Dynamic Pricing with Real-time Raw Material Forecasting To Improve Competitiveness of Blockbuster Products

At a large commodity chemical company, the lack of visibility into market & supply-chain driven cost of raw materials fluctuations was preventing sales from closing deals and hampering their competitive ability.

Challenge

Improve Deal Win Probability & Maintain Margins of Blockbuster Products

To enable the sales teams of the fast moving, competitively challenged blockbuster(commoditised) product groups, the following information was needed:

  • 1 to 3 month market forecast of base commodities (in the value chain of products)
  • 1 and 3-month recommended price for product factoring in raw material trend adjustment for desired contract period
  • Deal Win Probability & Pocket Margin for price points

Approach

Raw Material Price Forecast with Margin Recommendations

To start, elements in the value chain were identified, correlated, and analyzed for seasonal patterns and trends. Modelling techniques such as supervised machine learning methods and time-series analysis were used to Forecast Price movement of the raw material based on historical and forecast prices of linked base commodities.

On top of this, a predictive margin model was built that gives a margin recommendation on top of the fluctuating base commodities and raw material as a guideline to sales on what to quote.

Methodologies 

  • Supervised Machine Learning
  • Time Series Analysis

Results

  • Higher conversion probability of 32% on the monthly negotiated business
  • Ability to pass on increases and decreases in raw material price to customers swiftly
  • Forecast Accuracy of 97%, 94% and 91% for 1, 2 and 3 months, respectively