Over the past two decades, Toronto-based LyricFind has developed a global footprint in the business of licensing lyrics and in delivering to rights owners a new stream of revenue that was hardly tapped into by music publishers 20 years ago.

LyricFind was co-founded in 2004 by Darryl Ballantyne, who serves as CEO, Mohamed Moutadayne who acts as CTO, and Chris Book, who sits on the Board of Directors). All three were attending the University of Waterloo, and they started building a web site with lyrics. The story could have ended 20 years ago with the three partners sued into oblivion by music publishers.

But they did something unusual for a start-up — they went legit and started seeking licenses from music publishers in order to be able to license lyrics to third parties, in particular DSPs. At the time, sites offering access to lyrics were ubiquitous online, but not a single platform was licensed. At the same time, demand was high as lyrics usually appeared among the top three in Google searches.

“You have to remember that lyrics were not licensed before them,” explains Ted Cohen, the Los Angeles-based music tech specialist. Cohen recalls that in 2003, he spoke at Canadian Music Week in Toronto, and at the end of his session, a young man came to see him. They went for a coffee and Darryl Ballantyne asked Cohen, who was at the time dealing with digital at EMI, if he could intern at EMI. “He seemed to be a very smart and personable person,” says Cohen.

Eventually Ballantyne went to LA for an internship and the bond between him and Cohen grew stronger. Ballantyne went back to Toronto to finish his degree, and after he graduated, he contacted Cohen asking him if there was a job opening at EMI, but at the time Cohen had left EMI.

Ballantyne then told Cohen that he and his university pals had started a lyric site but wanted to do it as a business and asked for help and guidance. The idea was to aggregate all the lyrics possible from various music publishers and then sub-license them to third parties, and generate a new licensing revenue.

“I introduced him to EMI Music Publishing, They asked me what I though. I said — this is a brand new revenue stream,” says Cohen. EMI — now part of Sony Music Publishing — was first major publisher to commit, then they went to talk to Harry Fox Agency (and got a deal) and then to Universal Music Publishing and Warner Chappell.

Cohen remembers a major hiccup on the road to licensing when one of the major publishers said it had received $250,000 from a competitor to not license LyricFind. Eventually the publisher signed a licensing deal. “If the publisher had taken the money and not licensed LyricFind, that would have probably been the end of the project, but they got the deal,” remembers Cohen.

Now LyricFind has deals with hundreds of music publishers from around the world, as well as deals with collective management organisations, the latest being IPRS in India. The licensing successes of LyricFind owes a lot to the efforts of the licensing team, in particular Robert Singerman, SVP of International Publishing, and Nik McLeod, VP of Publishing.

For Cohen, LyricFind is a success story like there are few in this sector. “It’s been an amazing experience for me in that they created a new revenue that did not exist before, and they have never taken money from anyone, except a loan from Darryl’s mother. And they’ve been the good guys, literally.”

He continued: “I still think of him as the kid I met in a hotel lobby 20 years ago. Darryl has never been a dick. He is a lot more realistic today but basically, he is same person I have met 20 years ago. He’s somebody I would take a bullet for. He has a wonderful wife and two great kids. The company now has some 120 employees, and they collaborate, they communicate. He has created a culture that is collaborative and open. He has matured, but is still a good person, and he is certainly a better businessman.”

Creative Industries News sat last week with Darryl Ballantyne for a zoom conversation to talk about the company’s business. Here’s an edited version of the interview.

How’s the lyrics business these days?

Darryl Ballantyne: It’s good. I mean, it continues to just grow and be embedded in so many different places. And we’re still seeing tons and tons of growth every year. That is obviously something that we’re very happy about. We’re expanding rollouts with different DSPs. We have a number of DSPs that are working on implementing translations now. As you know, Deezer’s live with translations, and we’re going to see a lot more of that over the next year, including implementations of word by word synchronisation, instead of line by line, which really just makes the experience that much better. Plus the automotive side is really ramping up: we’re live in Mercedes right now and we have five different OEMs (original equipment manufacturer) launching next year. So it’s going to become that much more standard on the automotive world as well. So it’s been really interesting. The other fun thing that we’ve been working on lately that’s cool is that we launched an experimental kids project a couple of weeks ago. It’s called LyricFind Kids. Because of all the lyric video work that we’re doing, we wanted to just get a better understanding of YouTube monetisation and how that system works. So we recorded a bunch of children’s songs that are public domain songs. We did animations for them and then we’ve created regular videos, lyric videos, and lyric translation videos for them, under the LyricFind Kids channels. Initially, the purpose was just to learn, but it’s been live for about two weeks now and we’ve slowly been adding content to it. The reception has been amazing. There’s hundreds of thousands of views and the subscriber count is is growing really fast. It’s been fascinating to watch and it might actually turn into a real business line for us, like an actual revenue generator rather than an educational project. And it’s been a fun new experience.

