Data-Driven Marketing: a sentiment alalysis study in Nonprofit Marketing

Abstract

The technological expansion in recent years has created opportunities for competitive advantage by applying new Data-driven approaches to Data-driven Marketing practices. Large and small companies from diverse sectors can take advantage of the enormous amount of data generated to develop new business models, open new sources of revenue and initiate disruptive innovations. This dissertation seeks to analyse the use of different sentiment analysis models for data disclosed on social networks, specifically on Twitter, in a set of disclosures classified as Nonprofit Marketing. This study seeks to innovate in the contribution to knowledge by studying for the first time the relationship between message predictors and user Engagement on Twitter by Nonprofit Organisations. In this study, the quantitative method is used with the application of Machine Learning techniques for classifying tweets based on Positive, Negative and Neutral sentiments. The collected data were subordinated to statistical studies, namely Spearman's correlation. This research in the Nonprofit Marketing sector demonstrates that it is possible to predict which sentiment expressed in the message will have the best Engagement, thus generating innovative communications for followers that will lead to increased interaction. It should be noted that Positive messages tend to affect user Engagement negatively.