Applied Machine Learning: Promotions Forecasting
The focus of this talk will be an overview and a walkthrough of the iterative process of model building, testing, and deployment at a real-world retailer to optimize their promotions and coupons. Our speaker is Shareth Hariharan, Managing Director at Impact Analytics, where he leads a global team of data scientists solving some of the most pressing problems in omni-channel retail and supply chain.
This real-world case study describes the practical aspects of a data science implementation including the considerations and solutions to the challenges of:
- business requirement gathering
- data cleansing
- feature engineering
- back-testing and simulation
Targeting the right customers to promote and predicting the optimum price discount to provide is a challenging task for many retailers. Success or failure of the promotions depends on how accurate the forecast is, measuring the baseline and uplift, as well as secondary effects such as affinity/halo and footfall (or site visits).
In addition, complications arise due to seasonality, new product introductions, cannibalization, and the competitive landscape. We will explore how to tackle such issues through machine learning models and understand how this integrates into the end-product that is in use by this retailer.
About our Speaker
Shareth Hariharan is Managing Director at Impact Analytics, where he leads a global team of data scientists solving some of the most pressing problems in omni-channel retail and supply chain. Shareth has Master's and Doctorate degrees in Industrial Engineering and Management from Oklahoma State University where his focus was on Pricing and Revenue Management under Uncertainty. His work experience includes several successful implementations of advanced analytics solutions that combine machine learning and operations research techniques under different industry verticals such as retail, CPG and media, as well as digital and e-commerce analytics.
Shareth is a member of the Westlake Village Data Science Meetup and makes his home in Thousand Oaks.