When you started the company, what was your vision and almost 20 years down the line, did you fulfil that vision?

Darryl Ballantyne: Almost 20, yes! We can’t call ourselves a start-up anymore. I don’t know at what point you cross that threshold where you can no longer say that you’re a start-up, but wherever it is, we’ve passed it. (Laughs) The vision originally was just legalising the use of lyrics and having lyrics websites and streaming services have lyrics available legally where publishers and songwriters are getting paid. And that part we’ve accomplished. Lyrics are effectively in every meaningful DSP at this point. The lyric sites that remain are all licensed. And it’s turned into a real revenue stream for publishers and songwriters. So that part, we’re very happy about. And I never envisioned putting the lyrics in cars. I never envisioned doing lyric video creation or becoming a mini record label for kids content. So much of that stuff has all been different. The people that we get to work with and the parts of the music industry that we get to work with were not things that I ever expected to do, but they’ve also been a lot of fun. We kind of checked the box on lyric legalisation and lyric monetisation a while ago. So being able to get involved in different aspects of the music businesses has kept things interesting and allowed us to experiment and grow in different areas.

Did you have problems initially in trying to explain to music publishers what you wanted to do with lyrics?

Darryl Ballantyne: Yeah, yeah, yeah, definitely. (Smiles) Initially when we reached out to, like, the major publishers and others, they didn’t even know if they could license it. Lyric licensing at that time was all these one-offs and it was for a book or merchandise, and it was all done based on single approvals. There was no real ability for the publishers to license everything, all at once. But slowly that became the case. We even had one of the major publishers telling us when we reached out to them and told them we wanted to license all of their lyrics, ‘Okay, well, here’s our lyric licensing form. If you could fill that out, tell us the song that you want to license and what the purpose is, and things like that, and fax it to us at this number, then we’ll get back to you.’ And the form was just for a single song. This is a major publisher that owned like a million songs. And they wanted us to fill out a million forms and fax it to them! So there was a huge educational process at the beginning when we went to publishers and said ‘We want to license everything that you have. We want to monetise every song and create a revenue stream for you, and make sure that you’re getting properly compensated for all this use that’s out there.’ It was educational on both sides too. Obviously, I did not know much about music publishing and licensing. Fortunately, I had Ted Cohen helping out on a lot of that stuff. He knew everybody and everything. But I needed to learn a ton of things about how the industry actually works and how rights work.

Why and how did you get into this business in the first place?

Darryl Ballantyne: Originally we started off when we were in university back in 2000. You know, it was the first dot com boom. Everybody was starting something. Everybody had a start-up and making a huge pile of money in a matter of months. We decided that that’s what we were going to do too, starting a lyric website, just as a regular consumer facing product. When I was in university, I was the ‘name-that-tune’ guy. Everybody would come to me and ask like what’s this song or what’s that song and then I would tell them and Chris Book, one of my co-founders was trying to figure out the name of a song, came to me and asked me what’s the song, and it goes like this, telling me the lyrics. I didn’t know, so he went and tried searching online to figure out what the name of the song was, came back and said, ‘All these lyric sites that are out there suck, none of them let me search by the actual lyrics, let’s start our own’. So we started our own. It quickly became very popular because we were allowing searching by lyrics and we had this crazy notion of allowing people to comment on songs. At the time, it caused people to keep coming back and arguing with each other. It became really popular and we realised very quickly that we actually needed to figure out licensing. So we shut it down a few months after we launched it, tried to get licensing, realised it was a gigantic clusterfuck, for lack of a better term, and ended up shutting down completely. But that got us started on the lyric licensing side. Then, once we’d finished university, we came back to it and the industry had matured a little bit and was willing to work with us as an aggregator to monetise it all.

Unlike a lot of start-ups, you did not raise insane amounts of VC money. Why is that?

