The growth in electronic data collection via computerized self-administered questionnaires or Web surveys opens the door to incorporating data edits into the data collection instruments so that suspicious data entered by respondents may be corrected - or explained - at their source. The result should be improved data quality that more accurately reflects respondents' circumstances. However, incorporating data edits into the data collection instrument brings with it new responsibilities and new issues for survey designers. The U.S. Census Bureau has more than ten years experience with electronic data collection for establishment surveys, along with having conducted usability research with business respondents. This paper will share insights gained from this experience and research and will describe how editing strategies have evolved so that data quality is improved without sacrificing survey response.