Music industry problems – AI as the solution

November 29, 2019

Recorded digital music industry is where it all goes differently these days. With the total revenue of > $11 bln in 2018 it represents the fastest growing share of the entire music market. That is the reason so much pressure exists here.

Platforms look for ways to provide more content with moderate pricing to attract users. Look at Spotify premium subscription price of 9.99$ – 10 years ago you would spend the same amount to buy an album (or nothing by illegally downloading it).

Listeners scale and diversify music consumption. Number of music streaming subscriptions grew by 40% from 198.6 mln in 2017 up to 278 mln. They also add music to more modern entertainment categories like gaming streaming or Instagram stories soundtrack creating.

Photo by William Iven on Unsplash

Artists care about royalties hard enough to force U.S. copyright royalty board to submit a 44% increase this year.

Regulators are not any more chill. Consider recent case of Peloton in the U.S.: the fitness company was charged with a $150 mln dollars fine for using more than 1000 songs without permission.

Meanwhile, other jurisdictions do not leg behind. For instance, regulators of the EU introduced new copyright rules with Article 13 that came into force in June this year. These new rules shift responsibility for copyrights infringement towards business.

For music streaming services the major struggle is to achieve and maintain operational profits, which is now more complicated with greater royalties. For instance, in 2018 for Spotify only the fourth quarter reported operational profits. In Q1 2019 there was again a loss. And that happens with ~ $0.00473 per play. For Apple, the same pay is $0.00783, which makes it more attractive for artists, but does not move the company closer to profitability. (Though they do not strive for it) Whatsoever, listeners are barely ready to pay more, while all the streaming services platforms will have to spend more within the next 5 years. What if they could decrease the cost of their content? Music listeners do not always need their tracks to be prominent author content. Sometimes you just need something nice to sound as background, and that is the moment Mubert streams may come into play.

Video streaming and hosting services is the other market we look at. A million users watch Twitch streams game at any given moment with DAU of 15 mln. Many of them play background music. Some do this to replace pestered in-game music. Others want to customize their experience. To stream music publicly you need a license. Twitch has an elaborated policy with temporary and permanent bans as the main instrument for those, who ignore copyrights. To keep up with legislation users look for other sources of music.

They might pay 15$ a month to a bulk music provider such as Epidemic Sound or Legis Music. Else they can turn on copyright-free music streams you can find on YouTube. What if these streamers could simply select background music genre right within the App with much lower fee or no fee at all? And have it always fresh, versatile and with no time limits? Same goes to people who upload ready-made videos as similar rules apply to them. Content creators are usually limited to 10 – 30 seconds of playing author tracks (depends). Wouldn’t they be happy to be able to create 100% tracks that work best for their videos?

Photo by Kon Karampelas on Unsplash

We also consider social networks as one of the primary markets for us. Currently 2-3 bln people are combined DAU for social networks like Facebook, Snapchat, Twitter, WeChat, Instagram. Take the latter one. Its DAU estimated at 1 bln. And 0.5 bln for Stories. Meanwhile only 15-20% of the users (according to our approximation) have access to music in their regions. This is due to the need of purchasing different licenses for the same record in different countries.

Imagine now all of them have an opportunity to craft own track to accompany a story with a selected mood, pace, genre, etc. And all this does not require artists to do anything (apart from fulfilling our samples library). Now think of music for FB postcards. Or Snap video streams with some nice background music. Of course, even now users can find copyright-free music, crop it, download and attach to a story somehow. But isn’t it too complicated?

The least suffering, but nonetheless perspective market for us is voice assistants, especially when used with smart home speakers and other similar systems. (The one that grew up 82% from 2018 to 2019). More than 1/3 of all smart speakers’ users are DAU for music streaming services (~70% for MAU), which makes it the most popular feature for the segment. However, not in every region you can get connected to a streaming service. Apple Music works in 110 countries. Spotify has around 70 countries on the list. And even if you have access to a service you find appropriate to you, you still need to pay for the subscription. Simultaneously, quite often you don’t even need music to be distinctive. A nice rhythm that suits your home activity, let’s say yoga, would be more than enough.

Photo by Jonas Leupe on Unsplash

This isn’t an exhaustive list of segments we think will find generative music useful. It can be literally anything from public spaces like cafes & restaurants to fitness apps. With the existing technology we already provide, users can adjust music to genres, pace, mood they like. But the customization possibilities are infinite. We continue to work for functionality expansion.

Related Posts