October 25, 2006
At this point in time I am writing the conclusions of my thesis. During the journey of this research, my interest for topics that are related to my research is increased significantly. Therefore I keep track of some sites that regularly write about related topics, for instance TechCrunch of Michael Arrington. Yesterday TechCrunch posted about a new startup called iLike. It has some similarities with Last.fm and Pandora. Recently, many services started which can be compared to Last.fm and/or Pandora.
The reason I write about iLike, is because one of my conclusions it that recommendations are important in these kind of services. I did not try the service, but it is nice to notice that companies like this are there, and that they are related to my study. My first impression is that it can be compared to Last.fm, because they seem to support a community.
I think it would have been a service that could have been of value for my thesis, if it had just existed six months ago…
July 26, 2006
While I am writing my thesis, and reading literature on mass customization, characteristics of information products and other parent disciplines for this research, it is getting clearer that existing literature has a gap on customization of digital or information products. That makes my research interesting for me and maybe for some other people too.
As can be read on this blog, digital products are being customized and personalized on the Internet. Digital products in the form of music, or music stations, can be customized on Last.fm and Pandora. As I had contacted Frank Piller before on this research, he showed his interest in the subject. Now on his mass-customization weblog this topic is getting attention, as he has reviewed Pandora and some other services that allow music to be personalized.
It makes me curious if customization and personalization is getting more attention in scientific literature in the near future. But if Frank Piller is starting to mention it, I’m sure it will.
May 9, 2006
The interviews that I had scheduled so far are all done. If the respondents are reading this, I would like to say "Thanks again!". At this point, before I have analyzed the interviews, I have only transcribed them, I can already say their input is very valuable for my research.
As for now it has no use to go into detail of the contents of the interviews. First, I will treat the conversation confidentially. I respect the openness of the subjects, and will not misuse it. Second,like I mentioned, I yet have to analyze them. So I probably will give more detail after that.
For now, I can say my research is progressing as planned, and I will continue to update this blog on a regular basis. The feedback I get is motivating me to keep updating it.
May 3, 2006
Just one day before my interviews with Matt Nichols and Tim Westergren of Pandora, the system is getting an important update. They're extending their playlist generation with user input. It seems like an extra way to customize radio stations. When you let the system know the played song was not good, that feedback will be used for all listeners. It is called 'community feedback'.
Source: here and here.
April 27, 2006
May 4th will be a exciting day for me, as I am allowed to interview Tim Westergren, one of the founders of Pandora and now also Chief Strategy Officer. From the Pandora site: "A musician's musician, Tim Westergren is obsessed with helping talented emerging artists connect with the music fans most likely to appreciate their music. He founded Pandora and now serves as its Chief Strategy Officer. "
With this extra interview I probably get more input for my research, and enabled me to compare interpretations of Matt Nichols, who I will have an interview with the same day. I will try to find more people who I can have an interview with, it only can benefit my research.
At this point I'm preparing for the interviews. The book InterViews (Kvale, 1996) is helping me with the preparation, a very good book.
April 20, 2006
I am pleased to announce that I just made my second appointment for an interview, this time with Matt Nichols of Pandora. He agreed on having a conversation, it will also take place on May 4th 2006, just like the interview with Martin Stiksel of Last.fm. Time for me to go on with the construction of the interview questions.
I think Pandora and Last.fm are quite similar in their online digital product, despite they use different classification methods. The difference lies in collaborative filtering which is being used at Last.fm, opposed to the content recommendation system which is being used at Pandora. Steve Krause did a nice comparison between the two.
I hope to have the opportunity to have some additional interviews, but with these two I sure can complete my research.