But if it were an online bookstore, from a privacy standpoint, I’d be hosed. Going on silently, invisibly behind the scenes, is something called online profiling. It’s a technology used by many of your favorite online destinations as well as the companies who place banner advertisements there. As you click your way through a Web site, a powerful software program running on a massive computer is watching and taking meticulous notes. Says privacy expert Andrew Shen: “The offline equivalent of online profiling was if someone was following you around the mall all day, keeping track of what stores you went into, what items you looked at and tried on, which items you purchased, when you entered the mall, when you left. Everything.”
Online profiling may seem threatening, but it needn’t force you to give up the advantages of e-commerce. Like most technologies, it has good uses and bad. Advertisers put a positive spin on the practice, saying it helps them to better target ads, so that relevant sales pitches may find their way to your desktop–such as a low-priced airfare to Hawaii while you’re searching through a travel site for vacation bargains. Online stores say that it helps them improve the shopping experience.
No one is yet alleging that reputable e-commerce merchants are using this technology to spy on people or manipulate them–even in two class-action lawsuits brought this month against RealNetworks, a Seattle-based software company. RealNetworks has been accused of surreptitiously tracking the music-listening habits of its users through a program that 13 million people had downloaded for free from its Web site. Privacy advocates say that collecting data wasn’t the problem; it was the company’s failure to inform its users. To its credit, RealNetworks quickly issued an apology and a software fix.
Still, many people were unnerved by the discovery that the popular RealNetworks had the technological resources to gather the data without anyone’s knowing. “Online profiling is just so invisible,” says Shen, a policy expert with the Electronic Privacy Information Center in Washington, D.C. “Most people are not aware that this information is collected from them.” And if you don’t know it’s happening, the thinking goes, you can’t make an educated choice about whether you want give up your personal information in order to participate.
Data collection about your online behavior and personal identity happens in a number of ways on the Web. Most consumers understand that when they go to an e-commerce site and fill out a form (name, address, e-mail) those bits of information are stored somewhere in a database that belongs to the e-commerce merchant. How Web sites will use this information, as well as the below-the-surface traffic surveillance that might be happening, is what’s supposed to be covered in its privacy policy (chart). You can generally trust these policies, but watch for language that lets merchants change their minds in the future.
The real scourge, some say, is the data-collection practices of online advertising companies. The main way they get access to you is through the banner ads on Web sites. In a vast majority of instances, the Web site that you visit isn’t the source of the banner ad that appears at the top of the page. Yahoo!, for example, uses a company called DoubleClick to sell ads into the banner spaces on Web pages. DoubleClick performs this service for more than 450 Web sites. Using cookies, tiny coded identifiers deposited on your computer’s hard drive through your browser, DoubleClick builds a unique profile of you. Each time you visit a site that serves up a DoubleClick ad, your cookie profile is being updated.
The profiling is no differ-ent from what Yahoo itself might be doing. But have you ever heard of DoubleClick? Or Engage? How about MatchLogic? These are all companies that conduct online profiling on behalf of advertisers and they know a lot about you. Engage CTO Daniel Jaye estimates that his company, just one of several in the industry, has unique profiles for some 35 million Web surfers, or around 40 percent of the total online population, stored in 700 gigabytes of data in a server somewhere. Each anonymous profile contains 800 “fields of interest” which infer your personality from your online behavior. For example, if you’re someone who frequents recreational sports and parenting Web sites, your behavior suggests that you’re a 30- to 40-year-old male with kids. This profile might trigger a banner ad for the new GM minivan–perfect, the ad copy might say, for toting around your kids and your team’s softball equipment. Engage traces your exact footprints through a site but does not store that information in its profile database; instead it extracts a “score” that corresponds to a preconfigured field of interest. Nor does the company link personally identifiable information, such as names, to the profile. It also does not collect data on medical conditions, political persuasion or traffic through pornography sites. But could Engage do all of these things? “Absolutely,” says Jaye.
Privacy advocates worry about potential abuse. The threat to personal privacy is amplified by fact that on the Internet, previously isolated silos of data now can exist as a more or less desegregated information heap. Your marketing profile could conceivably be merged with postings you made in a newsgroup or open-forum Web discussion. “There’s a push toward merging databases, and that poses new threats,” says Austin Hill, a privacy fundamentalist with a knack for sketching truly scary scenarios. “Let’s say a company says, ‘We’re not interested in having gay and lesbian customers, so let’s not offer anyone with that profile this discount.’ All of those possibilites get opened up.”
EPIC’s Shen says it’s up to consumers to protect themselves. Hill’s company, Zero-Knowledge (zeroknowledge.com), sells software that can cloak your identity–from others and Zero-Knowledge itself–while you surf the Web. Anonymizer.com is a popular site for doing the same thing, with a lesser degree of anonymity. Other simple rules: don’t enter online sweepstakes; these are often fronts for companies that want to link your real name and e-mail address to cookie-based profiles. Learn about online advertising; 10 industry leaders have posted their privacy polices at networkadvertising.org. Most important, read the policies yourself.
The Federal Trade Commission last Monday convened industry leaders to explain the details of online profiling. Afterward, chairman Robert Pitofsky says he “wasn’t surprised” to hear that online profiling without consumer knowledge and consent was actually happening. “I had hoped that we would hear from the industry what their proposed self-regulatory fix would be, and that didn’t happen,” he says. Pitofsky says the FTC won’t act until the industry submits a proposal. Profilers may be watching us, but now we are starting to watch back.
