Pandora Town Hall Meeting Notes
Pandora Town Hall (photo courtesy of ashwinrao1)
Last week I went to a Pandora user talk at the Seattle Library. Pandora founder Time Westergren has been doing these since 2007 (I actually went to one of the first ones back then – check out my notes) and he’s done over 250 since. He shared some stats about the service, and spent most of the Q&A with the audience.
I’m amazed at how they’ve built such a passionate fan-base in short order (the talk was overflowing yesterday, and they are packed in every city). They have strong conversion/engagement on their website (1 in four songs are “thumbed up or down”) and great breadth across connected devices (Pandora is available across cars – through Ford Sync, radios, PCs, iPhones, TVs, etc.). Also, they are laser focused on simplicity. They have seen first-hand that adding features lowers overall site engagement. In fact, their lead designer used to make exhibits for a children’s museum ?, that’s how much they care about simple UI.
Tim’s personal story is also quite amazing. He spent 10 years touring the country as a struggling musician, and brought Pandora back from the brink on several occasions to make it what it is today, quite a household name. There is something to be said for endurance and sticking with something you truly believe in. Since some of you are probably Pandora fans (like me), wanted to pass along my notes from the town hall meeting:
- Launched Pandora 4 years ago as a streaming radio station, before that the Music Genome Project was active from 2000.
- $50m in revenue and pays $30m in royalties for music
- Almost went bust a few times! (dot com bust and then licensing litigation)
- 50M registered listeners
- 2M new users join the service every month
- 180 employees
- Genome started in 2000 – started online streaming in 2004
- Hires 20 musicians to classify songs along 400 attributes
- 5Billion thumbs so far, 1 out of 4 songs are thumbed!
- 750K songs in their genome adding 10k new songs a month
- Songs are played in 3-4 song sets within a given station
- 70% of artists not on major label – goal is to help the underdog
- Half of their music not on major label compared to 5% of radio being not on major label
- Big challenge is to not be too repetitive – that is their challenge
- Their lead designer’s previous job was to make exhibits for a children’s museum! They care that much about simple UI ?
- “Disocverometer” is one dial they think of adding (choose how much you want to hear new stuff vs stuff you know)
- When u add extra options, u lose more than u gain – it’s been proven for them…..less features is best
- iTunes and amazon are fulfillment methods for buying songs you like, when they added a third fulfillment method the overall conversion drops!
- Pandora sells more music on amazon and iTunes than anyone else and most of their stuff is ranked 100k (long tail artists)
- Commission they get is in single digit percentages
- They get commissioned on entire shopping cart from amazon – so if you click to buy a song and then buy a TV…they get a cut of that!
- 80% of stations launched from typing in artist names
- Audience mirrors population of US
- Vision is to build infrastructure to enable matching musicians and people, to help live acts find and serve their audience
- Pandora is 1% of all of radio – they want to be more of that!
- Only serves US market
- No prereqs for new artists, just need to be avail on amazon – they have a submission process
- They pay $.02/hour in licensing costs
- No API anytime soon – otherwise monetization would be tough since small changes in the interface have large impacts on conversion
- iPhone is a huge success, 35K users a day are added on iPhone added
- Half of radio listened in-car – they see huge potential there
- Pandora is on over a 100 devices, from cars to TVs
- Game consoles are big target incl Xbox – they want to get in that market