How to Pull Reports for Demand Forecasts

Walmart has two primary datasets in Scintilla that are used for understanding demand forecasting: 

• Store Demand Forecast
• Order Forecast

The Store Demand Forecast dataset is defined by a description of its action on the clickable bubble: “Review the projections of what Walmart expects to sell to customers through stores over the upcoming weeks.”

The Order Forecast dataset is defined by a description of its action on the clickable bubble as well: “Review the projections of what Walmart expects to order over the upcoming weeks.” 

The main difference between the way these two datasets works is the point of sale. Store Demand Forecast is meant to show a prediction of demand based on consumer sales, and the Order Forecast is meant to show a prediction of what the retailer will be ordering from the supplier. 

Demand Forecast Terms in Scintilla

Each of these datasets has specific business elements (see the Appendix) that are defined in-app. These function as “columns” in the report, and they show the main datapoints for these reports. 

There are only two business elements in the “Demand Forecast” section of the Store Demand Forecast dataset:

• Walmart Year Week Number: “Sale WM Week - WM week number of the demand forecast week. This goes all the way to forward looking week 104 WM weeks for store item forecast.” 
• Final Forecast Each Quantity: “It is the final forecasted quantity after making all the adjustments. Final Forecast Each Quantity = Sales Forecast Each Quantity + Total Adjustment Each Quantity.”

The Order Forecast dataset has five business elements with their own definitions, two of which are automatically selected for the dataset: 

• DC/Store Number (Automatically Selected): “Use this attribute whether pulling DC/WHSE forecasting report or Store forecasting report.”
  • Aliases: Store Nbr, Store Number, Whse Nbr, Whse Number, DC Nbr, DC Number

• DC/Store Type (Automatically Selected): “This attribute informs the user about type of business unit (DC or Store).” 
• Order Place Date: “The date that the order should be placed with the vendor.”
  • Alias: Plan Order Date

• Scheduled Arrival Date: “The shipment date recommended for the order to be delivered on time. This can be from the vendor or a DC.” 
  • Alias: Plan Receive Date

• Order Each Quantity: “Units ordered, derived from vendor pack order quantity and vendor pack qty.” 
  • Alias: Suggested Order Qty

There are also five additional business elements in the Order Forecast that are only available in Charter and that are designed to give more incisive predictive data to suppliers: 

1. Available Ship Date: “This is the date the item is available at the Distribution Center (DC) or Vendor.” 
2. Forecast Sequence Number: “The Forecast Sequence numbers are the unique numbers generated by the forecasting system for a given store/item combination on a given date.”
3. Need Arrival Date: “The date an item is needed to arrive at the store.”
4. Need Ship Date: “This is the date that the merchandise needs to be shipped by in order for it to arrive at the Walmart facility on time.”
5. Scheduled Ship Date: “The date an order is scheduled to ship. This date can be used for fetching forecasting values.”

Note that, in the case of this particular dataset, the Charter business elements do not have DSS aliases because they represent value adds that were not available in the Retail Link reporting app. 

How to Pull Demand Forecast Reports in Scintilla

Navigate to:

Walmart Scintilla > Report Builder > Dataset > New Report > Store Demand Forecast OR Order Forecast

For either of these reports, it’s best to click “Use this dataset” in the bottom right. Then navigate to the Columns tab, selecting all of the business elements in the Columns section most indicative of the report (i.e. “Demand Forecast” for Store Demand Forecast and “Order Forecast” for Order Forecast, respectively). 

From there, all the other sections and their business elements can be selected or not, depending on the sort of data analysis desired. 

From there, the Filters must be selected. The only Filter that is necessary to run these reports is related to Walmart time: 

• Forecast Walmart Week for Store Demand Forecast
• Suggested Order Date for Order Forecast

From there you can create the report and customize it just like any other report, saving it with a name, a cadence, a preferred format, and more. As with other reports, it is likely to take somewhere between 20 to 70 minutes to finish pulling.