TripAdvisor: The Travel Company That’s Really All About Data

2/8/10Follow @wroush

TripAdvisor, which was founded 10 years ago this month, is a study in paradoxes:

• The Newton, MA-based company’s family of travel-related websites attracts 35 million unique visitors each month, making it one of the most successful Web properties in the Boston area—yet few people even realize the company is headquartered in New England.

• TripAdvisor’s sites are all about helping consumers plan trips to far-away destinations—but the company’s founders and managers don’t consider themselves travel enthusiasts.

• InterActive Corporation acquired the startup in 2004 made it a unit of Seattle-based Expedia (which was later spun off as an independent company)—yet TripAdvisor operates almost autonomously and is still led by its founding CEO, Stephen Kaufer.

• TripAdvisor’s most valuable asset has long been the millions of hotel and restaurant reviews contributed by its users—yet it’s now branching into the one segment of the travel industry, air travel, where customer opinion (read: chronic dissatisfaction) seems to have little impact on ticket sales or airline operations.

To reconcile all these apparent contradictions, it turns out, you need to look at TripAdvisor’s history. It’s a very long one, at least in Internet years. Remember GeoCities, Excite, Go, Blue Mountain Arts, AltaVista, Snap, or Xoom? Neither do most other people—but they were all among the top 15 most-visited websites back in 2000, when TripAdvisor got its start.

TripAdvisor's hotel directory pageThe way Kaufer tells it, TripAdvisor began as a specialized search engine designed to help people locate first-hand reviews of travel destinations, but became a destination unto itself once the founders recognized the value of the reviews being written by its own users and figured out how to make money on them. Kaufer says the company was “a couple of months shy of going out of business” in late 2001 when it finally discovered that it could collect handsome fees whenever readers clicked through to hotels or other businesses to make reservations. That’s still the main source of its revenue today.

Almost since it started, then, TripAdvisor has been a lead-generation engine more than anything else: it attracts visitors with its huge collection of user-generated reviews, then funnels them to travel-industry sites to complete their transactions. “Think of it as a glorious Yellow Pages filled with all the information you might like to find about where you’re going, with contact information, candid photos, and reviews,” Kaufer says.

That means the company’s main challenge has always been building and maintaining a highly scalable website with a very fast back end and an intuitive, easy-to-navigate front end. Which means, in turn, that it’s full of people who were hired based on their engineering smarts, not their affinity for travel.

“We’re a fairly data-driven company—we’re always looking for ways to automate or improve something through algorithms,” says Adam Medros, TripAdvisor’s vice president of product. Both of Kaufer’s companies prior to TripAdvisor sold programming tools, a sector that “couldn’t have been further from travel,” Medros points out. So the evolution of TripAdvisor, he says, is “not necessarily an avid-traveler story but more of a problem-meets-solution type of story.”

Candid Camera

The original problem, in this case, was that Kaufer and his wife wanted to plan a Caribbean vacation. “I popped online to check out three resorts the travel agency had recommended, and darned if I didn’t find plenty of sites that talked about these resorts,” Kaufer told me in an interview last month. “But unfortunately they all said the same thing—the same boilerplate descriptions with the exact same photos. It wasn’t until I found a posting on an AOL member’s personal home page, with some candid photos and a very candid description of one of the resorts, that I realized this place was not what the brochure said it was going to be. We went elsewhere and had a great time, but my wife said, ‘Golly, there’s got to be a better way to do travel research on the Internet.’”

So TripAdvisor was born, in the form of a travel search engine. “We were taking the opposite approach from Google, which was at the time trying to surface search results that were highly authoritative,” Kaufer says. True to its search engine roots, the company simply provided links, then sent visitors off to other sites to read the reviews: “I wanted to surface the results buried on the back edges of the Web—just one person’s opinion about something.”

So it was perhaps to be expected, looking back, that the startup had trouble getting traction. “We had a site that we liked but we hadn’t figured out how to make money,” Kaufer says.

