Blog Archives

Industry Groove – Week 23

The two dominant topics this year – the development of a new streaming model and the influence of AI on the music industry – may seem like completely different areas of focus at first glance. However, upon closer examination, one quickly realizes that there are many more connections between these two topics than one would assume. The search for a new streaming model has been triggered by the incredible number of releases flooding the DSPs (and the associated declining market share of the majors). Since the production of new music has become even easier with AI tools, it is not a prophecy to predict that the number of new songs will only increase even more drastically. That’s why Universal CEO Lucian Grainge also demands an artist-centric model, where “artist” naturally refers to human creatives, explicitly excluding machines.

The latest announcements from Deezer demonstrate how closely the two topics are related. However, they also reveal various issues and make it clear that there are no simple solutions for now, even if some may proclaim otherwise.

Deezer to identify and tag AI-generated music on its service

  • This week, French streaming service Deezer announced that it is working on tools to detect AI-generated content and is also developing a system to label these songs. Users should know from the start what is “real” music and what is AI-generated.
  • Initially, the focus will be primarily on identifying songs that use the voices of well-known acts, such as the recent track “heart on my sleeve.”
  • It certainly makes sense to recognize and label (or potentially block, depending on legal considerations) such songs. However, it can certainly be debated whether such labeling is really necessary for other tracks that are entirely or partially created with the help of AI tools. For some, it may be interesting to find out whether the house track or the lo-fi tune was created with or without AI assistance. But do consumers really care?
  • Deezer is certainly trying to please the majors with this move. The company is already working closely with Universal Music, just like Tidal, in the search for a new streaming model.
  • Deezer’s long-term goal is to create a compensation model that distinguishes between different types of music creation.
  • This raises many questions as well. Does it only apply to songs that are entirely created with AI? And if so, can the system recognize this? How much human involvement will be necessary to receive the highest possible payout, and how will this be accurately measured? Should an artist really be paid less just because they used AI to program the drums, for example? Undoubtedly, there are many discussions ahead!

TIDAL launches Artist Home

  • Having a dashboard for musicians where they can learn more about the performance of their releases, understand their audience, and manage their profile has become a standard feature offered by platforms like Spotify, Apple Music, Amazon Music, Deezer, and SoundCloud. Now, TIDAL has followed suit with Artist Home.
  • Artist Home provides artists with the usual functions: they can customize their profile picture, add a biography and social media links, invite other team members, or report incorrect releases on their profile. Additional features such as fan insights are set to follow.
  • In an interview with MBW, Jesse Dorogusker, the CEO of TIDAL, explains that his company aims to position itself within the artist ecosystem, with streaming being just one part of it: “We will be an artist platform with music streaming assets.”
  • He describes music as undervalued and underpriced, which may come as a surprise coming from the CEO of a streaming service. However, it is a logical move, considering that TIDAL is actively searching for a new streaming model in collaboration with Universal Music.

Bonus Reads

  • An article that makes you think: Will AI erase human creativity? The current presence of deepfakes already demonstrates that AI threatens not only “bullshit jobs.” The author raises the question of what happens if the audience suddenly likes Fake Drake over the real one. Would the label even need the Real Drake anymore? The article argues that the prospect of being replaced by machines could impact motivation and, consequently, creative output.
  • This article argues in a completely different direction. Firstly, it reminds artists that they should have more confidence in their abilities, which sets them apart from AI-generated music. In addition, it warns us not to underestimate consumers (something the music industry is often guilty of). Consumers can indeed differentiate between a human-made song full of emotions and soulless AI-generated music; they don’t simply consume whatever is presented to them. However, the author also points out two false assumptions. First, that all AI-generated music is inherently bad and wrong, and second, that all human-made music is good and valuable. Let’s be honest, the majority of music created by humans is, at best, mediocre and serves only to make the truly exceptional songs sound even better.
  • Spotify is cutting 200 more positions in the department that has grown the most recently: podcasts. The podcast department is being restructured, including the consolidation of two podcast studios. Earlier this year, Spotify had already laid off around 500 employees. At the same time, Spotify announced that they now have over 100 million podcast listeners.
  • MBW had the opportunity to ask some critical questions to Ole Obermann, the Global Head of Music Business Development at TikTok. Among other things, the discussion revolved around how TikTok plans to generate a significant amount of money for the music industry through sync fees, Obermann’s stance on payments to labels, and his perspective on transitioning from fixed amounts to revenue-sharing. The interview also covers the plans for Resso (or TikTok Music) and SoundOn. For the labels that TikTok is negotiating with, there was also a very clear message: “If we were forced to take down a catalog or part of a catalog, we are pretty confident that we could still remain a compelling service for our users.”

Industry Groove – Week 22

The vacation is over, the inbox is somewhat organized again – it’s time for the latest Industry Groove newsletter and a look back at the second half of May. Fortunately for me (not too much work) and for you (not too much to read) it wasn’t excessively eventful.

The topic of artificial intelligence continues to generate so many headlines that it could easily fill its own newsletter. Many are still completely overwhelmed, overburdened, or thoroughly excited. Meanwhile, Timbaland believes he has found the solution to all the open questions regarding AI-generated music, while notable figures from the AI industry warn about the potential dangers of artificial intelligence, mentioning it in the same breath as pandemics and nuclear wars.

As usual, MIDiA offers a reflective perspective on things. In this article, they show how music is evolving from a static and passive experience into an active and dynamic form of consumption. The times when musicians would record and release their songs for simple listening are gone. Today’s generation wants to be actively involved and creative themselves. Artificial intelligence was the missing puzzle piece that now allows practically anyone to create music. For me, this is the must-read of the week. And now, let’s move on to the other topics that have been important lately.

US state of Montana bans TikTok

  • In Week 16’s newsletter, I reported that the Montana State House had voted in favor of a bill that would ban TikTok in the US state. The only thing missing was the signature of the Republican governor, which, as expected, was a mere formality.
  • The law is set to take effect from January 2024. However, as I understand it, it will only affect those who want to download the app anew. Those who already have it on their devices can continue using it.
  • Whether the ban will actually come into effect remains unclear for several reasons. For one, it is not certain whether Apple and Google will actually comply and prohibit app downloads at the state level.
  • Furthermore, two lawsuits have already been filed: one by TikTok users, who feel their freedom of speech is being restricted, and a second by TikTok itself.
  • Montana may be a small state, but people will be closely watching what happens here. As MIDiA correctly points out, this is just the beginning of many regulations that will not only affect TikTok.

TikTok integrates Apple Music, tests chatbot, and launches Music Hub

  • Apart from the ban in Montana, there is much more to report on TikTok. For example, the integration of Apple Music on its platform. TikTok is currently testing an “Add Song” button on videos with music, which will take users directly to Apple Music. The setup allows for seamless integration of additional streaming services as well. This integration is a great advantage for musicians, as it enables users to find songs more quickly. This could also indicate that TikTok Music is not coming anytime soon.
  • Another ongoing test is called Tako, a chatbot. It is designed to help discover relevant videos and creators based on the videos they have already watched and the questions they ask Tako. Currently, this is a small-scale test with a few users in the Philippines, but if it is rolled out more widely, it could potentially revolutionize TikTok. While others are busy copying the For You feed, TikTok is already working on something new.
  • Finally, there is also a new music feature on TikTok. The #NewMusic Hub focuses, as the name suggests, on new releases.
  • Or as TikTok announces it: “The #NewMusic hub allows fans to discover the best new music, often before it’s heard elsewhere, while providing a platform for emerging and established artists to grow their fanbases and connect with their audience.”

Now it’s already 120,000 songs per day

  • I admit, I had doubts for a long time about the claim of 100,000 songs being released per day. I thought it was an exaggeration, but now I believe it could actually be true. But wait, that number is already outdated. According to Luminate, an average of 120,000 new songs per day were uploaded to DSPs in the first quarter of 2023.
  • In other words, 10.08 million songs were uploaded to DSPs in the first quarter of the year. If this number remains stable, there will be a whopping 43 million songs by the end of 2023.
  • That is significantly more than in 2022 (34.1 million) and almost triple the number of songs released through DSPs in 2018 (16.4 million).
  • As we know, it has never been easier to release music, and with the many new AI tools, it has also never been easier to create music. The number is likely to continue rising rapidly, possibly fueling the demand for a change in the streaming system.

Instagram to launch a Twitter clone this summer

  • Although Elon Musk hasn’t completely crashed Twitter yet, there are still good reasons to abandon the platform. As has now been revealed, Meta wants to provide the platform to give those who want to leave a new home.
  • The app will be predominantly text-based and offered as a standalone app, but will also be integrated in parts with Instagram.
  • Anyone already using Instagram will be able to keep their username on the yet-to-be-titled platform. Additionally, all of your Instagram followers will receive a notification to follow you on the new platform.
  • According to initial reports, the text length will be limited to 500 characters. Furthermore, users will also be able to post photos and videos up to 5 minutes in length.
  • While there are already some Twitter alternatives, none of them have the market power that Meta brings. It will be interesting to see if this move pays off.

Universal teams up with AI company Endel

  • Universal Music has recently vehemently opposed certain excesses of the new possibilities brought by AI. However, they have also made it clear that they are not fundamentally hostile to the technology. This is now evident through their collaboration with Endel.
  • Based in Berlin, Endel primarily operates in the field of so-called “functional music”, which means music for relaxation, sleep, and similar purposes. The CEO had previously highlighted the potential for collaboration with major labels, and Universal seems to share this vision.
  • The exact goal of the collaboration is not yet fully clear, but Universal probably wants to grab an even bigger share of the immensely popular mood playlists.
  • Whether there will be functional versions of albums by superstars like Taylor Swift, as suggested by the Endel CEO, remains to be seen. Endel envisions the possibility of offering a relaxing and sleep-oriented version as a bonus alongside the regular album.

Bonus Reads

  • Unlike Spotify, Apple Music is not very transparent when it comes to the streaming numbers of individual songs (or many other figures, such as the number of subscribers). However, it was revealed that so far, not a single song on Apple Music has been streamed more than a billion times. The song that comes closest is Ed Sheeran’s “Shape of You” with currently 930 million streams. On Spotify, there are currently 419 tracks in the “Billions Club.”
  • This Billboard article highlights that many labels have become more cautious and are no longer signing everything that goes viral. Too often, success has been limited to a single song.
  • What will likely be “signed” more frequently in the future are virtual artists. This MBW post provides a good overview of virtual acts releasing through major labels and other big players such as HYBE or Tencent. The article also explores the history of virtual artists and demonstrates that this is not such a new phenomenon.

