NetTrends: What Web Sites Want to Know (About You)

By Dick Satran

SAN FRANCISCO (Reuters) - It's not just your imagination: Somebody is watching you.

In a report released by the Federal Trade Commission last week, the agency concluded that ``Web sites collect a vast amount of personal information from and about consumers.''

Its survey revealed that 97 percent of sites collect an e-mail address ``or some other type of personal identifying information.''

Speak your mind Discuss this story with other people. [Start a Conversation] (Requires Yahoo! Messenger)

For the paranoid, that's the bad news. The good news, for those who like to protect their privacy, is that the art of targeting and profiling individuals from Web site data is still at a primitive level.

The Internet has created an awesome technology for tracking individuals' habits -- from what they read to what they buy. But so far Web sites aren't doing much to use the data, and regulators and privacy advocates are mounting a campaign to keep them from ever doing so.

``Most online businesses just want to use the data to get to know their customers better and understand their needs,'' said Austin Hill, president and co-founder of Zero-Knowledge Systems, which sells privacy software.

Companies don't want to be seen snooping on their own customers, and there is an argument that companies are wise to keep clients' data private -- rather than letting it slip into the hands of a competitor who might use it to steal business.

But there is a huge temptation to get more out of Web sites, and since companies are investing heavily in e-commerce activities, they want to generate results.

``The default, when people don't know how or what data they are going to need, is to just go out and collect as much as possible -- 'In case we need it later,''' Hill said.

The grab-all-you-can ethos is one thing that's getting the attention of the FTC, as well as consumer groups. Even if it's not being used for much, there is access to more data, at less cost, than ever before, the agency says.

New Tracking Technology

``The technological developments that have made e-commerce possible,'' said the FTC in its report, ``have enhanced the ability of companies to collect, store, transfer, and analyze vast amounts of data from and about the consumers.''

So far, the information is being put to a relatively innocent use, said Internet analyst Henry Blodget of Merill Lynch, in a briefing about online privacy. It's not about disclosing Web surfers' identities at all -- but about putting advertising messages in the right ``context,'' although DoubleClick Inc. (NasdaqNM:DCLK - news) called off a program for large-scale ID-ing of Web surfers after being bombarded with protests earlier this year.

``Ninety-eight percent of is based purely on context,'' Blodget said, to better distribute Web advertisements for maximum impact.

That means, for example, putting a football ad onto the computer screen of someone who's just done a search on ''football,'' Blodget said.

Some marketers, though, are taking more aggressive stands, trying to track users' movements around the Internet and using it to compile demographic information and identify specific users without their knowledge. Such information will create ever more targeted and effective advertising, advocates say.

Objections To Data Mining

But this more detailed and invasive type of ``data mining'' is raising sharp objections from privacy advocates, even though it makes up a very small proportion of the Internet marketing surveillance effort.

The government, for its part, isn't trying judge how much data can be mined -- but is pressuring Web sites to make a full disclosure to users about what they compile, so there are no secrets out there. The present debate in Washington is over how much the sites should be required, by law, to reveal to users.

In opinion polls, most consumers say they don't want data compiled on them without their knowledge. But the surveys also show that being watched isn't consumers' biggest worry -- at least it doesn't rank with the fear of being ripped off. Identity theft appears to be the biggest concern, according to an Arthur Anderson survey released this week.

``People are concerned about people impersonating them,'' said Russ Gates, of Andersen's Technology Risk Consulting Practice, with more than four out of five saying they wouldn't want a site to have their Social Security number.

But the information they most fear giving out is readily available already to marketers. The contextual information -- what people think and do with their time and what are their preferences -- is the most-sought information on the Web. Being able to link specific preferences -- like ``What color do you like -- are you a pet lover'' -- to a specific person is a major coup that data miners see as most valuable.

``The theory is that having that level of target marketing is worth the price you have to pay to make it work'' said Anderson consultant Gates. ``It's a nice theory, but I'm not sure it works. ... People really can tune things out on the Web.''

For privacy advocates, it's a theory that they'd rather not see tested out in the wild.

Earlier Stories

• NetTrends: Halting Computer Viruses at the Network (May 17)