I’ve seen this mentioned a few times recently and it appears to be becoming a more prominent issue.

A few weeks ago I watched a great video by hadji called ‘The “Bad” Music of Instagram’ which is a decent dive into some examples and what’s going on:

And then a few days ago Anthony Fantano put out a video titled “Instagram Is Torturing Me With The Worst Music” which further shows the sort of content that’s being pushed.

I highly recommend checking out those videos, especially the first one, even if it’s just to get an idea of the type of thing we’re talking about. What we’re seeing here is an example of how algorithms are impacting what sort of music gets exposure on the platforms - and often not for the better.

We all know that engagement is an important metric when it comes to how much virality a piece of content gets. The more people engage with a post, and the more controversial it is, the more attention it attracts.

Music is of course subjective, but the content we are talking about is the stuff that imo 99% of people would place into the ‘bad’ category if they were asked to choose.

The problem with music as a form of content on social media is that there’s a real imbalance between its quality and the amount of engagement it tends to attract. Imagine a metaphorical ‘music quality spectrum’, and you might hope that the relationship between quality and engagement goes something like this:

But in reality, it actually looks a lot more like this:

At the top end of this spectrum, we have artists that have large audiences, and massive social awareness, who can generate high amounts of engagement when they put out new music. Die-hard fanbases, culturally significant releases, these sorts of things tend to pick up enough momentum to break through the noise and make it into people’s feeds.

In the middle we have a massive amount of artists who are creating decent music, very slowly growing their audience, but the content they put out is stuck in a sort of limbo, where they’re not getting enough engagement to break out, and they haven’t quite cultivated a big enough fanbase to move the needle.

Even though these are the sort of artists that we want to see pushed into our social feeds, the reality is the algorithm is stacked against them, with it being made even more difficult to compete due to the rise of the bottom end of the quality spectrum, which is starting to take over.

This ‘bad’ music is so pervasive precisely because of how bad it is. It drives engagement. As hadji mentions - “it’s the reels equivalent of stopping on the highway to watch a car crash”.

As a result, they drive a lot of engagement, with large amounts of (mostly negative) comments and an insane amount of views. The algorithm doesn’t know whether a piece of music is good or bad - it just knows it’s music-related content and it’s getting lots of comments and shares - so it goes and shows this to other people who have expressed an interest in music.

So in a counterintuitive way, the more awkward, cringe-inducing, and tone-deaf you can be, the more likely you are to have your music shared with people on the platform.

With the low barrier to entry, and every artist being told about how important social media is for promoting their music we’ll no doubt see this trend increase.

The one upside to this strange situation is that for now, this exposure doesn’t seem to be translating into engagement on streaming services - with the Spotify numbers for some of these artists nowhere near the levels that they are seeing on socials.

Let’s hope it stays that way 🤞

▶️ Artists Names Are Getting Stranger

An interesting post from Chris Dalla Riva looking at artist names over the years and how formatting, punctuation, and grammar have all changed. Inevitable really with the amount of people now creating music. It all starts to seem normal anyway once you spend enough time online 😆

▶️ California's music festival bubble is bursting

While this post talks specifically about California, it’s a trend I’ve seen discussed a lot this year. A good point raised in this Reddit thread makes a lot of sense - Millenials can afford to go but don’t want to, and Gen Z’s want to go but can’t afford it.

▶️ More Links

• YouTube Wrapper App ‘Musi’ Finally Pulled from the App Store — Free Lunch Ends for Millions of Users

• The Decline of the Working Musician

• Did That Song Go Viral on TikTok Organically — Or Was It Paid For?

• These Were The 40 Most-Played Songs At Pacha Ibiza This Season

‘Superfans’ has been somewhat of a buzzword in the industry lately, after Goldman Sachs got everyone looking for a way to extract the estimated $4.2b from the market that these superfans will supposedly spend.

While the investment pours in, and new tools and platforms are developed, one of the big questions is where these superfans will hang out.

One approach that we are seeing is labels building standalone apps, and also spaces where fans can hang out such as Weverse and Stationhead, but the amount of friction to onboard fans onto yet another platform means that this is only feasible for artists with a very large or dedicated audience, not to mention the amount of investment required to keep something like this updated.

The other approach is to build fan engagement into the platforms where fans are already hanging out - i.e. social media or DSPs. A good example of this is Taylor Swift’s in app experience on TikTok where fans can engage with certain tasks to unlock digital items and rewards.

While this approach massively increases the reach and reduces the friction of getting fans to engage, being able to do this is completely dependent on the platform and the tools it decides to provide to artists. I can’t envision this approach to fan engagement being rolled out to smaller artists any time soon.

