While it's nice to have a universal ideal when it comes to music tags and descriptors, having done a lot of research on the topic, every sync/production music company has a different interpretation and definition of music taxonomy. Ultimately, businesses need to be focused on how their audiences search for, find and use music and this needs to feed into the tags that you set. You must also be mindful that genres and themes are constantly growing outwards, so futureproofing your catalogue and ensuring it remains relevant in the constantly evolving chaos of music, is a smart strategy that all catalogues must consider.Â
If you’re only ever hitting the obvious genres and not getting any more granular, the gap between the user and the music they are looking for will continue to get wider as your catalogue grows, so before you automate your tagging process, the first step is to map out your ideal taxonomy.
Here’s some tips on where you can start. Â
Step 1 - Remove Unnecessary Tags
 If you've been asking musicians to add in their own tags as they upload their music, or you've tried to keep the same taxonomy as when you initially launched, there will be most certainly be a need for a refresh.
Firstly, note down every single tag you currently have in your catalogue and health check each tag. If there are less than 100 tracks against a tag, is it really worth keeping? If you have hundreds of tags, is it getting a little too complex? Use this as an opportunity to cut everything you don't need so that your users have a simpler way of navigating your catalogue.
You should also schedule regular reviews for your taxonomy so the next time a Lil' Nas X comes along and shakes up genres, you can integrate this as part of your next taxonomy review and update accordingly.
Step 2 - Add in New Sub-genres and Segments
Once you’ve listed out all of the tags that you definitely want to keep, then it's time next to look at the tags with large numbers of tracks. Believe it or not, these can be as useless as tags without many tracks. If for instance, you have 5,000 tracks tagged as Hip-hop, could you segment these into Old School Rap, Contemporary Hip-Hop and Trap?
Segmenting large genre sets is not only going to improve your user experience, but it will give more musicians a chance of being found.
Step 3 - Think Audience-first with Extra Categories and Themes
In the past, it's simply not been possible to tag every single part of a track manually with every single element, but with AI automated tagging, we can teach our AI to call out anything, as long at there is a common thread in the music. Are there any extra themes, or tags you could include that users are going to want to likely segment by? Male or female vocal? Language? Instrumentation? Live or electronic drums? The complexity of the playing? Use cases such as trailers or stings for sync music, or gym and party music for streaming?Â
Think about how your audience would like to search in an ideal world. Maybe for production music, they would love to search with descriptive words too. The more they can drill down with different categories, the better their search experience as a whole.
Step 4 - Map it and Tag it
Now you have your taxonomy mapped out, it’s time to start tagging. While you can take a manual approach, this often is a laborious task that takes a long time to process track by track. With Musiio Tag, we can custom train for any tag and taxonomy and can tag anywhere up to 1 million tracks in a day from just 0.05p per track, so reach out to us for a quote and let's get tagging.
Download our off-the-shelf taxonomy here for free and contact us if you have any enquiries, we’d love to hear from you.