But during the winter of 2001, he says, the company made a discovery: when it displayed relevant text ads next to its search results, click-through rates for those ads were much higher—about 15 percent, as opposed to industry averages closer to 0.2 percent. If a TripAdvisor visitor was looking at a search result page for the Westin Indianapolis, for example, he’d be far more likely to click on a text ad reading “Book a room on Expedia” than if the same ad appeared on some other page. In effect, TripAdvisor was providing travel companies with qualified leads. (This is, of course, exactly the same phenomenon that has generated such dizzying profits for Google and other providers of keyword-based advertising.)

In March 2002, TripAdvisor turned a $70,000 profit, and the company hasn’t looked back since. “From there, it was global expansion, covering the whole U.S. and other parts of the world,” says Kaufer. “A couple of years later it was local language expansion, into Spanish, German, Italian, et cetera. Eventually [in 2009] we branched beyond hotels, attractions, and restaurants into flights and vacation rentals. And along the way we built a media group that bought other travel sites, so we have TravelPod, CruiseCritic, SeatGuru, BookingBuddy, SmarterTravel, and three or four others.”

Actually, it’s nine others, and that’s not even counting the company’s 15 local-language sites. These days the company has 530 employees—about 300 of them based in Newton, with the 70-person London office as the company’s second largest. Its sites contain listings and reviews of 400,000 hotels and 500,000 restaurants across 70,000 cities worldwide, and in the travel sector its Web traffic is rivaled only by Expedia itself.

Mixed Reviews

While TripAdvisor’s massive database of reviews is its best-known asset, another interesting paradox is that the founders didn’t do much, at least in the beginning, to solicit or highlight them. “Originally it wasn’t really a big push,” Kaufer says. “Anecdotally, people told us that they loved the information they found on TripAdvisor, and that they’d like to post their own comments. So we added a ‘write a review’ button.” For a while, the company kept the links to outside reviews at the top of each listing, and buried its own readers’ reviews at the bottom, Kaufer says. “But when we looked at what users were actually doing on the site, they were skipping over the outside sites and going straight to our reviews. So we reversed the order.”

So while the “Write a review” option serves an important social purpose—giving visitors a chance to say whether they agree or disagree with the existing reviews, and to give back to the community by adding their own perspective—the decision to highlight them was, again, largely data-driven.

As TripAdvisor grew into a major repository of unmoderated reader reviews—meaning that angry and disappointed travelers are just as free to post their comments as happy ones—TripAdvisor ran into an interesting dilemma early on. Quite a few hotel and restaurant owners were put off by TripAdvisor’s openness. This was still the early 2000s, remember, before blogs and Twitter and Facebook had made the notion of global, instantaneous opinion-sharing commonplace. “A lot of folks in the hospitality industry were not ready for the concept of somebody being able to advertise their complaints to a worldwide audience,” says Kaufer.

In all fairness, Kaufer points out, it’s easy to see why a small-business owner like a bed-and-breakfast manager might be upset to see a 1-star review from a customer who was angry that the registration desk wasn’t open at 3:00 in the morning. “Our policy is, the customer stayed there, so they get a chance to voice their opinion. But that’s why we have always offered the management a response capability, so they can tell their side of the story.”

And over time, says Kaufer, the explosion of online customer reviews has pushed the travel industry to increase quality, and has allowed the best properties to achieve the success they deserve. “I still meet hotel owners who don’t like to shake hands with me because I’ve hurt their business, but far more often I get people thanking me for helping their brand shine,” he says. So a problem that originally had the potential to alienate TripAdvisor’s advertisers—hotels and restaurants themselves—turned into an asset.

