Teardown: Pandora Will Double Users from 50MM to 100MM and Do $125MM In 2010

These posts originally ran on TechCrunch.  You can check out my previous analyses of Groupon, Zynga, and Chegg.  And my interviews with Bambi Francisco at Vator.tv here and here.

Don’t call it a comeback
I been here for years
Rockin my peers and puttin suckas in fear
Makin the tears rain down like a MON-soon
Listen to the bass go BOOM
Explosion, overpowerin
Over the competition, I’m towerin

LL Cool J, “Mama Said Knock You Out” (1990)

Founded in 2000, Internet radio startup, Pandora, has had more name changes than Diddy, more styles than Madonna, and more near-death experiences than Ozzy.  Through it all, the company has systematically listened to and catalogued every song by hand along 400 different attributes to provide one of the most personalized music experiences in the world.  Based on recent growth rates, the release of Apple’s new iPhone operating system as well as Android’s emergence as a solid platform, I project the company will have 100MM registered users by the end of the year and generate $125MM in revenues in 2010 (at an average of just $.146 per user per month).  However, just like the aging metal singer, Pandora’s long-term health is far from assured.  The key question: Is Pandora scooping up barrels full of pennies in front of an oncoming Apple steamroller?

Company History and Finally Getting User Traction

While Pandora has raised more than $56.3MM in capital, the company subsisted on $12.3MM for nearly a decade, including $1.5MM over a four year span when no investor was willing to put in any new money.  This is a company comprised of passionate musicologists and savvy management experienced at putting the needs of the company ahead of their own.  After two previous business models failed to get traction, Pandora Radio launched in August, 2005 and garnered 2.5MM users in the first 10 months.  By October, 2007, 8.5MM listeners had registered and were joining at a clip of 500K new users a month.

Then the company’s prospects were forever changed with the release of Pandora’s iPhone app in July, 2008.  It was the top downloaded app that year and its registered users went from 17MM in Q3 2008 to 50MM at the end of Q1 2010.  Mobile (and other connected devices) now account for 30% – 40% of Pandora’s usage and this will continue to rise.

As you can see from the chart from Compete, fewer people as a percent of the user base are going to the website, preferring to listen to Pandora on their mobile device.

Pandora’s 5 Key Assets

Pandora competes in a highly competitive industry where traditional and satellite radio broadcasters battle with streaming music services, such as Rhapsody, and highly anticipated new entrants such as Rdio, Spotify, Google, and Apple.

Pandora has five key assets that are difficult to replicate at scale:

1. The Music Genome: Pandora has created a proprietary database of attributes that personalize and recommend music in the most effortless user experience today.
2. Preferences Database:  Pandora has a treasure-trove of more than 1B thumbs up/thumbs down/skips interactions to augment the Genome.  The company has said that listeners interact 7-8 times per hour to give feedback or configure settings.  This gives the company three advantages over traditional radio: higher levels of personalization, more opportunities to monetize, and additional demographic data.
3. Large Customer Base:  Advertisers buy reach, not music.  With 50MM registered users and 25MM+ listeners a month, Pandora is now an attractive advertising platform.   The large base allows Pandora to segment and geo-target to achieve higher CPM’s.
4. Distribution:  As the de facto Internet radio product, the company has been aggressive to partner with as many home electronics manufacturers as possible.  One can access Pandora on Sonos, Blue-Ray players, Radio’s, Digital Media Players, Home Theater Systems, and Car Radios.  They key driver, of course, is the iPhone app.
5. Experience and Clout with SoundExchange/Label Relations:  The biggest barrier to entry and the most significant cost of operating a music service is dealing with the record label and SoundExchange over royalties.  Pandora has years of experience dealing with this complex system and has so many users now that the company can strike more favorable deals than smaller players.

Pandora Makes Money, Pennies At A Time

Pandora has three tiers of service, 40 hours/month of free ad-supported music, $.99/month for unlimited listening after 40 hours, and $36/year Pandora One, and it makes a little off of affiliate fees from track downloads- but nearly all of its revenue comes from advertising.

