I saw this fun contest on Twitter called PerfectBusiness.  The rules were simple:  Boil down a great business idea into a single tweet.  Now here is a contest that I could do well with.  Here was my idea tweet:

@PerfectBusiness #micropitch : Pandora for unsigned bands. Seed real artists, listen to unsigned similar artists. Focused on discovery.

They re-tweeted and I replied:

(Them) RT @glenlipka: @PerfectBusiness #micropitch: Pandora for unsigned bands. (We’d listen right now.)

(Me) @PerfectBusiness #micropitch Pandora for unsigned bands – Build<1yr<1MM. Monetize band access,shirts,pay-what-you-can,etc. Open API.

I have a very low chance of winning, but wouldn’t that be fun?  Besides the nice trip, it would be awesome to build this business.

My big problem with Pandora is the lack of variety and surprise.  It’s fine to hear an old familiar favorite, but I really want to discover new musical talents.  Pandora has a good user interface, but the reliance on signed bands puts them in an awkward spot.  They have to follow the rules.  This means you can’t bookmark a song.

The experience I would create is the same as a street musician with an open guitar case in front of them.  You listen and if you like it, you throw a dollar into the case.  Maybe you buy their homemade CD or t-shirt.  Because it’s digital, I can even ask what kind of performer I want.  I want to hear people like Shakira or Green Day; BAM! There would be some no-name band that reminds me of those other bands, but its new and different.

Obviously, there would be folksonomy as well as self-described links to what bands sounds similar.  You might even say, “I like Shakira, but I am interested in learning about punk rock.  Who will help me bridge the gap?”

I think a system could be implemented quickly though.  Less than a year with some good engineers should do the trick.

The great thing about this is the business model.  You don’t need advertising.  Bands upload a logo and we partner with a custom t-shirt vendor.  Have the bands upload mp3s and partner with itunes or cdbaby to make the more popular songs available.  All the service needs to do is take a cut.  %10 sounds like a good number.  If a band is selling, they probably are popular and their bandwidth is paid for.  If a band isn’t selling, they probably use some mp3 storage, but that is cheap.

It’s fun thinking of little businesses like this.  I imagine a service like this would immediately get the people with music Otaku interested. People who might even have their own bands.  This is the kind of service that could replace the normal “sign with a label” model for bands.  I remember mp3.com had enormous potential, but finding an artist you liked was tedious and hard. CDBaby is ok, but you are listening to one song at a time, and have limited ability to find something unexpected.  Doing it right would unleash the potential.

Anyone got a million dollars to get it started?  Sir Richard Brandon maybe?