MARKETING RESEARCH CASE EXERCISE
AMERICA ONLINE: SAMPLING ISSUES ASSOCIATED WITH INTERNET USERS

America Online (AOL) serves over 12 million households, capturing over 47 percent of the Internet market. The audience that AOL targets consist of less technically sophisticated people, ethnically diverse people, women and families. AOL is planning to design its content to appeal specifically to these groups, as well as to people using high-speed modems and low-end computers (cost of less than $1,000).
The percentage of time AOL members spend online grew from 56.9 percent in July 1998 to 58.3 percent in December of that same year. This is almost three-fifths of all time spent in cyberspace from home. However, as new competitors enter the market, online usage will most certainly change.

Currently AOL estimates that its membership exceeds 11 million. In addition, when the company acquired CompuServe, the membership increased by an additional 2 million subscribers. While in the past AOL has targeted households and small businesses, the acquisition of CompuServe now has AOL moving into the big-corporation segment of the traditional markets, the company is planning to launch a research project to assess the changing needs and desires of its current users. More specifically, AOOL wishes to determine if its customers are satisfied with its current product offerings, and what changes in theses offerings are necessary in order to maintain its current level of subscribers. See the Web page of AOL at www.aol.com for complete listings of the offerings provided to AOL subscribers.

In developing this research project, AOL is faced with some unique sampling issues. Specifically, how does AOL determine and develop a sample for this research project? Assess the following issues:

1. What should be the universe and sampling frame used by AOL when selecting participants for this upcoming research project?

2. What type of sampling design should AOL employ (random, stratified, etc.), and why?

3. Given the unique composition of its membership, what problems, pitfalls, and biases should AOL expect to encounter in its sampling plan? How can it overcome these problems?