OPTIMALISASI CRM BERBASIS AI DALAM E-COMMERCE: ANALISIS PENGARUH TERHADAP LOYALITAS PELANGGAN DENGAN PENDEKATAN SENTIMEN ANALISIS

Abstrak

The purpose of this study is to analyze the relationship between the application of AI in CRM and customer loyalty. The results of sentiment analysis through Web Scraping are used to strengthen the interpretation of the results, not as a mediator or can be called triangulation or enrichment data. The population of this study were active e-commerce customers with a sample of 384 respondents obtained through purposive sampling techniques. Data analysts used the SEM-PLS and NLP approaches for sentiment classification. The study's results indicate that AI-based CRM has a positive and significant impact on customer loyalty. This is indicated by the T-statistic value of 29,415 > 1.96. The conceptual model is also very good, as evidenced by the SRMR value of 0.054> 0.050 and the NFI value of 0.907. Sentiment analysis collected through social media X from 3,213 tweets with the keyword skincare resulted in 74% neutral opinions, 13.7%, and 12.3% negative. Furthermore, the classification results of 7 aspects of positive and negative opinions such as aftersales, delivery, experience, order, packaging, product, and unknown, from positive opinions the largest percentage is in the experience (customer experience) 43.2% while for negative opinions the largest percentage is also in the experience aspect of 45.8%. This data is important for business owners to make improvements and improvements to create customer satisfaction. 45.8% of bad customer experiences will, if not followed up, make customers switch to other stores.