This chapter investigates to what extent the feelings and thoughts of consumers are effective in the process of new product development and improvement. Inspired by the user innovation theory, the study analyzes the online feedback of the users of a smartphone brand in Turkey. This analysis covers a sentiment analysis performed using the support vector machines, the random forest, and the recurrent neural networks algorithms. By studying 2005 reviews, the chapter concludes that two strategies are proposed for the firms. A first strategy is a hybrid approach: Given the imported input-dependency, we know that the cost of imported inputs matters for firms. The second strategy is repositioning the brand in the long term. This chapter attempts to contribute to the literature by providing new evidence in a developing country case whether user comments are effective on the modified versions of a high-tech product.
Top1. Introduction
New Product Development (NPD) and improvement process can be considered as a precondition for survival in competitive markets (Tripathy and Katyayn, 2021) and also a key for success for the growth of firms. A firm should allocate both its time and financial resources to this process which should have the right idea at the background to begin with. Perhaps, the foremost challenge is the satisfaction of the customer needs which is a key factor for marketing. Consumer feedback appears to be very critical for NPD. The inclusion of customers in NPD and improvement will increase the efficiency of the process (von Hippel 1978, 1988). For a successful NPD, strategies about the process should keep pace with customer needs (Hsu, 2018). So, customers' feedback is like a gold mine for the company in that they can improve the product.The feedback reduces the risks and the money they will spend.
This study tries to explore whether consumer emotions and thoughts have an impact on NPD and improvement process. The studies proposing that understanding consumers matter for NPD generally analyze the issue from the perspective of firms based on case studies (Goffin and New, 2001; Fundin and Bergman, 2003; Tzoka et al., 2004; Fang et al., 2008; Joshi and Sharma, 2004; Alli, 2018). We adopt sentiment analysis to analyze the issue from the perspective of customers in the framework of social networks based on User Innovation Theory (Eric von Hippel, 1978) Today, there are 4.5 billion internet users, 5.2 billion mobile phone users, and 3.8 billion active social media users in the World (We are social, 2020) and social media provides accessible user feedback of any products easily. The relevant data were obtained from two social network platforms for this study; i.e., Twitter, and a local e-comment website. Each social media platform possesses huge data of many different products. Therefore, we confined ourselves to a specific technology-intensive product, a smartphone brand produced in Turkey, as a case for which the market is highly competitive and NPD is inevitable. The NPD is perhaps more important in high-tech products than in other types of products because the product life cycle is shorter and more complex (Oh et al., 2015). 2005 user comments out of 4954 raw data related to these versions was collected starting from 2014, in which the smartphone was launched, up to 2019. We also chronologically ordered all versions of the product to sketch the product improvement line.
We utilized the text mining method using customer feedback in the case of a high-tech product for the analysis. The most widely used method in recent years is text mining to analyze the user comments in written text. There are few studies using sentiment analysis with NPD. Rathore et al., (2018) have built a new way of conversational pattern using sentiment analysis with the case of three models of smartphones (Rathore et al., 2018). Wu et al., (2019) have developed an extended social media analytics framework. Rathore and Ilavarasan (2020) investigated the emotion change pre- and post-launch period for three products. Finally, Giannakis et al., (2020) use social media as an information tool for every stage of NPD. Unlike other studies using this method sufficing with only acquiring accuracy rate, the paper connects the results of the sentiment analysis with the NPD process. We used python as the programming software.
The plan of the paper is as follows: the next section reviews the theoretical and empirical literature. The third section explains the methodology used in the study. The fourth section presents the findings. The last section is devoted to the discussion and conclusion.