Amazon visitors have not taken to tagging Amazon's books in significant numbers. With thousands of times the traffic, Amazon produced a tenth as many tags as LibraryThing. What's going on?
He goes on to talk about the power of numbers that inevitably feeds on itself; as more people tag more people find it useful which leads to a larger community. Passion also plays a part because the overwhelming number of active 'taggers' also have more books. Those that have less than 200 titles rarely tag books. (Since fees kick in over this number the users are also more financially committed to librarything).
Critical mass is important, even if we can't pinpoint the line. Ten tags are never enough; a thousand almost always is. Unfortunately, Amazon's low numbers translate into a broader failure to reach critical mass. With ten times as many tags overall, LibraryThing has fifteen times as many books with 100 tags, and 35 times as many with over 200 tags.
It is a very interesting article on the power of community and worth a read.
Lorcan Dempsey suggests these types of sites also represent examples of the 'Network effect':
Regular readers of this blog will not be surprised to hear me say that they also highlight a structural issue for libraries, how to provide services at the network level. These new services are network level services. They are aimed at the general user, not an audience circumscribed by region, or funding or institution. And, additionally, they provide an integrated service, moving the user quickly through whatever steps are needed to complete a task.
The other component of interest to me is that sites like librarything are becoming sources of viable bibliographic information and indeed the NY times article suggests that Librarything maybe selling or licensing some of its tagging data:
For example, he is in the process of selling some of his recommendations data, which is based in part on tagging statistics, to other sites that sell books and book information.
At Bowker one of our most important and costly editorial tasks was to assign subject classification to titles so that they could be found either in our own products or in the products of our customers. With well over 200,000 titles per year this is an expensive excercise and while it can be automated (and was) the process suffers from obvious limitations. Firstly, the subject classification methodology is quite rigid and is not always intuitive. Secondly, subjects change over time and books previously categorized could benefit from additional or changed subjects. Thirdly, the application of subjects is subjective and can pose a limitation on the subjects applied to the title. A subject expert wants to be as accurate as possible in applying the subject classification for relevancy and integrity. In BooksinPrint, we had an average of slightly more than two subjects per title over approximately five million records. Many had more and a few had more than 15 but the average was two. Many users of Librarything, myself included, have placed more than two tags against most titles.
The ability of a community to supply a range of subjects that reflect the work, what's important about the book and what becomes important in the wider world may provide a vast addition to traditional bibliographic database work. I believe the social network applications and the structured approaches will work in concert but increasingly it is the social aspect that needs to be actively solicited for inclusion into the traditional bibliographic databases.