I have been surfing Amazon and other sites trying to understand the scope of polls and ratings and how they are implemented and differ from one type of site to another. As I have been browsing around Amazon, I discovered the Collections area and Recommendations Explorer area had grown and had a listing of everything I had ever bought. It was kind of eerie to go through, but also kind of cool. I had been wanting the site to stop showing me items I owned as items I would like to own. I figured that the software developers at Amazon had to be creating something smart enough to keep track of what I had bought, therefore it should know not to recommend something to me that I already owned.

My discovery in the Recommendations area was that if you go through the list and rate it, it seems not to show up as a recommendation anymore. Maybe this has been here awhile, but I had never seen it. So I went through and rated the 300+ items I had previously bought. And it made me think about the value system we are developing into these ratings widgets. Are 5 steps enough? What if something is really a 10 or a 2 1/2 or can't be dissected into so few spaces. I found that I rated a lot of things 5 and 4 simply because 3 seemed too mediocre and 1 and 2 would beg the question of why did I still own it in the first place. Or perhaps I have crappy taste.

Anyway - today I was looking up a book that Christina Wodtke bought for someone else from their WishList (my wishlist)and I happened upon a variation of the ratings feature I had never seen before. Amazon is predicting how I will rate this book - assuming I buy it and read it - based on how I rated another book from my list. The assumption is this book is of similar style and content and that if I like one i would like the other. Kind of cool. Kind of creepy.

The interesting thing about all of these features is that Amazon is building up content without having to pay for it and they are creating "sticky" features that keep people coming back over and over again. I could spend hours browsing and reading peoples Lists and reviews of items and I might be sparked into buying something that I otherwise wouldn't have bought. Mulitply this and it may trigger a lot of extra purchases. Or not. No matter what it does on the e-commerce side of things, they have inadvertently created a community that leads people to follow and read the work and lists of other people with similar interests.