As information technology becomes more and more ubiquitous in modern service organizations, decision makers in these organizations need guidance in determining which service offerings should or can be automated, and which should not. This project seeks to provide such guidance, by empirically examining user reactions to both traditional, human-staffed service offerings and newer, automated service offerings. Specifically, five components of user satisfaction which have been established in the literature are measured both before and after an automated telephone information system is installed to replace a human-staffed customer service center. The field site is a large national financial services company, and two sets of 500 surveys each were mailed to users of the system. Results indicate that the longer the length of usage of a service prior to automation, the more satisfied the user. However, both the age of the user, and the user's previous experience with similar types of automation had no effect on satisfaction.