Inside the Mad, Mad World of TripAdvisor

Published March 13, 2015 12:00AM

The Koryo Hotel does pretty well on TripAdvisor, all things considered. The Internet never—and I mean never—works. The towels are “thin,” the sheet thread count low, and the milk powdered. Watch out for the “giant mutant cockroach snake hybrid“ in the shower. Guests even have to pay for the pool.

Still, the Koryo, a pair of dull beige towers connected near the top by a sky bridge, is the number-one-rated hotel on TripAdvisor for Pyongyang, North Korea. Despite its many faults, it garners three and a half out of five “bubbles,” in TripAdvisor parlance (so as not to be confused with the star systems that signify quality in hotels and restaurants), across nearly 90 reviews.

Perhaps this is no surprise. People come expecting the worst, and with expectations so dismally calibrated, something like hot water starts to sound pretty amazing. Reviewers carefully note that many of the hotel’s quirks—you can’t walk out the front door unaccompanied, for instance—are out of the manager’s hands. (You’ll have to take that up with the Supreme Leader.) It may not be the Ritz-Carlton, the sentiment goes, but considering the fact that you are staying in the marquee property in the showcase capital of the world’s most repressive regime, it may be best to, as one reviewer counseled, “just chill out, have some beers, some expired Oreos from the gift shop and make friends with the other tourists.”

The fact that so many people are so earnestly reviewing a hotel that they have not themselves chosen (accommodations are selected by government-sanctioned tour operators), in a situation in which management is hardly likely to care, in a country where the Internet-driven wisdom of crowds is a remote fiction, speaks to the curious power of TripAdvisor, which, in its decade and a half of existence, has changed travel as we know it. The reviews demonstrate the abiding urge to share and the faith that sharing—even for that one-more-grain-of-sand 13,786th reviewer of the Bellagio Las Vegas—will make someone else’s experience, or quite possibly everyone’s experience, that much better.

No matter your destination, you will, at some point in your research, visit TripAdvisor. The company, with the humble mantra “real hotel reviews you can trust,” has become—on a rising tide of 200 million user reviews and counting—a travel-industry Goliath, able to turn obscure hotels into sold-out hot spots, carry new flocks of visitors on digital word of mouth to quiet destinations, even rewrite the hospitality standards of entire nations. For travelers the impact has been equally profound. What begins as a simple search-engine query becomes an epic fact-finding mission that leaves no moldy shower curtain unturned, a labyrinthine choose-your-own-adventure—do you read the one-bubble rant?—in which the perfect hotel always seems just one more click away. For all the power of the service, it raises deep questions about travel itself, including, most pressingly, who do we want—who do we trust—to tell us where to go? “The future,” Don DeLillo once wrote, “belongs to crowds.” Are we there yet?

Many years ago, when the Internet was still just hobbyist bulletin boards in a handful of homes, I went backpacking with a friend across Mexico. Like everyone did back then, we carried a Bible-worn copy of Let’s Go: Mexico, which represented pretty much our entire universe. Its recommendations were usually decent but were based largely on one college student’s legwork. Who knows if he actually looked at the room or just hurriedly scanned the lobby? What was her standard for “clean”? The first night in Mexico City, our hotel had small creatures in the walls, unmentioned by the guide. Asking around, we eventually landed in a cheap and obscure place. That’s how it worked: We were in an information-poor environment. We gleaned recommendations from the gringo grapevine. You never quite knew what to expect, but wasn’t that why you were traveling?

Barbara Messing, TripAdvisor’s chief marketing officer, remembers those days, too. “There was that community of travelers in East Africa or South America who were circling the hotels in Lonely Planet that were really good or telling you what was closed or had good breakfast,” she recalls to me by phone from the company’s headquarters, near Boston. “That entire offline community got imported onto TripAdvisor.”

