(Photo, Nate Leath, the author, Sierra Ferrell in New Orleans, January 2020, end of the Before Times)
On my morning hike this morning I listened to a very wild recent podcast interview with Sierra Ferrell. She covers a dizzying range of topics: train-hopping, being homeless, collaborating with Zach Bryan and Diplo, she even talks about a near-death experience from a heroin overdose. I spent some time with Sierra a few years ago when we signed her to Rounder and made her debut album, but I didn’t get to know her all that well. Like many complex, highly creative people I’ve known, she isn’t super easy to get to know. Nevertheless, I always found her fascinating and otherworldly talented. If you haven’t spent time with her music, you’re in for a treat.
One thing in her interview really caught my ear. While discussing her creative process, she explained that knows she’s on to something when she gets goosebumps during songwriting. She knows when a song is great because she has a physical reaction. Maybe that’s true of all creative people and rarely articulated, maybe it’s the most obvious thing that everybody knows. I don’t know, but it’s definitely true for me.
The first time I heard Sierra sing (on a YouTube video my colleague/her producer Gary Paczosa showed me), I knew before the end of the song that she is truly great. I knew because the hairs on the back of my neck stood up. Within hours of first hearing her voice, Gary and I were strategizing to sign her to a record deal and work to develop her into what she has now become: a star who could redefine Rounder Records for the next generation. Our success as a label wouldn’t happen without artists like Sierra; when you find one, you GO before it’s on anyone else’s radar.
Someone asked me recently about my process for find great developing artists. I said I only trust one data point: those hairs on the back of my neck. Goosebumps. I can feel the music when it’s great. That’s true to a degree with my songwriting, I get very focused and energized when it’s going well. I don’t feel it until I hear it back on a voice memo, or maybe even a record. Once music gives me that feeling, I listen over and over chasing the feeling again and again. Connecting physically with music feels great, as I assume it does for most of us music fans. We listen to music to feel something.
My taste in music isn’t eccentric or experimental. I seem to love songs that lots of people love. I absolutely do not like all popular music - I can’t stand a lot of it. Nevertheless, I know if I like it, others will like it too. That’s why I have confidence in the hairs on the back of my neck or the goosebumps on my arms - they don’t lie. Either I feel it or I don’t. If I don’t, then I can’t be confident others will like it. It’s that simple.
Music works when it connects with people on a physical level. Music is good if it sounds good, is well-written, well produced, well performed, original, energetic, etc. Music can great even when it doesn’t tick every one of those boxes, but still causes my dopamine jets to fire and tingle on the surface of my skin. To me, good and great aren’t the same thing at all! I’ll hold out for great every time.
That raises an important point about how music marketing has evolved in recent years. In the old days, mediocre or even straight-up terrible acts got signed to deals1 all the time. Usually these projects lost momentum before anyone spent a fortune; but occasionally not-so-great music would get a multi-million dollar marketing push. Sometimes it even worked a little. In the monoculture, you could leverage relationships to get a project on radio or television, covered by print media, or added to a high-profile tour. Ultimately projects only succeed if the music is some version of great, so I can’t think of many mediocre projects that truly scaled.2
I have a theory how this happened. Before we had access to real time data on platforms, industry folks - “gatekeepers” - would make choices about developing artists through relationships or by chasing industry buzz. Clearly people working on the talent side of the equation, the process of A&R, development, and marketing, didn’t hold out to confirm that they actually understood how an act would connect with an audience. Maybe they sought to replicate the success of a similar act. Maybe a big time music lawyer or manager with a track record pressed them to sign the act. Maybe the act became the subject of a bidding war for reasons nobody fully understands.
Now we have data. Sure, we had different kinds of data back then - local radio airplay translating to ticket sales, record store sales, market research, etc. None of that would likely be available for a true development act. Now we have algorithms, platforms, measuring everything and available to everyone. Most of us have the most powerful individual marketing and research tool ever created - our smartphone - in our pocket. Why would a major label ever sign an act that didn’t have a compelling data story? They might as well just wait and see what pops up. They have sophisticated research tools to see what’s really moving. In a sense, that reliance on data takes power away from people like me, entertainment professionals who in another era would have shopped deals. There’s no point; nobody will sign an act without a data story demonstrating that the music connects with an audience. If I approach them to pitch an act, I can’t fake the date. Maybe they’ve made up their mind before they hear a note. Who needs goosebumps when you have real time audience data?
