We’ve come a long way since the wild west of lyrics, which are now easily searchable and provide artists with an additional source of income. Moving forward, lyrics will provide essential contextual metadata in the age of voice command.
__________________________
Guest post by Darryl Ballantyne of Lyric Find
Lyrics provide essential contextual metadata, information about tracks that can’t be extracted and discovered any other way. They can unlock a future when you can walk into a room and ask via voice command to hear sad songs about Paris in French or an upbeat song about tigers.
Already, lyrics top web and mobile search worldwide. Whenever lyrics are displayed via Google or on a lyric site, songwriters get paid, thanks to a licensing ecosystem that closed the gap between music services, websites, and music publishers. We started from scratch thirteen years ago, with an online Wild West of lyrics use. We’ve come a long way since then.
The road has been surprisingly rough to get here. First, there is the basic data-entry aspect for lyrics. You can’t rely on fans and get consistent and accurate results. There are no clear channels to gather lyrics from songwriters, artists, or publishers. We have learned that accuracy requires a dedicated team fluent in ten languages (and growing). This team manually transcribes and timestamps each and every lyric you see via search or synced to your favorite music service. When you consider the volume of content going up each and every day, across the world’s many languages, you get the daunting scale of a truly global legal lyrics ecosystem. Our database powers approximately 10 billion displays a year, and that number is growing fast.
Second, there is the complex web of relationships between platforms, services, and sites on one side and publishers, whose models don’t automatically harmonize with digital music companies’. We’ve had to create the relationships, build the technology, and get both sides to agree on the value of lyrics. Now the overwhelming majority of lyrics displays are licensed, and revenue for rightsholders is in the tens of millions–and also growing fast. The hundreds of publishers worldwide who have signed up for licensing earn more for a display of a lyric than they do from a stream on Spotify’s free tier. (LyricFind paid out many millions of dollars to rightsholders last year, and revenues have grown 50% year-over-year.) With the upcoming launch of LyricMerch, writers and publishers will see another huge lyric-based revenue stream turned on.
Lyrics are now proving their value in a different way. They make music discoverable, but without voice commands, a fan needs several steps to find that song on her computer or phone. By the time this fan has opened the right app, got it up and running, and is ready to search, she may feel seriously frustrated, especially if the song has a quirky title, uses alternative spellings, or is just hard to understand or recall.
Voice command removes those final steps from humming a few lines to hearing the song pour out of the speakers. It’s getting to the point when at last neophytes and technophobes can find whatever they want in the realm of music, simply by talking to a box. Tech giants like Amazon have mentioned explicitly considering voice and search as they shape their music products, baking in voice command as a way to differentiate from other streaming services. Lyrics are a requisite part of this, and more services will be following suit soon.
Like publishers, labels are starting to take note and get excited about this potential new use for lyrics, even though labels may have no direct financial stake in them. Lyrics are sticky, and they drive song plays. Now that they see a direct path running straight from lyrics-based search to revenue, rights holders (outside of the publishers) are realizing how powerful lyrics are.
There are more reasons to embrace this power, too. Lyric data, now incorporated as a basic feature into more and more music-related services, promises to do more than make it easy for fans to return to favorite tracks. Lyric data can help music fans filter material that contains profanity or explicit language, for family-friendly listening. It can also look for lyrics in a certain language, to set the stage for a trip or to enhance language learning.
Thanks to recent leaps in natural language processing, lyrics can push discovery in broader ways, training algorithms to find contextual connections–songs about a certain city or songs that capture a certain mood–that help listeners find the right music for the moment. This kind of contextual search would have been impossible five years ago. As machine learning and AI progress, it will likely get better and better, as we figure out how to teach computers to look for keywords in songs. They will be able to find associations that point to mood states, tone, or other categories. Here, the age-old tools of literary analysis and new technologies combine to enhance our listening experience.
This context-based lyric search across languages and feels has another upside: every licensed use of lyrics earns songwriters and publishers a royalty. This, when for many years, lyrics were loss leaders (when you had to print them in expensive liner notes) or completely divorced from sound recordings (with the dawn of the mp3 and file sharing). So, Alexa, play me happy songs about songwriters making money. Songwriters, publishers, labels, and listeners all stand to gain from the alchemy when lyrics data meets voice search.
Darryl Ballantyne is CEO and founder of LyricFind. Now the leading lyrics display licensing service, LyricFind was originally conceived in 2000 with Mohamed Moutadayne and Chris Book at Ontario’s University of Waterloo to be the largest, most accurate destination for lyrics on the Internet. Officially launched in 2004, the company pioneered the licensed digital lyrics space, successfully negotiating the first-ever mass lyrics licensing deal with EMI Music Publishing in 2005. Since then, Darryl has grown LyricFind to be the largest legal, licensed lyrics database in the world, with millions of licensed tracks available and hundreds of lyrics sublicensing clients. LyricFind has established licensing relationships with over 4,000 music publishers, The Harry Fox Agency, APRA/AMCOS, CSDEM, APEM, and many more.
The company now delivers services to hundreds of websites and music services that want to enhance their music products with lyrics, including Pandora, Deezer, Microsoft’s Bing, Amazon, Plex, Roon, MetroLyrics, Shazam, SoundHound, LyricsMode, SongMeanings, iHeartRadio, and many more.