Estimating the Economic Value of Incremental Data Assets

This article is a research paper that presents a methodology for estimating the economic value of incremental data in predictive modeling applications. It addresses the challenge of determining a fair price for purchased data, particularly in marketing scenarios where firms rely on third-party vendors. The methodology begins by evaluating how additional data influences model performance using classification metrics, which are then translated into economic terms. This allows firms to set data acquisition prices that align with expected returns on investment. The paper includes two case studies that illustrate the application of the methodology: one involving an Internet display advertising firm and another related to a charity's direct mail campaign. The findings indicate that the value of external data can vary significantly based on the availability of proprietary data, affecting the pricing strategy for data assets. The research contributes to the understanding of data valuation from a buyer's perspective, linking predictive modeling metrics with economic outcomes.