that I listened to 793 new artists this year.

 

What is my classmate's maximum willingness to pay for a service that he spent 23% of his year on? Is it more than the $5/month college discount he is on? Now you can make the argument that people have always listened to music and that music is a commodity that could be listened to through alternative mediums, but I think that argument becomes less compelling as Spotify continues to innovate with features like Wrapped Up.

[Arski]

It's that time of the year. Spotify released their "Wrapped Up" review today - showing users all sorts of statistics on their listening activity for the year. This has become a big social event, at least on my Instagram feed for the last two years (my feed may be biased as it is filled with college kids). People will post their listening stats to share with the rest of their followers. As a shareholder, I love it. It's free marketing and it instills a sense of social FOMO in the non-Spotify subs. The most common stat shared on my feed has been listening activity. I just scrolled past a classmate of mine who had spent over 123k minutes listening on the app. To put that into perspective - that is 2050 hours, 85 days, or 23% of his year. He is an outlier in terms of listening activity but even if the average user is someone like myself who had 46k minutes streamed, the engagement is staggering. My personal favorite stat was that I listened to 793 new artists this year.

 

What is my classmate's maximum willingness to pay for a service that he spent 23% of his year on? Is it more than the $5/month college discount he is on? Now you can make the argument that people have always listened to music and that music is a commodity that could be listened to through alternative mediums, but I think that argument becomes less compelling as Spotify continues to innovate with features like Wrapped Up.

 

I'm also subscribed to Sportify and it is indeed great. I also love wrapped up, that you see what you listened in the past year.

 

The average listening minutes of me and my friends are 50.000 minutes (18,19 & 20 year old college boys) and to be honest I am willing to spend more than the $45 per year I'm paying now for the service/platform. So I think they are looking quite monopolistic and are able to raise prices.

[Arski]

By the way, I'm no shareholder and are also not interested in holding it for the moment, because of the immense high price.

[Jurgis]

Doesn't any price raise automatically go 70% to music cos?

[Xaston]

Doesn't any price raise automatically go 70% to music cos?

 

No.  Money from increased user #s goes largely to the music cos, but not price increases.  The music companies are getting paid anytime a song they own the rights to gets played.  But if spotify can charge more for the monthly privilege of customers being able to stream those songs on their platform, without an increase in the amount of streaming the user does, all that price increase goes straight to spotify.

 

I think.

[Arski]

Doesn't any price raise automatically go 70% to music cos?

 

No.  Money from increased user #s goes largely to the music cos, but not price increases.  The music companies are getting paid anytime a song they own the rights to gets played.  But if spotify can charge more for the monthly privilege of customers being able to stream those songs on their platform, without an increase in the amount of streaming the user does, all that price increase goes straight to spotify.

 

I think.

 

Nope, you have to see it as all the money that gets collected in a country and that will be divided by the streamingsshare the artists have. For example Drake has a lot of streams and will of course have a bigger streamingsshare than most of the artists in the US. All that money is than divided between the artists, record companies etc., this will also differ per artist/record company because some might have different contracts than others, and spotify will keep a portion of it.

 

https://artists.spotify.com/video/how-spotify-pays-you

[D33pV4lue]

It's that time of the year. Spotify released their "Wrapped Up" review today - showing users all sorts of statistics on their listening activity for the year. This has become a big social event, at least on my Instagram feed for the last two years (my feed may be biased as it is filled with college kids). People will post their listening stats to share with the rest of their followers. As a shareholder, I love it. It's free marketing and it instills a sense of social FOMO in the non-Spotify subs. The most common stat shared on my feed has been listening activity. I just scrolled past a classmate of mine who had spent over 123k minutes listening on the app. To put that into perspective - that is 2050 hours, 85 days, or 23% of his year. He is an outlier in terms of listening activity but even if the average user is someone like myself who had 46k minutes streamed, the engagement is staggering. My personal favorite stat was that I listened to 793 new artists this year.

 

What is my classmate's maximum willingness to pay for a service that he spent 23% of his year on? Is it more than the $5/month college discount he is on? Now you can make the argument that people have always listened to music and that music is a commodity that could be listened to through alternative mediums, but I think that argument becomes less compelling as Spotify continues to innovate with features like Wrapped Up.

 

impressive. I listened to 197 new artists and 863 artists in whole. My time listened actually decreased by 33% from 2019. Love the wrapped-up segment and of course the company that makes it all possible.... Google...

