Case Study
We’re not a regular consulting firm. More than Data & AI suppliers, we are reliable partners. Both transparency and independence are at the core of how we work and guide the development of strong and reliable partnerships with our clients.
Our predictive pricing model, which utilizes Machine Learning and AI, has empowered our clients to establish an intelligent pricing governance. By merging historical customer negotiations, index prices, and proposal specifics, our model attains a 75% accuracy rate, serving as a valuable tool to optimize both revenue and margins for the organization.
A significant challenge with this model was our client's preference for linear relationships between features and the target. This constraint limited our use of more complex models. However, by employing certain feature engineering techniques, we enhanced the Logistic Regression model to a level where it became viable for the pricing governance.
Creating a Linear Algorithm that was able to predict the probability of winning a contract based on index pricing and past negotiations with a specific customer.
We've modeled the relationship by finding several lagged features that were able to predict how a customer would react to pricing changes. With this, we could model the probability of closing a deal along the different price curve, recommending a price that could strike a balance between optimal margin without losing the customer.
Although knowing what would be the best probability of winning the deal was important, it was also important to consider what was the margin generated by that specific pricing. Otherwise, our tool would just recommend the lowest price possible!
.png)
01
Collecting Data
02
Baseline Development
03
Feature Engineering
04
App Development
