Automating Market Analytics: A Data-Driven Framework for Competitive Analysis and Pattern Identification in the Automotive Sector

In the highly competitive automotive industry, understanding how vehicle pricing and incentives influence market performance is a key driver of strategic decisions. Nonetheless, translating this understanding into quantitative evidence remains a critical challenge for manufacturers seeking to optimize profitability and maintain competitiveness. This thesis, developed in collaboration with JATO Dynamics Ltd, focuses on automating analytical processes that support such evaluations and designing data-driven tools that can provide comprehensive and interpretable insights into vehicle market placement. The first part of this work covers the automation of the competitive analysis process, a core business operation used to compare vehicles within the same basket. The method reconstructs comparability by normalizing equipment configurations and computing specific indices that quantify differences among models in terms of price and value. The second stage explores the relationship between price competitivity and market share, aiming to assess how a model's price deviation from its benchmark group impacts its sales performance. To better capture temporal effects, lagged variables were introduced to account for the average delay between order and registration across vehicle models. However, given the heterogeneity and context-dependence of the econometric results, the study moved toward a pattern recognition approach, aimed at identifying analogous market situations based on a target scenario. This method allows the retrieval of the most similar cases from historical data, providing meaningful evidence even when predictive accuracy remains limited. Finally, the entire analytical pipeline was integrated into a LangChain-based AI agent, capable of leveraging multiple tools to automatically generate reports, visualizations and quantitative interpretations. The resulting system improves the accessibility of automotive market analyses, offering a flexible framework that can support both manufacturers and analysts in data-driven decision-making.