The Internet is Fundamentally Changing
How Research is Done

by Albert Fitzgerald

Editor's Note: Orginally published in the San Diego Daily Transcript, February 2000

Who is doing market research on the Internet? Just about everyone. Clients are "insourcing," and product managers are bypassing their market research managers. Programmers, advertising agencies, webmasters and traditional market researchers are all conducting research using this new research tool. As a data collection method it far surpasses the impact that computer-assisted telephone interviewing has had on traditional pen and paper studies. As a means of empowering clients it is revolutionary. The Internet has caused the largest paradigm shift in market research since the late 1950's when telephone surveying became commonplace.

There are currently about 90 million online users in the United States and Canada, 38 million in Europe and 27 million in Asia. The projected increase is 100 percent in the next four years. Currently, the majority of users are on the Internet two or fewer hours per week but as the services available increase, usage will also increase. The recent acquisition of TimeWarner by AOL illustrates the Internet's rapid ascent into mainstream communications.

Why are people using Internet surveys? The pace of business continues to accelerate and research issues and questions continue to increase in complexity. Clients are demanding faster and faster turn around with at least partial information available immediately. There has also been a continuing decline in response rates due in part to over-surveying and the increased use of voicemail. Internet surveys overcome the majority of these issues.

There are a range of options for using the Internet: You can email a "paper" survey and receive a reply by email, or record your survey and answers with a diskette by email using FTP downloads. You can have either static or dynamic HTML surveys using browsers. Utilizing Usenet, chat focus groups can be conducted online and newsgroups can be mined to find information and individuals. Message boards allow for dynamic real-time communication among numerous participants.

Examples include:

A high tech company emails every 1,000th customer with a URL that links them to a customer satisfaction survey site. The customer clicks and arrives immediately at the site. Results are immediately compiled and available for direct viewing or immediate follow-up by the company.

An insurance company emails employee satisfaction surveys to all employees. The employees email the completed surveys back. Information asked includes rating type questions on benefits currently in place along with open-ended questions which probe for their feelings on feature benefit plans.

A market research company is commissioned to conduct a new product concept test. In addition to standard ratings questions, choice model questions are asked and still photos or streaming video show actual products online. This allows consumers to assist in the actual design of new products.

An e-commerce site manager has a shopper take a web satisfaction survey when the shopper empties his or her cart. The information is available immediately to allow for any service problems to be addressed in a very short period of time.

An Internet portal conducts a site evaluation survey online. A split window on the computer screen allows for real-time survey completion and website evaluation. Website features are tested and assessed simultaneously.

The Internet provides a cost effective, timely method to reach a large group of respondents. It also provides an inexpensive method to search for hard to reach respondents. Internet surveying using the 24 by 7 capabilities of the web allows participants ranging from consumers to physicians to farmers to complete surveys at a time convenient for them.

Since the questionnaire is in plain view with an Internet deployment, it is critical that the survey sponsor ensures the highest quality possible. Respondent cooperation in completing the survey and their view of your company is dependent upon the look, feel and ease of use of the survey document. Researchers can no longer rely on skilled interviewers to help respondents through a survey. Any awkward or confusing questions cannot be immediately clarified.

Caveats:

While the trend has been for companies to implement research in house, several unexpected pitfalls lay in wait. A risk for the in-house research manager is that his or her internal clients are viewing their data as it is collected. While this causes some consternation for the researcher, i.e. will they make a decision on partial data which doesn't reflect the entire population surveyed, reviewing results on line is here to stay. Management needs information quickly and in some cases they just can't wait for dirty data to be cleaned, results to be verified, and insightful analysis to be performed.

The key to actionable research is comprehensive sample control. Using Internet data and projecting it to the population at large may not work. Consumer sampling in the United States worked because there is a known population of phone numbers. No such list exists for general consumer Internet users. Therefore, currently you cannot pull a random sample of consumer Internet users. This lack of sample control is the reason Internet research is currently not projectable to the general consumer population. For certain audiences such as high technology users, sophisticated methods exist to allow for valid sample frames.

As the Internet evolves, both as a medium and as a market, we will see new and innovative means of data collection along with dynamic methods for providing real time marketing intelligence to key decision makers. As with other high tech applications, it will be cheaper, faster, faster and of course, faster.

Return to Articles Home

Back to top