Technology company Song Sleuth recently agreed a new partnership with Warner Chappell Music.
It’s a sign that the industry sees a growing opportunity in revenue from user-generated content - a multi-billion dollar market that Song Sleuth is working to better monetise for rights-holders.
Song Sleuth is working for all Warner Chappell songwriters, covering more than one million musical works, to find UGC and derivative content that is currently unclaimed or facing other monetisation issues on such platforms.
The tech firm uses its AI and machine learning platform, UGSeeker, to find UGC that is currently being missed by existing detection systems on digital platforms - and therefore not being monetised.
Here, Song Sleuth CEO Jordan Gross opens up about his vision, their tech and the mission to find the money…
Missing or incorrect data has long been a problem in music - how important is fixing that going to be in the years ahead?
“Incorrect data such as missing or incomplete metadata, as well as, to take one example, ISWCs not matching work IDs is causing billions of dollars annually in lost or misdirected revenue, so fixing the data problem is an incredibly high priority for the industry. Because of this, we actively work with our partners to improve data quality and unlock tremendous value for them.
“When it comes to UGC, the issue is twofold: first, data is formatted and moved around in an unstructured and antiquated way. Secondly, there is no single central authoritative source to discover who owns what; rather there are lots of databases with conflicting ownership data. These factors, combined with ineffective methods of payouts, have created a perfect storm of issues that ultimately leave artists and rights holders out of pocket.”
Can you estimate what scale unpaid UGC royalties reach globally in 2023?
“The scale of the opportunity that we've identified so far reaches nine figures annually. In the twelve months between June 2021 to June 2022, the estimated value of payouts to the industry from UGC was around $2 billion on YouTube alone. So if you think of YouTube as being one of dozens of platforms online where there's UGC, we are looking at significant sums here.”
How is the company establishing itself in terms of music industry partners?
“We recently announced our first major deal with Warner Chappell Music and have a number of other deals in the pipeline, more of which will be revealed later this year. Integrating into our partners’ existing systems and processes is something that has to be done very carefully to ensure good data and billing flow - our process has been to run extensive trials before we enter full service with our partners.”
What do you think of the current debate surrounding AI and music?
“It seems like the industry can’t decide if AI is a threat or an opportunity. And, of course, the answer is that it’s both. We’re already seeing a paradigm shift in the creative process of making music. Everything from songwriting to recording, producing, mixing and mastering is already being impacted by AI and it’s easy to see how creativity is at threat.
“By the same token, at Song Sleuth, we use AI to trawl through vast amounts of UGC and make sense of it, including using machine learning models to automatically predict which videos include our clients’ content and the value of claiming that content. We also put a large focus on the quality of our models and, in particular, the training data used to fine-tune each one. We’ve put significant resources into building a team of operational music experts that manually review every piece of data that goes to our clients and as a result we have some of the highest quality datasets in the industry. AI is solving a problem that was otherwise effectively unsolvable through human intervention alone, so we are seeing it can be a real asset to the industry especially when paired with a little human touch.
We’re already seeing a paradigm shift in the creative process of making music
“Where things start to get really complicated around AI is copyright. Who’s allowed to do what? Who has the right to give the permission to do that? For example, if an AI system produces a new melody, a new track or even an entire song, who owns the copyright? Some say it should be those who developed the system, while others say it should be those who own the copyrighted material used to train the system. In the end, it will likely be some combination of at least those two parties.
“In any case, it’s exciting to see how copyright law evolves in this area. When we have big disruptions in the music industry, different elements move at different speeds – the technology advances at one speed, the legislation and business practices and infrastructure moves at a separate speed. These are complicated issues, but we will get there if there’s enough of a business opportunity there.”
How is the company financed, what plans do you have in terms of further investment?
“Song Sleuth is financed privately. We’re in a position currently where we have no plans to raise further money or raise money from venture capital. We want to be able to make decisions for the business quickly and be at the cutting edge, so we value a flat company structure, and that includes a small, agile board.”
What are the plans for Song Sleuth in 2023 in terms of expansion and new services?
“We’re looking at launching our flagship product, UGSeeker, on more social platforms such as TikTok towards Q3 this year, and we are always improving our algorithms and efficiency.
“We're going to be doing more work with artists and artist managers on Live Link, which is a service whereby we capture the content coming off of touring in near real time. By monitoring musical tours and performance dates, music rights-holders can claim and monetise the content from the moment that it is uploaded to YouTube - content that in some cases may live online for years before it’s monetised, if it’s even claimed at all. We're also looking at other data issues in the music business such as mechanical rights. We can't talk about them publicly but there's lots of exciting plans in the works.”