Creating Insights in Large Markets
Final Report Abstract
While markets of past decades consisted of relatively few firms and products, markets of the 21st century are much larger (e.g. thousands of products). What has not changed is the need of firms to understand the competitive environment in which they operate in order to make strategic and tactical decisions. However, such understanding requires the collection and analysis of large amounts of data (i.e. big data), which is not possible using traditional data sources (e.g. surveys) or methods. This project created insights into large markets. It developed new methods specifically designed to process big data and applied these methods in empirical studies to data containing searches and transactions of 100,000’s of consumers. The methods are illustrated on www.mapxp.app/LEDTV. Among others, the project found that cannibalization of new products is large because, on average, 51.15% of new products’ market shares come at the expense of cannibalizing same-firm products—with heavy heterogeneity in cannibalization rates (7.66% to 94.99%). On average, 64.51% of the cannibalized market share is acceptable but this value varies strongly. The project also developed a new model to examine the organic results of a search engine (here Google). The empirical study of the retail banking market identifies 591 firms that compete for search visibility across 8 search-submarkets. Only 27% of these firms actually produce financial products—and together they capture only 14% of the market’s search visibility. Using topic modelling, the authors find that most firms compete in more than one of the discovered search-submarkets, and that firms face different sets of competitors in each search-submarket.
Publications
-
“Visualizing Asymmetric Competitive Market Structure in Large Markets”, in Mizik, Natalie / Hanssens, Dominique (editors), “Handbook of Marketing Analytics: Methods and Applications in Marketing Management, Public Policy, and Litigation Support”, Edward Elgar Publishing, 431-447
Ringel, Daniel M. / Skiera, Bernd