Industry Groove – Week 19

For many, SoundCloud is still not much more than the platform where one can upload their music without a label or distribution. Of course, SoundCloud does not play in the same league as Spotify or Apple Music, and they also have a slightly different business model. However, dismissing the service as a niche product definitely falls short. This is not least because SoundCloud is not afraid to experiment and take steps that bring important insights for the music industry as a whole.

While many DSPs are stagnant or primarily focused on growth outside of music, SoundCloud continues to impress with innovative initiatives. Of course, their user-centric trial is noteworthy, as it can provide important data for or against the system. And now they have launched a tool that was actually long overdue, allowing artists to directly reach out to their biggest fans. Many artists will likely wonder why they don’t have this opportunity on other DSPs.

I hereby say goodbye for two weeks to a well-deserved vacation. The next Industry Groove newsletter will be released on June 1st and will bring you up to date on everything that happened in the second half of May.

SoundCloud allows you to contact your super fans

  • In the Week 13 newsletter, I mentioned that on Spotify for Artists, you can now view your Active Audience, which are your most active listeners. I commented that while this is nice, it would be even better if you could contact them directly. That option is now offered by SoundCloud.
  • The tool, called “Fans,” found on SoundCloud for Artists has already been used in beta by 10,000 artists. It provides artists with more data about their fans and allows them to send DMs to their superfans and attach (exclusive) tracks. This week, the option was expanded to an additional 50,000 Next Pro artists.
  • SoundCloud has data on super fans thanks to their user-centric experiment, which they call Fan Powered Royalties. They are now combining this data with interaction data and user reach.
  • According to SoundCloud, DMs can be used to promote new releases, exclusive songs, or live shows, ask fans for feedback, sell merch, or simply establish a more direct connection to the biggest fans. Of course, fans can choose not to receive such DMs if they wish.
  • In a blog post, Tracy Chan, Senior Vice President of SoundCloud, is quite clear about the competition: “Streaming services won’t tell you who your fans are. Instead, they run business models built on selling you access to your fans. And the streaming services aren’t alone – ticketing and merch platforms won’t tell you who your fans are, either. The dirty secret of the music industry is that these platforms expect you to give them content and sales to fuel their bottom lines, but they refuse to tell you who is listening to your songs (or buying your tickets and merch).”
  • After the initial Fan-Powered Royalties test was only available to artists who uploaded their music directly to SoundCloud, the test was expanded in a second step to include Warner Music artists. Now the labels and distributors associated with Merlin (including iGroove) are also joining in and participating in the user-centric trial.

Does the pro-rata model favor streaming fraud?

  • The answer to the question in the title is relatively simple: Yes. It is generally known that the distribution system currently used by virtually all DSPs is not suitable for curbing streaming fraud.
  • The topic resurfaced in recent days because it became known that Spotify has removed numerous songs created with the Boomy tool. Initially, it was mistakenly reported that the songs were removed because they were AI-generated music. However, the reason was simply streaming manipulation.
  • According to media reports, Spotify also deleted numerous other releases that were not related to Boomy. However, it has long been known that what Spotify and other DSPs detect is only the tip of the iceberg.
  • In an alarmist and perhaps somewhat hasty article on MBW, it is now demanded that the pro-rata system be buried immediately.
  • Of course, the article addresses important points. For example, that streaming manipulation could be abused for money laundering. The fact that criminals are part of the system is evidenced by the threats received by employees of the distributor Horus Music after they blocked uploads. It also highlights that honest artists are missing out on millions, if not billions. Pro-rata creates a (wrong) incentive to generate as many streams as possible rather than reaching as many listeners as possible.
  • The problem, however, is that there is currently no model that everyone can agree on. User centric has the most advanced tests, but the majors don’t seem to be warming up to it. Universal is proposing an artist-centric model, which is now being evaluated at Tidal and Deezer. It will take some time before a system change really happens, if all players really agree.
  • In line with this, a new study commissioned by Pro Musik on user-centric has been published. The study is based on figures from SoundCloud’s Fan Powered Royalty Program. It found that if User-Centric were used instead of Pro-Rata, 25.5% of generated revenues would be redistributed. Furthermore, 19% of the analyzed artists could at least double their earnings.
  • Discussions about a new streaming model will undoubtedly continue and I will keep you updated.

TikTok tests feature that connects musicians with influencers

  • TikTok is currently testing a feature called “Work with Artists” on a very small scale. It is described as follows: “Pick a paid music task you’re interested in; create a video with that song and post it; the more likes, shares, and views you get, the more you earn!”
  • This allows users to bypass agencies and directly connect with influencers who can then use the song in their videos.
  • In the example circulating on social media, the highest-paid influencer in the campaign received $120. Of course, such amounts are mainly aimed at micro-influencers, as larger influencers would not even turn on their cameras for such amounts.
  • As this is only a test that has not even been publicly announced by TikTok itself, it is completely uncertain whether this feature will ever be rolled out to all users. However, it could be interesting for musicians.

Who actually owns Spotify?

  • Of course, artists do not necessarily have to be interested in the exact ownership structure of Spotify. Nevertheless, it can be helpful to understand why certain decisions are made or why certain things are not implemented.
  • MBW has therefore compiled publicly available documents to find out who currently owns Spotify.
  • They have mainly identified two things: Firstly, the two co-founders still firmly hold the reins, and secondly, many of the major investors have remained the same despite the fluctuations that Spotify’s stock has undergone lately.
  • CEO Daniel Ek holds the largest share of stocks, namely 16.5%. The other co-founder, Martin Lorentzon, holds a share of 11.1%.
  • However, this only partially reflects the power structure at Spotify. When looking at the voting shares, the founder duo together holds 74.3%. Therefore, nothing happens at Spotify if Daniel and Martin don’t want it to. Consequently, they are also the ones who have to take responsibility for any mistakes.
  • After Daniel Ek, investment firm Baillie Grifford is the second largest shareholder with 14.5%. Baillie Grifford also owns 4.48% of Tencent Music Entertainment, which in turn holds 8.6% of Spotify shares.
  • Other notable shares are held by investment management firm Rowe Price & Associates with 5% and major bank Morgan Stanley with 4.4%. Additionally, 3.3% of Spotify belongs to Universal Music Group, a deal that has already received substantial criticism.

Bonus Reads

  • Spotify has announced that as of July 1st, all data older than June 2020 will be removed from Spotify for Artists. However, this does not affect the all-time stream count of your songs. The data can still be downloaded until June 30th. In exchange, there will be some new features on S4A.
  • For those who believe that Record Store Day is only an event for some vinyl nerds, they should take a look at the impressive numbers that Luminate has compiled. In the week of RSD alone, 1.809 million vinyl albums were sold in the US. This is the fourth strongest week ever since 1991, the year Luminate began collecting this data. Fortunately, 1.426 million of these albums were sold through independent record stores. Nice!
  • So far, the discussion about a new streaming model is mainly associated with Universal Music, whose CEO has additionally triggered the debate. However, Warner Music CEO Robert Kyncl also made it clear once again that he advocates for a change in the system. “It can’t be that an Ed Sheeran stream is worth exactly the same as a stream of rain falling on the roof,” and further punchy statements as well as his position on AI music can be found here. Perhaps he simply wants to distract a little from the mediocre quarterly results.
  • While Western DSPs may still have some reservations about AI, they are definitely not present in other parts of the world. Anghami, the largest streaming service in the Middle East and North Africa, has already used AI to generate music for its service. Now they are coming up with a world premiere: AI-powered podcasts, tailored to the user’s interests.
  • Three years ago, so-called fan packs, the combination of albums and merchandise, were excluded from the Billboard charts. Now they are making a comeback, but some rules are in place to prevent distortions. Why they will be reintroduced in July and what the guidelines are is shown in this article.

Deepfakes of the week

  • Many are still amazed by the deepfakes that are currently making the rounds. Well, we shouldn’t really be surprised. This video is already a year old and actually anticipates everything we are experiencing right now.
  • Snoop Dogg has Michael Jackson rap C-Murder’s “Down for My N’s” and comments on it with “This Shit is outta hand!“.
  • Timbaland never had the opportunity to work with The Notorious B.I.G. Now he has shared a track with AI-Biggie on Twitter and received not only love in return.

Industry Groove – Week 18

Let’s not fool ourselves: Artificial intelligence (AI) has been present in our lives for some time now. It suggests music to us, determines what appears in our feeds, may have decided whether we can take out insurance or get a job. If one wanted to, it could be easily ignored. Most likely, it was the same for most people as it was for me: until the end of last year, AI was present but only in the background.

Now it’s already normal for me to have ChatGPT or Notion AI optimize a text (although I still prefer to write them myself), use AI tools for translations, or create playlist covers with different tools for fun. I would also like to take the time to dive into the field of AI music production. I have already tried, which allows you to easily create song stems. One or two mouse clicks and I already have the a cappella that I was previously searching for in vain.

I have just scratched the surface and I’m sure that there are countless tools out there that will make my life easier, of which I don’t know yet. At the same time, I am convinced that there are also numerous offers that I consider to be completely unnecessary or even cause me concern. Because developments are progressing so rapidly that not only I, but also the music industry, governments and legislators need to sort things out, analyze the situation and carefully consider where the journey should go. However, we will hardly be granted this time, so it will come down to reacting rather than acting. The music industry already knows from bitter experience that this is unlikely to end well.

AI music: The legal situation is not as clear as claimed

  • Last week, I wrote that there is a certain degree of uncertainty in the music industry regarding AI music and deepfakes, as many legal aspects are still unclear. Interestingly, the industry wants to convince us that they already have more or less everything under control.
  • Universal Music, for example, firmly believes that their and their artists’ rights are being violated and that the laws already reflect this. According to Universal, this applies to both training AI with copyrighted music and to deepfakes like Fake Drake.
  • Meanwhile, the distribution company Believe, which includes TuneCore, is convinced that they can already detect and block music that was produced 100% by AI. However, some questions arise: Firstly, should all music that is completely produced by AI really be kept off DSPs, even if it does not violate any copyright? Secondly, how much music is already generated 100% by AI today? “Heart on my Sleeve” and many other currently most popular deepfakes are not. Thirdly, it is, of course, uncertain whether the tool really works smoothly and, if so, how long it is still possible to clearly detect AI music.
  • This informative article on The Verge demonstrates that the legal situation is not as clear as portrayed by Universal Music. “There’s no precedent for whether real Drake can stop robot Drake on the basis of copyright”.
  • According to the article, for example, an artist’s voice or flow is not protected by copyright. If at all, it is more likely to be a violation of trademark law. Depending on the case, personality rights may also be affected, especially if a Nazi text is put in the mouth of an artist.
  • I am not a lawyer and don’t want to pretend that I understand everything. What is clear, however, is that the situation is not as clear-cut as Universal presents it.

Resso ends free tier – will others follow?

