ML Works fixed trade promotions for a prominent CPG client

Sheryl was responsible for the revenue growth management (RGM) of a Fortune 500 CPG company that operates in 13 countries. Despite the deployment of machine learning models, trade promotions remained somewhat of a ‘black-box’ because of huge permutations of 1000 SKUs in 20 retail locations. Further, 156,000 models crossed with each other.

Sheryl struggled to optimize the impact of promotional activities on sales & profitability and provide accurate insights to the firm’s 56,000 trade promotion managers.

ML Works changed the trade promotions status quo

ML Works’ drift detection module helped Sheryl capture the efficacy & data drifts in models and restructure the trade promotion expenditures to achieve market diversification.

Here’s how:

  • Monitoring the models in production.
  • Capturing models’ accuracy & data drifts.
  • Flagging and alerting users on breakage(s) in the model pipeline.

Measurable benefits

  • $300M, otherwise at risks due to spending on costly trade promotions, are now better utilized.
  • Now, promotional campaigns are more transparent.
  • 56K trade promotion managers are happy with Sheryl. And Sheryl is happy with Tredence.