Sales forecast and electronic word of mouth: the power of feelings

Research has shown that word of mouth (WOM) among consumers is more effective in traditional decision-making than any marketing tool, such as personal sales or advertising media. With the development of the Internet, electronic WOM has replaced traditional word of mouth, becoming an informal and effective exchange of person-to-person communication between a perceived non-commercial communicator and a contribution regarding a brand, product, organization, or service company. The paper aims to improve the predicted sales using the sentiment component from social media posts and reviews. To this end, we have introduced a suitable e-WOM state-space model, inspired by the autoregressive sentiment aware (ARSA) models. Hence, we have tested our SSM on two different products: the sales of Gentilini's Osvego biscuits and the sales of electric cars in the US market. The sentiment has been measured using two corpora of Tweets, one in Italian for Gentilini biscuits and one in English for the electric car market.