And how. The site now has reviews of hotels, restaurants, and attractions in more than 45 countries, with contributors (all unpaid) adding their comments at a rate of 115 per minute. Some 890,000 hotels are listed on the site, and TripAdvisor boasts one of the largest collections of user-contributed travel photographs in the world. (The collection of shower-drain photos alone could fill a museum.) On its myriad forums, even the most banal query (e.g., “Does this resort have 110v plugs in its rooms?”) seems to excite a flurry of replies, often within 24 hours. Though the site sometimes seems like a place people go to air complaints, as Messing tells me, “In general, our reviewers are a happy bunch.” For hotels the average rating is above four bubbles. Because people use the site to plan their trips, she says, the ratings can be as much about “thanking the community for pointing them in the right direction.” In other words, reviews of TripAdvisor itself.

These days, you can hardly visit a restaurant in a beach town without seeing the TripAdvisor owl in the window or finish a bike tour without being implored, via follow-up e-mail, to leave your feedback on the site. “It hasn’t changed travel like jets changed travel,” says Henry Harteveldt, an industry analyst with Atmosphere Research Group. But “it has changed the satisfaction we can get from a trip and the ability to better understand the destinations.” Where travel is concerned, Harteveldt says, “information becomes instantly static and stale. With TripAdvisor, you know if a hotel that a few years ago was fresh and wonderful is still fresh and wonderful—or has become tired.” More important, he says, it “has empowered the consumer by making hotels and other related businesses far more transparent.” Sure, guests always had the ability to complain to the front desk or on one of the comment cards left on the nightstand, but that information went no further than the management. (As Heads in Beds author and former hotel employee Jacob Tomsky notes, “We used to confidently toss comment cards in the trash”—or, as they also referred to it, the “t-file.”) Now one’s smallest observation—the ecstasies of the rainfall showerhead, the disappointments of the room-service toast—has a global audience.

Those reviews carry demonstrable weight. A study by Cornell University’s Center for Hospitality Research found that for every percentage point a hotel improves its online reputation, its “RevPAR” (revenue per available room) goes up by 1.4 percent; for every point its reputation improves on a five-point scale, a hotel can raise prices by 11 percent without seeing bookings fall off. This has been a boon for smaller, midpriced, independently owned hotels. “Twenty years ago, the brands owned the sense of quality,” says Bjorn Hanson, a professor at New York University’s Tisch Center for Hospitality and Tourism. “If I stayed at a big-name hotel, I knew what I was getting.” That sense of confidence in quality, argues Hanson, has been supplanted by TripAdvisor. Not only can there be variation within a brand, but suddenly that quirky hotel that was once the obscure favorite of a single guidebook gets lifted to market prominence. Thanks to TripAdvisor, a formerly sleepy spot like the Magic Castle Hotel in Los Angeles—ranked number one in the city—is, says Hanson, “able to generate rates and occupancy levels that from a hotel-analyst point of view are quite extraordinary.”

Social media officially took over travel in 2010—or so the scientists say. That is the year, according to the Cornell study, that “guest experience mentioned in customer reviews” became the factor most frequently cited by consumers in surveys about the process of choosing a hotel.

For hotel owners, this meant that a luxury brand or pictures of nice-looking rooms on a website were no longer enough. Instead, believes TripAdvisor’s Messing, service became paramount. “That is why the number-one hotel in a major market is often not the hotel you would expect.” TripAdvisor’s ranking system, she says, operates like a value index: “For what I paid, how delighted was I?” The company closely guards its algorithm, but it’s engineered to emphasize the quality, frequency, and freshness of reviews.

If rankings were based only on the number of reviews, says Messing, “It would be purely a popularity contest or a size contest—the big hotels would always win.”

The impact of the so-called TripAdvisor effect can be enormous. When a group of researchers at Ireland’s University College Dublin examined hotel reviews in the Las Vegas market between 2007 and 2009, and compared the data with reviews in the Irish market (where, they noted, TripAdvisor was a more “recent phenomenon”), they found that while ratings in Las Vegas remained more or less constant during that period, in Ireland they rose from 3.6 to 3.8 bubbles. Their takeaway: as Irish managers scrambled to respond to the novelty that was TripAdvisor, hotels actually got better. Additionally, responses to customer reviews more than tripled over the two years of the study.

