Sales Forecasting

The Sales forecasting functionality allows you to plan product demand over a defined future period. It combines AI-assisted forecast generation with manual input, providing a flexible tool for demand planning.

Sales forecasting allows:

• Creating AI-assisted demand forecasts for products based on historical customer order data.
• Reviewing forecast data grouped by configurable time periods.
• Comparing forecasted quantities with previous periods and previous year values.
• Linking forecasts to the Master Production Schedule (MPS) for production and procurement planning.

Jump to:

1. Demo video: Sales Forecasting
2. Enabling sales forecasting
3. User access rights for sales forecasting
4. General sales forecasting process
5. Creating a sales forecast
6. Adding items to the forecast
7. Forecast overview
8. Forecast input table
9. Running an AI forecast
10. General and Detailed views
11. Linking a forecast to the MPS
12. Sales forecasting limits

Enabling sales forecasting

This functionality can be enabled in Settings -> System Settings -> Enterprise Functions -> Sales forecasting.

The following software sections become available:

• The Production Planning -> Sales forecasting page, where it is possible to create and manage sales forecasts.
• The Sales forecast row in the Production Planning -> MPS view, where it is possible to link a sales forecast to a product for production and procurement planning.

User access rights for sales forecasting

To work with sales forecasts, the user must have appropriate rights set at Settings -> Human Resources -> Users -> Rights of individual user.

Please note that even without the Read right for Sales forecasting, the user can see an item's forecast values in the MPS tab, provided they have the right to access MPS.

General sales forecasting process

Working with a sales forecast takes place in several steps:

1. Create a sales forecast by selecting the products and defining the forecast horizon.
2. Run the AI forecast to automatically generate demand predictions based on historical customer order data.
   1. If a product lacks sufficient historical data, forecast values can be entered manually.
3. Review and adjust the forecast values in the forecast input table, using the Detailed view to compare them with actuals and previous-year data.
4. Review aggregated data in the Forecast overview section, using configurable period groupings.
5. Link the forecast to the Master Production Schedule (MPS) to use the forecast values for production and procurement planning.

Creating a sales forecast

For creating a sales forecast:

1. Go to Production Planning -> Sales forecasting and click the Create button.
2. In the first pop-up ("Step 1 of 2: Create a new sales forecast"), enter the Name for the forecast.
3. Select the First month - the starting month of the forecast period. The default is the current month. Months up to two years in the future can be selected.
4. Select the Horizon - the number of months the forecast will cover. The available options are 3, 6, 12 (default), and 18 months.
5. Click Continue.
6. In the second pop-up ("Step 2 of 2: Select products"), select the products to include in the forecast.
7. Click Create forecast.

The forecast is created and ready for editing.

Please note that up to 100 products can be included in a single forecast. A maximum of 20 products can be added at a time. For more details, see Sales forecasting limits below.

Adding items to the forecast

To add products to the forecast:

1. Click the Add items + button above the input table.
2. In the pop-up, select the products to add from the item selection table.
3. Click Add.

Please note that up to 100 products can be included in a single forecast. A maximum of 20 products can be added at a time. For more details, see Sales forecasting limits.

When new products are added, the AI forecast is automatically queued for them.

Forecast overview

The forecast overview section in the forecast details provides summary statistics and a chart for a selected product.

To view the overview for a specific product, select it from the autocomplete field at the top of the section. The selection affects both the summary table and the chart.

The summary table displays aggregated forecast data, grouped by configurable time periods. The default grouping depends on the forecast horizon: 3 months for a 3-month horizon, 6 months for a 6-month horizon, etc.

The table shows:

• Forecast total – the sum of all monthly forecast values within the grouped period. Shown when a monthly grouping (3, 6, or 12 months) is selected.
• Estimated year total – replaces "Forecast total" when the calendar year grouping is selected. This metric provides a projected annual quantity by combining actual sales data for months that have already passed with forecasted values for the remaining months of the year. This allows comparing the full-year outlook across calendar years, even when some months have not yet elapsed.
• Change vs previous period – shows whether demand is increasing or decreasing between consecutive periods, expressed as an absolute quantity difference.
• Change vs previous period (%) – the same comparison expressed as a percentage, making it easier to assess the magnitude of change across products with different volumes.

For the first period, the change values are calculated using actual historical data, if available.

The chart displays historical sales data and forecasted values by month for the selected product.

Forecast input table

The forecast input table in the forecast details is where you enter the expected sales quantities for each product, month by month.

Each product is listed as a row, and each month of the forecast period is a column. To enter or edit a forecast value, click on the cell for the desired product and month.

The forecast input table supports two display modes, General view and Detailed view, which can be toggled using the tabs above the table. See General and Detailed views below for details.

Running an AI forecast

The AI forecast generates demand predictions for each product based on its historical customer order data. The system uses all available order history, specifically, the total ordered amounts and delivery dates of customer orders. Orders in statuses "Quotation" and "Cancelled" are excluded.

To run the forecast, click the Run forecast button above the input table. A loading alert is displayed while the values are being generated. Once complete, the forecast values are populated in the input table.

The Run forecast button replaces empty and previously AI-generated values with new predictions. Manually entered values are not overwritten.

Please note that if the forecast period is entirely in the past, the Run forecast button is disabled.

General and Detailed views

The forecast input table can be toggled between two views using the tabs above the table.

The General view shows only the forecast value row for each product.

The Detailed view shows additional rows for each product:

• Actual – the actual quantity of products sold in that month. Displayed for months that are in the past.
• Previous year – the sales quantity for the same month from the previous year's history.
• Year-on-year change (%) – shows how the forecasted quantity compares to what was actually sold in the same month last year, expressed as a percentage.

Linking a forecast to the MPS

Sales forecasts can be linked to the Master Production Schedule (MPS) to use forecast values for production and procurement planning. For details on how to link a forecast in MPS, see the Master Production Schedule (MPS) page.

Sales forecasting limits

Why is the maximum number of products in a forecast limited to 100?

Forecasting is a resource-intensive process. For every product, the system retrieves historical customer order data and runs AI-based calculations. Because of this, the maximum number of products per forecast is limited to 100, and a maximum of 20 products can be added at a time.

Why does a product show "Not enough historical data"?

The AI forecast requires a minimum of 3 months of historical sales data (from customer orders) to generate a prediction. If a product has fewer than 3 months of non-empty data points, the system cannot generate a forecast for it. In this case, the product row will be marked with a grey indicator and the tooltip "Not enough historical data to run a forecast with AI."

You can still manually enter forecast values for such products.