Value Recovery with Predictive Pricing

A multinational ingredient manufacturer was struggling with uncontrolled price variance due to lack of strategic price guidance, legacy deals & overlooked small customers resulting in low-price outliers and value leakage.This is a common feature at Business-to-Business companies. Smart pricing & data science methods can be applied to give structured price advice and recover value.

Challenge

Creating a data-driven and customer-centric way to price deals

After numerous efforts to clean the pricing estate and struggling from price erosion, business management was looking for a smart and effective method to recover value with pricing. The biggest challenge for business management was to come up with a smart, granular and customer-centric, approach to  price improvements. Earlier methods focused on bulk/blanket prices increase at regional-product levels which lacked the persuasion to convince sales and were rejected by customers.

Approach

AI-driven advanced techniques were used to provide price guidance at customer-product level 

Historical price-volume data was used for identifying trends in pricing as well as price outliers. Next step was to build a predictive price model wherein,  business management was involved in the design phase to ensure sound product – market dynamics were captured in the logic of the model. PCA, Clustering and advanced regression methods were used to identify price drivers, customer aggressiveness levels, product price sensitivities, regional sensitivities-and build a predictive model that gives a price recommendation based on willingness-to-pay for every customer-product combination.  Results of the model were validated by both sales and business management.

To guarantee price corrections  – A Value Recover Campaign – lasting 3 weeks and facilitated with insights, tooling, trainings & leadership guidance was scheduled. During this period sales went through each and every low-paying customer in their portfolio and committed to a to price revisioning (at account level!) that was then executed during annual negotiation period.

Methodologies 

  • Clustering algorithms
  • Supervised Machine Learning

Results

  • EUR 12 million of recovery potential identified
  • EUR 1.1 million of price increases committed within 3 weeks
  • New data-driven way of pricing that unites the business management and sales teams
To know more
Contact Us