In late December 2003, two of the men who invented the Web travel business caught up at dinner. Over nouveau American cuisine, former Travelocity CEO Terry Jones and Steve Hafner, who had just resigned as EVP of, talked about their successes and — as the wine flowed — their regrets. "We congratulated each other," Hafner recalls, "but we also discussed why we didn't enjoy using our own products."
Hafner and Jones had built their businesses with a common vision: to replace travel agents. But the sites, along with, merely became virtual agents. They only make money when visitors book travel. And visitors only do that between 2% and 6% of the time, according to Compete, an Internet market research company.
Dinner conversation eventually turned to creating a company that could deliver on the original mission. "The eureka moment was, What if we focused on giving the airlines leads instead of selling tickets?" Hafner says. They could collect money on a pay-per-click basis, the way Google does, and wouldn't have the customer-service expenses of Orbitz or Travelocity.
Within weeks, Hafner had found: funding from Joel Cutler, a VC who also attended the December dinner; a partner in Paul English, former Intuit VP of technology; and a chairman and resource in Jones. He also set a bold goal. "We wanted to make one site," Hafner says, "to rule them all."
Ironically, creating afor the travel business involves hours of time in the car. One day last spring, Hafner and executive VP of corporate development Keith Melnick invited me to trek two hours from Kayak's Norwalk, Connecticut, headquarters to the company's design-and-engineering office in Concord, Massachusetts, where they regularly spend the day discussing the site's features. Just after 6 a.m., they pile into the back of the company's stretch Chrysler 300 and scarf down Munchkins from Dunkin' Donuts while studying the site's performance on their laptops. So far, the numbers have been impressive. The site now has 35 million visits a month, and Hafner says Kayak has been profitable since August 2007. In December, Kayak acquired SideStep, a rival Web 2.0 travel-search site. The combined company produces $3.5 billion in annual travel sales, and it rakes in $140 million in revenue.
After a stop for breakfast at Babico's Cafe, a 1950s-style greasy spoon awash in neon signage, they hit the Concord office, where the first thing visitors see is a video screen flashing the number of searches that visitors entered in the last 60 seconds, the number of searches entered today, and the number of users in the last 20 minutes. "That gives us a sense of urgency," says English, who lives nearby and runs the engineering office. "It reminds us why we're here."
Today's agenda involves updating the site to let users look for flights by leg, as well as round-trips. Hafner's business model revolves around driving clicks. About half of Kayak searches end in a click on one of the results or on the ads that run to their right. Although Kayak doesn't make as much as Google does on an average travel-related search — Hafner admits that Kayak collects less than Google's 90 cents per click on travel keywords — the relative bargain makes Kayak a good deal for the ailing airlines, and the pennies add up. As airlines impose more cutbacks this fall, Kayak will likely be a more essential tool for both consumers (looking for deals) and airlines (fighting for customers).
The programming work for Kayak's new features has already been done, and Hafner, Melnick, and English sit in a conference room watching director of product management Lincoln Jackson present some design options. Should the feature be labeled as "new"? Which words should be clickable links? And should that be made more obvious by pop-up bubbles that say, click here to choose this departing flight? Or, as English suggests, are pop-up bubbles a sign that a design is too complicated?
English has an obsession with usability, to the point that he makes every Kayak employee answer emails that come in to the help address. "We got pushback on why a $150,000-a-year engineer rather than a $30,000-a-year support guy should talk to a grandmother from Florida," he says with some understatement. "But we're going to learn from her."
"Kayak has gotten where it is with word of mouth, and it's growing very quickly," says Gregory Saks, general manager of Compete's travel practice. "But they have to spend money to generate a bigger audience." Kayak currently attracts about 30% as many visits as Expedia, which generated $2.7 billion in 2007. That's one reason why Kayak bought SideStep: its audience. SideStep adds another 3 million unique visitors per month to Kayak's 4.5 million.
Kayak's marketing emphasizes the ways in which it is a better search engine. It offers more alternatives, flexibility, and airlines. For instance, Kayak includes, which doesn't make its flights available to online travel agents. Hafner has experimented with TV commercials, but Kayak relies mostly on the kind of click-through advertising it sells itself. (It's a top-five buyer of travel keywords.) That's more cost effective, although it only heightens the importance of Kayak converting its own leads into leads for the airlines and then getting those consumers to start at Kayak next time. Of course, few of Kayak's users understand this process, and the site doesn't really explain it. "If customers just see us as a travel site," Hafner says, "that's fine."
After a full day of meetings, the crew usually hits the Pleasant Cafe for pitchers of beer and games of Golden Tee. Then, Hafner and Melnick climb back into the Chrysler and put on a DVD for the ride home. Tonight, it's Pulp Fiction. Something about its underdogs searching for redemption resonates.
A version of this article appeared in the September 2008 issue of Fast Company magazine.