Adam Medros, who heads TripAdvisor’s global product division, told me that hotel owners have started thinking of the site as a marketing platform and are “embracing the feedback, even if it’s negative.” A recent internal TripAdvisor study found that owners who respond more frequently to comments are over 20 percent more likely to get booking inquiries. “Saying sorry is important,” says Medros, “but it’s not enough to say sorry you didn’t like the carpet.” Instead, the reply should be: “Here’s our plan for fixing that.”

Indeed, NYU’s Hanson says that TripAdvisor reviews now factor into the “capex,” or capital-expenditure, decisions companies make. “A general manager can go to the owner and say, ‘Look at the comments we’re getting about furnishings or parking lots compared with our competitors,’ ” Hanson says. “Instead of ‘I think we should do something,’ it’s ‘The reviews are costing us occupancy and average daily rate.’ ” TripAdvisor now goes “way beyond the individual traveler making decisions,” he says. “It’s influencing owners and brands. Even lenders will ask, ‘Are we putting our capital into a hotel that’s at risk because it’s getting poor reviews?’ ”

All this is a remarkable turn of events for a site that, back when it was founded in 2000, considered consumer reviews an afterthought. TripAdvisor was conceived as a meta-aggregator of travel information. Cofounder Stephen Kaufer, a Harvard-trained computer scientist who has served as the CEO and president from the beginning, directed employees to link to travel articles around the Web. But the site also let users post their own comments. Very quickly, TripAdvisor realized that the user reviews were getting all the traffic. As the Harvard Business Review described the situation in 2012, “They adjusted to focus on user reviews, such that fresh, authentic content was always available and didn’t cost the company any money to produce.”

In 2004, TripAdvisor was purchased by Barry Diller’s media conglomerate IAC for $212 million and was soon put into a group of travel brands under the Expedia umbrella. Over the next few years, it grew from a startup with $23 million in annual revenue into a multibillion-dollar global brand. In 2011, TripAdvisor was spun off as its own company again. In recent years, Kaufer has orchestrated a series of aggressive if rather quiet acquisitions, from booking company Viator to restaurant-reservations app La Fourchette to airline seating guide SeatGuru to VacationHomeRentals.com. Last year saw the introduction of a battery of new TripAdvisor products, including city travel guides you can use offline on your smartphone and a feature that lets you book an Uber ride to the restaurant whose TripAdvisor reviews you were just reading. The trajectory points toward the brand becoming a kind of always-there digital concierge.

While the big booking companies like Expedia and Booking.com also let users post reviews, no one matches the volume or weight of TripAdvisor, Hanson says. “If it were retail, we’d be calling it the category killer.” Those reviews have become part of a self-perpetuating, incredibly profitable cycle. Visitors go to the site to read reviews written by travelers, book a hotel based on those reviews (with a small percentage going to TripAdvisor), then write reviews of their own, and so on. As Inc. columnist Jeff Bussgang put it, describing TripAdvisor’s 98 percent gross margins, “For every dollar of revenue, the company is able to drop nearly half to the bottom line. I’m not sure the mafia could do better.” Perhaps not surprisingly, TripAdvisor is greatly expanding its headquarters outside Boston, and last year Kaufer took a 510 percent raise, to $39 million, making him the country’s fourth-highest-paid CEO. (Somewhat ironic for a site based on the promise of democratization and transparency, my requests to speak to Kaufer, and to visit the company headquarters, were turned down.)

Of course, when you are becoming a kind of middleman in the affairs of virtually every hotel in the world, there are bound to be complications. TripAdvisor has been the subject of numerous lawsuits that question the veracity of its reviews. When a hotel in Tennessee made a 2011 TripAdvisor list of the dirtiest hotels in America, the owner filed a $10 million defamation suit. Christopher Bavitz, the managing director of Harvard Law School’s Cyberlaw Clinic, which filed a brief in support of TripAdvisor, notes that federal law “essentially says that a website is not to be treated as the publisher or speaker.” It is a case, he says, of what is called “Internet exceptionalism,” and it dates back to the early days of the online world. As he explained it, “If CompuServe were held liable for every crackpot thing that someone said in a message board, the entire thing might grind to a halt.” The court ultimately ruled that calling the place the dirtiest hotel in the country wasn’t defamatory because people understood that it was hyperbolic.