That being said, we’ve seen mention from Spotify about ‘superfan clubs’ which may be aimed at a broader range of artists and could engage people directly in the app where they already consume their music.

Spotify also recently announced a partnership with AEG and Sabrina Carpenter to allow fans to get access to tickets early, as part of the Spotify Fans First program - a prime example of how interaction data can be used to segment and connect with devoted fans.

So, going forward will we see a new platform emerge to solve this problem, or will an existing platform start to integrate superfan features and become the dominant player in this space?

We’ll have to wait and see.

▶️ Swifties Vs Angels

There’s been a bit of friction between fans of Taylor Swift and Charli XCX after a Charli was denied a number 1 spot on the UK albums chart due to Taylor releasing UK-only editions of the Tortured Poets Department the same week (making it a total of 34 different versions of the album so far!).

Charli’s fans were upset by this seemingly intentional move, with chants of ‘Taylor is dead’ at a recent show in Sao Paolo, to which the singer responded later on Instagram criticising the behavior of the fans.

▶️ Suno and Udio Get Sued

The two biggest generative AI music generators at the moment are being sued by US labels body the RIAA. It seems clear that they have been trained on masses of copyrighted music so it will be interesting to watch this one play out.

▶️ Queen Sell Their Catalog to Sony for £1 Billion

The deal makes it the highest amount ever paid for a catalog - although Disney still owns the band’s rights in the US and Canada.

▶️ Why Are DJs Playing So Many New Dance Cover Songs?

Tapping into nostalgia has always been a successful move for artists and labels, this is especially true when it comes to dance music.

Glastonbury Looking Beautiful

@lesjoymusicAnd this was just a glimpse of it! #glastonburyfestival #glastonbury #glastonbury2024 #ukfestival #festival

Tiktok failed to load.

Enable 3rd party cookies or use another browser

More Links…

• Crazy Town Frontman Shifty Shellshock Dead At 49

• Deadmau5 Threatens to Remove His Music From Spotify

• HYBE subsidiary Supertone Co. launches virtual girl group SYNDI8

Sharing is caring

If you liked this email don’t forget to share it 🙏

UMG to Clone Artist’s Voices

Earlier this week Universal Music Group announced a deal with AI startup Soundlabs that will allow their artists to use the company’s AI voice model technology.

This essentially gives artists signed to the label the ability to train AI voice models that can be used in several different ways - a couple of examples being translating vocals to another language for songs released in different regions, and also turning their vocals into different types of ‘instruments’.

This also means that in theory these models could be used to create entire songs without the input of the artist in question.

I’m not against the use of AI in music, it’s just another tool in the box that artists can use, but things start to get questionable when you throw in the major label dynamics and how this could dramatically change the role of a major label artist. SoundLabs do state that artists will retain full control over the voice models, but as these tools become more commonplace will this always be the case?

We’re all too familiar with instances where labels have denied artists the ability to release music, but down the line could we see instances where labels will be able to create and release music that the artist didn’t even have a hand in making?

Of course, this is at the more extreme end of speculation, and no doubt AI clauses will be put into place in contracts to prevent this, but it’s not out of the question.

As well as translation for foreign markets, another application for this technology would be to streamline the process of creating reference tracks. Labels can take their reference tracks from producers/songwriters and immediately replace the demo vocals with the artist’s cloned voice to get a more accurate idea of what the song will sound like, allowing them to make quicker decisions on what sounds good.

However this ends up being used, it will be interesting to see which artists are empowered creatively, and which artists are further removed from the creation process.

Elsewhere…

▶️ After raising $125m, AI music generator Suno is now paying its most popular creators

There’s a prize pool of $100,000 shared between the top 500 songs in June (measured in plays, likes, and shares). Let’s hope they’re prepared for the bots 😆

▶️ Trying to Ban Phones in Nightclubs

James Hype tried to play a ‘no phones allowed’ night and ended up with a lot of phones in his face.

▶️ Ibiza festival criticised for charging DJs €5000 to perform

That’s not how it’s supposed to work, is it?

▶️ George Strait smashes US attendance record for ticketed gig

The Texas native drew 110,905 fans to Texas A&M's Kyle Field in College Station on Saturday 15 June

Sharing is caring

If you liked this email don’t forget to share it with your friends 🙏

As part of the sale, it was stipulated that the album could not be commercially exploited until 2103 - another 88 years - but it could be played at listening parties.

It’s an interesting experiment, designed to replicate what is seen in the contemporary art world, where value is placed on exclusivity. Inspired by the centuries-old practice of music patronage (where music was commissioned by wealthy patrons), it was always going to be interesting to see how the process would play out in a modern-day setting.

