Hallmark cards prides itself on its ability to translate societal trends into greeting cards. So a few years ago, when Hallmark's marketing group looked into its demographic crystal ball and saw 78 million baby boomers hitting age 50, the company created boxes and boxes of friendship, birthday, anniversary and thinking-of-you cards, all designed to subtly flatter the aging boomer's flagging middle-aged ego. Shipped to Hallmark stores in 2000, the Time of Your Life line of cards was displayed in its own section and featured active midlifers looking youthful as they frolicked on beaches and dived into swimming pools. "We had done a lot of research showing that baby boomers don't want to get old, but that if it's going to happen, they want to emphasize the positive side of aging," says Rachel Bolton, a spokeswoman for Hallmark. But Hallmark missed one tiny yet telling psychological detail: no self-respecting boomer wants to be seen shopping in the "old-people's cards" section. Faced with a choice between regular greeting cards and the 50-plus corner, potential customers avoided the cards as if they were a window display of geriatric diapers. In 2002 the Time of Your Life line was scrapped.

Knowing your customer well enough to avoid costly business goofs has always been an elusive goal of marketing strategists. Standard demographic data — age, gender, employment status, income, place of residence — are usually insufficient to forecast consumer behavior with any precision. That's why during the past five years market researchers have been developing more sophisticated tools to get inside consumers' heads. It's no longer enough for companies to know you are a 35-year-old white male making $45,000 a year and have a wife, 2.5 kids and a mortgage. To predict accurately what you'll buy and what you won't, marketers these days are more interested in whether you donate to Greenpeace or if you believe in creationism. Says Dawn Iacobucci, a professor at Northwestern University's Kellogg School of Management and editor of the Journal of Consumer Research: "Companies need to know what's on your mind. What's in your heart? What do you really want to do with your life?"

The answers, marketers hope, can be discovered through an emerging quasi-science known broadly as psychographics. Market researchers supplement conventional marketing data with informed assumptions about personality traits and human behavior gleaned from other disciplines, including psychology, sociology and probability theory. Using computers to organize and manipulate vast storehouses of such consumer information, they believe they are getting much better at sorting people into categories of like-minded individuals. And once the sorting is done, they are getting better at predicting how people are likely to behave.

A prime example is LifeMatrix, a psychographic-marketing tool launched in December 2002 by market-research giants RoperASW and Mediamark Research (both owned by NOP World). LifeMatrix considers hundreds of personal variables, including religious affiliation and political leaning, and uses them to sort people into 10 basic psychographic categories with jargon-rich titles like "priority parents" and "tribe wired." To each category is attached a battery of personality traits and purchasing preferences. Are you a working mom trying to balance job, family and cultural activities? LifeMatrix assigns you to the category "Renaissance women." But that's not all. LifeMatrix makes some rather intimate and sweeping assumptions about Renaissance women: they're involved, caring and optimistic; they're altruistic rather than hedonistic; they're likely to be heavy Internet users who enjoy museums and eschew radio and TV.

It may sound as if psychographics and the Psychic Friends Network have a lot in common. But LifeMatrix's proponents say the system isn't guesswork. A variety of inputs, including public- opinion polls and media usage, is used to create categories that accurately reflect personality types. Ed Keller, president of RoperASW, says companies applying LifeMatrix to their customer databases will have far greater success in predicting what those customers will buy. Keller says researchers using demographic data alone can correctly guess what kind of car an individual will buy only 18% of the time. But "when you combine people's attitudes, behaviors, life stages and values," Keller claims, "you can predict 82% of the time what car a person will buy next."

That's not to say psychographics is an exact science. In fact, there are numerous companies racing to build and sell tools similar to LifeMatrix, among them Monitor MindBase, offered by the market-research firm Yankelovich, and BehaviorGraphics, a joint venture between Simmons Market Research Bureau and Nielsen Media Research. All use different assumptions and psychological profiles to sort consumers into categories variously referred to as segments, clusters, affinity groups or passion groups and identified by such titles as "shotguns and pickups," "struggling singles," "band leaders" and "succeeders." MindBase, for example, extrapolates from a combination of attitudes gleaned from opinion polls and life-stage data: where you are in terms of marriage, children, home ownership. "Life stage is a much stronger way of looking at demographics," says J. Walker Smith, president of Yankelovich, "because age isn't really a predictor of everything. Knowing someone is recently married with one child is much more telling to a marketer than knowing whether she is 25 or 41."