This is not to say that an individual can’t be sued for defamation for a review he or she writes on TripAdvisor. But early precedent seems to go against such lawsuits succeeding: In December, the Ashley Inn, a hotel in Lincoln City, Oregon, dropped a suit it had filed against an anonymous reviewer (“12Kelly”) who’d said the hotel’s “rooms are nasty.” The reviewer, the judge ruled, was protected by media shield laws.

European officials have had a different take. In 2012, the UK’s Advertising Standards Authority decreed that TripAdvisor could not use phrases like “reviews you can trust,” because it couldn’t prove that its reviews were truthful—or had even been written by actual guests. Last December, Italian regulators fined the company more than $600,000 for, as The New York Times reported, “not doing enough to prevent false reviews.”

Indeed, bogus comments are a central challenge for TripAdvisor. The company has cautioned resorts that have tried to game the system by offering incentives to write positive reviews and notes that every single review goes through a tracking system, which maps where, when, and how (desktop, mobile, tablet) each was written. Automated tools designed to sense “algorithmic signals” of suspicious activity are backed up by a 250-person content-integrity team helmed by Medros and made up of staffers with fraud-detection backgrounds in police and intelligence agencies. (TripAdvisor declined my requests to speak to any of them.) If the team determines that manipulation is going on, a hotel could end up with the ultimate mark of TripAdvisor scorn: a red badge that warns users not to trust what they read.

But a more prevalent defense is crowdsourcing. The sheer scale of TripAdvisor creates, Medros argues, a kind of self-enforcing equilibrium: you would have to write a lot of fawning posts to sway the bubbles of the 13,000-review Bellagio. Even then, he says, you will only draw more people expecting that enhanced experience. “Even if you could beat the system, it’s like Sisyphus—the rock is going to roll down the hill.” Users, Medros insists, can “suss out” the outlier in a sea of positive or negative reviews. This speaks to what may be an even greater problem than fake comments: too many.

It begins with a simple message—“Plan and Book Your Perfect Trip”—and a search box. In the background beckons, at least on this day, a perfect coastal village in Greece. I was a few keystrokes away from paradise.

A number of months ago, I was looking for an easy winter getaway for my family. My wife and I wanted something to satisfy us (i.e., not Disney World) and our daughter (i.e., anywhere with a pool or ocean). I decided to plan the whole thing on TripAdvisor.

We eventually settled on Tulum, Mexico, the once ultra-bohemian outpost down-coast from Cancún that has become a kind of gentrified Brooklyn South, both for its East Coast proximity and its shared cultural vibe (read: beards, bike paths, yoga, organic everything). But with copious archaeological sites, spectrally lit underground freshwater pools, known as cenotes, and the sprawling Sian Ka’an biosphere reserve just to the south, it still seemed a worthy adventure.

As I came to understand while planning the trip, the more you hand yourself over to TripAdvisor, the more you get wrapped up in its worldview. I began to see Tulum less as a place unto itself than as a vast Gringolandia construct of projected desires and flummoxed expectations, worried last-minute inquiries about federales and narcos, a galaxy of unsmiling concierges and complimentary upgrades. I wondered how such a colossal gulf could exist between two humans having, at least on paper, the same experience. How can one place be “almost the perfect dream” and also a “horrible experience that no one should go through”? The language of the reviews started to affect me. I closed my eyes and saw phrases like “exceeded my expectations!!!!” silhouetted on my eyelids. In a vision quest of user-generated feedback, I reworked reviews into haiku:

And just when I thought I’d found a hotel, after I’d digested dozens of raves, I would come across that one review—the cockroach in the salad bar. Behavioral economists call this the Volvo problem, after a thought experiment in which a prospective car buyer, having read in bona fide sources like Consumer Reports how great a car is, hears from a neighbor that his is always in the shop. Suddenly, that aggregate of rational, emotionally distant information is overwhelmed by a much more narratively powerful, personally relevant source.