[Liberty]

I enjoyed this interview with Daniel Ek:

 

https://tim.blog/2020/12/06/daniel-ek-transcript/

[aglittell]

It's that time of the year. Spotify released their "Wrapped Up" review today - showing users all sorts of statistics on their listening activity for the year. This has become a big social event, at least on my Instagram feed for the last two years (my feed may be biased as it is filled with college kids). People will post their listening stats to share with the rest of their followers. As a shareholder, I love it. It's free marketing and it instills a sense of social FOMO in the non-Spotify subs. The most common stat shared on my feed has been listening activity. I just scrolled past a classmate of mine who had spent over 123k minutes listening on the app. To put that into perspective - that is 2050 hours, 85 days, or 23% of his year. He is an outlier in terms of listening activity but even if the average user is someone like myself who had 46k minutes streamed, the engagement is staggering. My personal favorite stat was that I listened to 793 new artists this year.

 

What is my classmate's maximum willingness to pay for a service that he spent 23% of his year on? Is it more than the $5/month college discount he is on? Now you can make the argument that people have always listened to music and that music is a commodity that could be listened to through alternative mediums, but I think that argument becomes less compelling as Spotify continues to innovate with features like Wrapped Up.

 

impressive. I listened to 197 new artists and 863 artists in whole. My time listened actually decreased by 33% from 2019. Love the wrapped-up segment and of course the company that makes it all possible.... Google...

 

Can you elaborate on Google's role?

[chesko182]

pretty sure he's referring to Spotify's use of Google cloud services.

 

It's that time of the year. Spotify released their "Wrapped Up" review today - showing users all sorts of statistics on their listening activity for the year. This has become a big social event, at least on my Instagram feed for the last two years (my feed may be biased as it is filled with college kids). People will post their listening stats to share with the rest of their followers. As a shareholder, I love it. It's free marketing and it instills a sense of social FOMO in the non-Spotify subs. The most common stat shared on my feed has been listening activity. I just scrolled past a classmate of mine who had spent over 123k minutes listening on the app. To put that into perspective - that is 2050 hours, 85 days, or 23% of his year. He is an outlier in terms of listening activity but even if the average user is someone like myself who had 46k minutes streamed, the engagement is staggering. My personal favorite stat was that I listened to 793 new artists this year.

 

What is my classmate's maximum willingness to pay for a service that he spent 23% of his year on? Is it more than the $5/month college discount he is on? Now you can make the argument that people have always listened to music and that music is a commodity that could be listened to through alternative mediums, but I think that argument becomes less compelling as Spotify continues to innovate with features like Wrapped Up.

 

impressive. I listened to 197 new artists and 863 artists in whole. My time listened actually decreased by 33% from 2019. Love the wrapped-up segment and of course the company that makes it all possible.... Google...

 

Can you elaborate on Google's role?

[D33pV4lue]

It's that time of the year. Spotify released their "Wrapped Up" review today - showing users all sorts of statistics on their listening activity for the year. This has become a big social event, at least on my Instagram feed for the last two years (my feed may be biased as it is filled with college kids). People will post their listening stats to share with the rest of their followers. As a shareholder, I love it. It's free marketing and it instills a sense of social FOMO in the non-Spotify subs. The most common stat shared on my feed has been listening activity. I just scrolled past a classmate of mine who had spent over 123k minutes listening on the app. To put that into perspective - that is 2050 hours, 85 days, or 23% of his year. He is an outlier in terms of listening activity but even if the average user is someone like myself who had 46k minutes streamed, the engagement is staggering. My personal favorite stat was that I listened to 793 new artists this year.

 

What is my classmate's maximum willingness to pay for a service that he spent 23% of his year on? Is it more than the $5/month college discount he is on? Now you can make the argument that people have always listened to music and that music is a commodity that could be listened to through alternative mediums, but I think that argument becomes less compelling as Spotify continues to innovate with features like Wrapped Up.

 

impressive. I listened to 197 new artists and 863 artists in whole. My time listened actually decreased by 33% from 2019. Love the wrapped-up segment and of course the company that makes it all possible.... Google...

 

Can you elaborate on Google's role?

 

Spotify uses GCP... this article came out after last years wrapped-up.

 

https://techcrunch.com/2020/02/18/how-spotify-ran-the-largest-google-dataflow-job-ever-for-wrapped-2019/

[valueventures]

Surprised the market hasn't reacted positively to this: https://www.cnn.com/2020/12/15/media/meghan-harry-podcast-spotify/index.html

 

I'm sure people will mock this deal, but obviously the royals have a massive following. Should only help accelerate the flywheel.

 

 

[Casey]

Anecdata: Joe Rogan fans are disappointed that there isn't a comments section on Spotify like there is on youtube, as this hurts a community aspect of the podcast/show that was part of the fun before.