Editor’s note: Pat Funes is vice president of operations at Answers Research, Inc., Solana Beach, Calif.

Emerging from its humble, low-tech beginnings in the 1950s, the segmentation study has evolved significantly. A tactically valid and valuable tool, segmentation research continues to help marketers find solutions to the most daunting high-tech product challenges

Combining the speed of today’s sophisticated computers and the technological breakthroughs of the Internet revolution, online segmentation research has emerged as a sleek, complex, and powerful marketing weapon for the new millennium. The speed, efficiency and cost-effectiveness of market research have all been drastically improved with Web deployment.

While impressive, online segmentation research does have its weaknesses. Because Internet penetration is limited (less than 50 percent in the U.S., 20 percent in Europe), straight online deployment tends to create flawed samples. Consequentially, the resulting data is not projectable to all but the most narrowly focused, tech-based markets. Clever market researchers are learning to overcome these sampling challenges and bridge the gap between the high-tech haves and have-nots by utilizing a blend of CATI and Web-based programming.

In the hyper-competitive 21st century game of business, segmentation research can be accurately characterized as a divide-and-conquer tactic. An innovative and complex mixture of art and science, segmentation divides a given market or population into smaller groups, or segments, based on common characteristics. Once divided, the marketer can then target these segments with enticing products, pricing or promotions designed to obtain greater market share.

Online segmentation research is powerful in part because it can efficiently and cost-effectively provide almost limitless ways to specifically categorize market data. The market can be segmented by product parameters, which include brand loyalty, attitudes, awareness, level of use and satisfaction. The market can also be grouped by demographic patterns, including gender, race, religion, geography and marital status.

Market segmentation allows the marketer to reach potential buyers in a given market using the most targeted, customized product or service possible. Segmentation can also increase a marketing campaign’s efficiency. Marketers who use segmentation research tend to make more efficient use of their resources by focusing on the best segments for their product or service. Segmentation should be used anytime you suspect there are significant, measurable differences in your particular market.

Online deployment benefits

Online deployment of segmentation research is a formidable weapon because it combines the benefits of speed, efficiency, and cost-effectiveness. Traditional segmentation research methods, such as paper-based surveys and disk-by-mail, move at a snail’s pace compared to the high-speed online deployment. Sophisticated computer programming facilitates the speediness of an online segmentation study. The study, programmed with a core list of survey questions, performs lightning-fast computations, with the data simultaneously computed, compiled, updated and displayed in easy-to-read graphs and charts right on the survey host’s computer screen. This real-time graphical representation allows marketers to analyze the data more quickly and efficiently than ever before. The good news is this speed and efficiency does not come at a price, it comes at a price reduction. Internet surveys are generally 20 percent cheaper than traditional methods, and the cost savings can be even greater for segmentation studies.

Additionally, we have found that offering a Web survey option significantly increases respondent cooperation and thus increases the efficiency of the sample and reduces non-response bias.

Online survey deployment of interviews during a segmentation study also eliminates the problem of interview bias. During a conventional telephone-deployed segmentation study, responses are affected by the subtle psychological influences exerted by the interviewer’s tone, inflection and speech patterns. Personal feelings and impressions of the interviewer, positive or negative, tend to reduce candor and produce tainted, less-honest responses. During an online deployment, voice-based personal interviews are replaced by written questions and responses on a computer screen. This tactic ensures that the respondents’ answers aren’t swayed, shaded, diluted or otherwise biased by an interviewer’s delivery or personality.

Marketers looking for extremely rich, candid responses would prefer to ask open-ended questions. Once again, online deployment comes to the marketer’s rescue. Typed responses allow the respondents to offer lengthier, detailed narratives at their convenience.

The ability to deploy one segmentation study simultaneously in several countries makes online deployment especially attractive. Let’s take a closer look at how an online deployment handles an ambitious, aggressive worldwide study with ease, compared to the logistical nightmares caused by a traditional, phone-based deployment.

To prepare for an online deployment in the United States, France, Mexico, Germany and Japan, the survey must be translated into each language. During translation, cultural differences are noted.

Once the survey is programmed, respondents complete the survey, and the results are compiled and translated simultaneously in real-time. The results are then converted into easily digestible graphical data, allowing the marketer to evaluate the findings instantaneously. During a conventional survey, the data formatting and compilation process is slow, complicated and costly. Multiple fielding agents from time zones all over the globe have to submit their files. Once submitted, these files have to be compiled and formatted, a process that is further complicated by differences in software. By comparison, an online implementation allows the marketer to complete the process in just a day or two. This accelerated process reduces the project’s schedule by half and dramatically lowers costs.

Drawbacks to online administration

Online deployment does have a few notable weaknesses. A significant disadvantage is the fact that online deployment may not be projectable to any target other than online buyers. The key to actionable, projectable research is sample control. Using Internet data and projecting it to the population at large may not work. Phone-based consumer sampling in the United States works because there is a known population of phone numbers, and market penetration is extremely deep (98 percent). No such list exists for general consumer Internet users, and market penetration is currently at only 46 percent. These numbers are considerably lower internationally. Therefore, you cannot pull a random sample of consumer Internet users. This lack of sample control is the reason Internet-based research is currently not projectable to the general consumer population.

Overcoming drawbacks

Acknowledging the disconnect between segmentation research and its projectability to all but the wired universe, clever marketers have found ways to bridge the gap. Researchers concerned with delivering a projectable sample can improve the projectability of their sample by exclusively using or supplementing with telephone recruiting, limiting the ratio of surveys sent to completes, quotas, and sample balancing. Additionally, until the penetration of Internet users is higher, researchers will use a combination of telephone and Web-based surveying.

Our firm recently conducted a blended administration of a segmentation survey of just over 600 completes among employees of Fortune 1000 companies. We contacted and screened employees at home or work by obtaining area codes and exchanges of these companies and, for home contact, mapped a seven-mile radius to obtain additional exchanges for a RDD sample.

Recognizing that some may not have access to the Internet at work or home, or simply prefer to complete a survey immediately over the phone, we gave respondents the choice to complete the 30-minute survey on the Web, over the phone when initially contacted, or over the phone later by calling a toll-free number that would connect them to a live interviewer. Of qualified respondents, only 16 percent refused to participate. Of the remaining, nearly 65 percent chose the Web option, while 25 percent elected the immediate phone survey and 10 percent said they would call back later to complete the interview.

Of course, not everyone who said they would complete the survey on the Web or call back for a telephone interview did, but both options yielded roughly an 80 percent completion rate. Compared to a traditional telephone administration, the overall response rate was significantly higher, yielding a more representative sample at a reduced cost.

When we compared results from the Web and telephone survey data, we found only a handful of significant differences. Of those statistically significant, most were related to an inflated importance of price and brand attributes of responses captured via the Web vs. telephone. This makes intuitive sense, as respondents are more candid via the anonymous Web than with a live telephone interviewer. Additionally, open-end responses were more plentiful both in number and content for the Web surveys.

Powerful weapon

By blending CATI and Web-based programming, we have discovered it is indeed possible to launch an online segmentation study that’s actionable and projectable across a much wider audience.

The evolution of segmentation research is not coincidental. The dawn of the computer era has improved the speed, quality and effectiveness of segmentation research. The Internet revolution has made online segmentation research one of the most powerful weapons in the arsenal of the 21st century marketer. And as Internet usage becomes more ubiquitous, online segmentation studies will become more valuable still.