Imagine that every time you walked into a store, the manager would look you over. He'd note down your age, your gender, whether you were well-dressed. You'd be free to shop and no one would ask your name. But day after day, every time you walked in, there he'd be with his clipboard. Farfetched? Not on the Internet. Companies are developing systems that, while guaranteeing anonymity, could track almost everything else about Internet shoppers. The technology is sophisticated enough that electronic merchants may get to know their client base more intimately than any store, magazine, or television network. If this electronic bean-counting disturbs you, think twice about divulging personal information before you go on an on-line shopping spree. Last week, the bean-counters chalked up another victory. A year-old San Francisco company, Internet Profiles Corporation, joined forces with Nielsen Media Research to market new Internet tracking systems. Nielsen already determines the ratings for TV shows. Now, it hopes to do the same thing for the Internet. Done correctly, the effort is commendable. Advertisers need to know the most popular spots on the Internet, particularly the fast-growing, graphical part called the World Wide Web. And the current method of measuring Web traffic is woefully inadequate. Internet Profiles and a handful of other companies have come up with statistical methods to make accurate estimates of total visitors. Joined by a company with Nielsen's credibility, Internet Profiles' methods may quickly become the standard. The sticky part is what happens next. Potential advertisers want to know more than the size of the viewing audience. They want to know its characteristics. Nielsen already provides that information for TV by surveying a random sample of viewers, who voluntarily participate. From the sample, it estimates the size and type of a program's national audience. Internet businesses could benefit from the same technique. But why stop there? If computers can track consumers more accurately than a survey can, then Internet businesses should be able to get specific demographic information about their customers. Internet Profiles, for example, has developed I/CODE, a kind of universal registration for the Internet. Consumers would fill out the form once. Then every time a Web business asked them to register before viewing its on-line material (an increasingly common occurrence), they could type in their I/CODE. So every time I logged into one of those businesses, a little trigger would go off in someone's computer: ''Alert! Professional, college-educated male employed by Boston-based newspaper is accessing information about Widget B and (oh WOW!) he's a Baby Boomer, so he probably has LOTS of disposable income!'' (Actually, all his disposable income fits in a coin purse, but you get the idea.) Of course, as Internet Profiles stresses, its system is voluntary. You don't have to fill out the form. And to its credit, the company doesn't require people to report their income. Nor does it give out names without specific authorization. But the on-line service CompuServe already is using incentives to get people to sign up. When Internet Profiles begins promoting I/CODE widely, probably next month, more companies will follow. And where will it end? Will other companies spring up to track the spending habits of my Internet code? Will the Web catalog service conclude that I'm a careful comparison shopper and give me less access than Mr. Impulse Buyer? Will Al's Online Oriental Rugs restrict shoppers who make less than $20,000 a year? The point is, technology is quietly boring holes in one more barrier of privacy. Before the wall crumbles, we ought to consider whether that's where we want the technology to take us.