Auto 60


Extreme demand volatility has become normalized for many tier one suppliers. Most take one of the following measures to mitigate the negative impact:

  • Overproduce to Ensure Demand is Met
  • Stress Operations to Respond to Last Minute Needs
  • Fail to Deliver On-Time Resulting in Poor Scorecard Results

Overproducing is the tactic most manufacturers use, but it negatively impacts inventory turns and cash flow. Stressing operations to respond quickly is common for many tier one suppliers, but it also negatively impacts employee retention, overtime, and on-time delivery. Most embrace these negative consequences to prevent deductions on their scorecard. A poor scorecard can hurt chances of getting future business, and can destroy a company’s reputation.

Our client faced the challenges mentioned above in the Automotive, Heavy Duty Truck, and Aerospace industries for decades, but 2020 presented a set of new challenges. The global supply chain shrunk and then quickly expanded resulting in bottlenecks at ports and demand outperforming supply for logistic companies. This exacerbated the existing challenges of shrinking customer lead times and growing global supply chain. There had never been a more important time to implement a solution to improve their demand forecasting.


Our client identified 117 high volume parts produced in Asia and sold in North America. We received three years of historical demand forecasting and actual shipment data that our AI software used to identify trends. Using a mixture of historical and third party data along with industry leading algorithms, we compared how their existing demand forecasts compared to our AI-optimized method.


The study spanned four weeks from September-October 2020 and used eight-week lead time data for the 117 parts. We measured the results using the absolute error of each forecast. The absolute error is the forecasted value vs. the actual number of products shipped to the customer.

  • Legacy Demand Forecasting: 50% Absolute Error
  • AI-Optimized Demand Forecasting: 20% Absolute Error
  • Actual Demand Forecasting Efficiency Improvement: 60%

Our client will realize improvements to the following areas by implementing our AI-Optimized Demand Forecasting Software:

  • Cash Flow
  • Inventory Turnover
  • Operational Efficiency
  • On-Time Delivery
  • Logistic Expedite Fees


Forecasting footer 01