  • Resso, the streaming service of TikTok’s parent company ByteDance, has informed its free-tier users that it will be discontinued starting May 11. Resso, which has been available only in India, Brazil, and Indonesia so far, will henceforth offer only a premium subscription.
  • Some commentators see this as a concession to the major labels, with whom ByteDance is negotiating intensively. It is an open secret that Resso (possibly also under the name TikTok Music) wants to expand into further markets. However, the rights holders, i.e. the labels, seem to be more inclined to apply the brakes, and one of the reasons for this is the poor conversion rate from free to premium.
  • This is not the only point of dispute between ByteDance and the music industry, but at least it is one less stumbling block that could stand in the way of expansion.
  • In an opinion piece, David Turner argues that ad-supported music streaming is generally broken.
  • Even before Resso, the Indian streaming giant Gaana had discontinued its free offering. Deezer has also ended its free tier in some markets, leading to an increase in revenue per user.
  • Turner analyzes the figures of Spotify, Pandora, and Tencent Music and shows why the current model should be reconsidered.

Social media platforms gain users but lose advertising revenue

  • Various social media platforms have released their first-quarter numbers and they all show a similar trend: while user numbers are rising, advertising revenues are stagnant or declining.
  • YouTube has already seen a decline in advertising revenue for the third consecutive quarter. It fell by 2.5% to $6.69 billion. However, revenue in the “Google Other” category, which includes YouTube TV, YouTube Music, and YouTube Premium, increased by a whopping 9% year-on-year, bringing in $7.41 billion.
  • Meanwhile, Meta has surpassed the mark of 3 billion users across all its services – Facebook, Instagram, WhatsApp, and Messenger – representing a 5% increase compared to the previous year. Revenue also grew, but at a much slower pace than before, namely by 3% to $28.65 billion in the first quarter of 2023.
  • Snapchat was able to increase the number of daily active users by 15% year-on-year, to 383 million. Not quite one percent of them, slightly more than 3 million, pay for Snapchat+. Revenue fell by 7% year-on-year.

Bonus Reads

  • MBW is closely monitoring the job postings of TikTok’s owner, ByteDance, and has discovered that they are apparently planning an app for AI music creation. The job posting refers to “an AI-powered tool that provides intelligent music creation and audio editing capabilities”. It further states that the vision is to “significantly lower the music creation barrier and inspire musical creativity and expression, further enriching the music content”. I will keep an eye on this topic.
  • Last week, I reported that Grimes is open to the use of her voice for deepfakes. This week, she has already launched the platform for this purpose in partnership with the development studio CreateSafe, inviting users to create AI-generated songs using her voice.
  • The decentralized streaming platform Audius is also open to new approaches. Artists using Audius can now decide for themselves whether releases that were trained with their music should appear on their profile. These releases then appear in a “Generated with AI” section on the profile. Audius is taking the first step here that other DSPs still have to take.
  • No matter how impressive and somewhat frighteningly good the current AI songs are, Bill Werde absolutely does not want to exaggerate this into a Napster moment. He sees more parallels with the Grey Album by the great Danger Mouse and the remix culture.
  • Despite talking big, Triller now has to dig deep into their pockets. The US company has to pay Sony Music $4.5 million for unpaid royalties. Triller generally has a rocky relationship with the music industry. They are also in court with Universal, and they removed the music delivered by Merlin (including that of iGroove artists) from their platform at the end of last year. Although Triller is trying to present itself as a patriotic alternative to TikTok, they are currently messing up with the music rights holders, who are extremely important for a short video platform.
  • I have written several times about how Spotify announced a HiFi offering over two years ago, which has not yet been launched. This article summarizes the whole debacle in detail and investigates who effectively killed Spotify HiFi in the end.
  • Chartmetric has analyzed in their usual data-based way what influence a performance at the Coachella festival has on streaming numbers and social media.

Deepfakes of the Week

  • Ghostwriter, whether it’s the same who created the Drake and The Weeknd song is controversial, brought together Bad Bunny and Rihanna for the song “Por Qué” and conquered TikTok by storm.
  • Deepfakes are not only suitable for creating songs by current stars but can also bring deceased artists back to life or make reunions a reality. For example, when Paul McCartney and John Lennon suddenly appeared on a song together.
  • Few current AI projects have received as much positive feedback as the AISIS album – originally an album by the unknown band Breezer, which they then uploaded to the internet with the AI vocals of Liam Gallagher. Oasis fans are celebrating it and even Liam seems to have no problem with it.

Industry Groove – Week 17

The strong growth of Spotify in the first quarter shows that the ongoing discussion about a new streaming model is solely an issue within the industry that does not affect consumers. This is an aspect that must not be ignored when searching for an optimized model. Because not everything that is good for the industry and musicians is automatically good for the fans.

AI and deepfakes continue to dominate most headlines. I have read countless articles and comments, and what one mainly feels is a strong sense of uncertainty, especially because much is still legally unclear and also because in many places it is difficult to estimate what AI can actually do and what it cannot (yet). This week, I would like to present some initial solutions and, above all, show that while the topic deserves attention, there is no reason to panic.

Should artists make their voices available for deepfakes?

  • Many in the music industry were not aware of how good deepfakes have become, so the Drake-The-Weeknd song may have been a real shock for many.
  • Now, after the initial shock has subsided, solutions are being sought. And some argue that we should not close ourselves off to the technology and should make our voices available.
  • Grimes is presenting herself as a pioneer, as she states that anyone can use her voice for deepfakes if the profits are split 50/50.
  • “Feel free to use my voice without penalty. I have no label and no legal bindings. I think it’s cool to be fused w a machine and I like the idea of open sourcing all art and killing copyright,” she writes on Twitter, forgetting that it is precisely copyright that ensures that profits are split 50/50 and that she can enforce this. It is also worth noting that Grimes, as a successful and financially secure artist, is operating from a very privileged position and can easily engage in such experiments.
  • However, even less secure artists should take the step, at least according to the Building Stans website. They argue that now is the ideal time for artists to consider launching their own APIs.
  • They argue that focusing solely on enforcing copyright is the wrong approach. Instead, new monetization opportunities should be explored and utilized, such as the proposed artist API, which should capitalize on their uniqueness.
  • Specifically, this could look like training an AI on one’s own music and voice. This high-quality and artist-approved AI would then be made available to fans and creators through an API access for a fee. They could use it to create songs and when these are released, the artist would receive their share.
  • For Building Stans, however, this is just one of numerous possible use cases for an API. They are also well aware that access to the API would need to be carefully regulated. It’s all still in the future, but at least worth considering.

AI music and deepfakes: No reason to panic

  • Now that it has become obvious what AI is already capable of and awareness is growing that the technology is rapidly improving, many in the music industry may have a queasy feeling. This is not entirely unfounded, but there is still no reason for panic. Therefore, I would like to give voice to the thoughtful voices and commentators.
  • For example, there is Stuart Dredge from Music.Ally, who points out that AI music can and will bring many advantages. We should not limit everything to the “dick moves” of certain people who want to clone the voices of artists without their permission and make money from it. He is right.
  • By the way, a certain Daniel Ek, perhaps known to some as CEO of Spotify, also shares this view. On the one hand, he is excited about the possibilities that AI offers for the music industry. At the same time, he is offering to collaborate with labels to prevent certain excesses. Allowing innovation while protecting creators will certainly not be an easy balancing act.
  • Furthermore, it must be noted that AI tools have made enormous progress and, even if they have not yet reached mass usability, they are likely on the cusp of doing so. This investigation by The Verge shows, however, that currently it is still rather cumbersome to produce a deepfake. As an example, another “Fake Drake” song called “Not a Game” is used, whose vocals are based on an a capella offered on the Looperman website. The voice of Drake was then overlaid on these vocals using AI technology. The song’s beat, meanwhile, comes from the BeatStars platform. Thus, artificial intelligence was only used to generate Drake’s voice here. Although all of this may not have been very complicated, it is far more than simply generating a song with a single prompt.
  • Jessica Powell of Audioshake doubts that “Heart on My Sleeve” is a purely AI-generated song, but rather that more manual work went into it as in the example above. She also raises the question of whether these deepfakes are simply comparable to remixes. This is not entirely wrong, especially when it comes to playful examples like Rihanna singing a Beyoncé song, which, apart from legal aspects, is harmless. The problem is that this technology can put any lyrics in an artist’s mouth. Powell, however, is convinced that ways will be found to use this technology and create new monetization opportunities.
  • Tatiana Cirisano believes she has identified an advantage of the circulating deepfakes. If an artist is unpredictable and their music is not interchangeable, it is much harder to produce a AI clone of them than, for example, with Drake, whose songs, let’s be honest, often sound very similar. Perhaps AI music can contribute to more individuality again.

Spotify grows stronger than expected and may increase prices

  • Spotify has released its Q1 2023 figures, which show greater growth than expected.
  • The number of premium subscribers increased to 210 million, which is 5 million more than in the previous quarter and 28 million or 15% more than in Q1 of the previous year. All regions have grown stronger than predicted, especially Europe and Latin America.
  • Overall, Spotify has 515 million monthly active users (+22% compared to Q1 2022).
  • However, the revenues did not quite meet expectations. Although they grew by 14% YoY to €3.04 billion, fluctuations in the advertising business prevented the company from reaching its goals. Advertising revenue amounted to €329 million, while subscriptions brought in €2.7 billion.
  • As almost always, Spotify still reports a loss. In this quarter, it amounts to 156 million euros.
  • Despite the impressive growth of premium subscribers, the ratio of premium to free users continuously decreases. In Q1 2019, 46% were paying customers, in Q1 2022 the number dropped to 43%, and now it reached a new low of 40%. This is relevant because significantly more money flows to the artists from the premium subscribers.
  • And then a small miracle actually happened: Daniel Ek announced that he wants to raise prices. However, there is also a “but”: this depends on negotiations with the major labels. At least Universal and Warner have already indicated pretty clearly that they consider the price increases overdue.

Indies secure 27% of the publishing market

  • According to new figures published by the Independent Music Publishers Forum (IMPF), the publishing companies affiliated with major labels hold a market share of 59.9%.
  • According to their calculations, the independent sector increased from €1.95 billion (2020) to €2.08 billion (2021), which corresponds to a market share of 27.1%.
  • And what about the other 13% you may ask? IMPF defines independent as having less than 5% market share. Therefore, Kobalt (6.8%) and BMG (6.2%) are not included. Now the math adds up.
  • However, even without these two, the independent sector is the strongest player, ahead of the largest major, Sony Music Publishing, which has a market share of 24.4%.
  • The entire publishing market was worth 7.68 billion euros in 2021.
  • The 2022 figures are not yet available – as most musicians know all too well, publishing always takes a bit longer.
  • Currently, around 80% is paid out for the master rights and 20% for the publishing rights. The publishing industry refers to this as a remnant of the CD era, when labels had high costs for the production and distribution of physical products. Now, in the streaming age, this distribution should be reconsidered.

