Last month, Spock Networks launched to the general public. It was in private beta for the first part of the year but I'm pretty excited to see it go live and am actively watching to see it grow. Spock is a startup company that promises to serve up accurate information on people including name, birth date, address and relevant articles, tags and links. The results were pretty impressive. I found my college roommate immediately with a search on her name. They had her picture, several books that she has written and a completely accurate tagging of who she is - ie Harvard grad, Indian, award-winning writer. It had a much harder time finding some folks from work who have managed to stay off the web and some of my friends who have hard names equivalent to John Smith. I did find myself though my entry wasn't that compelling, I'll have to go and try to edit it ;-)

But more seriously, Spock identifies three main search spaces (general web search, product search and people search). According to the company, 30% of all web searches are for specific people (e.g. Britney Spears, President of GM or the “vanity search” for oneself). The company contends that this type of search is not adequately addressed and optimized in existing web search engines such as Google. Given this outlook, Spock is specifically indexing people online using social networking sites, Wikipedia, blogs, photo websites and web searches. At launch they promise to have over 100Million people indexed.

Although the people-search space is crowded (see competitors in Wink, Streakr, ProfileLinker, LinkedIn, ZoomInfo, UpScoop, PeopleFinders, Reunion.com, ZabaSearch, PeopleSearch.com, and the list goes on and on...), Spock is worth a closer look. Unlike ZoomInfo, which is positioned more toward a business directory and requires an upgraded service to view names, Spock seeks to index the whole world of people and be a true search engine. Wink, the closest offering to Spock, does a simple crawl of the 3 major social networks and therefore has less reliable data. While the majority of its competitors focus on scraping information from social network websites (such as MySpace, Bebo, Hi5 and LinkedIn) or from public business directories, Spock takes people search a step further using probability algorithms to determine whether the profiles across networks and web sources reliably describe the same person. In addition, they will allow a user to “claim” their profile giving them more weight over the information for that profile. The profile collects images of a person, articles about the person referenced from Wikipedia and on the web and Spock also highlights relationships between people’s profiles. You can tag other profiles and view tags that other people have entered as well as add tags to your own profile and upload photos. The tagging features of Spock allow a search on “presidential candidate 2008” to return a list of the current candidates. A search on drunk driving returns Mel Gibson and Vice-President Cheney (that's entertaining, our blog has a bad language filter - I just found out after writing Cheney's firstname). It’s a fascinating “mashup” of functionality - de.li.cious + Wikipedia + LinkedIn + Flickr all centered on people, personal and contact information online and the relationships of one person to another.

Spock is addressing an area of weakness in the Google search engine and potentially defining a whole new search competency. If they manage to gain a foothold, the Spock profiles would aggregate your personal data across the web, determine popularity or more accurately, reputation, online through how someone is tagged and allow user voting to validate information on a person’s profile. Spock is interesting for social communities as a way to drive traffic when people do people-searches. Spock closed $7M in December 2006 and has been consistently generating a good deal of excitement from a demo at the Web 2.0 Conference in April 2007 to a number of blogs since their public launch. It may still be early to make a call on Spock, but they are a company to watch.