In a recent article in this Journal, S. Vargo, the service-dominant logic theorist, makes the unarguable claim that "What is needed is a <i>true science of service</i> (Vargo and Akaka, <i>Service Science</i>, 2009, p. 39, emphasis in original). To constitute a true science of service, services research must use valid measures. The <i>sine qua non</i> requirement for a valid measure is content validity, which comprises item-content validity and answer-scale validity (see Rossiter, <i>International Journal of Research in Marketing</i>, 2002, and <i>C-OAR-SE Construct Measurement for the Social Sciences</i>, Springer, forthcoming). The present article begins by criticizing the content validity of E-S-QUAL (Parasuraman, Zeithaml, and Malhotra [Parasuraman, A., V. A. Zeithaml, A. Malhotra. 2005. E-S-QUAL: A Multiple-Item Scale for Assessing Electronic Service Quality. <i>Journal of Service Research</i> <b>7</b>(3) 213-233.]), the principal academic measure of e-retailer service quality, which is probably the most important construct in contemporary services research. Rossiter's (Rossiter, J. R. 2002. The C-OAR-SE Procedure for Scale Development in Marketing. <i>International Journal of Research in Marketing</i> <b>19</b>(4) 305-335.) C-OAR-SE construct measurement theory is then applied to develop a new and more valid measure, called ER-SERVCOMPSQUAL. This new measure seeks potential customers' and current customers' judgments about the quality of the <i>components</i> of e-retail service (thus ensuring high item-content validity), which are rated in terms of behavioral answer categories (which have high answer-scale validity). The new measure clearly shows e-retail managers where to make improvements in their service quality. [<i>Service Science</i>, ISSN 2164-3962 (print), ISSN 2164-3970 (online), was published by Services Science Global (SSG) from 2009 to 2011 as issues under ISBN 978-1-4276-2090-3.]