Investments in Web3 massively decline

  • As Music.Ally correctly points out, many technologies have been declared the “Next Big Thing” in recent years. Examples include concert livestreaming, social audio, NFTs, the metaverse, Web3, and currently AI. Quite a few of these were soon dropped again.
  • Now, companies working in Web3 are also feeling the fickle interest of the public and investors. Venture capital funding plummeted by 82%. Specifically, it amounted to $9.1 billion in Q1 2022 and now, one year later, just $1.7 billion. The number of deals also fell from 770 to 333.
  • However, this may not necessarily mean much, as funding has declined in almost all sectors. In Web3, it has simply declined more strongly, partly fueled by various scandals and bankruptcies in the crypto sector.
  • Now that the hype is elsewhere, we can first weed out those who just jumped on the hype train, and secondly refocus on how the technology can actually bring value.

Bonus Reads

  • “How Did Hip-Hop Media Get So Bad?” – the title essentially says it all, yet it doesn’t do justice to the very reflective article. Bad journalism is one thing, but the proximity to very conservative circles and ideals in the best case, and far-right extremists in the worst case, is another.
  • Let’s stick with the hip-hop writers. Two weeks ago I linked to Complex’s list of the most powerful figures in the hip-hop media world. Now, Complex Mag has also dedicated an article to the pioneers of the hip-hop media landscape. Show some respect!
  • Frank Ocean’s performance at Coachella has sparked discussions and inspired Eleanor Halls to engage in some interesting thought experiments on the topic of “What do artists owe their fans?”
  • A study shows that compared to musical accompaniment, lead vocals have become quieter over the years. In addition, there are clear differences between different genres. I am always fascinated by the nerdy topics that some people choose for their studies. Awesome!

Industry Groove – Week 16

Assuming you haven’t spent the last few days under a rock on Mars, you’ve probably heard about a completely or mostly AI-generated song with the voices of Drake and The Weeknd that has been making waves. In addition, various “cover versions,” such as an AI-Drake performing “Munch” by Ice Spice, have been circulating. While the covers took place on social media, the Drake and The Weeknd collab uploaded by the mysterious ghostwriter made its way onto Spotify, Apple Music, YouTube Music, and the like, although not for long.

“im just getting started“, ghostwriter declared on TikTok, and of course, this is just the beginning. The music industry is faced with the big question of how to deal with it. Because it is clear: the technology is there, it will not disappear, and above all, it will only get better. The Pandora’s box is open and now a pragmatic solution is needed, as soon as possible. The first instinct, of course, is to put a stop to it all and ban everything. But wouldn’t that just be repeating mistakes from the past, when the music industry thought it could stop changes and then slid into a deep crisis? However, simply watching from the sidelines is also not an option, as thousands of songs like “heart on my sleeve” will flood the DSPs. There is certainly no easy solution, and the Industry Groove Newsletter will closely monitor any possible outcomes.

What’s next for AI Music and deepfakes?

  • AI has the potential to support many artists in their creative process. However, in the last few days, it has become apparent that the capabilities of AI go far beyond that and that bad actors have new tools for their fraudulent schemes.
  • It becomes undoubtedly dangerous when an artist loses control over their own voice. When you can imitate the voice of well-known artists with easy-to-use tools, we are moving on thin ice in many ways.
  • The current cover versions can of course be seen as harmless games, primarily problematic in terms of copyright. However, it becomes really dangerous when artists are given political messages to mouth that they would never express. It is already easy to make Drake rap a racist text, and as a consumer, one will soon no longer know what is real and what is not.
  • In the specific example of the Drake and The Weeknd track, it is not a cover but a new song. The entire copyright situation is therefore even more complicated, especially since ghostwriter, who released the song on the DSPs, did not mark Drake and The Weeknd as artists. Universal’s lawyers will probably have to work overtime.
  • Side note: ghostwriter seems to be on a revenge mission against the industry. On TikTok, he wrote: “i was a ghostwriter for years and got paid close to nothing just for major labels to profit. The future is here.” This can probably be understood as a threat.
  • However, the reaction from Universal also has a slightly threatening undertone: “The question arises as to which side of history all stakeholders in the music ecosystem want to be on: the side of artists, fans and human creative expression, or on the side of deep fakes, fraud and denying artists their due compensation.”
  • The music industry is now facing the difficult question of how to deal with this situation. The initial reaction, of course, is to demand a complete ban. It’s quite possible that Universal will now go after ghostwriters and also the distributor (it wasn’t us) who uploaded the song. However, history shows that such bans are neither useful nor fully enforceable. The mistakes that were made at the end of the CD era should not be repeated. But what is the alternative?
  • Some voices are calling for simply allowing it, but under the condition that it is clear that it is an AI-generated song. They advocate dealing with AI music like the Content ID on YouTube. Users can use popular songs for their videos, but the earnings go to the artists. Ironically, we would need AI to filter out the AI-generated songs.
  • However, there is a difference: with Content ID, the songs were created by the artist themselves, and they had creative control. It is very different with AI-generated music, where only the voice “belongs” to the artist, but they have no influence on the song. Would each individual song have to be manually checked? And if they are approved, will the artist be marked and will the AI song then appear on the official artist profile? If so, how can fans still distinguish what really comes from their favorite artists and what comes from an AI program? Yes, it is complicated.
  • It is also already suspected that the financially strong labels will soon be buying up AI companies en masse.
  • For those who want to do a deep dive into the topic of AI and deepfakes, I recommend this link.
  • A typically insightful analysis comes from MIDiA.

Which genres are dominated by major labels in playlists?

  • It is no secret that artists signed to major labels are overrepresented in Spotify playlists. I already showed this in this article.
  • Dmitry Pastukhov wanted to investigate this in more detail and therefore analyzed the 500 (!!) most popular genres and over 6,000 playlists on Spotify.
  • Among the most popular genres (pop, dance pop, hip-hop, rock, and EDM), the average major label share was 48%. The lowest proportion of major label songs in the top 10 genres is found in EDM, where only 27.5% of the songs were released by a major label. This also applies to other electronic genres in the top 100.
  • Looking at the top 100 genres, the major label share is still at 43%. Across all 500 genres examined, it drops to 27%. The less popular a genre is, the lower the proportion of major label songs on average.
  • Indies are particularly well represented in local repertoire, niche sounds, and new genres.
  • Of course, it is not necessarily surprising that majors are less present in niche genres, but the support with numbers is very informative.

The next step towards a TikTok ban?

  • It remains unclear whether TikTok will be banned in the United States and potentially in other Western countries. However, we are now one step closer to the ban.
  • The Montana State House has voted in favor of a bill that would ban TikTok in the US state. It still needs to be signed by the governor of Montana to become law.
  • The Republican governor has not yet commented on whether he intends to sign the bill.
  • If he does sign it, it is expected that the law will be challenged, which would then become a test case for a nationwide ban.
  • If the TikTok ban becomes a reality in Montana, it would be the only territory in the world, along with Afghanistan and India, to completely ban the Chinese app.
  • TikTok has 150 million users in the United States, which is 45% of the population.

The impact of influencers on viral TikTok songs

  • This article shows that viral songs on TikTok often feel organic, even though there is often a campaign and influencers behind them.
  • While smaller influencers are happy with a few hundred dollars, the big names are said to receive five-figure sums from labels or PR firms.
  • As law enforcement officials and TikTok itself place more value on transparency, the hashtag #MusicAd is becoming more and more visible in popular posts, albeit often hidden. Usually, only the truly big influencers follow these guidelines, as they have more to lose.
  • Smaller acts lose out because they cannot afford such campaigns. However, it should not be forgotten that expensive campaigns do not guarantee success, especially not with TikTok’s unpredictable algorithm.
  • That TikTok sometimes even ensures that songs go viral, they call this “heating” a song, I have already shown here.

Bonus Reads

  • “Link in Bio” is perhaps the most posted phrase on Instagram. So far, Instagram has only allowed users to post a single link. This changes now: Instagram now allows up to five links in the bio. However, most artists probably find that five links are by no means sufficient, and the tool cannot visually compete with the popular smart link providers. A step in the right direction, but not more than that.
  • It seems that Spotify is doing some spring cleaning – everything that is no longer needed has to go. After just recently shutting down Spotify Live, it has now been announced that they will also be scrapping the music quiz game Heardle as of May 5th. This is less than a year after they acquired it.
  • AI is the dominant topic this week, which fits with a survey conducted by music distributor Ditto among its customers. The survey comes to a surprising conclusion for many, that AI is already being used much more frequently than expected. 60% of respondents have already used AI for a release. Most of them used it for artwork or mastering, but 20.3% also used it for production and 11% for songwriting. A large majority also wants to continue using or experimenting with AI in the future. Only 28.5% said they would never use AI for their music.
  • As already reported, Universal Music is aiming for a new streaming model, one which they call Artist-Centric. David Martin, CEO of the Featured Artists Coalition, believes that if you want Artist-Centric, you should also talk to the artists to better understand their needs and desires.

Industry Groove – Week 15

Great to have you here! Welcome to the first edition of Industry Groove Newsletter. From now on, I will be summarizing the most important news from the music industry on a weekly basis and providing relevant links for those who want to dive deeper into a topic.

A brief history of how this newsletter came about: It all started in June 2022 with the idea of creating an internal newsletter for iGroove employees, so they wouldn’t miss any important developments in the music industry. Since I was already writing articles for the iGroove Magazine and constantly scouring relevant websites, it was clear that this would be my job. Soon, what I had secretly hoped for from the beginning came true: the feedback on the newsletter was so positive that we decided to make it accessible to everyone. And here we are now. By the way, all internal newsletters published in 2023 can now be read in the iGroove magazine.

Almost one third of 2023 has already passed, and in these first almost four months, two main topics have dominated the music industry. One of them is the discussion about a new streaming model initiated by Universal Music CEO Lucian Grainge. We are currently living with the contradiction that streaming is the most important source of revenue and the strongest growth driver, but at the same time, crisis symptoms are increasing. The urgency of the topic is evident from the topics covered in this newsletter alone.

The second dominant theme is, unsurprisingly, AI, or artificial intelligence. The rapid development in the field of AI has led to topics such as Web3, Metaverse, or NFTs taking a backseat. Of course, they are neither gone nor irrelevant, but the momentum is currently clearly not in the Metaverse, as the article on Decentraland shows.

But now without further ado, let’s get to the topics of the week.

What could a new streaming model look like?

