Trade Promo Optimization: 3Wk Pilot Implementation

Objective: Develop a scalable solution across brands and regions to study the efficacy and efficiency of trade promo spends coupled with a simulated environment to determine an efficient TO-BE state.

Key Challenges Addressed:

1. The current process is highly complex and non-transparent owing to increased competition, fragmented consumer segments, different trade terms with every retailer, non-binding and varying agreement structures.
2. Solutions built are custom to a certain region / brand and are difficult to replicate.
3. Data access / security
4. Time to execution

How do we address your challenges:

1. Multiple data sources (internal and external) are ingested and analyzed to develop a strong foundation of the current state.
2. Combination of on-premise and cloud elements to ensure scalability and region specific access control.
3. Use ML as a web service for faster code deployment and future enhancements.

Pilot Outcome: Scope: 1 Country; 2-3 PPGs; 1 Customer

ML models for Trade Promo Effectiveness; Price Elasticity curves; Promo KPIs: ROI, incremental volume / revenue

Implementation Plan The break-up of the implementation plan is as below: Week 1 - Conduct data discovery and sanitize required data elements Week 2 - ML models to extract baselines and elasticities Week 3 - Model refinement and Promo KPI computation

This implementation uses the following native Azure components: ADF pipelines ADLS Gen 2 Azure SQL Database Azure ML Power BI