Inferring Brand Proximities from Online Conversations

Journal of Brand Management, 19, 467-483 (April 2012)

36 Pages Posted: 2 Aug 2010 Last revised: 29 Aug 2012

Paul Dwyer

Willamette University - Atkinson Graduate School of Management

Date Written: July 25, 2010

Abstract

Consumer choice among brands to add to supermarket shopping baskets provided insights that fueled a long stream of research into market structure, that is, the distinction between brands that compete for the same purchase decision. However, the range of brands appearing in such shopping baskets was limited to those found on supermarket shelves. This study investigated whether consumer choice of which top brands to mention online, often called user-generated content, encoded a broader perspective on how brands are distinct in consumer perceptions. Brand distinctiveness is usually portrayed on a map where brands are plotted in a coordinate system or as a tree where brands are clustered into hierarchical categories. However, this study found that the conventional mapping and clustering techniques failed to depict coherent brand proximity from its data, while the seldom-used in marketing minimum spanning tree and force-based layout algorithms succeeded. Online participants generally organize brands around the expected market categories. However, there are sometimes unexpected inclusions of brands in otherwise well-defined categories, and interesting placements of brands relative to one another. Marketers may use these insights as brand-building feedback, and as a source of co-branding ideas to take advantage of the ways consumers organize brands.

Keywords: market structure, consumer-generated media, brand management

JEL Classification: D49,C69

Suggested Citation

Dwyer, Paul, Inferring Brand Proximities from Online Conversations (July 25, 2010). Journal of Brand Management, 19, 467-483 (April 2012) . Available at SSRN: https://ssrn.com/abstract=1648503

Paul Dwyer (Contact Author)

Willamette University - Atkinson Graduate School of Management ( email )

900 State Street
Salem, OR 97301
United States

Paper statistics

Downloads
62
Rank
290,727
Abstract Views
620