Optimized Inventory with Real-Time Forecasting for Leading Manufacturer
overview
The client lacked a modern and efficient way of evaluating open bids, or opportunities, their bid process and forecasting. Their business sells earth moving machinery through dealers and has access to thousands of open bids across the nation. The client’s current mechanism for pricing is reactive causing diminished margins and productivity loss. Incorporation of a predictive model enabled them to set the price in a more calculated manner, leverage many other factors and to produce insightful reports achieving a higher margin and securing more wins.
Download the case study brief here.
challenges
Lack of a bid assessment method resulted in decreased productivity and stagnant win-rate
The client lacked a method of quantifying the impact of the hundreds of variables that impact the result of a bid. They had limited understanding of which variables affected bid outcomes and by how much. The inefficient, manual process of collecting bid information led to a dependency upon “just-in-time” advice that frequently relied upon tribal knowledge.
technology used
Amazon Web Services (AWS)
Amazon SageMaker
Amazon Forecast
AWS Glue
Python
solution
Integrate targeted, filterable intelligence dashboard from current and former bid data to the existing CRM
Using a gradient-boosted decision tree model, PREDICTif took years of client data to develop a predictive model that consumes open bid/opportunity data to determine the likelihood of winning bids. The model is scalable, portable, and future proof with a retraining process that incorporates data. The dashboard integrates into the Client’s CRM for increased accessibility within the company.
outcome
Elevated efficiency, overall cost reduction, more accurate demand forecasting
The client has boosted productivity and efficiency by adding accurate demand forecasting. The use of AWS services assists in decreasing overall costs, boosting profits. They can now optimize quoting, focus on dealer benchmarking, and evaluate individual product strengths and weaknesses. Now they can easily acquire a significantly higher win rate without redistributing efforts.


