Demand planning is a collaborative and continuous process that always leaves room for improvement. Therefore, the quality of forecasts must be measured, challenged, & analyzed to identify the causes of discrepancies between forecasts and actual outcomes, and to refine the models used.

Data Quality: An Often-Overlooked Foundation

Every simulation relies on data. That’s why data analysis is essential before any simulation takes place. Proper data governance is critical to ensure that simulations and strategies are based on verified and actionable facts.

Working across multiple systems increases the risk of conflicting data. Therefore, it is vital to implement strong data synchronization processes between systems and to regularly monitor data accuracy.

In the context of demand planning, accurate data allows you to define priorities and identify sourcing and production strategies that are aligned with customer needs while minimizing exposure to uncertainty. This means focusing on supply lead times, production timelines, new customer launches, inventory positions, relevant sales, and more.

Data Accuracy: A Tool for Performance Evaluation

Once data has been cleaned and harmonized across systems, all stakeholders in the operational chain share the same knowledge base, making it easier to collaborate on improving forecasts.

This foundation makes it easier to establish ongoing controls on demand-related variables to ensure that the data in use remains relevant.

Tracking key metrics, such as production times, quantities produced, delivery times, and customer usage, ensures that safety measures (e.g., safety stock, buffer time) remain appropriate. This tracking can be dynamic, using tools such as Power BI, Tableau, or Excel, to monitor performance indicators, identify drops in forecast quality, and trigger necessary corrective actions.

This ongoing monitoring also becomes a strategic asset in contract renegotiations with suppliers, customers, or transporters by helping identify the actions with the highest return on investment.

Forecasting: Which Indicators Should You Use?

A variety of indicators can be used to measure forecast quality and identify the root causes of discrepancies:

• MAE (Mean Absolute Error) – a statistical metric that measures the average magnitude of errors between paired observations, commonly used in regression and forecasting to quantify the average difference between predicted and actual values.
• MAPE (Mean Absolute Percentage Error) – a statistical measure of prediction accuracy for forecasting models. It calculates the average absolute difference between predicted and actual values as a percentage of the actual values. 
• Bias (Forecast Bias) –  the difference between the forecast and actual sales.
• FVA (Forecast Value Added) – measures how much each step in a forecasting process improves or degrades the forecast’s accuracy and bias compared to a baseline.

Identifying the right indicators is important, but knowing where, when, and how to use them is even more critical to getting meaningful insights.

A deep analysis of these metrics can reveal trends by product, supplier, or category over time. This helps define actions to maximize customer service while minimizing risk, through the identification of seasonality patterns, market trends, temporary shortages, or alternative sourcing options.

Implementing forecasting tools like SlimStock, Blue Ridge, or even temporary dynamic solutions using Excel, promotes collaboration across all supply chain stakeholders. It also enables teams to focus efforts on high-impact products, thanks to relevant dynamic classifications.

At Avalon CSC

We support you in the implementation of tailored supply chain solutions aligned with your business needs and strategy. Our team is here to assess your requirements, standardize your data, and deploy the most suitable solution, while providing comprehensive support and training to all users as you transition toward industry best practices.