Fetch Technologies Inc.

Founded:

1999

Core Business: Precision search technology for business and government clients


Employees in 2007:

40


Employees in 2006:

30


Goal:

To becomes as much of a household name among target businesses as Google is among consumers


Driving Force:

The desire of businesses to efficiently gather and aggregate data from the Internet without having to keep a team of

programmers on staff

Selling businesses on the advantages of deploying artificial-intelligence search technology to probe the so-called “deep Web” wasn’t quite ready for prime time in 1999 when Fetch Technologies Inc. got off the ground.

Ironically, it was the more recent popularity of consumer search engines such as Google that helped the trio of computer scientist founders make the advantages of their precision extraction technology more apparent.

“Companies began to realize that they were missing things, because Google and Yahoo can go a thousand miles wide, but only an inch deep,” said Chief Executive Robert Landes, who joined the company in 2003. “Fetch may only go an inch wide, but it can go a thousand miles deep. And that’s what our clients need.”

Today, a variety of companies use Fetch to gather real estate listings and job openings, or even to monitor blogs and news sites for items about themselves and competitors. Farelogix Inc., for example, uses Fetch’s technology to integrate fares, scheduling and other information from airlines, hotels and other travel destination sites.

And officials at Menlo Park search company Zvents Inc. say contracting Fetch was essential to their goal of becoming a more complete aggregator of local meetings and community events around the country.

“Fetch has a significantly richer technology because it’s more of a power tool,” said Gordon Ra, Zvents’ chief technology officer, whose company vetted comparable software before signing with the company. “Instead of having to specify every single detail you want each time you search, the program learns what you want it to do.”

VIN numbers

Landes uses a car-buying analogy to describe the difference between the approaches taken by Fetch and a standard Web crawler.

Someone using a conventional search engine for information about a Honda Accord typically gets only general information about the car and perhaps links to dealers whose Web sites the user again would have to search manually. Using Fetch, the user could use the same search criteria to compile the VIN number, color, description and style number for every Accord for sale in the U.S.

Companies can license Fetch software and hire their own staff to run it in-house, with Fetch providing occasional technical support. The Fetch Agent Platform, which works like a software “wizard” to walk the user through the process, enables customers to efficiently construct their own searches with no programming required.

Despite that ease of use, the company’s most popular product for commercial users is its subscription service, in which an assigned Fetch customer representative constructs the searches and delivers the results.

“You could call us a Google for business,” Landes said. “We can get into places Google can’t, and automate the process to the extent that a company doesn’t have to keep a bunch of programmers on staff to set up the searches and run the program.”

A typical corporate subscription can range from $150,000 to $300,000 a year. Clients simply wanting to license the software can pay anywhere from $75,000 to $300,000.

The company also is in the early stages of a third service: to become a warehouse of commonly used data that a customer can access on a project-by-project basis.

Fetch engineers also are implementing the technology in creative new ways. A recent project melded Fetch with social networking software to determine the most connected, and thus potentially the most influential, opinion makers among a certain group of professionals, so that the client could better target their marketing message.

“Machine learning been around in academic circles for a number of years, but our special sauce is applying it to extracting information for commercial uses,” said Gary Minton, a company founder who continues to serve as chairman and chief technology officer. His partners Craig Knoblock and Yigal Arens still serve on the board but spend the bulk of their time on research at the Marina del Rey-based company ISI, where Fetch was first developed in the early 1990s.

Demonstrating scalability

The Defense Advanced Research Projects Agency, the U.S. Air Force and the National Science Foundation were among agencies that funded development of the early Fetch technology at ISI. Post-911, other security-oriented agencies began using the Fetch platform to rapidly import and integrate data from multiple Web sites and databases for emergency response, location intelligence and antiterrorism activities.

However, by the time Landes came on board in 2003 as chief executive, the El Segundo-based company still had only seven employees and no commercial contracts.

A year later, though, the company signed its first large corporate client, J.D. Powers & Co., the Westlake Village-based marketing firm best known for its automobile customer satisfaction surveys. That was followed a few months later by a deal with BizRate.com, the Santa Monica-based consumer comparison shopping Web site.

“That was the deal that really demonstrated that our product was scalable,” Landes said. As corporate subscriptions rose, Fetch’s annual revenues jumped from $1.1 million in 2004 to more than $5 million last year.

Fetch faces several challenges in transitioning from a startup to a mid-size corporation with hundreds of employees. Among them is handling the growth and maintaining a close-knit corporate culture.

The company has grown from nine to 40 employees over the past three years, and has moved at least three times and likely will outgrow its current facility on Rosecrans Avenue this year.

Because a customer representative can only handle a limited number of clients, there’s an ongoing challenge in the Los Angeles job market to hire enough technologically savvy employees with the right people skills, in addition to engineers and programmers to work on new applications. Landes notes that he has 17 openings for engineers alone.

“One thing that’s special about us it that we’re an L.A. company, not a Silicon Valley company,” he said. “Staffing is a challenge, but it’s a nice problem to have because it means we’re growing.