  • As mentioned in the intro, the topic of “new streaming model” has dominated the year so far. A lot has happened, and to provide an overview, I have summarized the most important points in this article.
  • The article sheds light on why the majors, especially Universal, are now suddenly also demanding a new streaming model.
  • Furthermore, it explains how the current system and the most discussed option, user-centric, work.
  • Finally, the article presents nine other possible models that are currently being talked about.

Price increases would result in significant additional revenues

  • Many artists and labels have been demanding for a long time that DSPs, especially Spotify, increase their prices. This demand is now also coming increasingly from the major labels.
  • Especially because growth in streaming is declining or stagnating in many places, it is demanded that Spotify follow the example of Apple Music and Amazon Music and raise its prices in the US and other important markets.
  • Barclays estimates that a 10% price increase at all DSPs would increase Universal Music Group’s earnings per share by 13% and generate €400 million ($430 million) in annual revenue and €240 million ($258 million) in gross margin. The same price increase would also increase Warner Music Group’s earnings per share by 21% and bring in an additional $256 million in revenue and gross margin of $158 million.
  • Analysts at JPMorgan Chase estimate that a price increase by Spotify would lead to annual additional revenues of 200 million dollars.
  • However, Spotify still seems to be focusing more on subscriber growth than revenue growth. It remains to be seen how long shareholders and labels will continue to watch this game.

Millennials want a different streaming experience

  • Recently, there has been a lot of discussion about streaming services struggling to reach Generation Z. Some of the latest adjustments by DSPs were explicitly aimed at young consumers.
  • However, according to MIDiA, there seems to be a problem with another generation as well: the Millennials. They were actually considered the foundation of the streaming industry.
  • In their latest report, they focus on 25-34 year-olds in the US. They show that this age group is the most valuable, but also the most threatened for the music industry.
  • Millennials are the generation that invests the most time and, above all, money in music. They are also the age group that witnessed the shift from physical products to streaming and is more attached to the old media than expected.
  • Nearly half of all respondents stated that they would rather cancel a music streaming subscription than a film subscription.
  • Additionally, more than half want people, not algorithms, to suggest music to them. They also want more social elements on DSPs, such as more tools for their profile pages or messaging features.
  • Unlike with Gen Z, one can sense that Millennials are resisting becoming completely passive music consumers, which is almost inevitable with streaming.
  • For DSPs, it will now be extremely difficult to reconcile the quite different needs of the two most important age groups.

Universal Music takes action against AI-generated music on DSPs

  • Universal Music not only wants to set the direction for DSPs in terms of streaming models, but also in the area of AI-generated music. This was revealed by the Financial Times.
  • Universal has reportedly urged DSPs in an email to block their catalog for all developers of AI technologies that use music to train artificial intelligence.
  • Additionally, Universal has sent numerous takedown requests, stating that this music was used to train copyrighted content without the consent of the rights holders.
  • Universal Music is thus getting involved in the already intensively debated copyright debate, which should have been conducted much earlier and will now likely occupy numerous courts.
  • DSPs have so far remained diplomatically silent.

Why AI-generated music is currently not a threat to established labels

  • Universal Music obviously sees AI-generated music as a threat and one must wonder how far it will go, especially when you look at how easy it is to produce a song with Kanye’s voice, for example, as seen in this tweet. What also doesn’t help is that some bank analysts have lowered their valuation of Universal’s stock, explicitly citing AI-generated music as a threat to their business model.
  • However, MBW analysts currently see no threat to established labels. The crucial question is not how much AI music floods the DSPs, but how often it is actually listened to (which is logical, but sometimes forgotten).
  • In 2022, the three majors plus Merlin (a consortium of numerous independent labels and distributors, including iGroove) lost 2% market share on Spotify. Since 2017, it’s been 12%, which of course contributes to the recent activism of the majors.
  • Although this is not a pleasant development for the major companies, one must take a step back and look at the big picture.
  • On average, 98,500 songs are uploaded to DSPs every day, and nearly 50% of the approximately 100 million songs on Spotify and other platforms were released in the last three years. A flood of releases will not come with AI-generated music, this flood is already here.
  • Only 4% of all released songs come from the majors and yet they maintain their market share more or less. In this sense, a loss of 2% is not much.
  • Because the music of the majors is listened to, which cannot be said of many other releases. 42% or 67.1 million songs were played less than 10 times on Spotify in 2022, 38 million or 24% were not streamed at all throughout the year.
  • Spotify therefore also explains that there are only 200,000 professional acts on their platform, which is only 2.2% of all artists on Spotify. These 2.2% are responsible for no less than 95% of the royalties paid out.
  • MBW therefore believes that AI-generated music currently, if at all, represents a threat to functional music, such as sounds for sleeping, exercising, or learning.

Decentraland: Interest continues to decline

  • In the intro, I already mentioned that developments in the metaverse are still progressing, but the hype has flattened out quite a bit. This is also evident in the latest news surrounding Decentraland.
  • One of their biggest events is the Metaverse Fashion Week, which also involves big brands like Dolce & Gabbana or Tommy Hilfiger.
  • Last year, 108,000 people attended the Fashion Week, but this year, there were only 26,000.
  • Another example: The sale of virtual land in Decentraland peaked in 2021 with a trading volume of one million dollars per week. This number has now dropped to $50,000.
  • In another article, Techdirt points out that interest among young US consumers remains limited.
  • 27% own a VR headset and only 7% plan to buy one anytime soon.
  • Sales of VR headsets in the US declined by 2% in 2022, and worldwide sales dropped by 12%.

Where are the numbers to evaluate and compare labels?

  • Nowadays, anyone can evaluate how successful an artist is based on numbers such as streams, monthly listeners, followers, likes, etc. And without the right numbers, it’s now almost impossible to get a record deal.
  • The article from MIDiA now asks where the numbers are to evaluate the performance of a label. This would help artists decide if a record deal is really right for them and, if so, which label is best for them.
  • What could be possible metrics? For example, the ratio between signings and successes, the loyalty of artists to a label, and whether they renew their contracts. How do labels perform compared to each other? How does the performance of indie labels differ from that of distributors?
  • Of course, we always hear about the successes of labels. But what about the failures, the dropped artists, the botched album campaigns?
  • Now the question arises of who could provide such data. First, they see Spotify or other DSPs, as they have all streaming data available.
  • Other options would be ROSTR or the companies that compile the charts. However, these are probably too closely linked to the majors.

Bonus Reads

  • Spotify has shut down its app Spotify Live, which is just another proof of the rapidly declining interest in social audio apps – a short-term hype triggered by Clubhouse. This article from MIDiA explores what’s next for the live audio sector, with the conclusion that live audio makes more sense as a feature than a standalone product.
  • Metadata may not be sexy and often overlooked, but as this article from Billboard shows, they could contribute to revenue growth.
  • In the first three months of this year, one trillion songs were streamed worldwide, breaking the trillion mark faster than ever before. If you add up all the users, 960,000 years of music were listened to.
  • Complex has compiled a list of the 25 most powerful figures in the (American) hip-hop media landscape.
  • Hootsuite presents us with an entertaining overview of the history of social media, starting from SixDegrees and Myspace to BeReal.

Industry Groove – Week 14

In this edition, we again focus on the topic of new streaming models, bid farewell to Spotify Live, and take a look at the music markets of the two most populous countries, India and China.

What could a new streaming model look like?

  • Since the beginning of the year, I have written a lot about this topic, and since many ideas are circulating, I have now summarized everything that has happened in one post.
  • The article examines why the majors, particularly Universal, are now suddenly demanding a new streaming model.
  • It also shows how the current syste and the most-discussed option to date, User-Centric, work.
  • Finally, I present nine other possible models that are currently being discussed.

Spotify Live is history

  • If any further proof was needed that social audio apps were just a short-lived Covid hype, it is now here. Spotify has shut down the app Spotify Live, formerly known as Greenroom.
  • However, Spotify explains that they will continue to experiment with Live Creator features on their main app, but a standalone app is not necessary for this.
  • They learned a lot from live activities, but they also paid 57 million for the app back then. There would certainly have been cheaper ways to learn those lessons.
  • Some of the live shows will now be continued as podcasts on Spotify.
  • Meanwhile, TechCrunch reports that Spotify is experimenting with new user profiles. As always with such tests, it is unclear whether they will ever be rolled out to all users.

New numbers from YouTube Shorts

  • We already knew that YouTube Shorts generates 50 billion daily views. Now, YouTube Music’s boss Lyor Cohen has released some new numbers.
  • The top 1,000 songs on Shorts generated more than 280 billion views in January alone.
  • Fan-created Shorts are said to increase an artist’s audience by more than 80%.
  • Artists who posted Shorts at least weekly received more than 50% of their new subscribers through Shorts.
  • This not only positions YouTube Shorts as an alternative to TikTok or Reels but also as an entry point to YouTube Music, long-form videos, or livestreams, all within the same ecosystem.

Two-thirds of TikTok users don’t know what they’re listening to

  • The latest article by MIDiA’s Mark Mulligan raises some general questions that should make everyone involved in music marketing think. He also presents some interesting numbers.
  • First, he reminds us that fandom is in crisis. Music is becoming more and more commoditized, and in the constant flow of new songs, artists are increasingly fading into the background. There are fewer and fewer super fans, and more and more passive consumers.
  • Partly filling this gap is TikTok, but the article shows that more people discover new music through TV than through TikTok (the numbers are all based on the UK). This, of course, indicates the increasing importance of sync placements.
  • Perhaps the most alarming statistic is that 64% of users usually don’t know which artist they’re listening to on TikTok. And only 19% of users who discover music on TikTok leave the app to listen to the song on a streaming service.
  • Mulligan also criticizes the fact that many labels build their marketing only on short-term success and not on the long-term development of artists, even though this would be much more lucrative in the long run.

1 billion lost annually through streaming fraud

  • It would be one thing if the fraudsters were only stealing money from the DSPs. However, every fraudulently earned dollar means one less dollar for an honest artist. That’s why we need to keep bringing up this issue again and again (which I’m doing and probably starting to annoy people).
  • The article by Billboard shows that the issue is receiving more attention again, and there are various new studies.
  • As already reported, according to a study in France, 1-3% of all streams are fraudulent, but these are only the ones detected by DSPs. Deezer estimates the number to be closer to 7%. The company Beatdapp even believes that it’s 10%. However, as a manufacturer of “Fraud Detection Software,” they naturally have an interest in making the problem appear as large as possible.
  • If we assume that it’s 10%, not less than $1 billion would end up in the wrong pockets.
  • According to a study, 80% of fraud does not aim to push a release into the charts but rather to increase long-term streams.
  • It’s also clear that fraud occurs at all stages of the career ladder, not just with smaller artists trying to launch their careers.
  • At the moment, it’s mainly the DSPs themselves, as well as distributors, trying to detect and prevent fraud. With moderate success, as it turns out.
  • Some voices are therefore calling for an independent institution to search for fraudulent streams. A worthwhile idea, in my opinion, as DSPs are known to have a limited interest in uncovering fraud.

