<p>This paper addresses the challenge of acquiring new customer data for targeted marketing initiatives. When such data must be obtained from external sources, it becomes crucial to identify the specific customer characteristics that are most desirable, in order to maximize the value and utility of the acquired data. To this end, we propose a novel approach grounded in the kernel density estimation (KDE) technique to construct representative profiles of preferred customers. Unlike traditional data acquisition methods, the proposed KDE-based approach effectively captures complex, nonlinear, and multimodal relationships within the data, where multiple peaks in the estimated density function correspond to distinct subpopulations exhibiting diverse purchasing behaviors. We present both analytical insights and a computationally efficient procedure for generating these customer profiles. These profiles are then leveraged to define precise acquisition criteria for sourcing prospective customer data from external providers. The effectiveness of the proposed method is demonstrated through an experimental evaluation conducted on a real-world dataset.</p>