Performance Measurement

Accurate and timely demand plans are a vital component of an effective supply chain. Inaccurate demand forecasts typically would result in supply imbalances when it comes to meeting customer demand and hence Demand Forecast accuracy at the SKU level is critical for proper allocation of resources. In designing a holistic measurement process, it is important to understand that the demand metrics are the key driver and hence ownership of these metrics require careful deliberation.

Web Presentation of Forecast Accuracy and Safety Stock Planning is a useful resource to define, structure and use demand forecast metrics with in a supply chain process.

Supply chain metrics include customer service metrics (fill rates), inventory metrics (inventory turns, dead and near-dead inventory, inventory coverage), manufacturing planning (production volatility, manufacturing schedule adherence), and demand metrics (forecast accuracy and forecast bias. Once defined, the question is who will own these measures. All the measures are inter-related. By observing the metrics, we can understand the effect of organizational bias.

Assessing the metrics process and creating a new value-enhancing process generally involves the following steps:

1) Value Chain Balance - Overview

Value Chain Balance - Understand the importance of a balanced Value Chain
Demand Chain and Supply chain
Metrics and Goal Alignment
Unbiased Demand Plan – the Holy Grail
Value chain Measurement Process

2) Review of Customer Service Metrics

Measuring Customer Service
First time fill Rate
Order Complete
Decomposing Service Failure

3) Inventory and Production Metrics

Inventory Turns and Inventory Coverage
Value of Inventory Coverage Metric
Slow-moving and non-moving inventory
Manufacturing Schedule Adherence

4) Forecast Performance Metrics

Measuring Forecast Error
Mean Absolute Percent Error
Effects of Demand Volatility
What are the alternatives to MAPE?
Forecast Bias and Attainment

5) Introduction to Exception Management

What do we learn from Forecast Error?
Decompose forecast error to understand major drivers of error
Define process and systems to understand exception management
Illustrations of sample exception reports