Using Lexical Semantic Analysis to Derive Online Brand Positions: An Application to Retail Marketing Research
Section snippets
Background and problem statement
If managers do a simple keyword search to assess what the online world says about their brands, they will usually get back thousands of hits from a search engine. They will then face the onerous task of sifting through the voluminous data generated by those hits. Currently, the usefulness of the hit information is limited by the fact that the listings contain no summary information about the content. Nor do the listings contain information on the “semantic” properties of hits containing textual
Lexical semantics—some initial ideas
Typing the name of a brand or a travel destination into a search engine will often provide thousands of hits, but the searcher does not gain information on whether these sites have positive, negative, or neutral things to say about the brand or destination (Turney 2002). To get a better insight into the meaning in the text, researchers in the fields of artificial intelligence and natural language processing are beginning to focus on evaluating the content of large text-based corpora.
One of the
Lexical semantics—the method
For the purpose of our analysis, we utilize two key properties of textual data: (1) consumers attach meanings to words and (2) meanings are inherent in the text, or more specifically in the adjectival expressions used by the author of the text. Understanding the valence or semantic orientation of a word (Hatzivassiloglou and McKeown 1997) can help describe its associated noun. That is, we can infer the evaluative nature of a sentence describing a noun by examining the association between the
Co-occurrence of brands and descriptors
In our applications, word probability p(x) will be calculated as the number of documents in an online database that have the word x mentioned in them, divided by the total number of documents relevant to that problem. Let's say we are interested in examining the associations between automobile brands (Toyota, Honda, BMW, etc.) and certain descriptor variables (reliable, efficient, expensive, etc.). In this scenario, one would calculate the PMI of “Toyota” and “reliable” as follows:
Study 1: Brand positioning analysis
Despite the central role of positioning in marketing, there is surprisingly limited empirical work on consumer-derived typologies that reflect brand positioning strategies (Blankson and Kalafatis 2004). Branding is particularly important in the retail industry given the industry's highly competitive nature (Ailawadi and Keller 2004) and the fact that many retail chains are introducing private brands to compete with national brands (Bellizzi et al. 1981). Brand positioning relates to how brands
Study 2: Online brand personality assessment
It is now common to refer to brands as having their own personalities (Aaker, 1997, Fournier, 1998). Aaker (1997) defined brand personality as “the set of human characteristics associated with a brand” (p. 347). Through her empirical work, she identified five distinct dimensions of brand personality: sincerity, excitement, competence, sophistication, and ruggedness. She further broke down these five dimensions into 15 “facets” that were themselves composed of 42 “traits.” In Study 2, we attempt
Study 3: Attribute-level brand comparisons
In study 3, we compare two brands on a given attribute. Specifically, we compare brand pairs on “reliability” for several home appliances. Brand pairs were chosen based on the results reported in Consumer Reports’ 2004 Buying Guide. Consumer Reports’ results were based on a national survey of tens of thousands of responses it received for its annual questionnaire about products bought in the previous five years or so (Consumer Reports 2004). Thus, the Consumer Reports listing of brand
Discussion
This paper demonstrates that massive search engine databases like Google can be successfully mined to provide information of interest to online retailers. The three studies reported here are based on lexical semantic analysis which yields valuable insights into online brand representations that are of considerable value for making strategic marketing decisions.
Our approach offers a promising potential for analytical examination of a freely available database. Because the web provides an
Conclusions and limitations
The untamed jungle of the textual content on the web, coupled with the ease and frequency with which people can add content and the automated web crawlers that are commonplace today, provides a rich and detailed source of content for consumer researchers looking to better understand consumers’ relationships with products and brands. We present a significant first step at making sense of this content for marketing analysis, while encouraging and inviting other scholars to generate a stream of
Acknowledgements
The authors would like to thank Aleksey Cherfas for his significant assistance with the programming for this project. We also gratefully acknowledge the financial support of the UMD Chancellor's Grant program and additional support from CRITO/Project POINT.
References (44)
- et al.
Understanding Retail Branding: Conceptual Insights and Research Priorities
Journal of Retailing
(2004) - et al.
Added Hometown Ideology and Retailer Legitimation: The Institutional Semiotics of Wal-Mart Flyers
Journal of Retailing
(2001) - et al.
Positioning Options for Achieving Brand Association: A Psychological Categorization Framework
Journal of Business Research
(April 2002) - et al.
The Influence of Online Product Recommendation on Consumers’ Online Choices
Journal of Retailing
(2004) - et al.
Added Knowledge Structures and Retail Sales Performance: An Empirical Examination
Journal of Retailing
(2000) - et al.
Exploring the Effect of Retail Sector and Firm Characteristics on Retail Price Promotion Strategy
Journal of Retailing
(2003) - et al.
Global Media, Local Metaphor: Television Shopping and Marketing-As-Relationship in America, Japan, and Taiwan
Journal of Retailing
(2008) - et al.
Effects of Online Communication Practices on Consumer Perceptions of Performance Uncertainty for Search and Experience Goods
Journal of Retailing
(2007) - et al.
eTailQ: Dimensionalizing, Measuring and Predicting E-Tail Quality
Journal of Retailing
(2003) - et al.
Understanding Product Migration to the Electronic Marketplace: A Conceptual Framework
Journal of Retailing
(2005)
Dimensions of Brand Personality
Journal of Marketing Research
Unspeakable Sentences
The Semantic Web
Scientific American
Consumer Perceptions of National, Private, and Generic Brands
Journal of Retailing
The Development and Validation of a Scale Measuring Consumer/Customer-Derived Generic Typology of Positioning Strategies
Journal of Marketing Management
Recognizing Subjectivity: A Case Study of Manual Tagging
Natural Language Engineering
Online Reviews: So Consumers Use Them?
Advances in Consumer Research
Word Association Norms, Mutual Information and Lexicography
Buying Guide 2004
Service Quality and E-Commerce: An Exploratory Analysis
Managing Service Quality
Mining the Peanut Gallery: Opinion Extraction and Semantic Classification of Product Reviews
WordNet: An Electronic Lexical Database
Cited by (44)
Unlocking the Potential of Web Data for Retailing Research
2024, Journal of RetailingPositioning through visual merchandising: Can multi-brand outlets do it?
2023, Asia Pacific Management ReviewRole of gender in the creation and persuasiveness of online reviews
2023, Journal of Business ResearchThe Future of Digital Communication Research: Considering Dynamics and Multimodality
2022, Journal of RetailingThe role of machine learning analytics and metrics in retailing research
2021, Journal of RetailingCitation Excerpt :The abundance of user-generated content (UGC), such as customer reviews on websites and tags on social media platforms, allows retailers to conduct market structure analysis, which is an efficient alternative to the traditional methods of focus groups and surveys. In a prescient article in the Journal of Retailing, Aggarwal, Vaidyanathan, and Venkatesh (2009) develop an early approach to understanding brand positioning using Google search data. Lee and Bradlow (2011) extract specific phrases from the text of online customer reviews and conduct cluster analysis.
E-service quality and e-retailers: Attribute-based multi-dimensional scaling
2021, Computers in Human Behavior