Darryl Ballantyne: We tried. We did try. In the earlier days, we did go out and try to raise funding, but I was terrible at it. I sucked at raising money, and I went about it in entirely the wrong way, and therefore was never successful in actually raising money. In hindsight, it turned out to be a good thing, because if we’d raised money in 2005, 2006, something like back then, we never would have hit our stride, we wouldn’t have been around to do it. We would have ended up needing to have some sort of liquidity event, so that the fund could close, and it would have just been a terrible outcome. We wouldn’t have hit the point where we were really starting to grow significantly and have meaningful revenue and profitability, and were creating more value. In hindsight it worked out really well, but it was tough in the first bunch of years. We were kind of living with our parents and not paying ourselves any real salary and bootstrapping everything as much as we could. In the end, we came out with a business that we still have complete control over and we can make our own decisions and not worry about a fund’s liquidity timeline or any external pressures to do things a certain way that we don’t agree with.

And if someone now came to you and said, ‘I want 51 percent of the company and here is X millions of dollars’, would you take the offer?

Darryl Ballantyne: I mean, there’s always a price, right? If it’s something that we think is the right fit and the right people and the right price, we’ll have a look at it. Not that anybody’s gonna offer this for us, but if somebody said ‘Here’s a billion dollars for 51%’, obviously, we’re gonna take it. But it’s it’s not something that we need to do. It’s something that we can decide to do when we feel like it’s the right time and the right fit for the company overall. We’re happy to do it either way. And we don’t have a pressure from investors or others to have that kind of liquidity event.

How do you make money? Where are the monetisation points in the whole process?

Darryl Ballantyne: It’s funny because a lot of people on the outside look at it and say, well, I’m not paying for lyrics. They’re free. So how are people making money? For us, it’s kind of the standard free adjacent monetisation models where services are monetising through advertising, then we’ll take a share of the advertising revenue. Or if they’re monetising through subscriptions, then we can take a share of the subscription price. Or a per display fee, or a per device fee. If you think about automotive, licenses, for example, were paid per car by the car manufacturers to have the feature in there. So, all of that is kind of customised to each individual implementation, based on what is the business model that makes sense for that service that is using the content. Generally it’s either a revenue share on advertising or subscription prices, a price per device for hardware things, or a price per display of the lyrics. These are kind of the common models that we use for monetisation. In all of those scenarios, the end user isn’t directly paying for lyrics. It’s getting bundled in with overall features that are available.

And from that pot, then you pay the rights holders, right? And do you pay them through the collective management organisations, or do you pay directly the publishers and the songwriters?

Darryl Ballantyne: All of the above. Usually the vast majority of it is direct to publishers or CMOs. We work with HFA, for example, or APRA AMCOS [in Australia] or CSDEM [France’s music publishers’ association], or we have CMO partners all over the world for different regions, but we also work directly with a lot of publishers. And we have artists that can sign up directly with us if they are not represented by a publisher. We can work with them that way. We try to encourage people to go through one of the aggregators as much as possible. We’re not huge fans of sending out micropayments or large numbers of payments for small amounts. It’s easier for us if it gets bundled in with other revenue streams that an indie artist is receiving. We work with DistroKid, for example, and DistroKid passes along licensing for the publishing where their independent artists control it. And then we pay the royalties to DistroKid and they include it in their overall payment. So it’s much more efficient for everybody to use those aggregators. It means that people get paid faster for the use.

One of the things about being a privately-owned company is that you don’t have to disclose your financial results, which is very frustrating for people like me because we don’t know how much money this business generates. Are we talking about millions of dollars or hundreds of millions of dollars? What’s the scale?

Darryl Ballantyne: Tens of millions on a yearly basis, if you look at the value of the lyrics licensing business, including us, our competitors, direct licensing from publishers, or the work that we do managing direct licenses. A major publisher will license a DSP for the rights, but we provide the content and the administration of those rights, so that licensing portion that is paid directly to the major publisher from the DSP, and doesn’t show up as revenue for us and our competitors. So in terms of the overall lyric business, it’s comfortably into nine figures a year in monetisation that flows through the industry.

And that’s money that did not exist in the past.

Darryl Ballantyne: Exactly. It’s all a brand new revenue stream that wasn’t getting captured.

This business also creates a certain volume of data. How do you deal with it?

