Can a survey move from paper forms to electronic data collection without diminishing data quality and response rates? How does a survey balance the preferences and business practices of the establishments with the need for additional data while also achieving cost-savings? The Occupational Employment Statistics (OES) Survey, a large voluntary establishment survey, wrestled with a decision to solicit electronic payroll records and web submissions instead of using 24-page paper survey forms. OES analyzed mode and response rate data and trends and reviewed other research regarding data quality and collection modes. Analyses identified electronic data collection as the mode of choice for establishments for most, but not all size classes. Electronic data collection also facilitates the collection of point wages and job titles which are needed input for BLS research efforts such as auto-coding. In this paper and presentation we examine the research and data analysis used in the decision-making, as well as the impact of eliminating paper forms on mode choices, response rates, and survey costs.