Leveraging Matching and Competition Information to Design Effective Paid Search Ads

Abstract

Paid search advertisers strive to optimize search relevance (through ad design) and competitive advantage of their ads (through bidding strategies) so as to more effectively attract consumers’ attention. Consumers’ interest in search ads that appear with search results may extend to ads which do not necessarily have a literal or exact match to the search term, but which are related to the information they are searching for. Their interest in relevant ads may also spill over to competing ads which often appear in close visual proximity to focal ads. Hence, different from exact matching and positional competitive advantages that have been well examined in the extant literature, this study develops a systematic framework of metrics of 'inexact' (but relevant) matching and 'non-positional' competitive advantages from an informational matching perspective.

We introduce the concept of 'semantic matching' to capture the semantic relevance of paid search ad content to the information searched for by consumers and also introduce two information-based metrics of competitive advantages. Empirical results using a unique dataset from Google China show that 'semantic matching' can effectively increase click-through as compared to 'exact matching'. Our results using competitive metrics provides empirical evidence for the positive externality effect of search ad competition predicted by the theoretical literature in economics and marketing, while showing that in the presence of directionality, the negative competitive effect dominates the positive externality effect. We study the variability of extant and proposed metrics with complexity of consumer search and also demonstrate the efficacy of the new matching and competition metrics in improving search ad design.

Keywords: Paid Search Advertising; Click-through Rate; Exact Matching; Semantic Matching; Online Competition