technology
People-Powered Internet Grows Up
On today's Internet, algorithms rule. But a handful of startups are using large-scale human participation to offer online services that computers alone can't deliver. Can human judgment scale with the Web?
The question of the next decade is: How can we find what we want -- the perfect job, just the right pair of shoes, exactly the news that's important to us -- amidst the maelstrom of information that's available on the Web? Google, of course, is the de facto answer, it's algorithms generating a ballpark guess at what we want when we type in a few search terms. But the burgeoning mass of data on the Internet is threatening to outmode such robotic tools. So a growing number of start-ups is putting forward another strategy for filtering the Web: Use human judgment first, computer power second.
Of course, human judgment is unreliable, inefficient, expensive and difficult to scale. It's also a relatively scarce resource compared to data, which grows online at an exponential rate. Here are four people-powered sites, and how they plan to keep their people-powered business models durable as the Web shifts and swells.
ThisNext
Sites like Amazon have been taking buyer reviews for years, but on ThisNext, user opinions don't all carry the same clout. If you're an inveterate product-suggester and people follow your personal recommendations (and therefore, your personal style), you achieve "Maven" status, a high-profile position that gives your opinions increasingly more visibility. Mavens don't get paid, but they do earn what CEO Gordon Gould calls "social capital." And social capital, Gould says, "is becoming an increasingly valuable asset in the Web 2.0 universe."
TheLadders
Marc Cenedella, CEO of the 5-year-old company, says the model scales well -- up to a point. If the site grew to, say, 30 times its current size, "you couldn't keep the community together," he says. "It really has to be a human-sized group. At over 1,000 people, you couldn't have the elbow-to-elbow passing of knowledge. That shared judgement pool is essential."
Wigix
For this to work, sellers' listings have to accurately reflect age, model name and make of every item listed. (If, for example, a seller lists a "Flatscreen TV," that TV-buyer's sale won't execute.) To solve this problem, Wigix asks users to apply online to become "category experts" who oversee categories and check that all the listings therein have accurate information and SKU (or stock-keeping unit) information. For that service, the category experts -- who are not in-house employees -- receive a commission of 1% of the sales in their category. The more accurate the catalog listings in their area, the more sales will execute, and the more money they'll make.
Reddit
"There's a core group of Reddit users that are submitting and editing," says Alexis Ohanian, Reddit's 25-year-old co-founder. "But some point we're going to hit the ceiling of nerds on the Internet who want to share cool, geeky stories. It's depressing, but we know that we will." So to keep the 3-year-old site growing at a brisk pace, the Reddit team--which says it receives about 4.3 million unique hits per month--recently announced a new function on the site that will allow users to create Reddits for their own online communities.





