Blinkx Plans Ad Model for Desktop/Web Search Tool
A 10-employee San Francisco-based search company is raising eyebrows with a tool that does what the big boys are still only planning: combining desktop search with Web search.
A 10-employee San Francisco-based search company is raising eyebrows with a tool that does what the big boys are still only planning: combining desktop search with Web search.
Blinkx offers a desktop application that examines the document a user is looking at, then suggests similar Web pages, blog entries, news stories, video clips, even documents from the user’s own hard drive.
Users can also type a search term into the application and receive results from all those sources. Though it’s early in the game, Blinkx execs say they plan to make money by offering ad placements, along with sponsored versions of the software that would incorporate e-commerce search of the sponsor’s site.
Google is reportedly working on a tool that combines Web and desktop search, while Microsoft execs have said they are also working on such a unified search offering. (HotBot unveiled a free desktop search app, HotBot Desktop, as part of its browser toolbar earlier this year.)
Co-founder and managing director Kathy Rittweger says the idea for Blinkx arose from her frustration with traditional keyword search.
“We wanted to sort of say, take the whole article and put it in the search engine and see what happens,” she said. “Why is it that we have been trained to accept that we have to use only up to about ten words?”
Rittweger shared her frustration with computer scientist Suranga Chandratillake and the two began to further develop technology he’d previously worked on at Autonomy. Autonomy released a similar-sounding product, Kenjin, with much fanfare in March of 2000. It even signed a partner, UK retailer Tesco, to use the Kenjin engine on its site. Blinkx is licensing some of its technology from Autonomy, but says the new product is very different.
One thing that’s changed significantly since the Kenjin days is consumer familiarity with the Internet and with search. In May 2004, Nielsen//NetRatings found the number of user sessions/visits per month to be 64 at work and 31 at home. In February 2000, just before Autonomy released Kenjin, those numbers were 39 and 18, respectively.
Much remains to be ironed out in regard to Blinkx’s proposed ad model. The company appears to be focused first on building the tool and getting it to users, and plans to figure out how to make money later. Initial versions of the software feature one sponsor spot per search, but the company hasn’t ruled out adding more. Blinkx plans a bid-for-placement auction model, in which advertisers would buy keywords. On local hard drive searches no ads will appear, to avoid possible privacy concerns.
The company is also looking into building branded versions of the technology for e-commerce players, who in turn could offer it to customers. Under that model, Blinkx would receive a percentage of the revenue generated through the tool. The company says it’s “in discussions” with possible e-commerce partners.