A smart alerting solution for one of the largest FMCGs through identifying OOS and Off-Sales anomaly


Client is one of the largest (Fortune 500) FMCG companies in the world. Client wanted to develop a system to generate smart alerts for merchandising teams at a store level based on off-sales trend

  • Identifying OOS Situation: Built ML-based models to identify the Out-Of-Stock scenario in a store at an SKU level taking into account the level of phantom inventory
  • Identifying Off-Sales Behavior: Built ML-based models to identify the off-sale behavior of an SKU in a particular store attributable to phantom inventory, stock less than presentation stock or improper operations within the store
  • Smart Alerts: Built a system that generates alerts for the store manager and merchandising teams to maintain healthy stock in the store and increase the revenue


The approach to address the client’s challenge included:

  • Analyze Store-SKU behavior and estimate phantom inventory
  • Calculate corrected inventory at the store to estimate reorder point
  • Generate OOS and zero scan alerts based on inventory levels and sales patterns at the store
  • Use advanced ML algorithms to forecast Store-SKU level sales and compare with actual sales to identify anomaly due to shelf mismanagement
  • Prioritize alerts based on business rules and $ opportunity

Key Benefits

  • ML-driven model to evaluate the impact of trade promotion spends
  • Scalable platform to understand the trade spend effectiveness across brands and regions
  • Visualization platform cum scenario planner was embedded to help category managers optimize trade spends


Icon Boost

Acting on 3% OOS results in an overall revenue boost of 4%

Icon Boost

Nudging merchandising teams to achieve higher alert reach resulted in an additional 1.5% revenue

Talk To Us