The psychographics movement is all about building better pigeonholes, and there is no single correct approach. But marketers agree that the old demographics-based categories are defective. "You can no longer use the previous generation as a model for today's generation," says Roger Brown, director of global research for the Principal Financial Group, the nation's largest firm that focuses on 401(k) retirement-plan administration. "Baby boomers are very different from their parents in terms of how they approach retirement. Even within this generation, there are key differences that are forcing us to look more closely at attitudinal and motivational information."

Indeed, consumers today have far more choices and, with their greater ethnic diversity and the blurring of gender roles, are far more difficult to stereotype. A woman who checked the "Hispanic" box on her census form could easily prefer Vogue magazine to Latina. "Not only is the world less homogeneous today, but even in the past, we assumed the world was more homogeneous than it actually was," says John Karlson, senior vice president of strategic development at Martin/Williams, an advertising agency based in Minneapolis, Minn. "People's identities are much more wrapped up in their hobbies, sports teams, political affiliations and attitudes toward the media, which radically affects the way marketers should approach them. If you don't go way beyond demographics and bring deeper insight into marketing, you're not doing your job."

Developments in database technology have made the job easier and cheaper for market-research firms to link databases, creating more detailed consumer profiles. When Hyundai decided to give the psychographic treatment to car buyers earlier this year, it chose LifeMatrix partly because that company's data are linked with information stored at Mediamark Research, which tracks what consumers read. The combination will provide Hyundai not only with profiles of the types of individuals likely to buy its cars, but also with data on what magazines those individuals read — enabling the carmaker to create effective ads and place them in front of the very consumers it is trying to attract.

There may be a dark side to the increasing precision with which marketers can locate and track their quarry. The Orwellian overtones of companies and market researchers' getting together to share vast databases of detailed, individual consumer behavior are hard to deny. Just the names of psychographic tools (Monitor MindBase, LifeMatrix) are enough to get privacy advocates worked up. But researchers say they are only putting to more effective use information that consumers surrendered when they used credit cards, registered on websites or responded to questionnaires. And marketers keep raising the stakes. Personicx, launched by Acxiom Corp. in 2002 and using information from public records, third-party research, product-warranty cards and other sources, is designed to be updated monthly to reflect such life-changing events as the birth of a child or a job promotion. The next objective is to be able to identify an individual's "velocity component"--whether he is moving up or down the socioeconomic ladder and how fast. "This may be the holy grail of consumer information," says Charles Morgan, Acxiom's CEO. "If you can figure out whether someone is on the up-and-up, then you can devote more money marketing to him than to someone going the other way."

Gerald Zaltman, a Harvard Business School professor and the author of How Customers Think: Essential Insights into the Mind of the Market, believes marketers should even delve into the unconscious mind. Clients like Procter & Gamble use the Zaltman metaphor-elicitation technique, which enables them to uncover deep metaphors that lie beneath people's conscious opinions on products or advertisements. Zaltman is experimenting with brain scans to see which parts of the mind are active during certain purchasing decisions.

But psychographics may not be that ominous after all. Scientific purists argue that market researchers aren't really getting inside people's minds — they're just manipulating ordinary shopping and demographic data to make it look that way. Carrie Hollenberg, a senior consultant at SRI Consulting Business Intelligence, says many companies have created psychographic types by merely "assigning a company's customers to various groups with colorful names and descriptions. That's not using true psychological traits to get into who people really are." Hollenberg's firm owns a personality-profiling system called vals (values, attitudes and lifestyle survey) that was developed in 1989 using research conducted at the University of California, Berkeley, and Stanford; she says vals gets the psych job done correctly.

In fact, it's likely the vagaries of consumer impulse will continue to befuddle the best prognosticators at least some of the time. No matter how much data suggest otherwise, a nervous mother of two may get roped into riding a roller coaster and discover a new passion for risk. A stilettoed fashionista might wobble into Wal-Mart. Some things will always be unpredictable.

• Pamela Paul