Based on revenue and average user numbers released by the company in 2008 and 2009, and using simple ratios, I estimate the company will average just over 70MM registered users for the year and generate close to $.15 per user per month, on average, or $1.75/year, up from $.138 and $1.66 respectively last year.  To be clear, this is not to say that Pandora will average 70MM listeners over the year.  This is clearly a scale business since royalties are variable with each hour listened and there is little margin for error.

Traditionally, CPM’s for Internet and radio have been the lowest compared to every other form of advertising, other than outdoor.  Online, audio, and mobile (another low CPM medium) are the three advertising channels available to Pandora.  While a home page takeover can garner a $20 CPM, an audio ad can be as low as $4, and a little banner on the iPhone $.50 CPM.

In an interview with Bambi Francisco last year, the company stated that they are getting average CPM’s of between $8 and $10.  But not all ads are created equal and early mobile advertising has been in the $1 to $2 range.  Depending on where a user listens, ad type and frequency, sell-thru rates, and frequency caps, Pandora is generating different CPM’s by the hour.  I assume the company will run two ads per hour with some being of “high quality” CPM’s of $10, some of “medium quality” of $7, and some of “low quality” of $4 across its inventory to get the following chart:

Based on those assumptions, I project the average Pandora listener who listens to 11.6 hours a month generates $.228 in revenue and $.0767 in gross margin.  Depending on the customer lifetime value (how many months the average customer listens), this does not leave a lot of money left over to spend on marketing to acquire new users.

There are passionate Pandora customers that easily surpass the free 40 hour per month limit.  Because of the chart above, you can see that if the company does not monetize its heaviest users smartly, it could literally go bankrupt from the variable royalty fees.  I assume that when a listener pays the incremental $.99 she is likely go to way over.  Based on my analysis, Pandora breaks-even on its royalty fees at 120 hours a month (4 hours every day).

For its premium customers that pay $3/month or $36/year, they would have to listen to 8 hours every single day for the entire year to breakeven on royalties.

This is the math that Pandora must do on a daily basis.  The company needs to be smart at predicting and monitoring its listener behavior.

Strategy: Pandora Platform

Pandora’s goal is four-fold: 1) to provide the most personalized music recommendation service, 2) to gradually add in as many high CPM advertising opportunities without degrading the user experience, 3) to carefully monitor and project listening hours by its heavy listeners, and 4) to be available wherever and whenever people consume music.  Because of the high variable cost of music royalties imposed on the business, Pandora is paying between 60%-70% of revenues to SoundExchange.  This leaves very little margin to spend on customer acquisition, so ubiquity is the only way it will grow its user base.

Pandora Is Still At Risk

Despite reaching nearly 100MM by the end of 2010, the company still faces many challenges:

1. Size Matters: Compared to all the other music streaming services, Pandora’s catalog is quite small.  Comprised of 700K tracks, the company has made a conscious decision to filter out many albums.  While this may not be a big deal to many listeners, it does mean that Pandora tends to repeat songs more than the others.
2. Limited Reach: Because of the restrictions placed upon the company by the royalty holders, Pandora is only available in the United States.  This situation will need to change over the next few years or Pandora’s upside will be severely impacted.
3. New Entrants: Rdio, Spotify, Google (http://www.dmwmedia.com/news/2010/06/15/cnet-google-music-service-could-debut-fall)
4. AT&T Data Plan Changes:  AT&T’s decision to limit data plans could be potentially disastrous for Pandora.  This situation will have to be watched closely.
5. Limited Listening Options:  While the simplicity of the product makes Pandora such a compelling, frictionless experience, customers may miss no rewind or repeat, limited skipping, and the inability to pick whatever they want to listen to.

If you go back and consider the 5 core strengths of Pandora, there is one other company that has them too- Apple.  Once Apple gets serious about streaming music (and it will), Pandora will face the biggest challenge in its history.  Therefore, I expect Pandora to generate alot of strategic interest from Google, Clear Channel, Sirius/XM, and traditional radio broadcaster CBS.  I expect Pandora to be acquired in the next 24-36 months.