• by Laurie Sullivan  @lauriesullivan, May 11, 2009

The search industry sits on the cusp of many innovations. The latest: the computational knowledge engine WolframAlpha expected to launch within a week. It's based on the software program Mathematica created by its 49-year-old namesake, Stephen Wolfram.

Wolfram's engine has been billed as a "Google killer" by many. Soleil Securities Senior Internet Analyst Laura Martin points out that Google, with its $128 billion market cap, has begun to attract real competition just as capitalism would suggest. "Plenty of venture capital money will become available to fund any potential threat to Google because the stakes are so high in becoming the next Google or displacing the current one," she said.

Google's revenue stream relies on the volume of clicks times the price, so anything that hurts, diminishes or threatens click volume is bad for business. Soleil Securities estimates that 15% of searches are done by people looking for answerable questions. If WolframAlpha could make the process more efficient, clicks on Google, over time, would gravitate toward the Wolfram's search engine.

Despite its potential, the WolframAlpha has challenges. Unlike Google, the search engine does not index Web pages (like this one -- so if you searched Wolfram instead of Google, you might never read this), but rather draws information from databases packed with information. It relies on humans to update data and algorithms to keep the information current.

It can't find stores that sell Jimmy Choos or Canon cameras, but quickly computes and spits out facts to questions such as how many Nobel Laureates were born on a full moon or the magnitude and location of earthquakes worldwide within the past 24 hours.

After gaining access to test the site, it's clear it can't answer all complex questions. Not yet, anyway. Questions like the "average housing price in Orange County, Calif." stump the search engine. From that query it tells you that "Price, Utah" is not in "Orange, Calif."

Knocking Google from the top of the mountain will require more than better search results. It means WolframAlpha would need to perform "so amazingly better it becomes shocking," according to Kevin Lee, chairman and CEO or Didit, New York. "People have habituated themselves to Google, and even changed how they search to make Google work better for them," he said. "WolframAlpha is another run at the holy grail of search, a natural language interface that taps backend algorithms that learns over time what people actually meant when they searched and what the best answers were."

The six degrees of separation between Google and WolframAlpha are tied to the company's founders. Google co-founder Sergey Brin once spent a summer interning for Wolfram. Some search engine optimization professionals have suggested the relationship could spur the integration of the WolframAlpha search engine into Google's backend. The integration could provide information Google doesn't offer.

Wolfram said during a Web demonstration last week that the information his search engine provides that Google doesn't might include comparing the height of Mount Everest versus the length of the Golden Gate Bridge, WolframAlpha seems to cast a serious shot across the semantic Web's bow to compliment, rather than replace classic search engine services, according to Marty Weintraub, search guru and aimClear president. "Most SEOs see it as a potentially indispensable research tool," he said. "Certainly, the service will be a 'go to' engine for certain academic endeavors; however, the ability to vet data seems far from a given. We're most excited to see the service will be available on a tiered set of APIs for mashups, making data available to blend with other reports."