Optimizing Competitive Positioning through Software Product Lines: An Industrial Application in Digital Marketing
José Manuel Sánchez Ruiz, Noelia Lopez-Duran, Miguel Ángel Olivero, Francisco José Domínguez Mayo, David Benavides, Carmen Macedo Figueroa · 2025
Entrepreneurial companies face the challenge of positioning their product in a market full of possibilities, where innovating and standing out from the competition is increasingly difficult. Understanding the products’ characteristics is key to highlight them against their competitors. However, comparing products is not simple, since gaining in-depth knowledge of all available market alternatives is highly complex. Product Lines, and their representation through Feature models are a relevant tool to identify, classify and compare different product regarding their features. SWOT matrix is a strategical artifact to better understand the differences among them. This study presents a software solution that enables the generation of a SWOT matrix through the configuration of a feature model and the use of LLMs to gather market information. This is the first study to combine software product lines and large language models to generate a product SWOT matrix based on in-depth insights about a product. It enables (i) structured product modeling via feature models using the Universal Variability Language and (ii) automatic generation of SWOT (Strengths, Weaknesses, Opportunities, Threats) analyses. Domain models are defined with expert input and transformed into interactive configuration forms, facilitating the definition of userdriven products without requiring SPL expertise. This approach allows to get a better understanding of market products and its differential value in relation to its competitors. Our results suggest that combining variability modeling and LLM-based reasoning provides a scalable and structured approach to automate market differentiation.