In a move that no one saw coming, the purchaser at the time was disgraced 'pharma bro’ Martin Shkreli (the guy who raised the prices of multiple life-saving drugs), which caused a fair amount of controversy and led to the members of Wu-Tang donating a significant portion of the proceeds to charity as a result.

After Shkreli was convicted of securities fraud the album was seized by authorities and ended up being sold for $4 million to a group called PleasrDAO, back when crypto music and DAOs were having their moment in the spotlight.

Last month it was announced that the album was going to be on loan to the Museum of Old and New Art in Tasmania for an upcoming exhibition, and of course, Shkreli couldn’t help stirring the pot by playing the album to thousands of listeners during a ‘spaces’ session on X the other day, as well as continuously talking about how he’s made multiple copies of the album that are out in the world.

He’s now being sued by PleasrDAO, which is probably going to be difficult given the nature of DAOs in general. Days later PleaserDAO has followed up by announcing a strange NFT drop:

For $1 fans can buy an NFT that gives them access to an encrypted version of the album that they can’t listen to (???), as well as a five-minute ‘never-heard-before’ sampler of the album. Every purchase also reduces the amount of time until the album can be released by 88 seconds, doing the maths that’s over 30 million sales required to get it released anytime soon!

To me this looks like a cash grab that’s likely to tarnish the album’s reputation as opposed to gaining any sort of cultural traction. The album sampler will no doubt find its way onto YouTube, and the encrypted album just seems like digital clutter.

It also raises questions about the initial stipulations of the sale, and from what I can see they seem to have relaxed the rules around making money off the back of the music to do this.

A Failed Experiment?

Given all that’s happened, how does all this reflect on the group’s initial goal to release music differently and draw attention to the lack of value surrounding music?

While it has undoubtedly succeeded in creating value around the music in one way (the price tag), it has also highlighted the downsides that come with this way of releasing.

You hand over sole control of the art to someone who has the money to afford it, at the same time preventing die-hard fans from being able to engage with it. Had the original purchaser been more philanthropically minded this could have played out in so many interesting ways, but instead, we’re looking at a strange attempt at squeezing money out of the project.

I still think there’s a better balance when it comes to the ownership model of music which is yet to be found. I wrote a couple of years ago about a type of generative music (not AI) as an example of how artists could create thousands of unique variations of their songs for private ownership, while still keeping the original in the public domain for everyone to appreciate, which I think is just one of many ways things could work.

Maybe things will change in the future, but for now, unlike the world of contemporary art, the ‘original’ in music commands little in terms of value.

Elsewhere…

▶️ Arena Tours Are Underselling

A lot of arena tours are underselling, with artists being forced to cancel or downsize. Things look unsustainable now that the post COVID bounce-back is over.

▶️ Is anyone making money on merch?

Samuel Hine has written an article for GQ on the decline of merch, but Ronnie Radke claims he’s taken millions in merch profits from Playboi Carti after he used the band’s without permission - so who knows?

▶️ Rihanna isn’t retired

The singer clarified that she’s not retired and is working on her new album. She hasn’t put out much music over the past few years but the singer still has more total sales than Taylor Swift which I didn’t realise!

▶️ Spotify put out their Songs of Summer predictions & a guide to new releases

You can check out the predictions for the songs of the summer, as well as an artist’s guide to marketing new releases.

▶️ Rick Ross has some strange stuff up for auction

Rick Ross is auctioning off some of the, in his words, “dopest, rarest luxuries”. Take a look and don’t ever question his taste again 🧐

Sharing is caring

If you liked this email don’t forget to share it with your friends 🙏

Whether it’s Drake’s bizarre new track confusing everyone, Daniel Ek’s viewing AI music as just more content to feed the machine, or the unexpected side effects of playing with AI music tools - keep reading to find out more…

Drake or Fake

A few days ago Toronto social media personality / rapper Snowd4y posted the song ‘Wah Gwan Delilah’ ft. Drake to his Soundcloud - it’s a remix of ‘Hey There Delilah’ by the Plain White T’s from years back.

It quickly garnered plenty of attention, and I, like a lot of people out there assumed it had to be AI generated. Even the Plain White T’s looked confused:

With Drake undoubtedly still a little bruised from the last few weeks, I assumed he wouldn’t release a track that further reinforces a lot of the criticism that has been sent his way recently.

But from what I can see it looks like there’s a decent chance it might be real, Billboard tried to investigate but the results were inconclusive, and Drake has raised further questions by posting support on his Instagram - although he merely acknowledges the track as opposed to verifying it’s him singing.

