Leap of Faith

Inexcusable waste of time. Huge disappointment. Crappy.

Those were some of the more politely expressed sentiments that greeted the Jan. 7 launch of Wikia Search, a new Internet search engine conjured by the minds behind Wikipedia.

Naturally, with the unflattering reviews raining down like morning e-mail spam, the company’s executives were eager to respond. Within minutes of a drubbing from the influential Silicon Valley blog TechCrunch, Wikia Inc. co-founder Jimmy “Jimbo” Wales stepped into the site’s very public square and fired back.

Sort of.

“Yeah, the search sucks today,” wrote the 41-year-old Web entrepreneur on the blog’s comments section. “But that’s not the point. The point is we are building something different.”

Different indeed. Wikia, working from modest headquarters in San Mateo, Calif., is attempting something at least as audacious as Wikipedia’s launch was in 2001: It is building a for-profit search engine to compete with Google. There are no legions of engineers. Instead, the endeavor is powered as nonprofit Wikipedia is: by volunteers. (Wikia and Wikipedia are different companies; all they have in common, besides a “by the people and for the people” philosophy, is their co-founder, Wales.)

As Microsoft’s and News Corp.’s recent interest inYahoo makes clear, search engines are big business. Last year, online advertising for search engines totaled $11 billion to $12 billion, according to Chris Sherman, executive editor of Searchengineland, an online news publication about search engines.

So it seems almost quaint that Wikia would jump into the fray with a low-tech strategy you might call faith-based.

Here’s how it works, so far.

Go to search.wikia.com and enter a term--”second president of the United States,” for instance. Early last month, that search yielded 54,582 results. The top three results were all “Ron Paul 2008--Hope for America.” Not one of the top 10 referred to John Adams. (By comparison, a Google search of the same term yielded 14,600,000 entries; nine of the first 10 were highly relevant.)

So how does anyone imagine such a start-up will one day compete with Google, or even Yahoo? The idea behind Wikia Search is that users around the world will work to improve the results each time they perform a search, adding the vital human touch that Wikia executives say is missing in the big search engines. Humans, not computers, voluntarily assign star ratings to results as they come up. The more stars a result receives, the higher up the results list it goes. The higher up the list, the more hits it gets.

“It’s an enormous amount of power--the decision where to send all that traffic,” says Wales, speaking by phone from Mumbai, India, where he’s just spoken with university engineering students about the search engine. “Humans should have a hand in it all.”

Take a look at Wikia Search results now and it’s not immediately obvious how, as a user-volunteer, you’d assign ratings. But roll the cursor over the results and star ratings will appear, one to five stars for each entry. If you find the entry completely irrelevant, give it one star; wonderfully relevant gets five. By mid-February, the system wasn’t quite there yet, though. Clicking on the stars in an attempt to rate the result gave you a message that said, “We’re currently using these just to test out the star rating and are not yet storing them permanently.”

Wikia’s headquarters may be just down the road from Google’s gleaming, mammoth Mountain View offices with its top-notch gym, infinity pools and 17 cafeterias at which employees eat gratis, but Wikia’s corporate culture couldn’t be more different. Located in a Main Street-like shopping district populated by banks, antique stores and cafes, Wikia inhabits the second floor of a building it shares with a furniture store, a European clothing shop and a prenatal imaging center.

Inside, it would be tough to seat more than two dozen people in the main conference room, with its metal folding chairs and mismatched IKEA tables. Wikia doesn’t employ a team of chefs; a converted office functions as part kitchen, part supply room. For recreation, the 10 employees don’t have a gym, but a massage chair and a big-screen television equipped with a Nintendo Wii system.

The mostly twentysomething staffers, dressed in jeans and T-shirts, are quietly hunched over laptops in a large L-shaped open room. But they have nothing to do with running the company’s search engine--they’re focused on other parts of the business.

Wikia was co-founded in 2004 by Wales and British Web entrepreneur Angela Beesley largely as a social networking site; this area still accounts for the most significant part of the privately held company’s business. (Wikia is backed by $4 million in venture capital funds, primarily from the widely respected Bessemer Venture Partners. It also received an undisclosed amount from Amazon.com.)

