Monday, March 10, 2008
There's still room for human-built search tools
Way back in internet history, the primary way to discover new sites on the web was through human-built directory sites. In the current Google, keyword search-dominated world, it's easy to forget this. While geeks like us know that a few people out there are trying to give human-powered search a second chance, it has been essentially forgotten in favor of machine-driven bots that crawl the web and produce results based on complex and ever-changing algorithms.
Honestly, I love machine-driven search. Way back in my Yahoo! days (I was there when the above logo existed) it was clear that hand built web directories could never scale - and this was when the internet was "small." But those directories DO have a place and I hope that some of these newer incarnations of the old Yahoo! idea succeed where Google fails.
How is Google failing, you ask? With billions in revenue, massive search share, and endless innovation, how could Google fail? Well, it starts with simple searches. I'm currently evaluating shopping cart technology and also looking for NetSuite consultants. Searching in Google for these vendors works to a degree, but never turns up an exhaustive and thorough list of vendors. I've found that after searching for shopping cart vendors for over a week, I'm still finding strong candidates who failed to make it into Google's top 30 results (and who looks at more than the top 10 anyway?)
What the world needs is a good set of lists. Lists that Google can find and present when users are doing these kinds of searches. The world can not be limited to the top 10 results for any particular query, but unfortunately that's how it works right now. For most businesses, if you don't crack the top 10 for your key terms, you might as well not exist.
I hope that Mahalo and Squidoo succeed for this reason. Google's style of search has a place, but it can't be the be-all end-all for finding information. Google needs to find a way to direct users to lists of things that are compiled by humans because, at the end of the day, it is human intelligence and filtering that provides the most value to the end-user.