Darryl Ballantyne: The data is a patchwork of many, many different sources. It can be the lyrics themselves that are being created by our in house content teams and teams all over the world that work on translations, for example — or the feeds of lyrics we get from indie distributors, wherever possible, or third party professional translation firms — this all helped build that data. And then the ownership data that we have comes from our CMO partners, our publisher partners, our own research, our direct publisher relationships. It all gets kind of merged together and matched to recordings and matched to different compositions across different publishers. But there’s not one simple way to do it. It’s just a whole lot of aggregation from multiple sources for pretty much every piece of our data set. We work with companies like Rovi, MusicStory and GraceNode and others to help us organise that data and try to have the recording side information to match publishing compositions too. It’s a complex process that gets more and more complex with all the different pieces of data that we add in, such as international rights from different sources. It’s a gigantic data problem.

What’s your appreciation of the state or the quality of the data itself?

Darryl Ballantyne: It’s definitely a lot better than it was 15, 20 years ago. Publishers in general are much more open with their data. When we were first getting started, we had publishers just refusing to tell us what they owned. Well, how are we supposed to use it and earn any money for you? That was definitely a challenge. Some people wanted to match usage after the fact, and so they’re effectively saying, ‘okay, tell us what you used and we’ll claim and tell you what’s ours’. Well, our answer was, ‘you’re telling us that we should go on and infringe on copyright and then you’ll just clear your own stuff? But how do we know what we have the rights to use in the first place?’ So those were some significant challenges early on. It’s definitely better, but we have a lot of challenges with people claiming things that they shouldn’t or people claiming things that they might control in their own territory, but when they claim it worldwide it creates conflicts in the systems. We have to clean up that data and figure out who we’re supposed to be paying in order to pay the proper owner. The more deals that we do and the more global we become, the more work it is to clean up those conflicts. There’s no global source of truth. It’s one of those things that we’re working on, and 30 other companies are also working on it. We’re all doing the same work. There’s got to be a better way. I’m sure there is!

How do you get the lyrics in the first place? Is it something that you do in house? Is it something you get from the publishers or from the songwriters?

Darryl Ballantyne: A chunk of the data, we get it from the artists and songwriters themselves when it flows through a digital distributor, for example. Taking DistroKid as an example, an artist will submit the lyrics to them and then they pass that along to us along with the licensing right when they have it. But for most of the popular content, we’re creating it ourselves. And we’re doing that through our content team that literally sits with headphones on transcribing the lyrics and synchronising them as well as working on the licensing information to make sure that they’re properly cleared. We have a big team here in Toronto that works on our core languages, but then we also have external teams all over the world. We have a team in Vietnam, in India, in Morocco, in Eastern Europe that works on it. A number of those teams all over the world working on their local languages to make sure that content is available and properly done in our system.

Is AI going to put you out of business?

Darryl Ballantyne: No, I don’t think so. We’ve obviously looked at a lot of those tools and we continue to experiment with ways to make the process more efficient. AI is something that can be a helpful tool in the process, but it’s nowhere close to being able to do a perfect transcription or a perfect synchronisation. It’s just not that easy for that type of content. Plus, the AI is terrible at figuring out ownership data, license management, negotiating rights and being a clearinghouse and aggregator for all of this content. It’s something that we view as a helpful tool, but not a threat.

And in general, do you see AI and the music sector as a good match or will it be a risky one?

Darryl Ballantyne: I think it’s both, right? At the same time, is it something that can truly be prevented? It’s a good match for some people and it creates some good content. It also creates terrible content. And it’s a licensing nightmare. For most people involved, when you look at training models and if you’re training it off of copyrighted material, how do those rights work? Is that a right that a label or a publisher, for example, can even grant? When does that infringe on moral rights or name, image, and likeness rights, or things that they can’t actually control themselves? How do you know what an AI model has trained on for sure? That cat’s out of the bag. There isn’t a world without AI content generation that is going to exist going forward. It’s here, and it’s here to stay, and the question is will it add value to users and to the ecosystem. There’s a clear value added from these AI models for many different purposes, whether it’s background music or music in gaming or commercials, where if you can just automatically generate it, it’s,a pretty good value proposition for people who need music. But it a licensing mess, and it’s gonna take a long time to figure out. I personally believe that if AI is using copyrighted material to create that value, then there should be some sort of compensation to the owners of that material. Those are massive challenges that the music industry is going to have to figure out, along with many other industries.

In terms of growth, you probably have covered pretty well all the Anglo-American repertoire and you are now expanding in other markets. You recently you did a licensing deal with IPRS in India. Is that your new frontier — Asia, Africa, Latin America?

