Current Volume 8
Cross-device advertising has become an integral component of digital marketing strategies, driven by the proliferation of devices and the non-linear paths consumers take before conversion. However, accurately attributing marketing performance across devices remains a significant challenge for marketers due to fragmented user identities, inconsistent engagement signals, and inadequate modeling frameworks. This paper proposes a unified cross-device attribution model that leverages deterministic and probabilistic identity resolution, combined with machine learning-based multi-touch attribution (MTA) algorithms. Using data from a global e-commerce platform, we examine the comparative effectiveness of rule-based, data-driven, and hybrid models in capturing true conversion paths. The study finds that hybrid models outperform conventional approaches in accuracy, flexibility, and actionable insights. Our findings have implications for marketers seeking to optimize budget allocation, personalize experiences, and achieve integrated campaign performance measurement.
Cross-device attribution, identity resolution, multi-touch modeling, conversion tracking, machine learning, digital marketing
IRE Journals:
Omolola Temitope Kufile , Bisayo Oluwatosin Otokiti , Abiodun Yusuf Onifade , Bisi Ogunwale , Chinelo Harriet Okolo
"Constructing Cross-Device Ad Attribution Models for Integrated Performance Measurement" Iconic Research And Engineering Journals Volume 4 Issue 12 2021 Page 460-476
IEEE:
Omolola Temitope Kufile , Bisayo Oluwatosin Otokiti , Abiodun Yusuf Onifade , Bisi Ogunwale , Chinelo Harriet Okolo
"Constructing Cross-Device Ad Attribution Models for Integrated Performance Measurement" Iconic Research And Engineering Journals, 4(12)