Flight to Quality

There’s a hoary old saying from the dot-com days: “Innovate or die.” Whether it’s true or not, it’s unlikely that TripAdvisor could have survived for a decade—much less grown to its current position as the 366th most trafficked site in the world, according to Web traffic reporting service Alexa—if it hadn’t kept adding new services. Not only has the company acquired a long list of other sites that specialize in information about cruises and other vacation opportunities, but in the last year it’s begun to go beyond its traditional listings of restaurants, hotels, and attractions by adding extensive information about flights and vacation rentals.

In August 2008 TripAdvisor acquired a majority stake in Cambridge, MA-based vacation rental listing and review site FlipKey, which had launched only five months earlier; it added FlipKey’s listings to its main site last spring. And that’s just a first taste of the new categories the company is investigating, Kaufer says. “We’ve got the biggest travel marketplace and the biggest travel audience out there, but they know us primarily for hotels,” he says. “I think the opportunity is there to teach everyone that we are good not just for hotels and flights and restaurants and attractions and vacation rentals, but eventually surfing lessons and limo services and tour guides and any service that a traveler might be interested in.”

But the company’s venture in 2009 into airline price comparisons is perhaps its biggest, most technically challenging, and riskiest move to date, as it represents a departure from the company’s traditional review-driven lead generation model. Since it’s not clear that customer reviews would make any difference in the market for flights—where customers make purchasing decisions based mainly on prices and schedules—TripAdvisor is forced to attract visitors to its new flight meta-search service, launched last February, based on the volume of its raw data.

Of course, this is exactly what many other flight price comparison sites try to do. To make its service different, according to Bryan Saltzburg, TripAdvisor’s general manager of new initiatives, the company makes a point of connecting to more flight databases than anyone else and offering a few unique bells and whistles, such as a “fees estimator” that attempts to calculate the full cost of flying, including baggage fees, airport taxes, and the like.

“Similar models would be Farecast or Kayak, but we have really just built a better mousetrap,” says Saltzburg. “We have more online travel agencies participating, better content, and the largest result set online, and we return the best price more often than any other well-known travel site out there.” Anecdotal evidence suggests Saltzburg is at least partly right: A search last week for flights from Boston to San Francisco in mid-summer returned 2,429 possible itineraries, with a low price of $348. The same search on Orbitz returned only 241 results, with a low price of $347.

Out of the Park

It’s too early to tell whether TripAdvisors’ flight search service will become a major part of its business—Saltzburg says “millions” of people have shopped for flights using the service, but the company hasn’t released more exact figures. But it may be the company’s record of success exploiting travel data that has persuaded its owner, Expedia (NYSE: EXPE), to keep its hands largely off.

It’s unusual for startup CEOs to stick around for two or three years after their companies are acquired, let alone six. But Kaufer says his autonomy at TripAdvisor has been nearly complete; otherwise he would have been out the door years ago.

“I agree it’s totally rare—there are very few similar situations in the industry,” Kaufer says. “When people ask me why I’m still here, they say ‘Do they have you locked up with golden handcuffs?’ and I say ‘No, I have enough money.’ I don’t need to work, period. I love the job and I love the autonomy that Expedia has given TripAdvisor and the whole media group. I’ve told my boss and I’m happy to tell the public that when the rules of engagement change and Expedia wants to start micromanaging what we do here, it’s their right. They own the company. But I won’t be CEO.”

In another 10 years, Kaufer may or may not still be in charge. But by then TripAdvisor is likely to look very different—if only because the travel industry itself is evolving. For example, the company is expanding aggressively into China, where it already owns major travel sites Daodao.com and Kuxun.cn.

“It’s just a matter of time before that represents the largest travel market in the world,” Kaufer says. “I want to be a serious player there. There aren’t many American companies that have succeeded with that objective, but I want you to be hearing two years from now that we were the local company that knocked it out of the park in China.” With TripAdvisor’s long experience making products that scale up exponentially, it may just have a chance.

Wade Roush is a contributing editor at Xconomy. Follow @wroush

By posting a comment, you agree to our terms and conditions.