IMPACT OF MATCH RATES ON COST BASIS METRICS IN PRIVACY- PRESERVING DIGITAL ADVERTISING
DOI:
https://doi.org/10.15662/IJARCST.2020.0304003Keywords:
Privacy enhancing technologies, Multi-party computation, Causal inference, Digital Ad platforms, Auction based RTB’s, Data encryption and retrieval, Randomized control trialsAbstract
This study explores the role of match rates in determining cost basis metrics for randomized controlled trials (RCTs) conducted through Privacy Enhancing Technologies (PETs), such as multi-party computation (MPC), Trusted Execution Environments (TEEs), and cleanroom- based protocols. These methods are used to measure the causal effectiveness of digital advertising campaigns while maintaining data privacy. We specifically analyze how variations in match rates influence incremental Return on Ad Spend (iROAS) and other cost basis metrics, emphasizing the challenges in deriving accurate estimates within privacy- preserving frameworks. Alternative estimators of match rates and their computability under PETs are discussed. Practical issues, including the inability to account for unmatched conversions, and potential proxies for mitigating these limitations, are examined [1][2].
References
1. European Union. (2016). General Data Protection Regulation (GDPR).
2. California Legislature. (2018). California Consumer Privacy Act (CCPA).
3. Abadi, M., & Andersen, D. G. (2017). "Privacy-preserving machine learning." Communications of the ACM.
4. Li, X., et al. (2020). "Trusted execution environments for secure computation." Journal of Privacy Technology.
5. U.S. Census Bureau. (2020). Population Estimates for U.S. and Canada.
6. Tadelis, S. (2016). "The economics of digital advertising."
7. Bonawitz, K., et al. (2019). "Practical secure aggregation for privacy-preserving machine learning." ACM SIGSAC.
8. Papadimitriou, C., et al. (2020). "Data privacy in large-scale platforms." Journal of Computing.
9. Hardt, M., et al. (2019). "Addressing biases in differential matchability." Advances in Privacy Technology.
10. Zhang, X., & Zhang, L. (2021). "Efficiency in privacy-preserving RCTs." Journal of Advertising Research.
11. Chaudhuri, K., et al. (2022). "Future directions in PET-based advertising." IEEE Transactions on Big Data.
