Prof. Dr. Miguel Brendl, Dr. Marco Biella

Marketing

Research Topic
In one popular type of advertising a celebrity like Michael Jordan endorses a brand like Nike. Brand managers choose a particular celebrity such as Jordan based on intuitive and practical considerations, one being that the endorser must (somehow) fit the brand. In contrast, we propose that this selection can be informed by more systematic evidence-based considerations, and that doing so results in counterintuitive conclusions. When Michael Jordan endorses Nike, they co-occur in one advertising, that is, they are paired. Behavioral experiments show that merely via this pairing some attributes of the celebrity get transferred to the brand. As a result, consumers will perceive Nike as more athletic and related to victory, even when Michael Jordan is not in sight. This well-studied mechanism is known in the literature as attribute conditioning. The marketing literature has confirmed managers’ intuition that in order for celebrity endorsement to work, the selected celebrity must “fit” the brand . Unfortunately, “fit” is never defined explicitly in this literature. This is a problem because fit may just be another word for one’s intuition that this endorser will be effective, boiling down to saying, “this endorser will be effective because she seems effective”. Such a fit concept does not contribute any insight. Practitioners currently implement a working model of fit based on attributes similarity between brand and celebrity (henceforth, similarity model of fit). However, this model implies that high-fit celebrities can shift brand perception only up to a point, limiting the extent to which they can affect brand perception. This is the main problem we want to address.

Statement of the Problem
To illustrate the problem, assume that a brand (i.e., for ice skates) can be perceived along only two attributes, elegance and strength. Customers initially perceive the brand as high in elegance but low in strength, and the firm wants to reposition the brand to be perceived as high on both attributes. According to the similarity model of fit, only similar skaters (elegant but not strong) are potential endorses. This implies that the repositioning can be achieved only via multiple campaigns executed in sequence. Indeed, campaigns t1 and t2 can successfully shift brand perception while campaign t3, relying on the same Endorser B as t2, cannot because B does not fit with the brand prior to the execution of t1. A very inefficient process. We propose a novel hypothesis that challenges the status quo: if the attribute that needs to be transferred is a defining one for the endorser, attribute transfer will be effective even for low-fit endorsers. If this is the case, the low fit is not a constraint and rebranding campaigns can rely on non-fitting celebrities. They have the additional benefit that their defining features align with the intended shift in positioning. Indeed, contrary to current wisdom, we predict that t4 would achieve the intended repositioning in a single hop. Moreover, such a strategy will draw attention because the celebrity will be surprising as an endorser of the brand. Prof. M. Brendl Dr. M. Biella 29 This should boost memorability and increase brand awareness. Indeed, even though tennis aficionados might remember which brand of tennis rackets Roger Federer endorses, newcomers will not. However, many customers would remember the brand endorsed by Dwayne „The Rock“ Johnson because the low fit is surprising. Of importance, this low fit should only be effective if it is a defining feature of the endorser.