Spock offers an ambitious “people” search engine

Updated

Spock, a secretive Menlo Park start-up incubated with $1 million from Clearstone Venture Partners, will unveil a search engine for people by the end of the year.

From a demo we’ve seen, we think it could be a powerful addition. Spock could take this in some interesting directions. Its main challenge will be to wean users from Google as a first stop, though more on that in a sec.

When Spock launches, it will have 100 million profiles of people in its database, by far the largest open repository of profiles anywhere. Spock delivers a mixture of facts and research on a people, but also opens a profile to social input, giving it a touch of Wikipedia.

This move is a no-brainer, and it makes you wonder why no one has done this yet.

LinkedIn, ZoomInfo and other people-contact related sites were built in different eras, and have focused on specific subsets of people (LinkedIn and ZoomInfo on business execs, for example). Spock, however, exploits all the latest tagging technology and the exploding number of public profiles on the Web since social network sites like MySpace became popular last year.

Scrubbing millions of profiles from the Web wasn’t an obvious thing to do when Palo Alto’s LinkedIn launched several years ago. LinkedIn began as a contact site, allowing people to request meetings through their layers of relationships. It has since tried to move toward a more open model. Indeed, LinkedIn is aggressively building out its people profiles even as we write. (Last week, it also kicked off a major expansion into Europe and Asia as part of a land-grab, with a German version to go live soon.)

Spock starts from the other end. Spock dispenses with the “contact” element of LinkedIn. It is an open site, for people seeking information about other people.

ZoomInfo, which you must pay a subscription for, has 29 million profiles. LinkedIn has about 9 million profiles, and wants to grow to 100 million by 2008. Spock’s 100 million, meanwhile, will only grow, according to co-founders Jaideep Singh and Jay Bhatti.

If Google is a place to find Web sites that are relevant for your search, and Amazon is place to find goods, then Spock wants to let you find people, they argue.

Here’s an example of how it works: If you type in “actress,” Spock returns results like Google — with listings down a page. In this case, the first entry is Felicity Huffman, who Spock’s engine finds as the most relevant for “actress.” (Now, if you type in “actress” into Google, you’ll see why Spock has a chance; there are few actresses in the results, except for the annoying site ActressArchives at the top). Moreover, as both Spock and LinkedIn make their profiles more popular, these will rank higher in Google’s results anyway.

Continuing with our “actress” example, you first get a photo of Huffman, but you also get a bunch of tags underneath telling you how she is relevant. For example, there’s tag for “Oscar nominee for best actress,” and “Desperate Housewives,” for which she is well known. There’s a “Wikipedia” tag. If you click on these tags, Spock will take you its relevant results for that tag. This gives users a way of searching for information related to the Huffman.

The tag font size gets smaller if Spock’s engine detects the tag isn’t relevant for the person. So if users create a “sexy” tag for Huffman, the tag may get larger or smaller, depending on how many people agree. Spock gives users an option of clicking on the tag and selecting “yes” or “no.” If they select no, Spock factors this into its database. Then, if you type in “sexy actress,” Huffman will have fallen slightly in the ranking. Spock has built ways to keep people from gaming the system. If you want to add tags, for example, you have register — one way for Spock to monitor usage.

Nicole Kidman is the second result under “actress,” even though she won an Oscar (Huffman was only a nominee). Why would an engine rank a nominee higher than an actual winner? Chief executive Jaideep Singh says Spock’s engine factors in hundreds of variables for its algorithm on determining relevance. This is Spock’s secret sauce, he says. We asked if his algorithm takes advantage of Google’s APIs. He said yes, but there are many other sources, he said.

Spock will make money by running relevant advertising beside the profile results.

Spock has seven employees in Menlo Park, two in India, and six more part-time.

Singh and Jay Bhatti met in business school. Bhatti has a background in consulting, having worked at Accenture, Deloitte and Microsoft. Singh was a VC at Clearstone and worked at WindRiver. Jeff Winner, VP of engineering hails from Friendster, eGroups and Netscape.

David Stern, the investor at Clearstone (who contributed an op-ed to VentureBeat here) said the investment is a return to his firm’s roots as in investor in consumer companies — eToys, Overture, PayPal, United Online, MP3.com and eMusic are among them.