Page 95 - 36-1

NTU Management Review Vol. 36 No. 1 Apr. 2026

knowledge (Stremersch et al., 2025).
Fourth, the interpretability functions embedded in the AI models (e.g., SHAP value
analysis) enhance managerial trust in machine learning outcomes. Managers can visualize
how specific value drivers and operational activities affect predicted PLCV, thus aligning

analytical insights with practical experience.
Overall, the empirical evidence supports the feasibility and managerial usefulness
of establishing a data-driven PLCV-AVM system. By integrating predictive analytics,

activity-based management logic, and interpretability, the proposed framework advances
financial performance evaluation, deepens activity-level value insights, and enhances
strategic planning capabilities in dynamic product and channel environments.

4. Research Limitations / Implications

Despite promising results, this study faces several limitations.
The empirical analysis relies on a single case company, which limits generalizability

of the findings. Future research could extend the sample across multiple industries to test
the model’s scalability and robustness. In addition, although AI models such as XGBoost
and Neural Networks deliver superior predictive performance, they require extensive data
preprocessing and expert interpretation, which may constrain adoption among firms with
limited digital maturity.

Another limitation concerns the temporal scope of PLCV estimation. Since the data
span only six years, the full effects of product evolution and market dynamics could not
be fully captured. Longitudinal studies incorporating post-2021 data may provide a more
comprehensive understanding.

The practical implications of this research are nonetheless substantial. The AI-driven
PLCV-AVM framework offers managers an integrated performance dashboard that links
cost efficiency with value creation, supporting more informed decisions on resource
allocation, product development, and sustainability investment. The study also implies

that management accountants must cultivate data analytics competencies and collaborate
closely with data scientists to leverage AI responsibly within value management systems.

87