Mon

Feb 26
2007

Brady Forrest

Brady Forrest

Comparing LibraryThing and Amazon's Tags

librarything.gif
Tim, the founder and CEO of LibraryThing, estimates that LibraryThing has 10x the number of tags in books than Amazon does despite having a fraction of the traffic (13 million vs. 1.3 million). Thingology, the LibraryThing blog, has his very thorough and well-thought out post comparing LibraryThing's tagging system with Amazon's.

LibraryThing is a virtual bookshelf that allows members to organize, tag, and share their books. I'm sure Amazon sells many of the books that end up getting listed on LibraryThing.

Tim sampled 100 ISBNs from each site to get these figures. Tim's post provides theories on why the disparity may exist and what to learn from it, but when it really comes down to it this is the key:

Tagging works well when people tag "their" stuff, but it fails when they're asked to do it to "someone else's" stuff. You can't get your customers to organize your products, unless you give them a very good incentive. We all make our beds, but nobody volunteers to fluff pillows at the local Sheraton.

Which is incredibly similar to a Joshua Schachter quote:
"You have to understand the selfish user" - user #1 has to find the system useful or you won't get user #2. Systems that only become useful when lots of people are using them usually fail, because there's no incentive for people to contribute themselves."

And is summarized quite succinctly by Jason Lefkowitz in the first comment on Tim's post:

People WILL tag things if the tags are useful to THEM.

People WILL NOT tag things if the tags are useful to SOMEONE ELSE.

Amazon's tagging system does not motivate and this isn't surprising. You can't share them (like you can on del.icio.us), there is no user ranking (like there is with Amazon's user reviews), and personally most of the time when I am on an item's page it is before I have read the book and I rarely go back afterwards (which is when I would be in a better position to tag).

Assuming that Amazon wants tagging to succeed what can it do?
Tim lists several ideas for how the the system can be improved (more prominent UI, better URLs, sharing, exporting, using a non-commerce site). I agree that Amazon will need to get this data from somewhere else. I am not certain that it will be able to do it on it's own site. It's too buy-oriented to trigger this need for organization. Amazon will need a more social site that is built around sharing or organizing that can gather this data. I can think of three places where it could get this data:

  • Amapedia - The product wiki that Amazon launched a while ago. If Amapedia opens up with an API and is able to build a community, then it may be able to gather tags in larger quantities. (Radar post about Amapedia's launch)
  • AllConsuming - A Robot Coop site that allows members to list, share, and tag products that they have consumed. They have an API and this data could be easily correlated back to Amazon's product pages via ISBN. I've written before about their relationship previously (1, 2).
  • Shelfari - Amazon just made an investment in LibraryThing competitor Shelfari. It's a site for members to share, list, review books that they possess. Tagging could easily be added and the could be shared with the Amazon product page.


tags: web 2.0  | comments: 1   | Sphere It
submit:

 
Previous  |  Next

0 TrackBacks

TrackBack URL for this entry: http://blogs.oreilly.com/cgi-bin/mt/mt-t.cgi/5269

Comments: 1

  J.O. From Urban MVP [10.11.07 09:58 AM]

Tagging books is definitely an issue with Amazon. Having the ability to share these tags I can also see some issues with user manipulation of tag results. Overall though tagging is an ongoing issue with many web 2.0 websites mine is an example. Keeping tag results accurate and on point has taken an immense effort. Also there is a whole problem with the tagging format due its inability to recognize partial searches. For example a keyword that may contain a dash in between them won't show up for a the same tag without the dash. Ie "keyword" "key-word".

Post A Comment:

 (please be patient, comments may take awhile to post)






Type the characters you see in the picture above.