Deep Learning-Based Customer Insight Analytics for Data-Driven Product Development and Innovation

Abstract

The rapid growth of digital commerce has generated substantial volumes of customer reviews and product-related information. Extracting meaningful patterns from these data can help organizations understand customer experiences, identify product-related preferences, and support evidence-based product development. Deep-learning techniques provide opportunities to analyze customer-generated information systematically and transform complex feedback into actionable customer insights. This study aimed to apply deep-learning-based customer insight analytics to identify meaningful patterns in customer feedback and examine product- and category-level insights relevant to data-driven product development and innovation. The study used the Brazilian E-Commerce Public Dataset by Olist, incorporating customer reviews, product information, order-item records, and product-category information. Customer ratings were organized into feedback categories, and the available customer-generated information was processed to examine feedback patterns and differences across product categories. A deep-learning-oriented analytical framework was applied to derive customer insights relevant to product development and innovation. The analysis showed a predominance of positive customer feedback, while negative and neutral responses provided additional information concerning areas requiring further attention. Customer responses also varied across product categories, demonstrating differences in review volume and reported satisfaction. These patterns indicate that customer-generated information can provide meaningful evidence for identifying product-related strengths, potential improvement areas, and customer preferences. Deep-learning-based customer insight analytics provides a systematic approach for interpreting customer-generated information in digital commerce. Integrating customer feedback with product-level and category-level analysis can support evidence-based product development, customer-oriented decision-making, and innovation planning.