Sound Search #5: Jason Feinberg, KOSIGN | Kobalt Music

Jason Feinberg learned to code at eight, taught by his father, a CPA and software developer. Musician, audio engineer, college radio DJ, software engineer: that combination rarely sits in one person, and it runs through Jason's whole career. Since around 2001 he's worked in digital marketing and strategy at labels including Concord, Epitaph, and Universal Music Group, and run artist marketing at Pandora. He also lectures in the Music Industry major at UCLA's Herb Alpert School of Music. Today he's Senior Vice President and Head of KOSIGN, Kobalt's publishing platform.

We talked to Jason about why music publishing is the most complex corner of the business, where the real innovation is actually happening, and why he thinks the industry's biggest competitor is a lack of education.

What's something about how KOSIGN works that people are usually surprised by?

It's not something we hide. It's the opposite: it's what we promote and talk about all the time. When you join KOSIGN, you get access to Kobalt's administration infrastructure. KOSIGN doesn't have its own tech stack. We don't have our own deals. We don't have our own administration, royalties, or registration teams. We're a portal straight into Kobalt's world-class infrastructure and deals, including access to AMRA, the Kobalt company that handles global PRO collections.

The same suite of publishing administration tools that Max Martin, Childish Gambino, Phoebe Bridgers, Paul McCartney, and every other Kobalt client gets are the exact same tools you get when you join KOSIGN.

What makes music publishing such a complex business?

Complexity is probably the single most defining word of music publishing. There's a long list of reasons, but the summary is that much of it runs on a 20th century framework trying to keep up with a 21st century streaming economy that generates millions of transactions every second. Given how much information, connectivity, and organization that system needs to work, it wasn't really built for the world we live in now. That's why a lot of publishing companies have invested in technology and ways of working that let them keep up and even innovate. I'm biased, but I believe Kobalt is the leader in that, and it's why Kobalt was founded in the first place. Twenty-five years ago, Willard saw that the 20th century framework wasn't going to keep up with the 21st century reality we're in.

If you think about why it's so complicated, especially compared to the recorded side, or live, or merch: within every song you have the recording and the composition. Publishing focuses on the composition, and a composition can have multiple writers, each with different publishing companies, each with very different tech stacks, deals, and infrastructures. So right away, the people behind it are far more complicated in arrangement than some of our other revenue streams. On top of that, within publishing there are multiple royalty types, and each has its own divisions. There's mechanical, there's performance, and performance has a publisher share and a writer share. Now you're drawing a forks-and-tree diagram of all these royalty streams that add up to the composition, then you divide it by the number of writers and their shares. One writer can have 2%. Another can have 91%. That alone adds complexity.

Then you have multiple organizations collecting royalties in different parts of the world. It's global, and they all have to be synced up and able to communicate. They each have different timelines, different systems, different terms. Then you get into things like black box accounting, where if we don't know who to pay, there's an entire separate way the industry approaches paying that out. And none of it works without proper data hygiene, the ability to organize information so it's accurate, comprehensive, and in the format each company needs. By the time you jam all of that together, you have an incredibly complicated system that even a company as sophisticated as Kobalt has to aggressively work at to keep up with how the industry is evolving.

What's the most exciting publishing tech flying under the radar right now?

A lot of the innovation in publishing tech, and in the industry overall, isn't what clients see or what surfaces in the final product. The innovation is happening in the ways of working, the operational excellence and efficiencies that happen behind the scenes.

We talk often about speed to revenue: collecting and paying clients as fast as possible so they can reinvest in their businesses. Any innovation that shortens time to revenue benefits everyone. So removal of friction is a major piece of this. KOSIGN has been particularly effective at making it as frictionless as possible. There's more friction to remove, to be clear, but we've removed a lot of it in how a client can apply, get accepted, get set up, and deliver their work before we hand it off to Kobalt for admin. We've gotten that process down to single digit days, which in this industry is revolutionary. A lot of the time you're talking about months of redlining contracts and a big process just to get set up. We productized it: a simple web interface, standardized terms and conditions, no red lines flowing back and forth. We boiled the service down to what we feel is right for the market we cater to, and the market's responding.

Another one is global micro licensing. Sync revenue is typically reserved for the bigger artists and the well-known, longstanding songs. But the idea of micro licensing platforms that are fast and almost self-serve, where someone can license a composition or recording very quickly with clear, customizable rights, opens up another revenue stream for writers. It's the same framework as the back office side: removal of friction, making it faster and easier to find what you like, pay for it, and use it.

And there's enhanced data matching using more modern computing methods. Instead of making claims based on limited information and hoping it doesn't go into conflict, there are now a lot of data sources and a lot of logic. Based on a little core information from the client, we can supplement it to make sure we're transmitting the necessary information in every format every partner around the world needs, so it's matched properly and we maximize the royalties attached to that work. A lot of works get under-monetized simply because the data is incomplete and not all systems can attribute revenue-bearing activity to the composition. Smarter matching technology has made a material increase to royalties collected, with no additional work from the client or from us. Every one of these things ultimately puts more money in a writer's bank account faster.

You talked about publishers and labels needing to get their catalog in order. How do they do that?

