RetailSynth: Synthetic Data Generation for Retail AI Systems Evaluation
Authors:
Yu Xia,
Ali Arian,
Sriram Narayanamoorthy,
Joshua Mabry
Abstract:
Significant research effort has been devoted in recent years to develo** personalized pricing, promotions, and product recommendation algorithms that can leverage rich customer data to learn and earn. Systematic benchmarking and evaluation of these causal learning systems remains a critical challenge, due to the lack of suitable datasets and simulation environments. In this work, we propose a mu…
▽ More
Significant research effort has been devoted in recent years to develo** personalized pricing, promotions, and product recommendation algorithms that can leverage rich customer data to learn and earn. Systematic benchmarking and evaluation of these causal learning systems remains a critical challenge, due to the lack of suitable datasets and simulation environments. In this work, we propose a multi-stage model for simulating customer shop** behavior that captures important sources of heterogeneity, including price sensitivity and past experiences. We embedded this model into a working simulation environment -- RetailSynth. RetailSynth was carefully calibrated on publicly available grocery data to create realistic synthetic shop** transactions. Multiple pricing policies were implemented within the simulator and analyzed for impact on revenue, category penetration, and customer retention. Applied researchers can use RetailSynth to validate causal demand models for multi-category retail and to incorporate realistic price sensitivity into emerging benchmarking suites for personalized pricing, promotions, and product recommendations.
△ Less
Submitted 21 December, 2023;
originally announced December 2023.
