The Idea That Wouldn t Let Go

The Idea That Wouldnt Let Go

When I was a freshman in college, somebody lent me Miles Daviss Kind of Blue, which I had never heard before, Jason says. It really moved me. I thought it was fantastic. I wanted to find more of this wonderful music, so I went to a record store and asked the sales clerk, ˜What can you tell me about Miles Davis? I really like Kind of Blue. And he said, ˜Miles Davis is listed under D.

This was an accurate answer, but it was clearly not the answer I was looking for. I thought there must be a better way to learn about music, that a record store has a higher calling than just to list albums alphabetically . Over the years , I thought about a variety of thingsabout clipping album reviews and sticking them to record store shelves , or wearing headphones in the store, and as you picked up a CD you would automatically hear sound samples from the CD. I wanted to find a better way to connect to the music, to appreciate it more deeply and to discover new albums and artists .

This problem was in my mind for six years. Every time I went to a music store I would think, ˜Hmmm, how can this be better? One day, I had a blinding flashthe answer was an online music store. It occurred to me that I could make a database of all the albums ever created, of all the reviews ever written, of all the biographies of the artists. I could put that on the Internet, people could connect to it, and they could learn about whatever they liked . They could go in and look up Miles Daviss Kind of Blue, learn about the album, look at recommendations, understand what else he has done, and see his career time line. They could really learn about and experience the music. I could support the system by selling the albums. That was the original inspiration, in February of 1994.

My twin brother, Matthew, actually took over most of the programming, although his background is in astrophysics. I quit my job, he came back from Philadelphia, where he had been pursuing a masters degree in neuroscience, and we both moved back into our parents house, setting up shop in their basement .

Matt had no programming background, but he taught himself to program in C. He took out the 1978 Kernighan and Ritchie monograph, the book that originally defined the C programming language, and he read that. The downside of that was that the 1978 version did not include a lot of the improvements, both stylistic and functional, that have occurred in the sixteen years between when the monograph was written and when Matt was reading it. But nonetheless it was a good start for him, and he learned some of the more elegant techniques over time.

We created a database, buying or licensing information from various books and magazines and other companies that could give us information or the data structures that we needed. So we built a compatible database of all this data that we got from reviews and bios and album titles. We built an interface that people could connect to over the Internetbasically a Web pageand we built a whole processing infrastructure so people could easily select the albums they wanted to buy and get their credit card information, and the CDs would be in a warehouse somewhere and would be picked out and sent to them.

The time it took from the initial conception to actually selling CDs online was six months. We started a business called CDnow and ran it. In our first month of operation we sold $387 worth of CDs; we had a gross profit of $14. We took the whole gross profit and floated it back into the business and went from there.