On Wikia, users interact in online communities on thousands of topics, including politics, television, movies, pets and video games. Revenue from that portion of its business grew by more than 650% from 2006 to 2007, according to Chief Executive Gil Penchina, who declined to give specific income figures. “So we’re doing just fine,” he says. “But the search business clearly has a much bigger potential market.”

With about 50 employees in its three main offices--in San Mateo, New York City and Poznan, Poland--the company lays claim to a million pages of content in 70 languages. The Polish outpost focuses on computer engineering, while the two U.S. offices concentrate on content. Wikia executives chose Poland as a location, Penchina says, because that country’s tech-savvy workforce is well-educated and accustomed to working with minimal budgets.

So where are the employees dedicated to Wikia Search? Other than Penchina, who doesn’t even have his own office (he sits in the open workspace with the rest of the Wikia staff), just a handful are paid to work on Wikia Search.

“The whole search team works from home,” half-jokes Penchina, a former EBay vice president. In other words, Wikia Search is relying on user-volunteers around the world to achieve its success.

A month into the project, Wikia’s Iowa-based “search architect,” Jeremie Miller, says Wikia Search has 1,000 to 2,000 registered users, “roughly double what we had upon launch.” They have registered about 100,000 star ratings and have created nearly 20,000 “mini-articles,” he says. (Mini-articles are essays about search terms that Wikia posts on its results pages.)

Still, Wikia needs many, many more volunteers to become a viable search engine. “It’s going to take 10 to 100 times [our current number] to get this thing up to scale,” Miller says.

Why would volunteers donate their time to help a for-profit company?

“Wikia Search is basically doing something unique,” says Mark Williams, an 18-year-old college student who lives on the south coast of England. He estimates that he puts in up to 20 hours a week rating search results for Wikia and writing mini-articles. “It’s changing the future of how people can search, so that they know how and why certain results are coming up, and if they don’t like the results then they can say so.”

Users such as Williams have faith that others will want to join in. There’s no doubt that the aura of Wikipedia’s astounding success--more than a third of Americans consult the citizen-generated encyclopedia, according to a report from the Pew Internet & American Life Project--sustains the morale of volunteers and principals in these tough early days of the venture.

Others aren’t convinced that Wikipedia’s contagious glory will translate to participation in a for-profit venture.

“That’s a huge flaw in their plan,” says Charlene Li of Forrester Research, a Cambridge, Mass.-based media firm. Li recently wrote a report advising businesses on how to handle the enormous influence of Wikipedia. “Wikipedia worked really well because there really wasn’t anything else, but Wikia Search is up against very tough competitors who are very, very good.”

The best search engines today--Google and Yahoo--operate on complex and largely secret computer algorithms to instantaneously mine millions of Web pages for results. Wikia Search utilizes algorithms too, but Wales and Penchina think their army of user-volunteers will create a better mousetrap.

Though cyberspace is home to several thousand search engines, only a handful generate significant traffic. Last year, the Google site alone accounted for 56% of searches in the United States. Of the roughly 113 billion searches conducted, Google performed nearly 64 billion of them, according to Comscore Media Matrix, which measures Internet audiences.

Wikia’s future depends on what is perhaps its most revolutionary and closely held belief: that people are “essentially good” and that this idea should be reflected in a website’s design and function.

One might think that would invite hackers or other online ne’er-do-wells to manipulate content for their own benefit.

“I like the restaurant analogy,” counters Wales, whose fondness for author and philosopher Ayn Rand led him to name his daughter after one of her characters. “Everybody in there has a knife and they could all stab each other, but we don’t think that way. If you did think that way, you would have some pretty sick social spaces. And the same is true online. If you start out designing a site thinking about all the horrible things people are going to do, it’s not going to be a very good site.”

Still, like Wikipedia, Wikia Search has mechanisms in place to protect itself from online evildoers and those who simply don’t play well with others. For instance, a merchant who deletes a competitor’s product in a search result can be blocked for a day by volunteer administrators, or in drastic cases, forever.

All this raises the question: Is it possible to improve upon Google? If so, what’s wrong with it that should be improved?

