In today’s high-technology markets, innovativeness and competitiveness are paramount for firm success. This article delves into the intricate relationship between research and development (R&D) investments, knowledge generation, and patent outcomes, with a specific focus on the pharmaceutical industry. Drawing upon existing literature and empirical evidence, we propose a novel theoretical model that captures the firm-level knowledge generation process, considering the intermediary role of patent applications. Unlike previous studies that overlook the patent application process, our model integrates this crucial step. Using a truncated Poisson model, we analyse a dataset comprising 137 independent European pharmaceutical firms to empirically test our theoretical framework. Our findings reveal that accounting for truncation significantly influences the estimation of innovation input elasticities, emphasizing the importance of considering the patent application process in assessing the effectiveness of R&D investments. The empirical results highlight the complexity of the innovation process, wherein R&D investments translate into latent knowledge, which in turn manifests into patent outcomes subject to the patent application process. By considering these intricacies, our model provides a more accurate approximation of the firm-level innovation process.