Decomposing Dynamics of Twitter Sentiments and its Relationship with Financial Performance: A Case of US Top Retailer

In the world of internet, online word of mouth is considered as wealth now a days but companies are still struggling with the outside in marketing capabilities (Berger et al., 2019). Organizations now interact through omni-channels and spend billions for this purpose to capitalize the financial performance. However, investigation of this data value in relation to financial performance still being compromised in research and practice. This study attempts to use Twitter data being converted into insightful variables in order to check its relationship with financial performance in the retail sector of USA. The first part of the research endorses the techniques to decode the features of tweets whereas the second part investigates their relationship with financial performance. Sentiment analysis, in the sphere of marketing has always been done using the classification into positive, negative and neutral sentiments. This thesis regards an approach of using emotions as part of tweet feature has also been used to measure the relationship with financial performance for the period of three years. The result of the investigation represents a positive relationship with some financial indicators and discuss relatability of the results with previous studies and in addition provides basis for future research directions. This thesis contributes to novelty of new methodology combined from the previous research which is an addition to theory and practice.