The first thing is that the mediums are very different. Music is a mostly linear audio product, whereas Mario levels are an interactive experience with audio and video. This affects a number of aspects, most notably the ways in which the items can be browsed. In a musical search, I can basically look for songs based on an artist, genre, track name, or album name, but that’s about it, and if I need to know more about a track (for example, to find if it’s a cover of another track with the same title), I pretty much just have to listen to it. With Mario Maker’s Course World, while there is no text-based searching option, I can scroll through levels to find levels that look good based on a small preview image, completion rate, stars awarded by other players, and more.
Another important distinction is that the way the content is being consumed is different. In Mario Maker, the content is fairly disposable-even if a level is really good, I’m probably not going to play it more than once or twice before moving on. In music, there are really two different consumption modes-one is finding music you know you already like, and the other is finding music that is new or at least new to you. Digital music services have historically done the first case reasonably well, but struggled with the second. Mario Maker is more analogous to the second case, so the question is why does this work there? Well, going back to the browsing aspect, besides the previously mentioned level information, one other aspect that really helps in level discovery is the ability to restrict results to recently created levels. Digital music services will generally have a front page with featured new major releases, but finding anything deeper is very difficult.
Based on this information, it seems like a music search with date filtering and detailed per-track evaluation and interaction would work well, and as you may already be thinking, such a service does exist. YouTube, despite being billed as a destination for visual content, has become the de facto destination for finding music, and it’s easy to see why. In addition to being able to find old and new songs quite easily (tags and description text really help), users can share thoughts on individual tracks (much like leaving comments on a Mario Maker stage). Personally, I love reading what folks have to say about each tune, and it really makes the experience richer in a way that digital services haven’t been able to match.
So, can a music service adopt this style? Sure, it’s just that for the most part, these services are still organized with albums in mind, which also drives their interaction model. Shaking free from that won’t remove all their disadvantages, but I think it will help them compete. We live in an age where consumers expect to be able to share their thoughts on creative works, and it’s not surprising that the service with the best tools has been the most successful.