by Bruce Crandall

E-commerce isn’t the only growing online trend. So is Internet research. And, apparently, with good reason. Businesses spend millions of dollars each year surveying customers on-site, over the telephone, or through the mail. While such feedback is valuable, the traditional ways of collecting it have become increasingly more expensive, time-consuming—and perhaps less effective. This is particularly true of telephone surveys. Rising labor wages, consumer frustration with telemarketers, and caller ID have all contributed to higher costs and lower production rates.

With the decline in telephone productivity, the Internet is becoming increasingly important among many marketing researchers as a viable data-collection source. With over half of the U.S. adult population now with Internet access at home or work, a substantial number of U.S. companies are tapping into this source of consumer feedback. Some marketing research companies have developed large consumer panels, where sizeable groups of respondents (100,000+ members) agree to take online surveys or to be part of online “focus group” discussions or in-depth interviews.

Such online panels are thought to be reliable data-collection sources because panel respondents are recruited not as a result of email spamming, but usually through a sophisticated combination of methods like telephone, mail, banner advertising, and print advertising.

However, many marketing researchers remain dubious of the Internet’s reliability, preferring instead more traditional data-collection methods. Indeed many companies conduct opinion polls and surveys every day with website visitors, but usually these represent only those respondents who are passionate about a certain topic or who visit a particular website more frequently than others. Some marketing research companies even have recklessly jumped on the bandwagon, promising Internet capability while delivering almost primitive survey programming and questionable respondent samples. This, along with a potential demographic bias of Internet users (i.e., typically younger, better educated, and more affluent), has contributed to the skepticism about Internet research.

While the Internet has its skeptics, there are strong reasons for considering it:

Internet Panels Allow For Large, Accessible, And Representative Respondent Samples

When respondent panels surpass hundreds of thousands of members, they allow researchers to overcome the Internet skew of younger, better educated, and more affluent users: the large pool of respondents allows for demographically balanced, representative probability samples to be drawn.

Depth And Breadth Of Quantitative Surveys

The Internet also allows a wide range of surveys to be fielded, including new product concept and advertising concept testing; awareness, trial, and usage (ATU); satisfaction; and segmentation, among others. Sounds, images, and video can also be incorporated into online surveys, enhancing the quality of product and advertising testing.

Qualitative Depth In A Quantitative Survey

The nature of Internet research, a methodology that allows respondents time to consider and quickly type in their answers, lends itself well to yielding almost qualitative depth from open-ended questions. Respondents can think through open-ended questions and type in as much information as they wish, offering it firsthand rather than through an interviewer.

Lower Cost

Internet surveys usually cost less than telephone studies and less than mall-intercept. Even qualitative (interactive) research conducted over the Internet—such as focus groups, in-depth interviews, or brainstorming sessions—typically costs less than its in-person counterparts.

Faster Turnaround

Internet studies are often faster than other quantitative surveys, due to the shorter data-collection cycle. However, while almost instant access to thousands of respondents is possible, a minimum window of time is necessary to randomly select from a base of qualified respondents; by cutting off data collection too quickly, the sample may become biased because it represents only those who discover the online survey first rather than those who might also qualify, but who may access the Internet less often or only at certain times of the week (or who may live in an earlier time zone and are, thus, more apt to see it before those in other, “later” time zones).

Quick Access To Data

As in computer-aided telephone interviewing (CATI), data from Internet surveys is available for review immediately following data collection. In fact, survey results can even be viewed online in realtime as respondents complete the questionnaires. Also like CATI, the data is available in topline or marginal format, but unlike CATI, open-ended responses in the form of “verbatim comments” can also be produced immediately after data collection.

Which Internet Research?

While online consumer research can be an effective data-collection method, not all Internet research is the same. Nor are the companies that supply it. So what should one look for in a company if he or she is considering Internet research? Besides the normal criteria used to select a marketing research supplier, here are a few other things to keep in mind:

Large Online Consumer Panels

As mentioned earlier, a large online panel provides the safest, most reliable data because the large pool of respondents allows for demographically balanced, representative and/or probability samples to be drawn.

However, one should always check how panel members are recruited (hopefully through a number of avenues), how often they’re surveyed, and how they’re compensated. For instance, those respondents who take more than two or three surveys a year for a given product category may lose their “amateur status,” becoming insiders and, thus, biasing the survey results.

Moreover, overworked and undercompensated respondents can become less willing to offer genuine opinions and often drop out of a panel completely. Since sampling reliability is key, how a marketing research firm recruits its panel members, how it treats them, and how often it uses them all go a long way toward producing reliable results.

Variety Of Data-Collection Methods

Anyone searching for an Internet research partner should look at the company’s track record and ability to use multiple quantitative and qualitative data-collection methods, such as telephone, mail, mall and on-site intercepts, focus groups, in-depth interviews, etc. A company’s core competence in a variety of traditional research techniques and methodologies says, at the very least, that it can probably conduct Internet research credibly as well.

Questionnaire Programming Capability

Companies providing Internet research should be able to create virtually any type of questionnaire (i.e., with skip patterns, rotations, etc.), using software that features Boolean and/or Algebraic logic. Questionnaire programming software should also be able to incorporate sounds, pictures, and video that are a vital part of any product and advertising testing.

Online Safeguards

A company’s Internet system should not allow respondents access into a survey without safeguards like user IDs, passwords, and security access codes; nor should it allow someone to complete more than one online survey for a given project. Optimally, all communications to and from the Internet server should also be encrypted, with the survey data safeguarded behind a protective firewall.


Despite the legitimate sampling concerns, despite the perceived limitations on survey programming, and despite concerns with the safety of respondent data, online research is coming of age. Internet research does have its skeptics, but it continues to grow at a furious pace, providing believers with a valuable alternative to traditional data-collection sources.

Contact Decision Analyst

For questions about this article please contact Clay Dethloff, Senior Vice President, (cdethlo@decisionanalyst.com), call 1-800-262-5974 or 1-817-640-6166.

Copyright © 2000 by Decision Analyst, Inc.
This article may not be copied, published, or used in any way without written permission of Decision Analyst.