Generalizing spillover effects of goods and service brand extensions: A meta-analysis approach

The paper examines the generalizability of forward and feedback effects of durable goods and service brand extensions through a main study and four replications. The replications address the generalizability of empirical findings across characteristics of parent brand (luxury versus non-luxury), extension type (e.g., goods extending into goods versus extending into services), and country (Spain, U.K. and Italy). To establish the generalizability of results across the studies, the first instance uses a vote counting procedure-counting the number of significant and non-significant results for each hypothesis. This procedure reveals some potentially useful patterns, such as the extension type as a possible intervening factor. Overall, however, establishing conclusive results about the characteristics of the parent brand and the country is difficult. Researchers then adopt a meta-analysis approach by integrating all effect sizes obtained in each experiment. Results show that fit between the extension and the parent brand is consistently the main factor when consumers evaluate brand extensions and the extended brands. However, brand image is more vulnerable to the perceptions of fit for parent brands associated with durable goods than for parent brands associated with services. Hence, when launching brand extensions, goods brands may face a higher risk to image than service brands. Another notable finding is that the effect of brand image on extension attitude is lower when the extension is in a different sector than the parent brand. Finally, neither the characteristics of the parent brand (luxury versus non-luxury) nor the country affected the results. (C) 2012 Elsevier Inc. All rights reserved.