What is it about?
Retailers commonly share point-of-sale data with market research companies such as ACNielsen and Symphony IRI. The data include retail sales, prices, and promotions at the Stock Keeping Unit (SKU) level. These data are given by the retailers with the expectation that the identity of the store will remain anonymous. The market research company sells the data to users like manufacturers of branded products, who use the data to measure price elasticities and promotion effects. However, data intruders could potentially use the data to identify the stores via statistical analysis. Therefore, it is important to think of ways to protect the data from such intrusion. This paper proposes a statistical method to modify the original data such that the data remain useful for their intended purpose, but it becomes a lot harder for intruders to use the data in unintended ways. The paper also proposes metrics to measure whether the modified data provide good protection from unintended use. Comparisons with other ways of data protection in real data show that the proposed model does well.
Featured Image
Why is it important?
Market research data gathered from subjects, whether stores or individuals, are widely used and very important for marketing decisions. If subjects don't feel their identities are protected, they will be reluctant to provide data.
Perspectives
Read the Original
This page is a summary of: A Flexible Method for Protecting Marketing Data: An Application to Point-of-Sale Data, Marketing Science, January 2018, INFORMS,
DOI: 10.1287/mksc.2017.1064.
You can read the full text:
Contributors
The following have contributed to this page