Profit-Based Counterfactual Explanations for Product Improvement: A Case Study of Manga Sales in Japan
06:00 · July 3, 2026 · arXiv cs.AI RSS
Counterfactual explanation (CE) is widely used to enhance the interpretability of machine learning models and support data-driven decision-making based on model predictions. However, existing CE methods typically require two exogenously specified inputs: a desired output value (target) and a distance function that quantifies changes in explanatory variables. In regression settings, neither the validity of target specification nor the practical interpretation of the distance metric has been sufficiently addressed. Furthermore, most existing CE methods focus on altering predictions rather than optimizing a decision objective, even though real-world decision-making often requires explicit objective maximization. To address these limitations, we formulate CE as a profit maximization problem in management and marketing contexts and propose a framework termed profit-based counterfactual explanation (PBCE). PBCE eliminates the need for exogenous target specification by directly maximizing profit as the primary optimization objective. Concurrently, the distance term is reinterpreted as the cost of modifying product attributes, providing a clear and economically grounded interpretation.
Summary
Counterfactual explanation methods typically require users to supply an exogenous target value for the model output together with a distance function that penalizes changes to input features. In regression settings these requirements leave open questions about the validity of the chosen target and the economic meaning of the distance term. PBCE addresses both issues by recasting the explanation task as a profit-maximization problem in which the model’s predicted demand, price, production cost, and attribute-modification cost are optimized jointly.
The resulting objective directly maximizes revenue net of production and adjustment costs, eliminating the need to pre-specify a target output. At the same time the conventional distance term is reinterpreted as the explicit cost of altering product attributes, giving the metric a concrete managerial interpretation. The framework assumes a monopolistic market and a demand function that decreases monotonically in price, conditions that align with observed pricing behavior in the Japanese manga sector.
The authors validate the approach analytically, through simulation, and with an empirical study of manga sales data. In the case study, PBCE yields attribute and price adjustments that increase predicted profit while respecting realistic modification costs, demonstrating how counterfactual reasoning can be aligned with an explicit business objective rather than an arbitrary prediction target.
Why it matters
This research is highly relevant for Dutch AI researchers and practitioners focusing on Explainable AI (XAI) and business optimization. By aligning counterfactual explanations with economic objectives, it offers a practical, transparent tool for Dutch enterprises and SMEs to improve products while adhering to EU principles of interpretable AI.
