Before You Get Burned (Summer 2004 Trust Magazine article)

Bill McGee isn't traveling anywhere, but Priceline.com doesn't know that. It doesn't even know he's Bill McGee. On this summer day, the online ticketing site thinks he is a Charley Bell, in quest of a cheap flight from Boston to Tampa a few days hence. Moments later, McGee morphs into Robert Bell, seeking a Priceline deal to Phoenix from Baltimore, followed by Nicholas Bell, who wants a seat from Pittsburgh to New Orleans. Tapping at a computer at an office just outside New York City, McGee creates personas as he goes, enjoying the white-lie names. "The Bell family," he says, "is very large."

Simultaneously, at keyboards in California and elsewhere in New York, five more people are posing as travelers they are not and seeking tickets they won't use, picking the same dates and destinations as the ubiquitous Bells and other characters McGee conjures. None of the companies realizes it, but six giants of Web travel booking--Expedia, Hotwire, Orbitz, Priceline, Quikbook, Travelocity--are going head-to-head to see which offers the most attractive deal when given identical parameters. 

As each fictional request is made and each site replies with possible fares and flights, testers print the screen pages. "So we have a record later," McGee says, should the companies raise doubts or questions after the findings are published. It is a contest months in the making, not only to ensure comparisons are apples-to-apples, but also to guarantee that no Web site can, in mid-test, trace the requests to their true source--Consumer WebWatch, a project of Consumers Union. 

Operating from Consumers Union's Yonkers, N.Y., headquarters, the non-profit research organization that is its parent, WebWatch has a modest full-time staff--just four (McGee is a consultant)--but mammoth ambitions. With no more than publicity and persuasion as tools, the project is shedding light on Web practices and prodding sites toward more openness about who they are, how they operate and what deals they make with third-parties for preferred placement or links. As WebWatch succeeds--and the road is long--consumers may feel more comfortable when they head to cyberspace to find a hotel room, buy a book, research an illness, plan for retirement or do any of a thousand tasks they can now do alone, without telephoning or visiting anyone. 

Getting the best fare to Hawaii, for instance, used to mean calls to multiple air carriers. It meant minutes on hold, listening to orchestral versions of The Carpenters. It meant giving the same travel details to numerous agents. Now a buyer needs only a computer, a modem and a Web address and, in seconds, can compare a plethora of fares, flights, dates and times rounded up by one of the online services. 

Except . . . how do those services get that data? Are they searching all airlines, or only those with which they have relationships? Are they displaying dates and flights in the order that most closely matches the request, or giving priority to carriers that have paid them a fee? Are they making the lower prices clear, or burying them in the displayed results, deeper than most consumers go? Is there, in one way or another (in McGee's words), "disturbing evidence of bias"? 

"We're seeking to pierce the veil of sites that say they give objective information," says Beau Brendler, WebWatch's director. 

The airline test is but one way WebWatch has been trying to "pierce the veil"of the Internet since the project began in 2001 with $4.8 million in support to Consumers Union from the Pew Charitable Trusts, the John S. and James L. Knight Foundation and the Open Society Institute. Much of the early emphasis has been on encouraging Web businesses to adopt standards of good behavior, some as simple as stating who owns the site and where it is. But the project has also commissioned studies to help businesses understand how consumers use the Web and how they want to be treated there, and it has delved deeply into hotel-booking, auto insurance and health sites to determine if consumers are getting good deals or data. 

WebWatch does not dispute that the Internet is a marvel of democracy, where anybody who can pay a Web server can have a Web site, and anybody with a computer can read it. Alongside the bricks-and-mortar world, there is now one that exists on desktop or laptop, where services are offered, products sold and advice is given, making it far easier to scratch something off a "to do" list or get smart about a problem or issue. 

Many sites bear the familiar names of real-world companies or organizations, which have merely opened online outlets. Consumers know their reputations and know where to find their stores or offices. Oft-times, though, it is a mystery who is selling a product or giving advice, what expertise they have or how they make money. It can be difficult, too, to know if a site sells email addresses or other data to third parties, which might send forth spam. And there might not be an address for returning merchandise and a phone number for asking questions. Consumers online lack "access to molecules"--humans and buildings, says Denise Caruso, a former technology writer who, as a Trusts consultant, did the original work that gave birth to Consumer WebWatch. 

Imagine, Brendler says, that "I'm a 17-year-old boy with a ‘social disease.' I don't want to tell my parents about it. I can't really go to my doctor without my parents finding out and, depending on local laws, the doctor might have to inform my parents anyway. 

"So I read a bunch of sites, try to figure out what I need for my problem, then go to one of these growing numbers of sites that sell prescription drugs without a prescription. Maybe I think I'm fully informed by reading the Web about drug interactions, proper dosage and so on. But I'm able to find a site that can sell me what I need. And so maybe I get an antibiotic. But it's been manufactured in Mexico or India with second-rate ingredients and factory oversight. And so on and so on. 

