How to Select the Best Demand Forecasting Model

Selecting the appropriate forecasting model is crucial for accurate inventory planning in Planster. Here’s how to choose the best forecasting method based on your sales channels, available data, and business needs.

Planster offers several forecasting methods to accommodate different product types and sales channels:

• Prophet: Great for data with clear seasonality or historical sales trends.

• Linear: Suitable for stable, steadily growing sales.

• Exponential Smoothing: Best for smoothing out irregularities and identifying short-term trends.

• ARIMA: Ideal for datasets with sufficient historical data to identify complex patterns and statistical relationships.

• Top-Down: Useful for overall revenue forecasting, with product-level demand derived proportionally.

• Bottom-Up: Effective for detailed, store-specific or retailer-specific forecasts.

Consider the following guidelines to determine the optimal forecasting model for your business:

• Bottom-Up Forecasting:

  • If forecasting for retail stores, use this model to leverage store-level data (store count x shelf facings x velocity).

  • Ideal for established retail channels where detailed information is available.

• Top-Down Forecasting:

  • Use this model when overall sales targets drive inventory requirements.

  • Helpful for new products or promotions where detailed historical data is limited.

• Prophet:

  • Select when clear seasonal patterns or historical promotional impacts exist.

• Linear or Exponential Smoothing:

  • Use for products with steady growth or consistent historical performance.

  • Exponential smoothing is preferable when occasional irregularities or short-term fluctuations appear in historical data.

• ARIMA:

  • Choose when you have robust historical sales data (at least 2+ years) to capture complex patterns accurately.

• For new products without history, start with Linear or Top-Down, then adjust based on actual sales.

• For products with clear seasonality, Prophet typically provides the most accurate forecasts.

• When historical data is extensive and stable, ARIMA and Exponential Smoothing models are highly reliable.

Regularly evaluate your chosen model’s accuracy:

• Compare forecasted versus actual sales periodically.

• Adjust your selected model based on observed forecasting accuracy and business conditions.

Reach out to Planster Support for personalized advice on selecting and optimizing your forecasting models!