Big Web Sites to Track Steps of Their Users

August 16, 1998

Big Web Sites to Track Steps of Their Users

By SAUL HANSELL

ome of the largest commercial sites on the World Wide Web have agreed to feed information about their customers' reading, shopping and entertainment habits into a system developed by a Massachusetts company that is already tracking the moves of more than 30 million Internet users, recording where they go and what they read, often without the users' knowledge.

Forum
Join a discussion on Online Privacy

The agreement calls for the participating Web sites to track their users so that advertisements can be precisely aimed at the most likely prospects for goods and services.
But while this system guarantees the anonymity of individual users, the underlying technology disturbs privacy-rights advocates, who have long worried about the growing ability of online companies to collect and store personal data about people who use the Web.
Many individual Internet services have begun to amass detailed records of who uses their sites and how they use them. But this new industry cooperative represents the most ambitious effort yet to gather disparate bits of personal information into a central database containing digital dossiers on potentially every person who surfs the Web.

Anonymity is promised, but privacy concerns are raised.

Participating sites will include the Lycos-Tripod site, which was visited by 14.8 million people in July, according to Relevant Knowledge, a market research firm, and the Geocities virtual community of more than 2 million personal Web sites that attracted 14.2 million visitors last month.
The system's proponents extol its promise for delivering precisely directed, sometimes personalized, ads. For example, an Internet user who looks up tourist information about England on a travel site in the network might be fed ads for airlines flying into Heathrow Airport and for hotels in London as he checks sports scores.
"If someone comes to your bookstore the first time, you can find out if they are interested in mountain climbing, organic gardening and tennis; you can present them books related to their interests immediately," said David Wetherell, the chief executive of the company behind the Internet system, CMG Information Services of Andover, Mass.
In addition to a few large sites, CMG has attracted a host of smaller participants like NBC Videoseeker and Ticketmaster.
CMG's system and a dozen other similar efforts under development are rooted in the same marketing needs that have prompted direct-mail companies to assemble mailing lists using nearly every publicly available scrap of information on people, from their auto registrations to their vacation habits.
But while mailing-list companies are limited to identifying people for mailing lists by broad interests -- for example, subscribers to fishing magazines -- Internet-based systems can find a person who reads articles about fishing even if the Web page he is visiting is part of a general news or recreation site.
The Internet systems can also tighten their focus by, say, sending an ad from a charter operator to someone who has spent time reading about deep-sea fishing.
Wetherell argues that CMG's system, known as Engage, protects people's privacy in ways that mailing-list companies never can. In particular, he says, Engage does not record the name, street or e-mail address or credit-card numbers of the people it profiles.
Instead, it places a unique identifying number on the computer hard drive of every person who visits one of the participating sites. That way the system can keep track of all the sites visited by that computer, regardless of the identity of its user.

What’s Collected, What's Not
A new service is working with many of the biggest and most popular Web sites -- like Lycos, Geocities and NBC's Videoseeker -- to track what people are looking at on the Internet. The information is gathered without the explicit consent of users by reading an identification number placed on each person’s computer when they visit the participating sites. To protect people's privacy, the service will collect only some kinds of information.
INFORMATION COLLECTED
USER INFORMATION: Age, sex, income, ZIP code and number of children, gleaned from filling out on-line forms.
INTERESTS: More than 100 sites have been classified into over 800 topics, including classical music, fishing, European travel and sports cars.
INFORMATION NOT COLLECTED
USER INFORMATION: Name, address and birthday will not be stored.
INTERESTS: Sites on sexual or health-related topics.

"We took the highest road you could possibly take with respect to privacy," Wetherell said. "We think you can learn a lot more about someone from their behavior than from their name and address."
Moreover, a user can choose not to have his or her surfing observed, by visiting the company's Web site and selecting an option that will remove the identification number, known in the language of the Internet as a "cookie," from their computer. Users can also set their Web browsers not to put any cookies on their computers, but this can complicate access to some sites.
Some privacy advocates agree that Engage's promise of anonymity could help protect Internet users from hackers and commercial or government snoops.
"The big long-term concern about privacy is the surreptitious compilation of every site you click, every page you download, every product you order into a single database," said Marc Rotenberg, the director of the Electronic Privacy Information Center in Washington. "Anonymity is like solar energy. It's a way to produce what you want without the unpleasant byproducts."
Yet Rotenberg and others also say they are concerned about whether Engage and all the participating sites will strictly maintain this promise of anonymity.
"Engage has done many good things to protect privacy, but my worry is they are firing the starting gun in the race for the bottom," said Jason Catlett, the president of Junkbusters Corp., a Green Brook, N.J., privacy consulting firm. "The worst actors will be left to use the most sophisticated surveillance techniques as they please."
Indeed, last Thursday, in the federal government's first enforcement action to safeguard privacy on the Internet, the Federal Trade Commission accused Geocities of selling personal information about its members. Geocities said it did nothing wrong, but changed its notification to members about how data about them would be used.

