Follow

# Forecast Structure

## Overview

CoV and ADI will help the system apply the proper forecasting algorithm to the level and nodes where the forecasts are generated.

The first step is to establish the different types of demand pattern categories to forecast.

To determine the characteristics of a forecast entities demand history, two coefficients are used:

• The first coefficient is the Average Demand Interval (ADI), it measures the regularity of a demand in time by computing the average interval between two demands. This is new.
• The second coefficient is the square of the Coefficient of Variation (CV²), it measures the variation in the demand quantities. In our case we will just use CV not CV²

To calculate ADI, we measure the average interval between two demands over the last 12 months of history. Per the following example:

The calculation for ADI is calculated as:

Per above, demand occurs on average every 1.83 periods.

## CoV

Then we measure the coefficient of variation which is calculated via the following formula:

## Demand Profile

To compute the coefficient of variation, we will only consider the non-zero values of the demand history. In the example above, the average quantity equals to 10.57 while the standard deviation equals to 6.43.

Based on these 2 dimensions and the thresholds for these values set within system settings, we classify the demand profiles into 4 different categories. This will run with each new planning period as a prior bucket period passes. This analysis will be calculated in the forecast analytic page.

• ADI Threshold: 1.32 (this will be the default setting that will be overwritten)
• CoV Threshold: 0.70 (this will be the default setting that will be overwritten)

The above will fill in the thresholds to categorize the history that will be the basis for the forecast algorithm family selected:

1. Smooth demand (ADI <= (ADI Threshold Value) and CoV <= (CoV Threshold Value)): The demand is very regular in time and in quantity. Apply exponential smoothing models - single, double, triple
2. Intermittent demand (ADI > (ADI Threshold Value) and CoV <= (CoV Threshold Value)): The demand history shows very little variation in demand quantity but a high variation in the interval between two demands. Use Croston's (chosen best from Croston/Croston Standard/ Croston Syntetos Boylan) if demand does not exceed 25 units per period or Croston Syntetos Boylan model if greater than 25.
3. Erratic demand (ADI <= (ADI Threshold Value) and CoV > (CoV Threshold Value)): The demand has regular occurrences in time with high quantity variations. Use Syntetos Boylan model only.
4. Lumpy demand (ADI > (ADI Threshold Value) and CoV > (CoV Threshold Value)): The demand is characterized by a large variation in the quantity of demand and in the interval between two demands. This is quite impossible demand to reliable forecast, no matter what forecasting tools and methods are being used. Use bootstrapping.

In addition, if the forecasting history contains <= 2 buckets with values > 0, Simple Moving Average will be applied.

## History used for ADI, COV, Demand Profile

History used for forecasting is based on "Forecast history length (months)" and "Forecast history based on" system settings.

For the ADI, CoV and Demand Profile which are resulting in the algorithm selection, only last year of the forecasting history (12 months or 52 weeks depending on the forecasting bucket type) is used.

Leading zeroes are removed only if they are part of the leading zeros in the whole forecasting history. Example - current month is July 2018, system setting forecast history length is 18. History used for the calculations is July 2018 - June 2019.

Thereafter, we will use the forecastability metric and the business importance, to recommend a planning approach for each SKU. This will be done in the forecast analytic section.

1. Collaboration: If the item has an A classification and CoV > (CoV Threshold Value) then the planning method should be collaboration
2. Statistical: If the item has an A classification and CoV <= (CoV Threshold Value) then the planning method should be statistical
3. Min/Max: If the item has a B or C classification and CoV > (CoV Threshold Value) then the planning method min/max
4. Statistical: If the item has a B or C classification and CoV <= (CoV Threshold Value) then the planning method statistical