Spotify grows strongly in India

  • Spotify told Billboard that they were able to triple the number of users in India in the last two years. They didn’t want to give exact numbers, but estimates put the number of monthly active users (MAUs) at 55 million. This makes them one of the top 5 countries in terms of number of users.
  • Around 10 billion streams are generated per month.
  • Where they don’t belong to the top 5, of course, is generated revenue. Pay-per-stream is significantly lower in India than in the world’s largest music markets (a subscription also costs only around $1.45). Many users also use the free offering, which should not be surprising given the rampant poverty in the country. India ranks 140th out of 190 countries in terms of per capita income. Less than half have a smartphone or internet access.
  • This is also reflected in the fact that India, soon to be the world’s most populous country, only has the 17th largest music market ($219 million in revenue).
  • There are an estimated 300 million MAUs in India (for comparison: the US has 219 million). With 1.4 billion people, there is still potential here.
  • When Spotify started, consumption of international music was between 70-80%. Meanwhile, it has reversed, with 70% of the songs listened to coming from local artists. That’s the highest share in the world, as Spotify explains.

The great potential of the Chinese music market

  • According to the IFPI, China is for the first time one of the five most important music markets in the world.
  • Last year, “Recorded Music Revenues” grew by an impressive 28.4%. Not less than 89% of these revenues were contributed by streaming.
  • There are approximately 100 million paying subscribers in total. This is a huge number, but considering China’s population, there is still plenty of room for growth. Especially when compared to video streaming, where there are a whopping 700 million paying customers.

Bonus Reads

  • The market shares of the majors are continuously declining, and as I reported last week, only 4% of uploaded songs come from the three majors. I already pointed out that the giants are now likely to push their independent distributors. The latest MBW article goes into more detail on this topic.
  • According to a research institution dedicated to transparency in politics, ByteDance has invested $13.4 million in lobbying in the US since 2019. $5.3 million of this was invested in the last year alone. This would even surpass Amazon, Meta, and Alphabet (Google). Currently, it seems that it hasn’t helped much, but perhaps the ban on TikTok can still be avoided. Meanwhile, government employees in Australia and New Zealand are also not allowed to install TikTok on their business phones due to security concerns.
  • This article aims to debunk 10 social media myths, such as the idea that you should spread unique content on each platform or that you have to post multiple times a day.
  • The Federal Association of the Music Industry (BVMI) and its umbrella organization IFPI have won in court against YouTube-DL, a site that allows users to rip music from YouTube. As around 30% of music consumers also listen to music illegally, this is of course only a small victory, but one can hope that this precedent will have a greater impact.
  • This announcement is only indirectly related to the music industry, but the news that Disney is shutting down its Metaverse division does have a signaling effect. It seems to me that more and more companies in the areas of Metaverse and NFTs are slowing down, which probably does not bode well for their imminent mass suitability.
  • The latest media articles suggest that the music industry company Utopia based in Zug, Switzerland, is in dire financial straits. After buying one company after another last year, they are now experiencing setbacks. They have now sold the British publishing platform Sentric to Believe, who want to build their publishing business.
  • The author of this article wonders why metadata isn’t handled as well everywhere as it is at Apple Music Classic. If this were the case for other DSPs, for example, one could search for Pharrell and not only get his own songs but also everything he produced. It probably interests very few people but would likely be something for a more expensive subscription that appeals to super fans and nerds like me.

How could a new streaming model look like?

What you'll learn in this post:
  • Why suddenly even the majors demand a new streaming model
  • Nine possible options for such a model
  • Why we shouldn't focus solely on streaming optimization

Unsatisfied musicians and smaller labels demanding a new streaming model and more fairness is nothing new. For DSPs, it has always been a bit annoying because they like to have the artists on their side. But it wasn’t much more than an annoyance that couldn’t be soothed with some nice words, glossy brochures, and pretty landing pages. Now, however, the criticism is suddenly not only coming from the artistic community, but also from the biggest and most powerful music label: Universal Music. Until now, it was actually assumed that the major labels had managed to arrange themselves quite well with the current system – after all, they either built it or at least gave their approval for it for many years, and above all, they profited from it for a long time. Now, the supposed opponents – on one side DIY artists and indie labels, on the other industry giant Universal – suddenly find themselves on the same side. At least at first glance. What is certain, however, is that since Universal CEO Sir Lucian Grainge spoke out and demanded a new streaming model, the debate has taken on an urgency never before seen.

What is Universal criticizing?

For most observers, it is obvious that the shrinking market share of the major labels is responsible for the sudden demand for a new system. Grainge sees the main reason for this as the incredible number of new songs being released. He refers to the figure of 100,000 songs being uploaded to Spotify and similar platforms every day. This number could even multiply in the foreseeable future, should DSPs be flooded with AI-generated music.

Grainge is particularly targeting so-called “fake artists” and “functional music.” The latter refers to releases for mood playlists for relaxation, workouts, or sleep, as well as releases based solely on sounds such as rain or hairdryers to calm screaming children. Grainge believes that such music cannot possibly be paid the same as releases by creative musicians (whether major or indie) who have invested a lot of time and love into their songs. However, Grainge ignores the fact that DSPs have become audio providers offering all kinds of content, and ultimately, users decide which ones they want to use. Nowadays, this includes jogging playlists, LoFi music for working, and the sound of rain for falling asleep. The streaming system has turned even passionate music fans into passive music consumers. Moreover, market shares are increasingly shifting from superstars to the middle ground, and it is in this middle ground where the most exciting and innovative sounds can be found, as well as a lot of functional music.

There is no doubt that there is a lot of fraud going on on the DSPs, and it would be important to prevent this as much as possible with a new system. However, just because functional releases have hijacked the streaming system and sometimes even taken it to the point of absurdity, does not necessarily mean that it is fraud. Rather, it clearly satisfies a need. Even Sir Grainge, probably the most powerful man in the music industry, must accept this and should not presume to decide what is good music and what is not. After all, doesn’t everything that does not cheat have its justification, even if Sir Grainge does not like that whale song is paid the same as Taylor Swift?

First tests are already underway

Universal didn’t just stop at words. They have already partnered with the DSPs TIDAL and Deezer to explore a new streaming model (or several). It is certain that it will not be the already partially explored user-centric model, as this also brings too many disadvantages for Universal. Sir Grainge and Universal have an artist-centric model in mind, but without going into detail about what that could mean. Let’s therefore search for a new model that could replace the already buried pro-rata model.


To briefly summarize once again, how the pro-rata model works, which is currently used by almost all western DSPs. Often a big pot is talked about, into which all revenues flow (subscriptions and advertising, depending on the DSP) and are then distributed proportionally. If there were 1 million streams in a month and you had 100,000 of them, you would receive 10% of the generated revenue (minus the 30% the DSP keeps for itself). That’s how pro-rata works in a simplified way. In reality, it is much more complex, as many different factors determine the value of a stream, such as its origin, the type of subscription, the number of subscribers and generated streams, etc. There are therefore different pots for different countries and subscriptions. The crucial point is that everything that has been listened to for longer than 30 seconds counts as a stream, regardless of whether it was listened to for 31 seconds or 5 minutes. A user’s money does not go directly to the artists he or she has effectively listened to.


The user-centric model is supposed to change this. The revenues generated by each user are only paid out to the artists he or she has actually listened to. Previous studies suggest that certain niche genres would benefit greatly, while hip-hop in particular would suffer significant losses. It is also unpleasant that low-listeners are rewarded. Users who listen to very few songs and artists per month distribute much more per artist than heavy users who check out a lot of music and thus distribute their subscription costs to many more artists. There have been encouraging studies so far, but also some negative points. It is difficult. In case of doubt, however, the user-centric model seems to have some advantages over the current system.

Is there a perfect and fair system for all sides? This remains to be seen, but it is highly doubtful. Let’s now look at some possible alternatives.

9 possible new streaming models

Various subscription tiers

One of the most easily implementable and understandable options is to reconsider the existing subscription systems and break away from the one-size-fits-all model. The existing system could still work for occasional listeners, but new tiers with additional features and content should be introduced for heavy users. Those users who are willing to pay more should also have the opportunity to do so. In addition, artists should finally receive the tools to monetize their super fans with exclusive content or other specials. Another option would be to offer certain fans the album earlier for an additional fee. One could learn a lot from Asian DSPs or providers like Bandcamp here.

Rewarding active listening

Recently, Spotify for Artists introduced the concept of an “Active Audience”. It shows which artists have been directly listened to through the artist profile, album, or release page, or whose tracks have been streamed through the library or personal playlist. The idea of rewarding active listening with a higher pay-per-stream rate and paying out a reduced rate for passive consumption through algorithmic playlists or the radio function has been around for a while. However, the majors may not be willing to cooperate since they also profit from passive consumption. It would certainly not be easy to distinguish between active and passive listening. Once a good compromise is found, this would be a fair improvement that treats all content equally, whether it is Taylor Swift or whale song that is actively listened to.

The multiplier model

The CEO of Warner, Robert Kyncl, advocates for this model, which goes in a similar direction as the idea of favoring active listening. Again, the actions of the user should be separated from those of the algorithms. Additionally, Kyncl suggests that the artist who a user listens to first should receive additional income. Artists who users actively search for should thus receive more than just their pro-rata share. At first glance, this sounds fair, but upon closer inspection, one naturally asks why only the first heard artist should benefit additionally and not also the second or third. Perhaps the idea is simply not fully developed yet.

Payment based on listening time

Currently, a stream is counted once a song has been listened to for 31 seconds. Whether you skip after those 31 seconds or continue listening for several more minutes, it has no influence on the payout. It would therefore be only fair to receive more payout for longer listening. For example, the payout could be increased every 30 seconds. This would probably result in rain songs suddenly lasting 5 minutes instead of just 31 seconds. The fairest option would be a precise accounting system based on seconds.

Threshold for payouts

This is a controversial but nevertheless worth considering option. It involves introducing a threshold for payouts. There would be different options for implementation, such as setting a minimum of 1,000 streams per release per month before payout or only paying out once a song reaches 10,000 all-time streams. This would affect many artists, especially those who only receive a small amount of money anyway. However, since there are many of these small artists, there would still be significantly more money available to pay established musicians. The administrative effort behind this would probably discourage DSPs, in addition to the many dissatisfied artists.

Top-Up subscription

Another interesting idea, which is unlikely to be implemented, is the following: Each stream is valued at a fixed amount, for example, $0.01. Once a listener has reached the $9.99 cost of their subscription (i.e. after 999 streams), they would have to pay extra to be able to stream more songs for the remainder of the month. This would generate more income that could be passed on to the artists, but it would not necessarily be user-friendly.

