Today I had an idea for a service that can make a difference on how we buy books. Actually, it can make a difference in us buying anything at all. Notice I say buying rather than selling. I see that Amazon recommendation engine is more a way for them to sell than for us to shop smartly.
So let’s say you’re not part of a book club, but you know some people that read good books and review them online on their blogs, other services, privately or on twitter. So you have this information available and this information is there for lot’s of books, but many people out there. The idea is to mix the reviews with the reputation and trust I have with my closest friends in their book reading habits and mix them to get a service that will show me books I might be interested in based on this.
So let’s say my group of friends that I’ve decided I trust and assigned trust levels to, read Crush It by Gary Vaynerchuk, then some of then reviewed it, some others just tweeted that it rocked, some other friends quoted a lot from it, etc. All this information is captured and stored in their respective category (I see a potential use of some variation of microformats to make the algorithm recognize what is what). Then I have all this information on books, such as Crush It available to me for when I would want to shop for a interesting book.
Let’s say that it’s a Sunday morning and I want to shop books. Instead of going to Amazon I’ll go to this service, let’s call it BookLogue. You have your settings already created. For example, for books you’ll just want to browse in shops that deliver items to your town in no more than 5 days and that offer free shipping for purchases of 12 dollars or more. Also you have some preferences for publishers that use recycled paper and books that are no longer than 500 pages.
You have all this settings and preferences and now you have the files of reviews and recommendations you “book club” has decided to share with you. In here you will have reviews you’ve already read, reviews about the book from people you trust highly and ratings. Now you search for Social Media Business Books. The algorithm will now browse for books that qualify in the category only in the shops that fulfill all the shopping settings we already entered the system. Then it will crawl through all the ratings according to the relevance you’ve given to all the members in your “book club” and to the ratings and reviews they’ve made. Finally the search results will show books that your “book club” has read. In this case Crush It! will have the higher ratings and the results will show some reviews that will help you decide whether you want that book or another one. The next result will have a lower rating & number of reviews and a less positive sentiment found inside them and so on.
What do you think of this model? What changes would you make? Or something you’ll like to add?