So, whether this is Drake doubling down with a Toronto-inspired inside joke that hasn’t landed too well, an attempt to move on from the beef by releasing something polarising, or just him playing along with the parody of it all, I’m sure we’ll find out over the coming days.

The more interesting aspect of this whole thing is how it highlights the grey area that now exists between unofficial and official releases, and how AI’s abilities can sow seeds of doubt without official sources to clarify the details.

AI can act as sort of a shield, in this example hypothetically allowing Drake to put out a song that’s a gamble, judge the public’s perception, and then decide whether to release it officially or play it off as a joke.

This raises the question - going forward, how will we be able to tell what’s real and what’s not?

There’s no question that we’re going to see an industry where more people are remixing artists’ work, cloning their voices, tweaking songs, and then getting everything cleared if it gets any form of positive traction.

This isn’t a new thing - there’s always been user-generated content in the form of remixes - but with the tools available today, stem splitting, AI voice cloning, and the ease at which collaboration can happen on social media - we’re likely to see the emergence of artists where a large chunk of their catalog is collaborative with their user base.

We’ll also continue to see shouts of ‘AI generated!’ more frequently when artists release music that falls outside the scope of their usual style or even release process.

The lines are becoming blurred, and it’s interesting to watch.

Daniel Ek Backtracks

After posting last week about the ‘cost of creating content being close to zero’, the Spotify CEO was forced to clarify his remarks amidst a flurry of criticism, with many artists criticising his lack of awareness and the realities of being a musician today.

It’s easy to take shots in this situation, and this wasn’t the best choice of words, but Ek is raising an important question, despite it being drowned out by the backlash.

Whether we like it or not, there’s going to be a lot of ‘music’ in the world going forward.

How do we define what is worth keeping and proliferating, and what should fade away into obscurity? Does it even matter?

So much art and knowledge has been lost to the world - how different would things look had it not been? Are there songs, books, or paintings that faded out of existence that are objectively better than the ones we regard as timeless today?

For the majority of human existence, culture has defined which works of art are preserved and revered, propelled by the stories that accompany them. But today we have algorithms that are shaping the culture and determining what stays relevant.

So who do we trust more with the responsibility of passing on the most important music for generations to come - humans or algorithms? What’s the right balance, or is there a better way entirely?

This question is a thought experiment in itself, with no right answers, but most definitely worth consideration.

Gambling on Music

An interesting thread was posted on Reddit detailing a situation that I could never have envisioned (thanks to Ryan at GarbageDay for the find).

The poster in question has become so obsessed with perfecting their song on the AI music generator Suno that they’re spending thousands of credits to do so, and are at the point where they need to stop ‘before it gets out of hand’, likening the process to gambling.

Is addiction going to be a strange side effect of generative music for a subset of people? You put your prompt together, click a button, and 30 seconds later you’re handed a brand new track, each time not knowing what it’s going to sound like.

It’s just another example of the same small dopamine hits that are scattered throughout our lives, and for some people, this will be enough to get them hooked.

And is this ‘gambling’ going to be a part of the music-making process going forward? When people talk about the amount of work that’s gone into making a track, will they be talking about credits and clicks? Hopefully not, but you never know.

If you’re intrigued by the 2500 credit song in question, the OP was kind enough to post a link - you can listen here 😬

Elsewhere…

▶️ Music festivals are dying in the UK, which seems to be part of a broader downturn in consumer interest, but elsewhere Electronic Music festivals seem to be doing great, along with the dance music industry in general.

▶️ Bring Me The Horizon marketed their new album with hidden clues to an ARG style website, which ended up getting hacked.

▶️ Is Taylor Swift’s success going to have any noticeable impact on the industry for smaller artists? And an interesting look at what it’s like to leave the cult of Taylor - you can’t please everyone I guess!

▶️ The Metric That's Ruining Hip-Hop? Does the genre place too much weight on first week sales versus long-term cultural impact?

▶️ Twitch launches a new DJ program to make it safer to livestream DJ sets.

Sharing is caring

If you liked this email don’t forget to share it with your friends 🙏

The last couple of months have been extremely chaotic for TikTok, and the fallout has left many artists wondering where TikTok should fit in their marketing strategy going forward.

It all kickstarted back when Universal decided to remove their artists’ music from the app citing unfair remuneration, resulting in millions of tracks being removed from the platform, and artists being caught in the middle with no control over the situation.

This bold move caused a lot of friction, with hundreds of artists expressing discontent. Still, despite all this, Universal stood firm in its position, even when Taylor Swift undermined them and put her music back on TikTok, able to do so due to her deal giving her ownership of copyrights.

