Blog Music

Gen AI music platforms are niche among consumers — is being *too* easy the problem?

Cover image for Gen AI music platforms are niche among consumers — is being *too* easy the problem?

Photo: Mary Eineman

Photo of Tatiana Cirisano
by Tatiana Cirisano

We all know how hype cycles work: a new technological advancement arrives, and industries rush to apply it to anything and everything. In these situations, it is useful to ask: What problem or need is this technology actually addressing? Usually, the answer is where the real opportunity lies (or, depending on where you sit, the real disruption). 

Generative AI music tools pose the same question. The needs-to-fill for businesses and creators are becoming clear. But what about for consumers?

MIDiA’s Q1 2024 consumer survey indicates that on the whole, the top reason consumers use AI tools is to get information, followed by creating content like images and writing. No wonder, then, that ChatGPT is by far the most-used generative AI service. Meanwhile, despite Suno’s eye-popping funding round, it is important to remember that music generators are still niche consumer offerings. According to the same survey, only 5% of consumers have ever used gen AI to create lyrics or music. This is likely to grow somewhat, but the question is how much? Founders of consumer-geared gen AI products envision a world where everyone is a music creator, but a recent Billboard article questions how strong consumer demand actually is, echoing a wider debate at play in offices across the industry.

The “IKEA” effect

AI music generators could serve some functional needs for consumers, like generating a song for a friend’s birthday or making memes (an essential means of connection for humans in 2024!). Let us also assume that at some point in the future, generating songs becomes a regular means of self-expression and communication, akin to posting a photo of your breakfast on Instagram Stories. We expect music to have its ‘Instagram moment’, but even if this use case grows relative to today, it may not get much bigger relative to other gen AI use cases, like text, video, and images. 

Unlike with professional creators, who are focused on finished products, gen AI music companies are ultimately looking for consumers to create for the sake of creating, which requires giving them a sense of creative fulfilment. This may be harder to achieve than it seems. A good reference is the so-called “IKEA” effect. In a 2011 study, Harvard researchers found that consumers tend to value objects they played a role in creating more highly than objects made by others. The same theory has been used to explain why a Betty Crocker “just add water” cake mix in the 1950s was outsold by a version requiring customers to additionally crack an egg. The historical accuracy of that case is murky, but there are myriad other examples, from Build-A-Bear to HelloFresh.

Of course, consumers are far more strapped for time and attention today than in 1950 or even 2011. They may be more likely to just pay for a finished product — just ask the hundreds of TaskRabbits who get paid to build other people’s IKEA products. Yet the theory could hold some weight when it comes to gen AI music. Maybe what gen AI products need is the musical equivalent of Betty Crocker having customers crack an egg.

All play and no work?

In other words, maybe consumers do not get enough fulfilment out of text-to-audio generators, because their near-complete lack of effort makes it harder to feel invested in the thing they are creating. Eventually, the act of prompting is likely to become an art in itself, but so far, it may not be sticky enough for consumers to see AI music generation as more than a novelty.

Meanwhile, AI-powered tools that allow users to modify songs — such as changing the speed or vocals, or remixing the stems — might. Here, users feel like they are playing a tangible role in the thing they are creating, not to mention building atop the cultural context that those songs already hold. This may explain why consumers are around three times as likely to be interested in most audio modification tools as to have ever used a music generator (MIDiA Research Q1 2024 consumer survey).

This does not mean that music generators cannot be creatively fulfilling for consumers. But for this to happen, generators need to nail the balance of making the creation process feel simple and easy, while still giving users enough control to feel invested in the thing they are making. It is clear that today’s consumers expect to go 0-100 with their ideas quickly, but just how fast? That is the question that consumer-geared, music AI products should be looking to answer.

The discussion around this post has not yet got started, be the first to add an opinion.


Add your comment