"The point is the Web is a wonderful convenience. But it can remove a number of checks, balances and safeguards." 

In the two weeks after the anthrax attacks of 2001, 23 Web sites sprang up offering cipro, the best-known treatment. None required a prescription, Brendler notes, and there was "no guarantee the stuff you're buying hasn't been sitting in a warehouse for 10 years." Not long ago, Caruso, who lives in San Francisco, ran out of a seasoning she had brought home from a trip to Boston. She found Web sites for it. But they were "just these weird, no-name boutique-y gourmet ‘shops' that didn't bother to tell me where they were, or who they were, or anything that would give me a clue whether I should type-in my credit card number," says Caruso. 

Moreover, the line between unbiased data and promotional slant can smudge on the Web. Newspapers, magazines and television have developed traditions of demarcation between news copy and advertising. (They seem to push that line more and more, though.) But a Web consumer can come across a book review that suggests clicking here to buy the book from an online dealer. What the consumer doesn't know, Brendler says, is that for every copy sold via the link, the dealer gives the reviewer seven percent of the sale. "Don't you think that might compel me to (a) only review books that I'm fairly sure people are going to want to buy, and (b) only review good books or give good reviews to every book?" 

Such hidden arrangements and fees are routine. Researching a disease, a Web user might come across a page of sober medical discussion that notes in passing that a certain drug has proven effective in treating the illness. But there is no indication the page and the drug have a tie: Both are creations of the same pharmaceutical company. Looking for a good car rental price, a consumer might obtain a host of choices from a site. "What they're really giving you," Brendler says, "is the lowest price among a group of favored business partners." 

In one study WebWatch reported that 17 volunteers who regularly use Internet search engines were unaware the sites often display results based on fees. For the study, the volunteers asked 15 search engines for health, travel and other data, and selected 163 of the displayed links. Almost all of their choices were on the first page of results, and 40 percent of the choices were labeled "sponsored" or "featured site," meaning the sites had paid the search engine. But, Webwatch said, "very few participants" noticed the label, and "each participant expressed surprise" when told what it meant. One woman said, "Oh, I'm such a sucker!" 

Jakob Nielsen, a member of WebWatch's 26-member board and holder of numerous patents for making the Web easier to navigate, says too many sites are "just not treating people reasonably, and it's very confusing and frustrating for users to be in this world." It is "an unpleasant environment where you never feel safe and you always feel hustled." In a survey WebWatch commissioned called "A Matter of Trust: What Users Want from Web Sites," Princeton Survey Research Associates found that only 29 percent of consumers trust sites that sell products, and only 33 percent trust those that provide information about purchases. 

It's not that most Web sites seek to deceive or confuse, Nielsen and others say. They just don't realize consumers might want to know about relationships with sponsors or be able to find an address or privacy policy. "A large number of companies are reasonably ethical," Nielsen says, and when asked to change a bad habit, "they'll obey." 

Beyond that, sites face "tremendous pressure to get the advertisers into the message and make that line as blurry as possible," so money can be made, says Michael Daecher, the vice president for content for the research site About.com. And the technology makes getting ads into editorial copy as easy as clicking on a link. Whatever the cause of problems, there has been "a chilling effect" on the Internet's information and commercial potential, Brendler says. 

Maybe not very chilling, though. 

According to Forrester Research, online sales rose 48 percent in 2002, to $76 billion, and then proceeded to rise 51 percent in 2003, to some $114 billion. It cited one analyst who said online sales were growing far faster than catalogue sales did in their day. WebWatch's own Princeton Associates survey found, moreover, that "online users' low ratings of Web site credibility do not stand in the way of people going online." In fact, 75 percent told the survey they shop on the Web and 73 percent provide personal data on the Web, whatever their expressed distrust of the medium. 

That is not, of course, the only chasm ever to appear between what people say and what they do. Technology consultant John Hagel told The New York Times that online shoppers might aver that a privacy policy is important, but then not check a Web site for one. What matters more to them, Hagel says, is what always matters in business: Does the company deliver a good product or service? 

Capitalism, indeed, works in Darwinian ways with bad practices, whether or not there are projects like WebWatch to highlight them. But even if market forces would whip the Web into shape, Caruso says, "Wouldn't it be nice, for once in the history of the world, if we didn't have to wait until a million people got burned?" 

The Federal Trade Commission, for one, is not waiting. In June 2002 letter to several major search engines, it warned them to make sure that they provide adequate disclosures when they have paid placements; the letter cited Webwatch's research at length. WebWatch's consumer surveys have been "very useful to us," says Heather Hipplsey, associate director of the FTC's division of advertising practices. "We very much embrace efforts that are being made for best practices," she adds. "We can't do everything." 

