Smart Promotion
Context & Background
The client is a multinational consumer goods company and wanted to optimize their trade promotion as part of the Local Store Marketing for various retailers across different countries.
Project Objective
- Given a country; For each retailer and each product, predict number of weeks to run the promotion and at what optimal price to run.
- Impact of different Key-Performance-Indicators(KPI) as constraints and objectives
Approach
- All required sales data, financial data, range of price & promotion weeks and the KPI constraints are collected from the client.
- Leveraged the best optimization tool in python to model the problem statement.
- Given the non-linear optimization problem and complex formulation in the model, adapted the technique of warm-start to the solver.
- Collect the required values of decision variables from the model.
Business Benefits
- The Key-Performance-Indicators have seen a desirable improvement within minutes.
- The optimizer model is scalable that more retailers and products can be added.