Theuring, Jonathan
(2019)
Examining Brand-Related User-Generated Content on Twitter: Machine Learning Approaches to Topic Analysis and the Investigation of Social Motivations.
[Dissertation (University of Nottingham only)]
Abstract
Creating and sharing user-generated content (UGC) online is one of the principal means by which consumers can take part in shaping a brand’s image. Brand-related UGC is frequently shared on social networking sites (SNSs) such as Twitter, generating vast volumes of data containing consumers’ concerns and opinions about brands. Easily accessible to marketing practitioners, the big data of brand-related UGC tweets can inform brand communications, targeted advertising, and a whole host of other marketing processes. The salient issue is with the interpretation of this data; due to the large quantity of brand-related UGC available on Twitter, it is impossible for researchers to analyse all of the tweets by hand. Building upon previous research, this study employs a text mining technique known as latent Dirichlet allocation (LDA) to analyse the topics discussed in UGC tweets relating to brands in a range of industries that have not previously been examined in this way (the fashion, gaming, restaurant, soft drink, and supermarket industries).
The first objective of this research is to determine the differences in the topics being discussed in UGC relating to brands in these industries via machine learning methods. Topics concerning products were the most prevalent in the UGC relating to brands in the soft drink and restaurant industries, whilst those pertaining to issues of service were the most common in the UGC relating to brands in the supermarket, fashion, and gaming industries. The second research objective explored in this study relates to consumers’ motivations to create and share brand-related UGC on Twitter. Given that a range of researchers have identified the desire for social interaction as a significant motivation behind brand-related UGC, this research examines evidence of social motivations in UGC tweets relating to brands in each of the industries. Thus, the second section of this study determines the relative significance of social motivations as the driving force behind the creation and publication of UGC in the five industries under investigation. In decreasing order of likelihood to elicit socially motivated UGC tweets the industries were: soft drink; gaming; fashion; supermarket, and restaurant.
The findings of this research have important implications for marketing theory and practice. The first part of the study demonstrates how LDA-based analyses of brand-related UGC can inform the positioning of products through brand communications, reveal brand-specific issues that need rectifying, and assist with the process of identifying and managing public relations crises. The findings of the motivation-based section provide insight into the ways in which marketing practitioners can promote positive brand-related UGC more effectively – if a brand already elicits a large proportion of socially motivated UGC, then marketers working for that brand should encourage consumers to publish brand-related UGC through competitions that necessitate social interaction (and vice versa). Finally, this study corroborates the work of previous researchers who have found that social motivations are a significant driving force behind the creation and sharing of brand-related UGC. Not only does this research identify the relative influence of the desire for social interaction as a motivating factor behind UGC relating to brands across a range of industries, but it also provides evidence of the different tiers of intrinsic and extrinsic social motivation that have been proposed by previous academics.
Actions (Archive Staff Only)
 |
Edit View |
Tools
Tools
![[thumbnail of 4338007 N14B63 Examining Brand-Related User-Generated Content on Twitter- Machine Learning Approaches to Topic Analysis and the Investigation of Social Motivations.pdf]](https://eprints.nottingham.ac.uk/style/images/fileicons/application_pdf.png "4338007 N14B63 Examining Brand-Related User-Generated Content on Twitter- Machine Learning Approaches to Topic Analysis and the Investigation of Social Motivations.pdf")