Investigating the Trade-Off between Data Granularity and Consumer Trust in Federated Marketing Analytics Using Differential Privacy Techniques

The growing adoption of federated learning in marketing analytics reflects an industry-wide shift towards privacy-preserving data strategies, yet it also presents a critical trade-off between data granularity and consumer trust. This paper explores how differential privacy techniques affect the fidelity of insights drawn from consumer behavior data while mitigating privacy risks in federated environments. We examine the extent to which granular data can be retained without compromising consumer anonymity, and how varying privacy budgets impact model accuracy and stakeholder trust. Through an interdisciplinary approach combining computational experiments, privacy risk modeling, and consumer perception analysis, we evaluate how organizations can balance the utility of detailed marketing analytics with ethical data stewardship. Our findings reveal that while fine-grained data significantly improves personalization and campaign targeting, consumer trust declines when privacy guarantees are weak or opaque. We propose a calibrated differential privacy mechanism integrated with federated learning to optimize this trade-off, offering a framework for achieving both regulatory compliance and strategic marketing outcomes. The paper contributes to ongoing debates on responsible AI, data governance, and trust in digital ecosystems.