Research purpose. The aim of the paper is to create a model that allows building an optimal brand portfolio, allowing an organisation to achieve its goals. The created model is based on the bivalent programming theory. A mathematical model of optimum brand portfolio is created based on linear programming with restricting conditions being the maximum acceptable risk level and budget. The basic types of resources and basic types of relations between brands are explained, which are part of the process of brand portfolio optimization.
Design / Methodology / Approach. Knowledge and many years of experience of mainly economic disciplines were used for the selection of characteristics for brand portfolio specified in this article. Our assumptions were based mainly on project portfolio management, operational analysis and linear programming as well as tools and methods of graph theory.
Findings. Brand portfolio management such as creating, planning, organising and then maintaining a successful brand is a costly and long-term process involving effective marketing strategies and decisions. The prerequisite for brand portfolio creation is deciding on the number and type of brands. A properly constructed brand portfolio is a prerequisite for achieving business goals.
Originality / Value / Practical implications. Brand portfolio optimisation requires sufficient attention; however, rather than the selection of the highest number of brands, it should be based on compilation of a set, according to pre-defined priorities, which would provide the best possible means to meet the company’s goals for the current limitations. It should be implemented upon objective rules (in our case maximum allowable risk level and available budget). Frequent changes in the brand portfolio structure are not beneficial since they reduce the ability for the company to achieve its targets and represent excessive use of resources. In addition, qualitative brand characteristics have to be respected in the brand portfolio management, but this was not covered in our research.
