Abstract
The successful registration of a color as a trademark requires the courts to assess whether the color has developed secondary meaning with respect to a specific brand. Survey evidence often is used to support distinctiveness claims but frequently carries little weight because of criticisms of the methodology used. Cadbury's application to register the color purple in New Zealand, for instance, provided a context for comparing three methods of establishing brand/color distinctiveness: a traditional approach and two novel approaches involving a color wheel and choice modeling. All three methods revealed strong associations between Cadbury and purple, but the new methods tested are potentially more robust and less susceptible to challenge in the courts.