Can one predict when a regular consumer will be purchasing a car? Even better, what kind of car? Such a feat would be manna for auto makers and car dealers but seems as impossible as predicting the next major California earthquake. Apparently, though, it may not be as impossible as it sounds.

comScore Media Metrix has just released the results of its Audience insite Measures (AiM), a system that supposedly predicted car buyers’ purchasing habits within the next six months based on their Internet profile and usage at various sites. According to the results, consumers’ age 35 to 44 years of age have a greater than 15 percent chance of purchasing within the next six months. Children, interesting, increase the likelihood that the browser will become a buyer.

I have to admit the information sounds impressive. However, I saw a few problems. The information measures folks’ answers based on their intention, not their ultimate purchase. More useful information would be to follow survey takers and find out how many actually purchase their stated vehicle (or not). Also, the information is gathered from the major “advertised” sites which may not necessarily be the most visited sites for that particular demographic. Where, for example, is Kelley Blue Book? Also, is regional information provided? Buyers on the East Coast will factor in weather in their purchases than those, for example, in California or Arizona. I remember an article lambasting GM’s Cadillac commercial showing the vehicles driving through snow during the winter months? Why was that a problem? The commercial was shown in Florida which is hot during that time. Such information is important if AiM is to help websites focus their ads.

Overall, though, I think comScore’s moving the right direction and I’ll be keeping an eye on its future stats