Critics have complained that Google’s search function is a “black box.” No one knows precisely how search results are calculated, and that continually arouses suspicion. Google also struggles with something called “disambiguation”--the process by which its search engine determines which results to show for a word with multiple meanings, such as “jaguar.” (It could be a car, an animal or a computer operating system.)

“Language can be very ambiguous,” says Searchengineland’s Sherman. “Wikia Search could offer some real editorial guidance to the Web in one of the few areas that Google doesn’t excel in yet.”

Wikia executives have another criticism of Google. They look at its inability to control spam from its e-mail service, Gmail, and they see a parallel with the search engine: Google has been unable to entirely eliminate irrelevant search results.

“I’m constantly surprised by how much money, how many computers, how many algorithms go into preventing spam, and yet every day I look in my mailbox and I go, ‘Dude, that’s not right,’ ” Penchina says. What’s striking, he says, is how much better people can do than Google’s software in filtering out unwanted e-mail. In the same way, he says, volunteers can do that with search results.

Then there’s the issue of transparency, which is hugely important to Wikia’s core philosophy.

“If you’re suspicious of a search result on Wikia,” Penchina says, “you can go in and look at the code and go, ‘Oh, I found that tricky little thing they did and now I’m going to tell the press about it. Right now, I go to Google, Yahoo or Ask.com and I don’t know how they came up with their decisions. Sometimes I see their products with links to other parts of their sites, and I wonder if those were the result of algorithms or if somebody got involved.”

If somebody got involved. What Penchina is alluding to is paid placement. (On its website, Google assures users, “There is no human involvement or manipulation of results, which is why users have come to trust Google as a source of objective information untainted by paid placement.”) At Wikia Search, built-in transparency ensures that paid placement would never be allowed to stand.

Legions of naysayers have already lined up against Wikia’s quixotic quest. Critics have little confidence the search engine will survive in a fiercely competitive arena--no matter how successful its co-founder might have been with the once equally improbable Wikipedia.

“I think that’s kind of a silly way to look at it,” says Wales, who lives in Florida and spends almost half the year abroad. “I just like doing stuff that seems cool. I don’t really think in those competitive terms.”

Many have blasted Wikia for releasing a product that is so obviously inferior by today’s marketplace standards. But Penchina counters that the criticism ignores the company’s ethos and misses the point. He says that the old business model dictated that “people spend a lot of time behind closed doors, building something, refining it, cleaning it up and making it just perfect and then going, ‘World, here it is! Look at this amazing thing I just built, and it’s tremendous and it’s all mine and you can’t have it.’ And we come from a world where you start small and you ask for help. Someone else might say it’s a piece of crap, but we look at it and go, ‘Yeah, but look at the potential.’ ”

The site’s early volunteers don’t seem troubled by its poor performance to date. “It’s going to get much better in the future,” says Williams, the volunteer in England, “and if my children use it in the future then I will be proud to say, ‘I was part of that.’ ”

Meanwhile, all the Wikia-bashing has also resulted in some goodwill, even sympathy, from critics. Sherman of Searchengineland was critical of Wikia Search’s launch, but he says he would like to see the underdog pull off an upset. “I’m kind of rooting for them because of the David versus Goliath aspect,” he says. “Google has started taking on aspects of the Microsoft evil empire--it’s too big, too dangerous. And isn’t it cool that we’re going to have an upstart that will give us an alternative?”

Others warn it would be a mistake to underestimate Wikia Search based on its early performance. “Anyone who is selling Jimmy Wales short over the launch is going to be in for a surprise,” says John Palfrey of the Berkman Center for Internet & Society at Harvard Law School. “He’s leveraging many of the same things that made Wikipedia a global force. I think Wikia can have a huge impact on search engines over time.”

The $64,000 question in business self-help books asks, “What constitutes success?” Like the results of an Internet search, the answer to that question is highly subjective. But to Wales, it’s obvious.

“Success is if we can push forward toward our political goals and I keep having fun,” he says. In hard numbers, that translates to: “Hey, we’d be ecstatic if we get 5% of the searches. That’s a dream.”

How big a dream, time will tell.