Forecasting Solutions Project Process

The Project Process
Examine data for underlying trends, seasonality
*If technically needed or
requested by client
Customisation*
1
step
2
step
3
step
Sales Forecast System

Decomposition of the sales data based on the observed influences, setting «coefficients»
1
step
Base Sales
Volume Trend Decomposition
(x1000)
Sales Forecast System

* e.g. “weather” scenarios (avg t° from the last 3-4 years , avg + or – 2°C)
Forecast the economic and other category variables
2
step
Inflation
# of SKU
Weather
Sales Forecast System

Sales Forecast
Variables Forecast
Forecast base sales by segment using the past base sales
Then, we applied the coefficients from the model to the forecasted influencers
Volume Trend Decomposition
(x1000)
Baseline forecast and decomposition of forecasted sales based on forecasted variables
3
step
+
Sales Forecast System

Scenarios simulations
4
step
We change influences forecast at Step 2 and repeating Step 3 and can see how category sales will react to this change and compare trends
What if during
next summer
temperature is
+5oC higher?
Example of
scenario:
Weather Forecast Current
Sales Forecast Current
Weather Forecast Scenario
Sales Forecast Scenario
Quarter 1 Quarter 2 Quarter 3 Quarter 4
Sales Forecast System