And TripAdvisor reviews are nothing if not narratives, often bursting with details that stray far beyond thread counts and shower sizes—a roster of travelers’ previous trips, the fact that a couple was celebrating their 16th wedding anniversary, or the reflection, offered by one reviewer, that “sometimes I stay up late playing my guitar.” They also, as TripAdvisor’s Messing noted, tend to present things on the sunny side. Camilla Vásquez, an applied linguist at the University of South Florida and author of The Discourse of Online Consumer Reviews, has thoroughly analyzed a corpus of TripAdvisor comments. She notes that, within a “categorically negative review,” there will usually be “an effort to work toward some feature of the hotel that’s positive.” Even in the most withering one-bubble critique, 30 percent of her sample included “some kind of positive statement.” As Vásquez explains, “We all know those people who will complain no matter what, and reviewers work hard to present themselves as not that person.”

As I read up on Tulum, it became more a game of parsing reviewers themselves (are they like me?) and trying to decode the more cryptic comments (“Found the owners to be quite French”). This is why, according to Philippe Brown, of the luxury tour operator Brown and Hudson, TripAdvisor has not necessarily proved the lethal threat to travel agents that it was first assumed to be. “There’s too much info, and so many hidden agendas, like national or regional biases,” says Brown. Visit TripAdvisor, he says, and “you’ve probably learned that a hotel is filthy, according to Italian people, but people from England think it’s great.” There is, he says, “loads of info, but no insight.”

TripAdvisor claims that it’s trying to solve the information-glut problem: instead of giving you 12,000 reviews of the Bellagio, how can it provide, in the first page or two, the half-dozen that are most relevant to you? When I spoke to Medros, he explained that the company was busily beta-testing a number of personalization algorithms. “Some are based largely on past behavior. We had one case of a user who had written tons of reviews of Best Westerns,” says Medros. “When that person looks at a new city, and the first hotel we show them isn’t a Best Western, that’s just wrong.” Medros envisions a kind of Pandora-style travel genome—if you like hotel X, you will love hotel Y. Of course, this works only if the reasoning makes sense: early on in my research, before I had selected a destination, when I looked at one property—“#33 of 48 hotels in Managua”—I was told that I had been sent there because I “researched similar hotels in Captiva Island.” Really? How similar are hotels in an upscale Gulf Coast enclave and a gritty Central American city?

TripAdvisor has also rolled out Facebook integration. Suddenly, those reviews from Facebook friends, and friends of Facebook friends, pop up first (if you’ve given permission). I found that this helped, but only to a point: my family’s TripAdvisor account was linked to my wife’s Facebook page, and often the reviews would note that some “friend” had stayed at a particular hotel—the only problem being that we had no idea who the person was and if we should trust them any more than Bob from Saskatoon. But Medros calls it a “starting point.” “Even if my friends are idiots,” he says, “I still want to know what they thought of it.”

Driving down the dusty, mangrove-lined track to Hotel Jashita, in Soliman Bay, Mexico, I passed a man on foot who, as I’d learn a few moments later at the reception desk, was the property’s co-owner. While we talked, a fragment of one of the hotel’s TripAdvisor reviews floated into my consciousness. “Lots of smiles,” it said, “but he came across as … stand-offish.” I actually wasn’t getting that vibe, but it raised the question: How much can one person’s experience match my own?

Quite a lot, in fact, at least on more cut-and-dried matters. On our second night, my five-year-old daughter, settling in for bed, spied a lizard on the ceiling. As this proved altogether too novel, I was implored to dispatch the hapless creature. Deciding to outsource, I made a motion to call the front desk. Then I remembered, from a TripAdvisor review, that there were no phones in the rooms! I suddenly realized that I was reliving someone else’s inferior service experience. Then it occurred to me that I actually like the absence of phones (not to mention televisions) in rooms. And that lizard? It evaded the best efforts of the night man. My daughter, naming it “Lizzie,” made her peace. The subjectivity of travel, like life itself, balances on a knife’s edge: I could just as easily have disparaged the creature-infested, amenity-lacking room as rhapsodized about its Zen-like simplicity and proximity to nature.

