📝 SummarXiv

Abstract

The finite lifespan of advertising effectiveness, commonly known as "creative fatigue", presents a significant challenge in computational marketing. Identifying the onset of performance degradation is critical for optimising media spend and maximising campaign returns, yet traditional methods often lack the sensitivity and scalability required in modern advertising ecosystems. This paper introduces a novel methodological framework for detecting creative fatigue by applying path signature analysis, a sophisticated technique from stochastic analysis, to performance time-series data. We treat an ad's performance trajectory as a path in two-dimensional space and use its signature as a rich feature descriptor. By calculating the distance between the signatures of consecutive time windows, our method identifies statistically significant change points in performance dynamics. We further translate these statistical events into a direct financial metric by quantifying the opportunity cost of continued investment in a fatigued creative. Through synthetic experiments and case studies, we demonstrate that this signature-based approach provides a mathematically principled and complementary framework for analysing creative fatigue patterns. To the best of our knowledge, this represents the first application of path signature methods to the advertising fatigue detection problem, opening new avenues for geometric approaches in marketing analytics.