Abstract

Theory and empirical evidence suggest that the dimensions of business-to-business customer service differ to individual service. In response, Gounaris (2005a; b) developed the INDSERV business-to-business customer service scale, the sub-scales of which are potential quality, hard process quality, soft process quality and outcome quality. INDSERV seems to be a valid and superior scale for business-to-business settings. However, Gounaris treats the four internal dimensions as indicators of a single latent “B2B service” factor. This paper argues that researchers should model the INDSERV dimensions as a process, rather with potential quality being an antecedent of hard and soft process quality, and all three being antecedents of outcome quality. This argument is tested through structural equation tests both of original Gounaris (2005c) covariances and a new dataset from 170 supply dyads in South Africa. The results generally support the alternate conception of business-to-business customer service. Modelling customer service in this way allows more complex and interesting model generation. Allowing direct relationships between the indicators of INDSERV, and differential relationships of these with other variables, allows for broader hypothesis development and therefore greater expansion of the theory. Various implications arise for marketing and other practices and researches.

Key words: Business-to-business customer service, B2B, INDSERV, structural equation modelling, South Africa.