The Influence of the Congruence Between Brand and Consumer Personality on the Loyalty to Print and Online Issues of Magazine Brands

Abstract

An important development in the theory of brands has been the concept of a brand personality as part of the brand image. The construct of congruence relates the personality of brands to the self-concept of consumers. It is assumed that congruence between the recipient's perception of his or her own personality and his or her perception of the brand personality leads to a higher degree of loyalty. The congruence construct has been proven to be a loyalty-promoting factor in various studies for consumer good brands. Whether the influence of congruence also exists with regard to media brands has, however, not yet been examined. As a step to close the gap, this research explores the effects of congruence on consumer loyalty in the media, taking the leading German news magazine brands— Focus, Spiegel, and Stern—which are published in print and online, as a case. The study uses a structural equation model, which is estimated by the partial least squares method, based on a questionnaire answered by 736 respondents. The model was tested for all respondents and those with print and online preference separately. As a result, the influence of congruence on loyalty could also be detected in our media case. Differences of the impact within the models for print and online give rise to the assumption that a differentiated approach to brand management in these two fields has to be taken into consideration.

Notes

1. The importance of using media-specific operational definitions for research that replicate conventional branding studies was pointed out by McDowell (2006, p. 242).

2. Chan-Olmsted and Cha (2007) proposed to compare news brands across different media platforms and to examine the perceived differences in brand personality, if any, of particular news brands between their offline and online outlets.

3. We thank an anonymous reviewer, whose valuable remarks significantly contributed to improving this article.

4. Congruity is used as an interchangeable expression for congruence.

5. We also had to ensure unidimensionality, meaning that indicators relate only to one construct (see also footnote 9).

6. Meyer (2004) designed a structural model that explains, among other things, the use of newspapers as a function of credibility. A version of Meyer's (2004) model was applied by Oyedeji (2009)to the cable network environment. Here, Oyedeji also assumed an indirect influence of credibility in Meyer's (2004) narrow sense on loyalty. Concerning the cross-media background of this study, Rössler and Ognianova (1999) could prove that recipients transfer the existing credibility bonus from the classical media to the new media.

7. The corresponding traits are in parentheses.

8. We directly measure the indicators of credibility on a 5-point Likert scale, and not as a difference between ideal and actual evaluations of the indicators, as it is done to measure functional congruence by Kressmann et al. (2006) for economic reasons and in the assumption that the direct measure highly correlates with a measurement by the differences.

9. Tests have been performed applying the following test statistics and thresholds: (a) formative constructs: Discriminant validity is tested by the correlation of indicators within the constructs < 0.9, and multicollinearity is tested by the variance inflation factor (VIF) < 10 (indeed, all VIFs are smaller than 2, except for integrity and competence in the case of ideal congruence, where they are 2.5 and 2.8, respectively); and (b) reflective constructs: Covariance validity is tested by the average variance extracted > 0.6, reliability of the construct is measured by the explanation of the variance of the indicators by the construct > 0.7, discriminant validity is tested by the Fornell Larcker criterion to be met (Fornell & Larcker, 1981, p. 45), and predictive validity is tested by Stone-Geissler's Q > 0 (cf. Fornell & Cha, 1994, p. 72), as well as unidimensionality, meaning that indicators relate only to the construct they are supposed to reflect by a confirmatory factor analysis.

10. Concerning the test of the overall and the other structural models, which are discussed later in this article, we did not detect multicolliniarity for the constructs that influence endogenous variables, tested by variance inflation factors (VIF) < 10 (indeed, all VIFs are smaller than 2). Stone-Geissler's Q ² indicated predictive validity with values > 0 for all endogenous reflective constructs (cf. Fornell & Cha, 1994, p. 72), except credibility in the print model and brand relationship in the online model, which were at values slightly below 0, equaling about –0.1. This indicates a lack of predictive validity for these constructs.

11. The resulting ambiguity of the advice to avoid vitality, as well as stability, vanishes when analyzing the single print brands separately. In this case, the Importance-Performance-Matrix leads to a more plausible interpretation concerning the positioning of indicators for the single brands. To present these findings is beyond the scope of this article.

12. In addition, the causal model has been separately tested for all brands (i.e., Focus, Spiegel, and Stern; and Focus Online, Spiegel Online, and Stern Online). This research discovered interesting differences between the brands. To show these results is beyond the scope of this article.