Darryl Ballantyne: Certainly. A lot of those countries are big value generators. Brazil is, for example, our number four country right now. So it’s a huge area of focus for us. India, obviously the publishing rights there have been extremely messy over the years, but it’s a huge area of focus for us and it’s an area where our clients are asking for more and more coverage and more and more availability there. That’s an area that Robert [Singerman, SVP of International Publishing] and Nik [McLeod, VP of Publishing] have been really focused on this year, delivering the IPRS deal and a number of other deals to make sure that all the rights are in place. China is another big area of focus where the traditional copyright regime there has been kind of non existent, but they’ve done a lot of work over the last years to clean that up and turn it into a market that a rights holder can actually thrive in. We’ve been doing a lot of work there to build up the rights and the licensing. And for us a key market is the automotive sector. Those other areas are major frontiers for us where we still believe that there’s a lot of growth to be had and a lot of traffic to capture particularly if you look at India and China. Just population wise alone, the scale there is incredible and we’re only tapping the surface of that at the moment. We think that there’s a lot of growth that we can have there and to a lesser extent in other international territories, outside of North America and Europe. There’s certainly more growth to be had places like Africa as more and more people come online and services get larger. International is a big area of focus for us. And translations are something that will have more and more value on the international stage than in the the US or Canadian markets, for example.

What’s the lyrics business of tomorrow going to be? More direct licensing to the consumer or there will still be third parties like the DSPs or others to provide access to lyrics?

Darryl Ballantyne: It’s going to be indirect because the key with lyrics is that you want them at the point of consumption of the music. When you listen to a song, you want to be able to follow along, see the translation, see the synchronisation. Music services are going to continue to thrive. They’re going to continue to serve all of that need. Lyric consumption is going to end up being wherever you’re listening to music. And it’s not something that people can handle directly.

What feedback do you get from your conversations with DSPs?

Darryl Ballantyne: Pretty much everybody now knows that lyrics are a feature for a streaming service. So there’s no real question anymore about why should I have lyrics. They know that it’s key. For a while we were kind of in that in-between period when we’d go and talk to DSPs that didn’t have lyrics, and without exception, they would tell us that it was their number one or number two requested feature from their users. There’s always been a lot of demand, even pre-DSPs, based on the huge amount of traffic that lyric websites received back then before there were more sources of lyrics available. So the demand has always been there. Lyrics was Google’s number one search term for a long, long time. DSPs have recognised that, and it’s something that is really key to create really great experiences. Lyrics and, and lyric translations become that much more important in a global music world where we’re listening to all different things in all kinds of different languages and trying to understand that. So there’s just that much more value that can be added to the lyric experience and the user experience.

Where do you see LyricFind in the next 5 to 10 years from now? Are there new businesses that you want to start?

Darryl Ballantyne: Our big areas for growth, as I mentioned early, are translations and word by word synchronisation. That’s a big area, but we also look at the video side. We’ve been doing a lot of work on the lyric video tool, and I think we will be expanding into more video formats and more platforms, and doing integrations. That’s an area that we’re really heavily investing in and is going to become more and more important. So much more of the entertainment experience for music is built around video now, so being able to service that part of the ecosystem becomes a really key strategic plan moving forward. With automotive, we see a lot of growth there within five years. It’s going to be fairly ubiquitous. Obviously, the automotive development timelines are a little longer than what we’re used to in regular tech and DSP builds, but the groundwork has been laid there and that’s becoming much more of a standard process, with every car now is connected effectively. So that makes that much more of a possibility. And, who knows, maybe we’ll have a giant kids record label in five to ten years. that’ll be a whole different division. We’ll continue to focus on translation and synchronisation, video and further metadata services, like Lyric IQ and analysis that we can do to help people filter and discover music in a world where there are millions of songs being released in a year.

Are you still having fun doing it?

Darryl Ballantyne: Not every day, but most days. Anybody who tells you that they have fun every single day in this world is not being honest, but I have way more fun days than not. And being able to do different things in different areas of the music business really helps with that. I’ve been able to get to know a lot more people on the label side and building the video side has been really enjoyable for me.

Longevity must give you a sense of actually belonging to the ecosystem. You’ve weathered quite a few storms, so that probably gives you some peace of mind, doesn’t it?

Darryl Ballantyne: Absolutely. You know, we feel pretty secure in what we’re doing and in how the business operates. With time and resources, we’ve been able to build a team that has so much expertise, in so many areas that are really integral to making sure that we don’t do anything exceedingly stupid. It definitely helps being able to look at the whole music business in a holistic and knowledgeable way and understanding our place in that world.