Suno AI Long-term Test – what to do with the (many) songs?

100 Songs per Month – No Problem!

1000 Songs per Month – Not Really.

The flood of AI-generated music is impressive. Anyone active on social media regarding this topic is confronted daily with countless, often poorly thought-out AI tracks. The dissatisfaction there is high, as the sheer volume usually causes songs to get lost. Essentially, every AI user and musician tries to grab streams, which rarely creates a genuine audience. I am no exception; streams depend almost exclusively on one's own subscribers, followers, and friends, who first need to be gained. By the end of January 2025, 10,000 AI music tracks were published daily on Deezer alone, and this number is likely to increase significantly with quality improvements, potentially stabilizing at around 5 million commercially released AI tracks per year.

The problem of oversupply is not AI-specific; it essentially affects every artist and especially musicians with their own songs.

How much should an AI-generated music track cost, or generate in streaming revenue long-term?

If you calculate the effort of creation, which currently stands at about one hour per minute of final AI music content, and you assume an hourly rate of 85 EUR net, then commissioned work per track for shorts would today be around 100 EUR, and for songs with vocals between 300 and 500 EUR. This would cover licensing costs, time for creation, mastering, cover, lyrics... So, this can certainly be offered as a service for exclusive use (individual commissioned work). Most AI music creators will not be able to make a profit from this.

The Fascination and Challenges of Suno AI & Co.

Anyone using Suno AI, Udio AI, and similar tools is initially thrilled: a song is provided within seconds – something even experienced bands couldn't produce so quickly. With each new version, the quality noticeably improves. Nevertheless, known weaknesses such as “Shimmering,” cut-off songs at the beginning and end, incorrect word emphasis, etc., remain a problem that requires careful control. However, with enough time and credits, fragments without these errors are repeatedly created.

As already described in the article “Music Workflow,” the real challenge is not the creation of songs, but their finalization.

Metadata and Distribution – A Question of Workflow

In my workflow, I ensure that the music files contain all relevant data in their ID3 tags. This means that important metadata is delivered directly. But how do you distribute the songs effectively? The “Music Workflow” also offers approaches for this.

DistroKid and Feiyr – Music Distribution Services

If you want to distribute your songs professionally, you can consider services like DistroKid or Feiyr. Both platforms offer ways to publish music on streaming platforms like Spotify, Apple Music, and others. While DistroKid relies on a low annual fee, Feiyr operates with a pay-per-release model, which can be more flexible for certain projects.

Long-term Goals – An Own nuonu-Radio

Long-term, I plan a nuonu-Radio where all tracks play on a loop. To implement this, however, at least 50 hours of original material would be required. This vision opens up exciting opportunities to build a consistent audience and present one's own music world.

The Platform as a Business Model

The development of the platform itself is a business model for me. In addition to the free songs, I am working on plugins and apps that could be interesting for many users. These are constantly being developed further and will be presented in detail under the Code menu item.

Additional Revenue Streams through Training and Commissioned Work

Additionally, training in the AI field and improved customer loyalty through high-quality content are conceivable. Smaller commissioned works for creating specific music tracks could also represent a source of income. Therefore, the mindset of having to earn money directly with AI music tracks is not crucial for me.

Inspiration for the Reader

The implementation of these ideas is feasible for everyone, provided one is willing to invest the necessary resources. It is important to understand that, in addition to time, licensing costs and technical tools are needed to achieve high-quality results. Nevertheless: with the right approach, anyone can successfully start and implement their own music projects.

Monetization / Streaming

Of course, a small monetization would be great. However, this requires many subscribers, which are difficult to acquire purely with music. With low reach and short songs, it's very hard to meet the necessary minimum criteria. This would be much easier with decent videos for well-attended events or even political commentary. So, it needs topics that interest people, something they enjoy watching repeatedly and for longer periods. In earlier times, we had many 100k views but no monetization. Nearly 1000 fans from that era are unfortunately not enough today. But I'm consistently staying on track and giving social media a chance. I've long been critical of them, mainly because I don't get recurring visitors to older content there. A thread runs out within a few days, and posts expire. Therefore, I am now focusing on quality content, collections, apps, plugins... less competition and yet products that many users need.

Free Downloads as a Strategy

I have therefore decided to make all tracks available for free download. This strategy lowers the barrier to purchase and offers the potential to build a small fanbase. Should someone wish to use the music commercially, they can acquire a license at any time.

If a track actually proves to be a hit, there remains enough room to act accordingly.

Creative Commons as a Sensible Alternative

As musicians, we have relied on Creative Commons content for many years. There, too, different uses are distinguished, so it seems logical to license quickly created AI music under such a license. Platforms like jamendo.com are very well suited for publishing and monetizing music and CC. Licensing and CC cannot be published simultaneously with ContentID. Buyers choose CC content because they can then monetize their videos themselves. So, it's either one or the other; otherwise, you will sooner or later receive a warning from the platform operator. Those who create film scores are usually better off with CC.