Editor’s note: Rudy Nadilo is president and CEO of Greenfield Online, Inc., a Westport, Conn., Internet-based marketing research firm.

Imagine surveying thousands of your most important customers without having to pick up the phone or send a single piece of mail, and still coming in well under your original budget. Now consider conducting a focus group during a blizzard and all the participants show up. Such feats are entirely possible with on-line marketing research.

On-line research is an efficient way to survey individuals with higher levels of education, slightly higher incomes, and greater probability of spending. These "consumers who spend" (i.e., better-educated, higher-income individuals) are people with the money and the mind-set that make them the important buyers of most product categories, including packaged goods products. The Internet is proving to be the best route into the minds of these consumers. A well designed, thoughtfully executed, Internet-based marketing research sample delivers your most important consumers, faster, better and cheaper.

Consider the following example: a major international cosmetics manufacturer recently introduced a new beauty product which achieved a satisfactory level of first-time buyers but a low rate of repeat buyers compared to its chief competitor. The company knew they needed to act immediately to change their strategy before consumer habits solidified in their competitor’s favor.

The company faced two research hurdles: low product incidence and a very specific research target group (frequent buyers of high-end cosmetics who had tried their product but rejected it). They worried that a traditional scanner panel would not produce enough relevant respondents. Furthermore, their budget was low, and they needed detailed results in less than a month. The company’s research director decided they were ready to try their first on-line research project.

The on-line research project began with a pool of several hundred thousand households. During the space of a week, participants were carefully screened for matches to the company’s research criteria. Over the next week, participants logged onto the research Web site to answer nine different questions about the two competing cosmetic products and their reasons for buying one or the other and then not buying again. By the end of the third week, the data had been analyzed and delivered to the client.

The results were surprising.

Conventional research done before the product launch had suggested that the number of male consumers of the beauty product would be negligible. However, the scanner panel had not contained sufficient numbers of "consumers who spend." On-line research showed that male "consumers who spend" had a much higher trial rate (approximately 12 percent) than the general population of men (less than 1 percent), but that these males were switching to the competitor’s product because its product seemed less feminine.

Conventional research had led the company to overlook a significant portion of their most important consumers, because the "technically" representative research pool of the traditional sample diluted the impact of the product’s actual buyers. The on-line research clearly demonstrated the importance of specific types of male consumers. The company immediately began targeting more of its marketing to men. Detailed analysis demonstrated that many of these men fell into two distinct groups: male college students and gay men - groups which are notoriously difficult to reach with conventional research but which are easier to access on-line. This finding led the company to begin very targeted and successful marketing campaigns in these communities.

Within months, the company’s product had surpassed its competitor’s product in overall market share, largely because it had captured more than 80 percent of the male market, nearly all of whom came from the group of "consumers who spend." The company’s research director said that in the future he planned to use scanner panels only to determine the components of product purchasing (penetration and buying rate), but that on-line research was the best way to figure out the "who and the why" behind the penetration, particularly among "consumers who spend."

These consumers are important to marketers because of the 80/20 rule of thumb: 20 percent of customers drive 80 percent of volume. This principle is fairly accurate across nearly all consumer markets and is the main reason consumer panel research companies analyze the 20 percent of heavy buyers so carefully. Currently, much of that important 20 percent is on the Internet; conventional methods no longer offer the most cost-effective means for reaching them. An on-line research panel with hundreds of thousands of households can yield 10 times more heavy buyers than a similarly priced conventional scanner panel.

One of most important keys to the success of the previous cosmetics research project was the fact that it was done using computers. Because of the stigma attached to men who use beauty products, the company doubted that face-to-face interviews or telephone interviews (which might be overheard by the male participant’s family or friends) would elicit honest answers. Many men might not have even admitted to using the product at all. Recent research published in the journal Science suggests that people are far more likely to answer honestly to a computer than they are to a person. Computers offer participants a feeling of privacy and anonymity.

Studies show that research conducted by live researchers (whether in person or over the phone) can lead participants to give answers that they think the researcher would prefer to hear defeating the entire purpose of marketing research. Computerized, on-line research simply delivers more accurate results.

Furthermore, on-line research typically generates more detailed responses. People love to offer their opinions, but dislike the intrusion of telephone interviews and the overwhelming length of paper questionnaires. In contrast, on-line research is done at the participant’s convenience, when they are most enthusiastic about contributing. By removing the need to bring participants together in person, or to contact them via mail or the phone, on-line market researchers also provide their clients with efficiencies of scale. Finally, on-line research is simply more fun. Participants can be exposed to sound, color graphics and video, maintaining their interest and stimulating their responses.

Almost every sort of quantitative or qualitative research done conventionally can now be done on the Internet. Examples includes tracking customer satisfaction, focus groups, package testing, copy testing, and nearly all other kinds of marketing research. As more consumers begin to use the Internet, more research will be conducted on-line. Companies that start taking advantage of on-line research now will be better placed to use Internet research to reach the even larger population of future Internet users.

Unfortunately, not all on-line research is of high quality. A number of firms take too many shortcuts as they simply take all the Internet users they can find to obtain a large enough pool of on-line research participants. Another common mistake many researchers make is limiting their research pool to just one Internet community or worse still, just one "opinion" Web site. Worst of all are research "spammers" who compile databases from responses to unsolicited e-mail.

As they hurry to utilize the Internet, some firms wander from good research techniques and use methods that would be unacceptable in conventional research. When used appropriately, by people who know marketing research, the Internet offers a great way for marketers to learn from their most relevant consumers more quickly, with greater quality and lower cost than traditional research.