with Auren Hoffman
June 16, 1997

Marketing with Psychographics
Demographics are poor indications of buying habits

by Auren Hoffman

What happens after you build a database full of information? What can you do with this information? How do you aggregate all this information? The solution involves data gathering, data consolidating, data drilling, and data mining.

Let's take GuestiMate, for instance. Here you might gather data on thousands of people that visit your web site. You might ask them to give you information about who they are, their interests, their life, and other miscellane ous information. But how exactly is this information going to help you?

Well, you can do a simple query on this information, like in the GuestiMate example. You can find all the people that match a certain demographic criteria. But how does that help? If you are selling a product, you want to target the people most likely to buy it. If you are selling life jackets to canoe lovers, how do you know that white women over age 50 who live in Colorado and ordered a catalog in the last 2 years are you most likely customers?

The fact is that demographics are poor indications of buying habits. Latino people are all different, listen to different music, eat at different restaurants, and drive different cars. An Asian male, let's call him Mike, who lives in Los Angeles, will p robably not like 80% of the same things as another Asian male in Los Angeles. But Mike might be a 85% match with a black woman who lives in Britain. To determine a person's buying habits, companies are slowly starting to analyze psychographics ra ther than demographics.

There are many different theories as to how best to analyze someone's psychographics but the general idea is to type an individual's personality based upon past data on the person. If person A strongly likes eight different movies and person B also stron gly likes the same movies, there is a strong correlation between person A's and person B's movie tastes. Firefly does exactly this. Firefly is setting up kiosks in Tower Music and Tower Video. Here people enter in the music and movies they are interested in. Firefly builds a big database of people from around the world and scores people relative to one another. Since person A and B (above) are very similar, the next time person B goes to rent a movie, Firefly might suggest the movie that person A just rented the week before.

There are many services like this. Another service, GroupLens, asks its users to rate newsgroup articles after they read them. Then, as every article appears, GroupLens will assign it a weighted aver age from all its users with the responses from people who usually agree with you given more weight and the responses from people who rarely agree with you given little weight.

In both Firefly and GroupLens, data is collected on an individual and then matched against other individuals using sophisticated technology to create a profile. This profile can then help this person to make purchasing, reading, or other related decisio ns. In using psychographics, rather than demographics, a company is more likely to market the right product at the right time to the right person.

Previous Summations: