Using social media data and decision-making tools for marketanalysis in furniture industry

In the contemporary market, digital platform evaluations and online reviews serve as primary reference sources for consumer purchasing decisions. This study integrates text mining and MCDM methods to identify selection criteria for children's furniture and analyze causes of customer dissatisfaction. The research was conducted in three distinct stages. First, 2.000 Amazon reviews and expert opinions were analyzed using the AHP method, identifying safety, material quality, and ergonomics as the most critical purchasing factors. In the second stage, 1.811 complaints from the "Şikayetvar" platform were coded simultaneously by human researchers and the ChatGPT-4o model using MAXQDA. The high correlation between these results demonstrated the efficiency of AI in large-scale data analysis. Key issues were identified as service quality, product problems, and manufacturing defects, with the criteria structure optimized through multidimensional scaling and hierarchical cluster analysis. Finally, 3.677 user reviews of core children's room furniture modules collected from Amazon were examined. Sentiment analysis was performed on this data using three different methods, and the reasons for consumer satisfaction or dissatisfaction were identified on a product category basis. Subsequently, the TOPSIS method was applied to the criteria derived from the dataset to rank the factors influencing purchasing decisions. Python software was utilized as a tool throughout various stages of the study. Consequently, digital consumer feedback provides companies with the opportunity for data-driven planning for future product development and marketing strategies, rather than just retrospective evaluations. This integrated approach serves as a strategic model for optimizing quality management in the children's furniture industry.