One Web site that is a vision of good practice is--no surprise--Webwatch's own, wwe.consumerwebwatch.org. It clearly states the project's goals, sponsorship, privacy policy, physical address, phone number and other components of the WebWatch "pledge," composed of five broad guidelines. The most important of these is "advertising and sponsorships," and urges sites to "clearly distinguish advertising from news and information" and to disclose "paid result-placement advertising, so consumers may distinguish between objective search results and paid ads." Caruso says the idea is not to end financial relationships, but to give context. Is an airline paying a search engine to have its fares displayed first? "Fine," Caruso says, "just tell me you took money for it." 

The guidelines, however, confront a devilish problem. WebWatch has no power to force businesses to adopt them. It can only praise those that do. Once in 10 newspapers and twice in only The New York Times and The Wall Street Journal, the project has purchased ads that listed sites that have "taken the pledge" and passed muster. Getting the imprimatur is an incentive, because business sites know consumers follow recommendations of Consumers Union, the parent of WebWatch. Others are probably following most or all of the guidelines without even knowing that WebWatch has developed a formal list. 

Its other major efforts have been consumer surveys and investigations similar to McGee's. Along with ConsumerReports.org and Consumer Reports magazine, WebWatch reported in October 2002 on auto-insurance sites. Testers posing as a couple from Santa Monica, Ca., were given price quotes that "were generally higher than those found on the California Insurance Department's Web site." Using WebWatch's five guidelines, the investigators gave "e-ratings" to six insurance sites, finding that only InsWeb.com was "useful." The others either took time to deliver results, operated in a limited number of states, offered a limited number of quotes or poorly disclosed business relationships and other data. 

McGee also led a team that found pluses to booking a hotel room online. Using Expedia, Hotels.com, Lodging.Com, Orbitz and Travelocity, the testers found that 85 percent of the time, at least one of the five services came up with a better room rate than the travel agent system. Travelocity in particular did well, coming up with the lowest rate in 29 percent of the tests, comfortably higher than any competitor. 

Of course, that means that in 71 percent of the cases, Travelocity's hotel room rate wasn't the lowest. A consumer who shopped there and nowhere else would pay more than necessary most of the time. Given the same dates and cities, some of the five booking sites found no availability, even as competitors found 11 or 12 possibilities. That suggests commercial arrangements: If a Web site's hotel partners had nothing available, the search concluded there were no rooms, because it wasn't looking outside its universe. Moral? Consult multiple sites. That wasn't supposed to be essential, given the Internet's ability to gather data in one place quickly. "It's quite clear that the consumer is not being offered the maximum benefits of such technology," McGee wrote. 

WebWatch publicizes much of its work in conjunction with Consumers Union, whose magazine has four million subscribers and its Web site a million. The project also relies on coverage by newspapers and television. But Scott Silverman, executive director of shop.org, a division of the National Retail Federation, says he does not hear much chatter about WebWatch among his members. Nielsen, a project advisor, suggests the project be likened to a tugboat pushing a supertanker: "The Web is really big. It's going to take a while." 

About.com is one of the companies that has taken the WebWatch pledge and been publicized in newspaper ads. But the site did not detect a surge in "hits" after those ads appeared, says Daecher, the vice president for content. In fairness, the About site gets so many visitors--16 million a year--it would take a great deal of new traffic to produce a spike. But About's experience suggests that placing an occasional ad in newspapers does not produce waves of response. 

An Internet analyst says Web change will come only if substantial numbers of consumers shun companies that have bad practices. But consumers are not demanding that sites to follow the WebWatch guidelines, the analyst said, so businesses "don't have any incentive." 

They may get more. Consumer Reports has incorporated Webwatch's guidelines in its e-ratings, used to rate more than 175 Web sites in 19 categories. (Consumer Reports Online has the greatest number of paid subscribers of all Internet media sites.) And the organization is considering applying Webwatch's systematic mix of research, guidelines, ratings and outreach to improve consumer practices among "offline" businesses. 

In addition, Webwatch has developed into a successful convener. One meeting brought together some 130 industry representatives, federal policymakers, journalists and advocates to discuss Web credibility. And, with the American Accreditation HealthCare Commission, Webwatch gathered online health leaders to identify ways to improve the effectiveness of health searches on the Internet. 

Each day, the WebWatch site gets 1,300 "unique" visitors, or nearly half a million annually, Brendler says. And he believes businesses are sitting up and taking notice. "Every time we do one of these reports that is performance based--11 so far--we get a bigger and bigger response," he says. In addition, the list of companies that have taken the pledge by January, including Barnes & Noble, The New York Times, Travelocity and Orbitz. As Brendler notes, "Somebody's taking us seriously." 

Webwatch's site, www.consumerwebwatch.org, contains news, cautionary tales and resources pertaining to Web credibility, consumer protection, privacy and other related issues. It also lists organizations and government agencies that accept complaints of Internet fraud.