OK, even though I really love the power of tagging things and think *auto tagging* will quickly becoming a cornerstone of how the world will effectively managing the maelstrom of web content effectively, I’m not really feeling the new launch of Delicious – the site and application that in an important way is the grandfather – which in web years that means you have at least passed out of drooling infancy – of content tagging.
What? I have not even reviewed the new site or APIs? Yes, I’m blogging before enough thinking again (sorry, Sarah Lacy!)
Delicious’ basic tagging idea is very good – users tag stuff and share tags with people and the main site then has a body of information that can be used to determine the sites most appropriate for various tags. The challenge of course is that noboby much wants to spend that extra 10 seconds or so tagging stuff, unless of course you are a search engine optimizing person in which case you are going to be *far too willing* to tag stuff. This disparity in tagging enthusiasm can easily distort the results, especially for popular commercial terms. This is why I would argue that the best tagging is automatic or based on simple behavior observations rather than direct user feedback. Google, MSN, Yahoo alll have this type of massive behavioral data stream and I want to see them process it to improve the search experience. For example if, after millions of searches, 4 out of five people who do a lot of dental searches click to the site “dentistry.com” and stay there for several minutes after a query for “dental information”, you can be fairly confident that the site is a good one for that query. This is a simple equation but data can be processed in far more complex fashion to reveal a lot more about how others are searching and finding things. Generally this will give us a lot of insights.
So, I’m hoping a lot of folks use Delicious and tag like crazy, but I’m not holding my breath…