In situations of information overload and complexity, consumers consult their existing knowledge regarding brands as a guide in consumption decisions. This knowledge manifests as brand association networks (BANs) in consumers’ minds and reflects what the consumer thinks of when being confronted with a brand stimulus. BANs therefore characterize a brand’s image that determines consumers’ attitudes and behaviour. BANs serve as diagnostic instruments to explain a brand’s success or failure and to plan or control marketing activities. Traditionally, BANs are elicited directly from consumers utilizing survey-based instruments. However, in a dynamic and interactive environment, user-generated content (UGC) is increasingly relevant for a brand’s image and thus should be exploited for the elicitation of BANs. However, established elicitation instruments either follow another elicitation paradigm (i.e. surveys or interviews), or are unable to cope with volume, velocity, and variety of UGC as a big data source (e.g. content analysis). Hence, exploiting UGC for BAN elicitation requires the development of new, computer-supported instruments. Following a design science research approach, we contribute a novel methodology as our artefact to extract BANs from UGC using text-mining and net- work analysis. We evaluate our solution and demonstrate its utility for brand management on a study of automotive brands.
Egger, Marc; Volkmann, Gloria; and Schoder, Detlef, "Designing a Methodology for Marketing Intelligence Systems – The Case of Brand Image Diagnostics" (2018). Research Papers. 195.