Step one is to collect and organize information as early as possible. Every day that goes by where you're not tracking the terms of an agreement, who owns what, who's on the track, that basic information gets harder to gather. Think about the recording studio: how much information is technically already available at that exact moment. There are software platforms meant to be used in the studio or the business affairs office, so that once something is recorded, written, or agreed to, you capture it officially right away. Because every day, unfinalized terms get fuzzier, attention goes elsewhere, and people move on to the next job. All of it makes it harder to go back retroactively and get comprehensive information about who did what, who owns what, and who's owed what. That's the fundamental information royalty collection needs. The earlier everyone documents it, whether that's metadata or contractual terms, the better everyone does, and the faster to revenue.

Part of it is also about systems. Google Sheets, for all its uses, is not sufficient to run a music enterprise asset management empire. I've seen it, and I've been part of moving people to something more robust. It makes sense for smaller enterprises, but at scale there's no reason not to move from messy, disparate data to a centralized, organized single source of truth in a real data management platform. They exist at every complexity level and every budget level. Maybe 20 years ago, even 10 years ago, this information wasn't as necessary and portability wasn't as much of a daily demand. At this point there's no reason not to collect information early and store it in an official, portable platform that makes plugging into new opportunities seamless.

You've sat through plenty of music tech pitches. What questions do you ask to spot one that actually delivers value?

I've been on both sides. I've made the pitch, and I've received a pitch a trillion times, especially at places like Universal, where everyone wants to get involved. There's an immediate fork in the road at the start of any of these conversations: is this a solution that solves a problem, or a solution that's looking for a problem?

The platforms that catch and keep my attention, and I'd argue the market's attention, are the ones that solve, or at least contribute to solving, a very real problem. There are very real problems to solve every single day in this business. But the nature of modern technology is that anyone with an idea can build a solution, which on one hand is fantastic. On the other, it rapidly accelerates interesting ideas that don't actually solve any real problem. So the pitch goes out looking for the problem the technology can solve. That's not always ineffective. Plenty of great businesses started with an idea and then found who needed it. But I hate to say it: the majority of pitches I've seen in my career are solutions looking for a problem, without actually knowing that this client has this issue and my solution solves it in a way that's feasible, can be implemented, and aligns with client needs and business realities.

The other question is: does it scale? Do the economics remain viable, or ideally improve, as the business grows? That can make or break an implementation. You might have a solution to a problem, but if it doesn't scale cleanly, if every new plateau means throwing more resources at it and the economics get worse, then I'm not incentivized to scale. So it comes back to: have you put yourself in the shoes of the ideal use case? Are you user zero of your own product? Have you lived the experience of someone believing in and implementing this long term? If you have, and it scales, and it works, and it solves a problem, then call me up. Because then I'm listening.

What's the most common mistake music companies make when they buy or build tech?

Solving yesterday's problem is very common. Or even solving today's problem: I have this urgent thing, I just need it solved, versus thinking about a solution for the longer term. Sometimes you have no way around it. If the thing's on fire, you put it out. But a lot of the time, people align what they buy or build with the business they're running today, and things move so quickly that by the time you've built or bought it, it's already becoming out of date, and now you have to solve the next problem.

Then there's not designing with the user in mind. We've all had tech forced on us that's supposed to help our function and it's terrible. It doesn't solve our problem, it doesn't work the way we work, and there's a better solution out there. The user flow, the onboarding, all of it is against the realities of who's going to be using it. To me that one's inexcusable, but it happens all the time. It's a communication breakdown, a workflow and process and accountability and ownership issue.

The third one is not being able to separate hype from value. I talk about this in my class a lot. Hype is the narrative around why something is interesting and valuable, and often it's true. But hype and value aren't directly related by nature. Something can be hyped up and have no value, and something with a ton of value can have no hype around it. What we often see is that something works for one person or one campaign, the hype explodes, and everyone thinks that's now the solution to their problem. Maybe it is. Or maybe it doesn't provide you actual value. If you're buying or building based on hype without value, you end up having to resolve that problem immediately.

If you could put one change on the agenda for everyone running publishing or catalog work right now, what would it be?

Be educated, and learn more every single day. Never consider your education or your understanding finalized. Things move too fast and they evolve, and opportunities to solve longstanding issues still remain. Music publishing is one of the most complicated aspects of the industry, if not the most, but it's also a key revenue engine. A lack of understanding costs real dollars from day one: not understanding royalty streams, not understanding the data infrastructure you need to collect and maximize them, not understanding what happens to your money if you don't do these things, where it goes, and how long you have to collect against it. These are just a few of hundreds of examples where a lack of education is the fundamental problem we face.

The complexity is minimized when you have someone like Kobalt or KOSIGN doing your administration. We handle 99% of that complexity, as long as we're handed the right information. But to get handed the right information, and to come to us in the first place, the client has to be educated on how this works, what they own, and what they're trying to accomplish. That applies from a songwriter just getting a little traction all the way up to sophisticated companies with multiple layers of rights. I often say to my team that although we have competitors in the marketplace, our biggest competition is a lack of education. Because once someone's educated, I firmly believe our value is crystal clear.

This interview has been edited for length and clarity.

Sound Search is AIMS' interview series with music professionals on how technology is changing the way we discover and work with music.

Have someone in mind we should talk to? Reach out to us on LinkedIn.