Harnessing the Power of Azure Machine Learning to Improve Pricing Recommendations on New Equipment

Client

The client is world’s leading company that develops, manufactures and markets elevators, escalators, moving walkways, and related equipment.

Challenges

• The discounts provided by a sales rep is based on their experience or expertise, resulting in a loss of margin and revenue
• To develop a model to optimize the discounts provided to improve the margin and revenue; A range of discount minimum and maximum should be provided to the sales rep
• With quick and improved discounts provided, the sales rep should be able to increase the margin and revenue. Also, they would have more time for fieldwork

LTM Solution

• Designed and developed a model to recommend the minimum and maximum discount
• The best features were selected using feature importance
• The top features are rise, speed, load, installation area, sales category
• The model has been developed using Azure Auto ML
• The model has been deployed as a realtime REST API on the Azure Kubernetes cluster (AKS).

Benefits

Margin

For France, Margin was a concern - margin improved by 3% with the total increase in revenue by 45% and sales by 12% YOY

Revenue

For Germany, ML Model improved booked units sold by 4.5%, and margin dip by 3 % and revenue increased by 23% YOY

Future Work

The initial proposals process will be handed to the back office team. Thus the sales rep will have more time to focus on the market