Related Article
FTC Says GeoCities Violated Privacy Rules
(Aug. 14)
Navy and America Online Settle Case on Gay Privacy
(June 12)
Law That Helped Sailor in Privacy Case Has Major Loophole
(June 19)

It is not illegal for Internet services to sell personal information about their customers, and there are few laws protecting consumers' privacy in cyberspace. The Clinton administration's policy is that businesses engaged in electronic commerce should police themselves.
Yet there have been several recent instances in which companies have either lied outright to their customers or otherwise failed to live up to their own rules.
For example, America Online Inc., in theory, offers users the chance to shield their actual identities behind pseudonyms known as screen names. But the U.S. Navy recently forced the retirement of an 18-year veteran, Master Chief Petty Officer Timothy R. McVeigh, after a customer-service representative for America Online violated the company's policy and identified McVeigh to a Navy investigator as the owner of a screen name with marital status listed as "gay."
In order to avoid such problems, CMG executives said they would not track some online behavior that could be especially sensitive.
"We decided to avoid sexual preferences, adult content and medical information, because they are controversial," said Daniel Jaye, the chief technical officer for Engage.
Engage is the most elaborate system so far for monitoring where people go and what they do on the Internet. But dozens of other systems have been created to learn things about users in an attempt to improve responses to advertising.

Internet technology can supply site owners with some information about visitors to their sites -- the area code of a dial-up visitor, for example, the type of computer used, and, in the case of an office computer, the name of the company.
But these data are unreliable. For example, a computer with the same address might be used by two or more people.
For that reason, Web sites are increasingly seeking more detailed information. They do this by registering people for contests or services like free e-mail.
At each step of the way, the information collected is saved in databases that can be used to aim ads.
In one recent example, a quarter-million people filled out a survey at Walt Disney Co.'s ESPN site in return for a chance to win tickets to the NCAA Final Four tournament.
Other sites, including The New York Times on the Web, require users to provide their names, zip codes, age and ranges of income. All information except the customer's name is then used to compile profiles to help advertisers direct messages.
In general, advertisers say that such targeting techniques can increase as much as fivefold the percentage of viewers who click on a given ad. The Web sites, accordingly, charge a premium for delivering ads aimed at certain users.
But another frontier for Internet marketers -- one that alarms privacy advocates -- is the combining of information gathered from people online with vast stores of data on these same people kept by companies that compile traditional mailing lists.
Adforce Inc., a company in Cupertino, Calif., is developing a system in cooperation with Metromail, one of the largest mailing-list companies, to do just that.
Adforce executives are seeking to persuade Internet service providers to give them the name and address of each visitor as he or she surfs. Adforce would then instantly retrieve demographic and buying-habit data kept by Metromail about that person and use it to display advertisements aimed at him or her.
Like Engage, Adforce says it will not provide the names and addresses of the users to the advertiser. Charles Berger, the chief executive of Adforce, argues that this approach is less intrusive than the Engage system because it uses only information gathered off line in its advertising system.
"I feel my privacy is more violated if someone follows me around and watches what I read than if they look up that I have a Volvo in my garage," he said.
While Adforce does store some of the information it collects, Berger said neither Adforce nor Metromail planned to use the data about which sites users view.
So far, however, Internet service providers apparently have been reluctant to sell information about their customers to Adforce. As an alternative, the company is now asking individual sites to share the postal addresses provided by users who register for various services. Once a user's name and address have been captured by one site in the Adforce network, all Metromail data about that person would be used to select advertising on other sites in the system.
Berger declined to say which sites had agreed to provide such registration data.
Among the companies that have chosen to take a cautious approach is America Online, which has amassed the biggest repository of data about its customers. After being sued by McVeigh, America Online introduced a new, tougher privacy policy. Among other things, the company said it would not use data about customers' online habits to aim advertisements.
"AOL has gone from having one of the worst records in the industry to having one of the best privacy policies," said Catlett of Junkbusters.
America Online, however, has revenue from a regular monthly fee charged to each user, while most other Internet services depend entirely on advertising.
"If the ads aren't effective, these services are not going to be around," Wetherell of CMG said.
What is more, amid the ever greater cacophony on the Internet, companies are increasingly desperate to find ways to reach the few people who are likely to be receptive to their products.
"Advertisers and publishers want a better eyeball, and a better eyeball is a more targeted eyeball," said Richard Baumer, the president of Venture Direct Worldwide. His New York-based company sells ads on behalf of about 100 Internet sites, ranging from Golf.com to Universal Studios. Venture Direct has agreed to participate in the Engage system.
Baumer argues that customers will be better off when they see ads that are most relevant to their needs.
"Marketers with information that use it wisely produce a better relationship between consumer and supplier," he asserted. "As in any medium, crossing the line and abusing the information is objectionable. This technology is potentially more useful and potentially more harmful than we've ever seen before."

Related Sites
Following are links to the external Web sites mentioned in this article. These sites are not part of The New York Times on the Web, and The Times has no control over their content or availability. When you have finished visiting any of these sites, you will be able to return to this page by clicking on your Web browser's "Back" button or icon until this page reappears.

CMG Information Services
Engage
Lycos
Tripod
Geocities
Adforce

Metromail
Venture Direct

Help/Feedback | Classifieds | Services | New York Today