In most ways, I argue, algorithms are better gatekeepers than humans. AI-powered algorithms are programmed to do something highly specialized, such as measure audience reaction. Human gatekeepers have egos, complex relationships, distractions, and conflicting incentives. Algorithms are never wrong, but people are wrong all the time! Human gatekeepers may still screw things up by failing to understand the data, or mistrusting the data. The data, however, is reliable every time.
What that means is that no project hits the mark unless it is supported by data. Once the data tells the right story, major labels focus on scale. It’s already working - now it has to work better in more places. I recently spoke with the co-head of a major label and I asked about their role in scaling hit projects internationally. “To truly scale globally,” he said, “you need a fucking army.” Think of a major label as a fucking army of scale. They do not, however, do much in the way of early stage artist development. That’s on the artist and whatever team and resources they can amass and afford.
On the one hand - especially considering the tens of millions of tracks that are utterly ignored on digital platforms - it’s truly depressing. But on the other hand, it’s damned exciting. That means all the opportunities in artist development are hiding in plain sight, outside the reach of research algorithms. There are future superstars and paradigm-shifting artists currently on platforms with their music being largely ignored - at least for the moment. I don’t mean to suggest that everything that captivates an audience and scales on music platforms is great, or even good. I don’t mean that only great music connects an audience. Mediocre or even terrible music can scale for all sorts of reasons. What I’m saying is that algorithms make it possible for great music to find its way out of the ocean of mediocrity of music streaming platforms.
Am I so arrogant as to believe that my goosebumps are all I need to dive into 100 million tracks currently on Spotify and find a winner? No, absolutely not. I know what I like when I hear it, and I’ve learned (for the most part) how not to deceive myself. I have my own version of research, mostly just using my ears. I know how to read the data and that’s helpful. I have a growing network of people who send me music to assess. It’s on me to parse the mediocre from the good from the great. Basically, I have to pay attention. It pains me to remember a handful of times I missed the goosebumps that came with later listens - after I passed - because I was multi-tasking while listening. Sorry, Allison Russell. You deserved my undivided attention.
When I find something, do I send it to the Fucking Army of Scale? Usually not right away. The Fucking Army doesn’t want to hear from me until it’s time to scale. I’ll hear from them when it’s time, oh boy will I ever. Until then, it’s a grind of trying over and over to create connection and fan engagement, mostly with digital “content” on digital platforms. But here’s the key: It only works when people care. Maybe you can still make it work with bad music if your marketing is clever. But the only way I know how to do it is with great music. If the music is great, it’s worth the slog. Don’t try to fake it; it won’t end well. People can tell if the music’s authentic, algorithms can tell if the audience engagement is authentic. Fake ‘til you make it at your peril in this space - it’s easy to trash your reputation trying to fake consumption data.
Thank you, Sierra, for reminding me of what I already knew. Goosebumps are the key.
For example, Butt Trumpet got signed by EMI in 1994. I still want someone to explain that one to me.
This is tricky, but it’s important to parse music you don’t like from truly mediocre music. Some music I think is terrible, such as the boy band sound of the late 90s, was enormously popular. It doesn’t speak to me, but it speaks to a lot of people. I’m talking about music that doesn’t create that connection with large numbers of us.
In addition to being my favorite singer in the world right now, I don't think Sierra Ferrell gets enough credit for the visual aspect of her game. Her videos, even as an unknown, seemed to leave an indelible impression on all who caught them...the settings, her outfits, and always an obviously live vocal nailed perfectly...they're positively spellbinding. People tend to think of music videos as quaint 80s/90s relics, but there's a way to integrate it into current cellphone/social media culture as an attention-grabber seeing as they're just as accessible as an audio track, and boy has she blazed a trail that way.
As an emerging musician I’m trying to break into the algorithm with my new song and hoping to make the right connections to promote my song ‘Time Is Running Out’. It’s a duet I wrote, a candid conversation between two people. One person saying they have nothing left to give and the other person trying to encourage them that they have so much more to give. I hope you can take some time to give it a listen and share it. Thank you.