How to create reproducible brand personality scales

By means of personality traits, brands can be characterized in a concise and comprehensible manner that predestines a brand's personality for management with the assessment of brand characteristics and comparison with competing brands. To be able to do comparisons, the respective personality model has to be reproducible. The differing measurements must be invariant across brands, time, and, if needed, cultures. This reproducibility, however, is in question for existing brand personality scales. Recent studies could not replicate several of the existing solutions, neither in other countries nor for other brands than those in the respective studies in which the scales were generated. This study examines potential causes for invariance problems and identifies a more stringent application of the psycho-lexical approach as a remedy. The study traces back to Galton's (Fortn Rev 36: 179-185, 1884) thoughts about synonyms. When factor items possess substantially more pairwise synonyms with items within this factor than with traits outside that respective factor, the personality model will turn out to be reproducible. Surveys were conducted in the USA, Germany, and France to test for cross-cultural invariance. Implications, particularly for global branding, will be discussed.