Background
A leading US retailer wanted to drive incremental sales of denims in its stores. The retailer decided to run an experiment by placing different style and sizes together for better selection and hypothesized that it will improve sales and customer experience.
Challenge
Before implementing Trial Run, the retailer used to compute lift by comparing impact of the experiment on test stores with all the remaining stores which maintained status quo. This was not an optimal method and led to an inappropriate assessment. The retailer was also not able to bifurcate the impact by styles, which ruled out the possibility of analyzing cannibalization and halo phenomena.
Infrastructure and services used
AWS EMR, S3, Aurora, ElastiCache, Elastic Beanstalk, VPC
Download the case study to learn how Trial Run helped in designing and analyzing the experiment, and eventually helped the retailer avert potential loss.