In fact, TripAdvisor looms so large in the hotel’s success that when Jashita, which formerly had occupied the number-one spot for Tulum, was grouped into a new category for Soliman Bay, which is about five miles north of Tulum, its business plummeted. They started working with agencies like iEscape to publicize the hotel. While their bookings and ratings have recovered—Jashita showed up as TripAdvisor’s number two in the whole Riviera Maya region when I was there—a representative pointed out that “Nothing would bring us the business that used to be brought from TripAdvisor when we were number one in Tulum.”

After a few days, we traded our private plunge pool for a more affordable hotel, Don Diego de la Selva, in Tulum proper, which, like Jashita, was a TripAdvisor Traveler’s Choice. Charles Galligani, who moved to Tulum ten years ago from Paris, opened Don Diego just as TripAdvisor was gaining steam. “We followed the revolution,” he told me. For Galligani, the reviews are a testament to the vagaries of taste and a way to fine-tune amenities. “The beds are a bit too firm for the Americans and, on the contrary, for the Europeans they are not firm enough,” he says. (He has added a few softer beds.) “South Americans prefer the upper floor,” he said. “They say, ‘I don’t want to stay in this jungle, it’s too dangerous!’ ”

How did TripAdvisor perform outside of hotel reviews? We had a delightful dinner at El Asadero, an Argentinean-style steak place in Tulum that was ranked number four of all restaurants. Restaurare (number nine), a vegetarian spot on the beach, was tasty if a little rustic—the generator-supplied power blinked out briefly, although bonus points went to the hostess, who gave us homemade mosquito repellent. The taqueria Antojitos La Chiapaneca came highly recommended, but that may have had more to do with price and location than anything else—frankly, I’ve had better tacos in Queens. Some of the best food we had—the grilled boquinete at the beach shack Chamico’s in Soliman, the zesty enchiladas verdes at the Loncheria El Aguacate—were recommended by people we’d met but were many pages down on TripAdvisor, so who knows whether we’d have ever found them.

Wherever I go, one of my highest priorities is getting on a bike, and a TripAdvisor forum pointed me to iBike Tulum, located just off the cycling path on the way to the beach. The owner, Arturo Ramirez, not only hooked me up with a Specialized mountain bike, but also turned me onto a feisty Sunday-morning group ride (replete with stranded tourist rescue and crocodile sighting) into Sian Ka’an. The next day, we visited the lagoons of Sian Ka’an on a Yucatan Outdoors kayaking adventure, which was ranked number five of Tulum activities—and, it should be said, deserves its five-bubble rating. In between showing us boat-billed herons and offering my daughter tasty bananitos, owner Antonio Arsuaga told me that TripAdvisor is essential to his business. “It helps us to be more stable,” he said. Hotel employees who might normally recommend his outfit come and go, Arsuaga said, “but in the virtual world it stays.”

Leaving the cave tour at the cenote-and-zip-line complex Aktun Chen, where the guide’s joke-filled patter was as smooth and gently worn by time as the stalactites, we saw a huge sign, decorated in TripAdvisor green, with the brand’s signature owl, imploring visitors to post reviews. Spotting the owl became a game for my daughter: at the palateria in town, on the gate at Don Diego. By the time we were at the Cancún airport, passing by a small room that a TripAdvisor sign claimed was the best “fish spa” in Mexico (the fish nibble at your feet, offering a unique sort of pedicure), I began to feel a rash desire to partake in some activity that was not on TripAdvisor, an experience that had not already been mediated by the leveling winds of mass opinion—a rathole restaurant or fleabag hotel where I didn’t already know the front-desk clerk’s name. I wanted to have no expectations, either exceeded or unmet.

But all this was my young backpacker self speaking; as a harried dad, I needed some assurance that things would work out. I’d been wise to give in to the crowd.

Inside the Mad, Mad World of TripAdvisor

So huge is the “TripAdvisor Effect” that it can impact entire countries. One study found that as Irish hotel managers scrambled to respond to reviews, their hotels actually got better.

I made a motion to call the front desk. Then I remembered, from a TripAdvisor review, that there were no phones in the rooms! I was reliving someone else’s inferior service experience.