Leveraging Data Science to Estimate True Lift, and Optimizing Pricing and Trade Promotions

[vc_row type=”image” padding_top=”0″ padding_bottom=”0″ el_class=”innerBanner Csbanner” bg_image=”3172″][vc_column][vc_column_text]

Case Study

Leveraging Data Science to Estimate True Lift, and Optimize Pricing and Trade Promotions

[/vc_column_text][/vc_column][/vc_row][vc_row type=”image” bg_repeat=”repeat” bg_cover=”false” el_class=”customer-sec common-sec ds-container common-h2 bChallenges” bg_image=”350″][vc_column][vc_row_inner el_class=”customer-content risk-content”][vc_column_inner][vc_column_text]

Business Objective


Our client is a US OpCo of a global CPG manufacturer in the F&B category. In a joint workshop we conducted a couple of years back with leaders from across functions, one of the top most priorities came out to be: we spend more than a billion dollars on Pricing and Trade Promotions activities, but feel we still do not get the best impact for us and our consumers.

Major portion of their trade spends was around EDLP (Every Day Low Price) and TPR (Temporary Price Reduction) events. While the promotions delivered a sales lift on the promoted items, overall lift across products x retailers was not adding up, leading to a suspicion of cannibalization between products and retailers. Between the two, while cross-product cannibalization was being measured to an extent, cross-retailer cannibalization had proven to be a hard nut to crack.

Coming out of the workshop, we put together a multi-functional taskforce with participation from business and data science teams of the client and ourselves.

[/vc_column_text][/vc_column_inner][/vc_row_inner][/vc_column][/vc_row][vc_row type=”image” bg_repeat=”repeat” bg_cover=”false” el_class=”data-approch-sec common-sec common-container common-h2 challengeContent” bg_image=”1909″][vc_column][vc_row_inner el_class=”approach-content”][vc_column_inner][vc_column_text el_class=”busi-balck-content”]


  • Account teams had different ways of measuring promotion effectiveness. i.e., baseline vs lift
  • There was a need to validate results with a 3rd party tool which was already being used by client which provided a partial view of cross elasticities within-box alone
  • Estimating impact and attributing them to the promotion was challenging in an aggregate data that was available at Week x Product x Retailer x US level through syndicated data sources

[/vc_column_text][/vc_column_inner][/vc_row_inner][/vc_column][/vc_row][vc_row type=”image” bg_repeat=”repeat” bg_cover=”false” el_class=”data-approch-sec common-sec common-container common-h2 solutionMethodology” bg_image=”350″][vc_column][vc_row_inner el_class=”approach-content”][vc_column_inner][vc_column_text]

Solution Methodology 

  • Prepared modeling data aligning each Retailer vs Rest of Market to be able to measure cross box cannibalization
  • Generated features like EDLP, TPR, and Base Price etc.
  • Conducted exhaustive data exploration to understand the product groups cannibalizing each other during promotions, as well as halo effects
  • Leveraged multivariate regression models with exquisite parametric controls to understand the impact of own price effects, distribution, other within-box and cross-box effects on sales volume for every Promoted Product Group (PPG)
  • Processed model outputs to estimate the true sales lift from the gross lift after accounting for within-box and cross-box cannibalization
  • Designed an end-to-end optimization work-flow which suggests an optimized trade spend that maximizes true sales lift at a Portfolio level
  • Furthermore, developed an interactive Scenario Planner for business users to simulate various price promotional activities and estimate the impact on metrics like Gross lift, True lift, ROI etc.
  • Optimal Price and Promotions calendar module is also being developed as a final leg of this program.

[/vc_column_text][/vc_column_inner][/vc_row_inner][/vc_column][/vc_row][vc_row type=”image” bg_color=”#161717″ bg_repeat=”repeat” el_class=”data-approch-sec common-sec common-container common-h2 bImpact”][vc_column][vc_row_inner el_class=”approach-content”][vc_column_inner][vc_column_text]

Business Impact

  • Initial estimates in 2018 indicated potential ROI Improvement of 10% on a USD150MM Spend.
  • Solution is now being rolled out across multiple retailers and product categories for full-scale impact.

[/vc_column_text][kleo_button title=”Download Case Study” target=”_blank” style=”custom” icon=”0″ el_class=”dcsBtn popmake-11944″][/vc_column_inner][/vc_row_inner][/vc_column][/vc_row]

Copyright © 2023 Tiger Analytics | All Rights Reserved