Pandora: a 3-1 Lead in Music Streaming

Today’s episode and memo is about Pandora. Ten years ago, Pandora was in pole position in music streaming with over 80 million monthly active users. But the company had a rocky road to get to that point and an even rockier decline in online music consumption.

I’m joined by Tati Cirisano from MIDiA Research for this one. You can listen here or read below for a few highlights from our chat.

Technology in search of a solution

In 1999, Savage Beast Technologies was launched and soon evolved on the back of the Music Genome Project. Savage Beast Technologies founder Tim Westergren and his team developed a manual coding process to identify up to 450 “gene” components of a song. Attributes included the gender of the lead singer, the instruments used in the background, the structure of the bridge, and more. Each attribute went into a predictive model to determine which songs you want to hear next. It was groundbreaking, slightly ahead of its time, but was clearly where music was heading.

But despite the impressive work and the “cool” factor, the Music Genome Project didn’t have a substantial problem it solved at the time.

The technology was shopped around to music retailers like Best Buy, Tower Records, and Barnes & Noble via in-store recommendation kiosks, but it didn’t gain much traction. Plus, the company launched at the height of the dot-com bubble, which made matters worse. Fundraising became much, much harder. Westergren says over 300 VCs rejected him.

The company ran out of money, the credit card bills piled up, and Westergren had to give this speech to convince the team to stay. The most critical line of the speech was:

“When this thing finally finds its home, it’s going to change culture. How many times in your life do you have a chance to do that?”

It bought the time company time, which led to new investors, new leadership, a new name ‘Pandora,’ and a newfound “home” for the Music Genome Project—internet radio.

In the mid-2000s, internet radio was attractive because it brought the technology directly to consumers. Broadband speeds were steadily increasing, the music industry was steadily declining, and Pandora Radio was an opportunity to close that gap. And if Pandora was treated as “radio,” then it would be subject to the same low royalty rates as terrestrial radio.

Unfortunately, it wasn’t that easy. Pandora was subject to 50% royalty rates, much higher than terrestrial and satellite radio. Westergren and the Pandora leadership team fought numerous battles over royalty rates.

Looking back, Pandora was launched as technology in search of a solution. It moved from record store kiosks to internet radio to iPhone apps. Today, the influence of the Music Genome Project is seen in playlists like Spotify’s Discover Weekly, and other similar versions from the digital service providers.

Westergren was right. The technology eventually found its home. The music streaming algorithms have gotten better and better. Pandora may not capture all the value it inspired, but it’s hard to tell the story of music streaming without the company’s technology.

You can listen to the episode here or keep reading for more highlights.

Pandora and Spotify: two distinct approaches

Pandora’s leadership team fought the music industry lawmakers and tried to rally support from both musicians and fans to lower its rates. Pandora even bought a South Dakota-based FM radio station to help its cause. But the court of public opinion didn’t side with the company.

In 2013, Pink Floyd wrote an open letter in USA Today to speak out against Pandora. Meanwhile, Pink Floyd willingly gave its music to Spotify. The NAACP wrote a letter to push back against Pandora since its goal would lead to further under-monetization for the Black artists signed to record labels like Motown Records.

Technically, Pandora prevailed against ASCAP to lower rates, but it missed the more significant opportunity.

Spotify beat Pandora for two reasons. The first and obvious reason is on-demand streaming. Pandora’s personalized listening streams were a cool feature, but any Pandora user has reached their maximum six skips per hour and likely wanted more. Much like MTV’s decision to move away from music videos, the shift to on-demand streaming was a reminder that consumers want choice. Consumers want what they want, when they want to hear it. Pandora Radio was one step in that direction, but it didn’t go all the way (until much later).

But the less obvious distinction is how Spotify and Pandora approached the music industry and its royalty rates. Pandora pushed back loudly in the early days and ruffled plenty of feathers. Meanwhile, Spotify took the industry’s less-than-favorable rates early on and let the labels take an ownership stake in their company. But Spotify built its user base, grew its leverage over time, and flexed its power more and more. Many of the current debates today on streaming pricing, algorithmic plays, and payout models are because of the leverage that Spotify now has.

By 2015, around 25% of Spotify’s users had signed up for its premium plan, compared to just 5% of Pandora’s (which cost half the price of Spotify). Pandora’s paid product had a value prop challenge and lower adoption. The company later introduced on-demand streaming and made a series of acquisitions—Rdio, TicketFly, and Next Big Sound—but it wasn’t enough.

Mobile attraction

In 2016, SiriusXM started to pay more attention to Pandora. The satellite radio company had done exceptionally well in car stereos, often partnering with the manufacturers. Anyone who has bought a relatively new car in the past 15 years has likely been pitched on a satellite radio upsell.

But SiriusXM had yet to corner the mobile phone market. Pandora, like Instagram, Uber, and Airbnb, had accelerated in growth thanks to the adoption of the iPhone. In 2007, Pandora’s team was jailbreaking iPhones to build a native Pandora app in anticipation of the App Store’s launch.

In 2011, Pandora’s IPO valued the company at $2.6 billion, compared to the $1 billion that Facebook bought Instagram for in 2012 or the $60 million that Uber was valued at in 2011. Pandora was the darling of consumer tech for a hot minute there.

SiriusXM likely came in a few years too late, but if the company had been doing well before, then SiriusXM may not have been able to buy it off of the public markets.

In the five years since SiriusXM’s acquisition, there have been several attempts to make waves. Like its partnership with Marvel Entertainment for original podcasts or its creative partnership with Drake, but despite the big names, it wasn’t quite enough to move the needle.

Listen to the rest of the episode for more on:

– the pitch to keep 50+ employees at Pandora
– Pandora’s M&A deals
– the Podcast Genome Project

Listen to the episode here.