Add Thesis

A Quantitative Study of the Impact of Social Media Reviews on Brand Perception

Written by N. Joshi

Paper category

Master Thesis


Business Administration>General




Master Thesis: Jeong et al. (Susarla and Tan, 2008) in their paper "Inspecting the diffusion of user-generated content in online social networks" focused on the dissemination of user-generated content in order to increase the impact of social media on non-specific types of videos Diffusion of innovation is a theory that attempts to explain how, why, and at what speed new ideas and technologies are spread through culture (Wikipedia). This research studies the impact of social media networks among like-minded individuals on social learning and conformity. In contrast, our paper examines the impact of content creators who are responsible for generating content that helps expand their social networks. We specialize in the specific attributes adopted by content creators on YouTube that make it possible for like-minded individuals to succeed in social media networks. If considered in this framework, our research focuses on the reasons for the diffusion of innovation. The YouTube statistical study by Cheng et al. studied the correlation between the videos posted by content creators and the recommended videos generated by the YouTube algorithm (Dale & Liu, 2008). Although the research focuses on the importance of content creators’ choice, it emphasizes the snowball effect of YouTube rather than the influence of these content creators on viewer choices. In addition, this research also focuses on measuring the total number of views generated by videos as the most important popularity indicator. In contrast, as we will discuss later in this article, we prove that the number of views alone does not measure the popularity of the content creator, because the total number of views does not cover the affected audience. Our research uses the number of likes and other parameters as key indicators to study the performance of content creators. Comments and views are used to predict these parameters by using linear regression and R model analysis, and the impact of video content on general categories of views, comments, and video sharing is studied. A paper by Siersdorfer et al. focuses on comments on published videos and the meta-ratings of those comments. The study also used the sentiment analysis tool SentiWordNet Thesaurus (Siersdorfer et al., 2010) to analyze the sentiment in these comments. In addition, the study predicts the community's acceptance of comments that have not yet been rated in the future. Richier et al. proposed a model to study the influence of the popularity of videos and their categories based on the evolution of video views (Richier et al., 2014). Their research focuses on the number of views generated by YouTube users and how the predictive model works hard to determine the number of views in the future. Predictive models and the importance of content creators Research on the impact of content creators on audience participation is very limited. In their work, broadcasting themselves: Knowing YouTube uploaders, Ding et al. claimed that they had never done research on the importance of content creators before (Ding et al., 2011). Although they do focus on content creators, their work has not received substantial and relevant support for quantitative analysis. Although their research provides a different way to understand YouTube comments, they only proposed one hypothesis. Previous researchers focused on the length of the video, the age of the video (similar to the content published so far in this paper), and the specific content (topic) discussed in the YouTube video without special attention to genre or industry, and the current literature lacks Research on the attributes used by content creators and how brand managers use the technical knowledge used by these content creators to capture audience engagement. Previous work on is in Chapter 3. We validated our approach by analyzing restaurant reviews on Previous research focused on sentiment analysis of YouTube comments, focusing on positive and negative words (Hicks et al., 2012). However, previous work did not capture the performance of specific outlets close to specific target audiences. Our paper is unique in two respects: First, we focus on specific locations within a 10-mile radius of a densely populated university campus. Secondly, we specifically evaluated the performance of Tex-Mex restaurants in each U.S. state with unique cultural diversity. It reveals unique insights for outlet employees (the carrier of restaurant experience in terms of food and service) and reviewers (restaurant diners), which can help brand managers effectively invest their resources to obtain a better return on investment. Step I: Data Collection Crawl YouTube Comment Attributes In order to evaluate comments on YouTube, we divide the data into two parts: independent variables and dependent variables. We collected independent variables and dependent variables, as shown in Table 4 and Table 5. Kimono labs ( software is a free web crawler used in this paper to process comments. A web crawler "is a program that visits a website and reads its pages and other information in order to create entries for search engine indexing" (Rouse, 2005). We use Kimono Labs to crawl's reviews of Tex-Mex restaurants near selected locations on university campuses in selected towns in the United States. Kimono Lab uses their technology to help users create their own application programming interface (API). These APIs are a set of requirements that control how one application communicates with another application (Prooffitt, 2013). Read Less