DSPs differentiating themselves

Currently, it doesn’t matter much which streaming provider you have a subscription with. They all offer roughly the same songs at identical prices. However, DSPs could differentiate themselves more and position themselves more clearly. For example, Spotify could remain the provider that caters to the broadest audience, while other DSPs could focus more on audiophile listeners or super fans and adjust their offerings accordingly.

The majors building their own DSPs

Some voices, led by Dan Runcie from the newsletter Trapital, believe that the majors should build their own DSPs or buy existing ones. From the majors’ perspective, this could indeed bring some advantages, but it would be associated with a high effort. Whether artists, especially those who are not under contract with the respective major, would also benefit is more than questionable. Moreover, there would be a high chance of market fragmentation, as we know it from video streaming. Would users be willing to pay for multiple music streaming subscriptions? Therefore, many questions remain unanswered.

Artist promotion model

This proposal comes from the independent sector and would of course be fiercely fought against by the majors. The idea goes like this: The more streams an artist generates, the less they are paid for each additional stream. The idea is simple: money should be redistributed from superstars to emerging artists. This degressive model sounds appealing, but is of course completely hopeless.

No party is really satisfied

The current system doesn’t make any of the involved parties truly happy. The artists and songwriters rightfully complain about the meager payouts. The rights holders, i.e. labels, distributors, publishers, and so on, are fighting for a larger share. Meanwhile, streaming services, especially Spotify, report losses practically every quarter because they operate with incredibly thin margins. This is where the problem lies. You can’t increase payouts indefinitely without causing the streaming system to collapse. Any shift in favor of one of the involved parties has negative consequences for at least one of the other two. Therefore, it will hardly be possible to find a new system in which all parties involved are truly satisfied.

Of course, despite this, it’s still important to look for ways to optimize streaming, given its incredible market share. But even more important is that artists, managers, and labels look for alternative sources of income and don’t rely too heavily on streaming.

Industry Groove – Week 13

This week, (almost) everything revolves around a new streaming model and the question of whether streaming can work as a business model. On one hand, the latest numbers show that more music has been uploaded to DSPs (digital service providers) in the last three years than ever before. On the other hand, a shockingly high percentage of existing music is rarely or never listened to. After reading these articles, it becomes clear once again that changes are absolutely necessary.

The second major topic continues to be AI, and prominent representatives of the tech industry are calling for a little restraint.

Millions of unheard songs cause huge costs for Spotify

  • MBW has once again delved into the numbers of Spotify, although this time the numbers were provided by “Insights Provider” Luminate.
  • And these numbers reveal to us that 67.1 million songs on Spotify were played less than 10 times throughout 2022. These are not less than 42% of the 158 million songs on Spotify!
  • Even crazier is the next number: 38 million tracks or 24% of all songs received not a single stream in 2022. Not. A. Single. Stream!!! So not even the artist themselves or their mother felt it was necessary to play the song at least once.
  • Of course, the problem is not entirely new. About a year ago, I reported that more than half of all songs on Spotify have less than 500 all-time streams.
  • This leads MBW directly to the topic of costs that Spotify incurs for cloud storage (money that flows to competitor Google). At the end of last year, I reported in this article on at least 130 million annually that Spotify spends to store many songs that no one listens to.
  • Meanwhile, MBW sees these costs at around 150 million. A nine-figure sum that has a very direct impact on margins and payouts to the artists who actually generate streams.
  • The possible solution: artists with few streams must pay to keep their songs on the platform. The problem: that would cause quite a bit of bureaucracy with the various distributors. Does Spotify need to allow direct upload to their platform again?

Nearly half of all songs on DSPs were released from 2020 to the present

  • Another article from MBW, based on numbers from Luminate, shows that almost half of all songs on DSPs were released in 2020 or later. A very surprising number, in my opinion.
  • Since 2020, 90.4 million separate ISRCs have been loaded onto streaming services. Most of them were in 2022 with 34.1 million. These are thus 46% of all tracks in Luminate’s database (196 million).
  • Luminate also confirms that there are actually almost 100,000 songs (98,500 to be exact) being loaded onto DSPs every day. Or as they say, songs with a new ISRC (so if the distribution changes but the ISRC remains the same, it doesn’t count).
  • Of these 98,500 tracks, only 4% come from the three majors. But these companies still upload 3,940 tracks every day, which is quite a lot for just three companies!
  • MBW suspects that the majors will now push their “independent distributors” InGrooves, The Orchard, AWAL, Santa Anna, ADA, and Level Music further to regain market share.

Why Spotify’s business model cannot succeed (and what Apple has to do with it)

  • Regardless of how impressive the numbers that Spotify presents are (such as distributing $40 billion to the music industry), no one is really happy. Neither the artists nor the investors.
  • This article argues that their business model practically makes it impossible to make truly significant payouts to artists, and at the same time, it is practically impossible for the company to ever become truly profitable.
  • Since its founding, there have been hardly any quarters without a loss. By the way, the first time this happened was in 2018, 12 years after its founding!
  • Since they pay out 70% of their revenue (which is still too little), their gross margin is simply too thin. Especially if their fixed costs are so high (see article above).
  • The article reminds us that it was Apple who introduced the one-third share with their iTunes store, which became the basis for today’s streaming economy.
  • Over the years, this has cost artists billions. The only real winners are the consumers who get 100 million songs for a pittance.
  • The author believes that Spotify will never be able to compete with the tech giants, i.e., their direct competitors, and will sooner or later be bought out.
  • Although the article does not offer any new insights, it summarizes them well and, above all, shows that the streaming system cannot be fixed, so alternatives will have to be sought.

Spotify now shows “Active Audience”

  • Through S4A, Spotify now provides artists with insight into their active listeners.
  • Active in this case means that the song was listened to via the artist profile, the album or release page, or that the user played the track through their own library or playlist.
  • Furthermore, active listeners are divided into three categories: super, moderate, and light. Super are those who have listened to an artist particularly frequently in the last 28 days, moderate have listened to the artist multiple times, and light are those who have only played the artist’s music once or twice in the last month.
  • According to Spotify, active listeners will listen to an artist four times more frequently in the next six months compared to passive listeners who only consume an artist through editorial or personalized playlists.
  • What is now missing is a tool to actively engage this active audience.

Tech leaders demand a 6-month AI stop

  • Artificial intelligence (AI) offers incredible opportunities and is undoubtedly here to stay. Understandably, it not only generates enthusiasm but also fears, especially among many creatives such as songwriters.
  • To take these fears seriously and to give governments time to figure out how to regulate this area (which no government really has a plan for yet), over 1,000 people from the tech and science industries are calling for an immediate pause of all training of AI systems that are more advanced than ChatGPT-4 for six months.
  • These are not AI critics, but people from Apple, Google, or my good friend Elon Musk.
  • The first signatories write, “AI systems with human-like intelligence can pose profound risks to society and humanity, as extensive research by leading AI laboratories acknowledges.”
  • They further write, “Contemporary AI systems are now competing with humans in general tasks, and we must ask ourselves: Should we allow machines to flood our information channels with propaganda and falsehoods? Should we automate all jobs, even fulfilling ones? Should we develop non-human entities that could eventually surpass us in number, intelligence, and replaceability? Should we take the risk of losing control over our civilization? Such decisions cannot be delegated to unelected technology leaders.”
  • With the pace at which the field is currently developing, taking a break may not be the worst idea.

Bonus Reads

  • So far, the layoffs primarily affected the tech industry, but now the first major label is also affected. Warner Music is laying off 4% of its employees, which corresponds to 270 employees in concrete numbers. Warner has a total of around 6,200 employees.
  • The TikTok CEO was grilled by US politicians for five hours, and after Shou Zi Chew’s not necessarily convincing performance, Billboard sees a ban of TikTok as increasingly likely. They also summarize what impact this would have on the Billboard charts here.
  • This article shows how Fortnite could become a new source of revenue for musicians and labels.
  • I have already written several times in the newsletter that niches are becoming increasingly important. Spotify now wants to serve them even more and is doing so with their new “Niche Mixes.” Some of them are available now under “Created for You” – many more, even quite obscure ones, can be found through the search function. There are supposed to be tens of thousands of mixes in total, personalized for each individual user.
  • Spotify has recently launched or expanded some marketing tools. The overview and deep dive can be found in this paper by Music.Ally.
  • YouTube announced that in the meantime, 100,000 creators have linked their online shop with their YouTube channel. This is definitely an interesting option for musicians, especially since the shopping area is also being expanded for Shorts.

Industry Groove – Week 12

Actually, I wanted to write an extensive intro in which I draw a connection between the downfall of Credit Suisse and the music industry. However, I’ll refrain from doing so and am simply glad that I haven’t parked a single penny with big banks that are hardly distinguishable from criminal organizations for many years now. In good times, they dole out billion-dollar bonuses, and in bad times, the government, aka us taxpayers, have to shoulder the burden. Thanks for nothing!

Will the USA force ByteDance to sell TikTok?

  • Sometimes I wish I had a crystal ball to see into the future. For example, in the whole TikTok ban saga, which seems a bit like an (economic) crime novel. Will there really be a ban, or is it just power games?
  • The latest development: The Biden administration has apparently resurrected an old idea from the Trump era, namely that ByteDance must sell the US business of TikTok.
  • As expected, TikTok doesn’t think much of this and also questions the benefit to national security, which according to many US politicians is being threatened (which may even be true, but then the question arises of how the security of the rest of the world is doing, which uses the products of US tech giants and happily feeds the US intelligence agencies).
  • More and more countries are banning their government officials from using TikTok. Meanwhile, the UK has followed suit including the venerable BBC.
  • But the much more decisive question is: would the United Kingdom or the EU follow suit if the US were to actually ban TikTok?
  • TikTok CEO Shout Zi Chew will even appear before the US Congress (today, to be exact), where he will defend and clarify that his company has no connection to the Chinese government. He also mentioned in passing that TikTok now has 150 million users (and 7,000 employees) in the US, who would of course all lose their access to the app if there was a ban.

Meta ends NFT experiment and lays off 10k more workers

  • Mark Zuckerberg has declared the year of efficiency, which surprisingly includes Meta withdrawing from its crypto projects. The integration of NFTs will therefore not be continued for the time being.
  • It will no longer be possible to mint and sell NFTs on Instagram, nor will you be able to show off your NFT collection on Insta and Facebook in the future.
  • This is surprising mainly because the metaverse continues to have a high priority and NFTs are actually inevitably associated with it.
  • Meta is also bringing more “efficiency” to its employees. After 11,000 workers had to leave at the end of last year, the layoff of 10,000 additional workers has now been announced. In addition, 5,000 open positions will not be filled.

