Investigating value propositions in social media: studies of brand and customer exchanges on Twitter

Abstract

Social media presents one of the richest forums to investigate publicly explicit brand value propositions and its corresponding customer engagement. Seldom have researchers investigated the nature of value propositions available on social media and the insights that can be unearthed from available data. This work bridges this gap by studying the value propositions available on the Twitter platform. This thesis presents six different studies conducted to examine the nature of value propositions. The first study presents a value taxonomy comprising 15 value propositions that are identified in brand tweets. This taxonomy is tested for construct validity using a Delphi panel of 10 experts – 5 from information science and 5 from marketing. The second study demonstrates the utility of the taxonomy developed by identifying the 15 value propositions from brand tweets (nb=658) of the top-10 coffee brands using content analysis. The third study investigates the feedback provided by customers (nc=12077) for values propositioned by the top-10 coffee brands (for the 658 brand tweets). Also, it investigates which value propositions embedded in brand tweets attract ‘shallow’ vs. ‘deep’ engagement from customers. The fourth study is a replication of studies 2 and 3 for a different time-period. The data considered for studies 2 and 3 was for a 3-month period in 2015. In the fourth study, Twitter data for the same brands was analysed for a different (nb=290, nc=8811) 3-month period in 2018. This study thus examines the nature of change in value propositions across brands over time. The fifth study was on generalizability and replicates the investigation of brand and customer tweets (nb=635, nc=7035) in the market domain of the top-10 car brands in 2018. Lastly, study six conducted an evaluation of a software system called Value Analysis Toolkit (VAT) that was constructed based on the research findings in studies 1 - 5. This tool is targeted at researchers and practitioners who can use the tool to obtain value proposition-based insights from social media data (brand value propositions and the corresponding feedback from customers). The developed tool is evaluated for external validity using 35 students and 5 industry participants in three dimensions (tool’s analytics features, usability and usefulness). Overall, the contributions of this thesis are: a) a taxonomy to identify value propositions in Twitter (study 1) b) an approach to extract value proposition-based insights in brand tweets and the corresponding feedback from customers in the process of value co-creation (studies 2 - 5) for the top-10 coffee and car brands, and c) an operational tool (study 6) that can be used to analyse value propositions of various brands (e.g., compare value propositions of different brands), and identify which value propositions attract positive electronic word of mouth (eWOM). These value proposition-based insights can be used by social media managers to devise social-media strategies that are likely to stimulate positive discussions about a brand in social media.