Promotion Uplift Prediction for the UK based Global Leader in FMCG

UK based Global Leader in FMCG
We used Analytics to predict the effect of price promotions on sales for a UK based Global Leader in FMCG.

Business Objective

The esteemed client is a major retailer of Fast Moving Consumer Goods. Having global ambitions, the company aimed to expand its sales through planned trade promotions. However, those promotions often failed because their outcome wasn’t predicted accurately. In order to pursue their expected growth, the client required a more accurate uplift prediction model that used in-house data and covered all product lines.


Challenge

  1. The client depended on external vendor data which resulted in inaccurate uplift metrics.
  2. Incumbent uplift modeling served only 40% - 60% of the top products.
  3. The client’s available historical sales data was incomplete and insufficient for advanced analytics.
  4. Trade promotions were strongly budget-constrained.

Solution Methodology

  1. We developed an uplift prediction model using statistical modeling.
  2. We drew on two years of historical forecast data, actual sales. and historic promotions (On Shelf, Display, etc.). 
  3. This resulted in an accurate promotion uplift prediction - classified by product, volume, location, and season.Transfer learning was used to cover the remainder of the products while maintaining high prediction accuracy.

Impact

  1. The uplift prediction coverage increased from 60% to 100% of products.
  2. The prediction accuracy increased from 50% to an average of 80%..

Impact in Numbers

No items found.
Download Full Case Study

Want to know more ?

Drop us a message and one of our team members will get in touch with you.
Get in Touch
x
Please fill the form to download the case study
Thank you for the submission, download the file below.
Download the File
Oops! Something went wrong while submitting the form.