Global Recorded Music Revenues grow for eighth consecutive year

  • Every spring, IFPI releases the latest numbers on “Global Recorded Music Revenues” – i.e. the revenue generated by the music industry, excluding publishing and live music. Also, every year MIDiA publishes its figures on the state of the music industry a few days before IFPI. Now both are available.
  • According to IFPI, global revenue was $26.2 billion, up 9%. MIDiA sees growth at 6.7% to $31.2 billion. This is significantly less than in 2021, where MIDiA saw growth at a whopping 24.8%. However, MIDiA analysts note that the 6.7% growth is still a success given the difficult economic environment.
  • Streaming generates the most revenue, of course, according to IFPI, with $17.5 billion (up 11.5%), representing a market share of 67%. MIDiA sees growth at 8.3%, which brings them to $20 billion, representing 64.1% of total revenue.
  • Globally, 589 million people are said to pay for a streaming subscription. Last year there were only 523 million.
  • Physical sales have also grown, by 4% to $4.6 billion, representing 17.5% of total revenue.
  • Sub-Saharan Africa grew the most at 34.7%, followed by Latin America (25.9%) and the Middle East and North Africa (23.8%). North America (5%) and Europe (7.5%), on the other hand, are almost stagnating.
  • MIDiA also estimates market share. They see the Universal Music Group at 29.5%, while indie labels and artists who work directly with a distributor come together at 34.6%.

Albums are getting longer

  • Streaming is causing albums to get longer or, more importantly, to include more songs. The calculation is simple: more songs = more streams.
  • In 2022, the top 10 albums in the Billboard charts had an average of 19.1 songs and lasted 69.9 minutes.
  • The trend is clear: in 1977, there were 10.3 songs and 45.1 minutes, in 1992 11.9 songs and 51.1 minutes, and finally in 2014, 13.2 songs and 51.9 minutes.
  • From 2014, streams counted for the Billboard charts, and the length skyrocketed.
  • In this article I had already discussed whether overly long albums contribute to the steady decline in interest in the album format.

Fewer DIY artists earned over 10K on Spotify

  • Two weeks ago, I presented you with the latest figures from Spotify’s Loud & Clear. Now, MBW has delved deeper and discovered an important figure.
  • Specifically, the number of DIY artists who received more than $10,000 from Spotify decreased from 15,140 to 14,700.
  • Although 440 is not a huge drop, it is surprising because the numbers are otherwise steadily increasing and more DIY artists are uploading their music to Spotify year after year.
  • As mentioned above, DIY artists are generating more revenue overall. However, it seems that this revenue is being spread out among a large number of artists.

Spotify HiFi is coming – just don’t ask when

  • It has been over two years since Spotify announced its HiFi offering. Since then, there has been silence, and there were no updates at the Stream On event either.
  • Now, Spotify Co-President Gustav Söderström has broken the silence with The Verge, but didn’t say much.
  • Söderström said that the industry changed after the announcement, and as a result, the launch was put on hold. However, HiFi is still coming: “We will do it, but we will do it in a way that makes sense for us and our listeners. The industry has changed, and we had to adapt.”
  • From a technical perspective, the feature has apparently been ready for over a year and is already being used by Spotify employees.
  • However, the plan was to offer HiFi at an extra cost, which became difficult after Apple Music offered it for free, and Amazon followed suit.
  • Offering it for free is still not an option, so it will likely become part of a new premium offering – someday.

Can AI-generated music have a fanbase?

  • As mentioned multiple times before, AI is currently dominating the music industry.
  • However, Tatiana Cirisano from MIDiA believes that soon, no one will care whether AI was involved in a song, just as no one cares if a track was made on a laptop or in a studio.
  • But it’s clear that AI music will create even more competition in an already oversaturated market.
  • One question that keeps coming up is whether AI-generated music can touch people in the same way that music created by humans does. Or, in other words, can people become fans of AI-generated music and AI artists?
  • Tatiana Cirisano thinks it will probably be possible because, as important as an artist’s personality is, ultimately, it’s the label’s marketing team that makes someone a fan. And they can do that for AI music as well.
  • The main takeaway is that fandom will become even more important, especially because the music may be less distinguishable in terms of quality. So, standing out through the ability to build a fanbase will become more critical. But of course, AI knows that too…

Bonus Reads

  • A study by New York University found that the first 5 seconds of a song are crucial in determining whether someone likes a track or not. This is even shorter than what was expected feared.
  • Not only is Meta laying off workers, but Amazon has also announced a second wave of layoffs affecting 9,000 people in AWS, Twitch, and other units. Last year, 18,000 people had to leave, and now another 9,000 positions are being cut. The Amazon-owned Twitch will also be affected by the layoffs, with 400 jobs being eliminated.
  • Vinyl pressing plants are fully booked due to the continuously increasing demand, resulting in long waiting times for artists and fans. Metallica alone pressed over 900,000 vinyl records in 2022, even without releasing a new album. Their solution to keep their loyal fans supplied with new vinyl? Simply buy a pressing plant themselves.

Industry Groove – Week 11

In last week’s newsletter, the many changes at Spotify were more of a breaking news story, so let’s take a step back and look again at the Stream On Event. Was this really an innovative breakthrough or do the things that haven’t moved still outweigh it?

Undoubtedly, the countdown page and expanded use of short videos are interesting, but their effective use will only be seen when they are widely used and not just by the Ed Sheerans of the world setting up a countdown page. The new Discovery Feed has so far apparently been received critically by users (I don’t know about you, but everything still looks the same to me) and the expected criticism of the controversial Discovery Mode did not take long to arrive. And of course, one can certainly question whether the effect is increasingly wearing off as more artists use the tool and want to push their songs to the forefront of the queue.

So for the moment, what hasn’t changed weighs more heavily. The refusal to comment on HiFi after two years seems almost childish. Of course, the criticism of the still missing price increases is even harsher, which will be a topic further down as well.

Warner CEO calls for higher subscription prices

  • Robert Kyncl, CEO of Warner Music, spoke clearly and directly about the topic of price increases at a public appearance, and even though he did not name Spotify directly, it was clear that his criticism was primarily directed at them.
  • Kyncl stated that due to inflation alone, a subscription should cost $13.25 today and not $9.99.
  • All companies that have not raised their prices are partially responsible for undervaluing music. Music is the cheapest form of entertainment with the lowest price per hour.
  • After Universal has already made a clear statement in favor of price increases, one must really wonder how long Spotify can withstand the pressure from the industry.
  • A more nuanced analysis on the topic is provided by Dan Runcie from the Trapital newsletter. He also points out the very small margins of DSPs and that it is therefore not surprising that most of the major DSPs are owned by tech giants and primarily serve their customer acquisition.
  • He also mentions again the problematic relationship between DSPs and major labels and their favoritism.
  • Therefore, he also brings up his idea again that the majors should build (or buy) their own DSPs, which I still think is unlikely.

US music industry grows, but more slowly

  • In recent days and weeks, the most important music markets have released their 2022 numbers and across the board, growth is apparent. Now the world’s largest music market has presented its figures.
  • The US market grew for the seventh consecutive year, by 6.1% from $15 billion in 2021 to $15.9 billion. However, the growth rate is the lowest since 2015. In recent years, growth has always been double-digit except for the COVID year 2020, where it was still 9.2%.
  • Streaming accounted for 84% of generated revenue ($13.3 billion), up from 83% ($12.4 billion) the previous year. This represents a growth of 7.2%, whereas the growth was much higher at 23% in the previous year.
  • Meanwhile, there are 92 million paying subscribers in the US, which is a 9.6% increase from the previous year.
  • Downloads fell by 20% and generated only $495 million, which still represents 3% of the market.
  • For the first time since 1987, vinyl sold better than CDs (previously, vinyl only surpassed CDs in terms of revenue). 41.3 million vinyl records were sold, compared to only 33.4 million CDs. Overall, the physical sector grew by 4% and generated $1.7 billion.
  • Sync revenue increased strongly by 24.3% to $382.5 million.

Will the Spotify Discovery Feed bring about major changes?

  • An interesting analysis of Spotify’s TikTok-inspired Discovery Feed is provided by Tatiana Cirisano of MIDiA. First, she reiterates that Spotify has lost a lot of cultural relevance, and that Gen Z in particular desires a more active experience and is therefore much more present on platforms like TikTok.
  • The Discovery Feed was clearly developed to change this. According to Cirisano, however, it could have other side effects as well.
  • She believes that the Feed could reverse a trend that Spotify itself started with its playlists: the ever-shorter songs, the hooks right at the beginning of the track, the lack of intros, bridges, and so on.
  • In the Feed, only the most striking part of the song is presented, but the complete version of the song could have more rough edges.
  • However, the question arises as to whether the rather passive Spotify users will even use the Feed. In addition, Spotify is now competing more directly with TikTok, and this battle is hardly winnable. For the artists, the question also arises as to whether they want to fill yet another platform with videos.
  • Nevertheless, there is a possibility that Spotify could close the gap where TikTok Music was supposed to be. This is not least because, through the Feed and the video posts of the artists, they could collect more information about user behavior that goes beyond musical taste.

Only 755 years until a million artists can make a living from Spotify

  • “Our mission is to unlock the potential of human creativity by giving a million creative artists the opportunity to live off their art and billions of fans the opportunity to enjoy and be inspired by it.” This nice-sounding motto was issued by Spotify, and MBW wanted to find out if Spotify is on track to enable one million artists to make a living solely from Spotify money.
  • Currently, as reported last week, there are 57,000 artists who earned more than 10k, and 1,060 artists who hit the million mark.
  • MBW has now compared the current figures with those from 2020 and 2021 to find out about the growth. They focus on the artists who generated more than 50,000, since it is difficult to live off a lower amount.
  • From 2021 to 2022, only 1,300 artists surpassed the 50k mark. From 2020 to 2021, there were 3,100.
  • If the number of artists continues to grow in the same way, it will take only 755 years to reach the goal of one million. What MBW means by this is: find a new slogan!

Bonus Reads

  • Artists from the electronic music world were among the first to experiment with NFTs, so it’s only logical that Beatport has now launched its own NFT marketplace in collaboration with Polkadot. is now online.
  • In this article, Billboard wonders whether downloads will ever make a comeback like vinyl did in recent years. Probably not, but the article also shows why downloads can still be important (and lucrative) in certain cases and are a sign of a strong fan base.
  • As one gets older, one discovers less and less new music. This is not just a suspicion, but has even been scientifically proven. This article shows at what age one is most receptive to new sounds. I’m glad to be more of an exception in this case, even though my taste in genres is established, I am constantly on the lookout for new artists, albums, and tracks. Keep on diggin’!