This move highlighted the different perspectives that artists and labels have towards the app. Labels clearly have a much bigger interest in getting paid more for their catalog of already-established music, especially with a lot of older hits going viral.

Artists on the other hand view it as more of a promotional tool for their new music, accepting (begrudgingly) the low rates of compensation with the hope that any exposure carries over into other areas.

While this move from Taylor might have helped strengthen TikTok’s position, it all changed with the recent senate bill passing to either ban it or force a sale.

This places massive pressure on TikTok and it’s the sort of thing that has knock-on effects on public perception, with a recent poll showing that 58% of people in the US now believe it is used to influence public opinion.

This move is also likely to influence the perception of the app in other countries, with it already facing investigation in the EU for its impact on children’s safety and mental health.

Given all this, it’s not too surprising that last week we saw TikTok and UMG reach a new licensing deal including ‘improved remuneration’ for artists amongst other things.

Without being privy to the specifics, I’d hazard a guess that UMG managed to score a massive win given the amount of leverage they had.

TikTok had its back against the wall and has vowed to fight the ban (looks like they’ve already started), they will want to do everything they can to ensure the app retains dominance in a marketplace worth $16 billion to them, so paying higher royalty rates to get UMG on side is a no brainer.

While this agreement brings a powerful ally back and keeps the app as engaging as it has always been, it still leaves the future of TikTok up in the air and raises a question about how the music industry should approach the app from now on.

The bill gives ByteDance until 19th January 2025 to sell or face a ban - they have stated they would rather close the app than sell it.

This has got a lot of artists wondering if this should change their approach to TikTok, and whether it makes sense to be investing effort into building an audience that could disappear in a year.

While this may seem like a new problem, it’s actually one we’ve seen over and over again - and it all boils down to platform risk.

Negating Platform Risk

Just like when MySpace lost relevancy and artists lost access to fans, or when Facebook changed its algorithm to limit audience reach, or when Twitter shut down Vine. We’ve seen this film before.

Whether TikTok gets banned or not, this whole situation should highlight the importance of reducing dependency on one platform or channel.

When artists are too reliant on one channel they risk having the rug pulled from under them with little control over what they can do about it.

The best way to reduce this risk is by directing fans to channels they have ownership of, with the most obvious being an email list, in addition to diversifying activity across a few different platforms.

The good news for artists is that diversifying across channels is becoming easier to do. With every social platform striving to add on the features of its competitors, essentially becoming the same apps, repurposing content becomes way more viable.

Rebalanced Priorities

The changes these last few months should hopefully rebalance the priorities of artists and their teams. For a couple of years, TikTok was the darling of music marketing, turning unknown acts into superstars overnight. Marketing efforts were focused on understanding what format worked on the app and doubling down on whatever worked.

But when something works well, everyone starts doing it - and when everyone’s doing it, it suddenly doesn’t work as well anymore.

Everyone became reliant on TikTok, but we’re starting to see the signs of discontent - the results aren’t as consistent or scalable.

The resulting impact of this looming ban will be an increased hesitancy around overfocusing on TikTok, and as the industry places more focus on superfans, a natural side effect of this will be the development of tools and strategies to help move people along the funnel from casual listeners to engaged fans.

For artists, the focus going forward should be on how to use social media to gain exposure, while simultaneously ensuring that exposure is directed into owned channels that offer direct access to fans.

Other Stuff

Instagram Algorithm 2024 Changes — More Original Content, Better for Small Creators

In what seems like a move to make the app feel more like TikTok, Instagram is making a change to its algorithm to favour the creators of original content over content aggregators that repost content. This gives more of a chance for smaller content creators to blow up randomly in the same way they do on TikTok.

Universal Music Group sees a future where 1 in 5 streamers pays for a ‘super-premium’ tier

While the major labels keep talking about the priority of superfans, it will be interesting to see how this plays out. This ‘super-premium’ tier suggests this would exist inside a streaming service, which still raises the same issues that come with not having full ownership of your channels.

The False Drake v Kendrick Rumour

There was a rumor doing the rounds about UMG getting involved in the Kendrick vs Drake beef, alleging that they were making demands to attempt some damage control - which has subsequently been denied by Universal.

While the drama has been undeniably entertaining, and is responsible for a serious number of streams, there are surely some nervous heads at Universal wondering quite how all this is going to play out once the dust settles.

Having struck a massive deal back in 2021 the label won't want to see the longevity of Drake’s catalog impacted negatively - no doubt the crisis management team is working overtime rn.

Did AI Just End Music?

AI is moving quickly, it feels like it’s all so new, so it was interesting to see some examples of people trying to create the same tools way back.

Subscribe now