2. Demand data loading

This training presents the data structure used by Demand Planning and guides the load of each dataset in the Community Edition. The eight files follow the same process sequence, from 01 through 08.

Both tutorial languages provide identical original system exports. Scenario IDs and configured descriptions are in Portuguese; technical headers remain as exported by the application. The English text explains this same dataset.

Demand Planning overview

Demand Planning aims to generate a sales projection that combines two components:

• a statistical foundation for sales projections;
• collaboration among Planning, Sales, Marketing, and other teams in building the plan.

The demand plan is essential for make-to-stock operating models, particularly in long supply chains such as those that depend on imported products or components. Because this type of operation relies on a future sales projection to make inventory, production, distribution, and productive and logistics capacity decisions, the demand plan becomes the main input for the tactical and operational decisions that follow.

Data structure

In the Community Edition, the Demand Planning module has a straightforward data structure, consisting primarily of the master and transactional data shown below:

---
config:
  layout: elk
  elk:
    mergeEdges: false
    nodePlacementStrategy: NETWORK_SIMPLEX
---
flowchart LR
    %% Global wrapper keeps the three registration groups stacked and pointing to Sales on the right.
  subgraph registrations[" "]
    direction TB

    subgraph materialRegistration["Material Registration"]
      direction LR
      materialCharacteristics["Material<br/>Characteristics<br/><span style='font-size:10px;color:#64748b'>E.g.: Family, Category</span>"]
      materialCharacteristicValues["Characteristic<br/>Value by<br/>Material"]
      materials["Materials<br/><span style='font-size:10px;color:#64748b'>Finished Goods, Semi-Finished,<br>Raw Materials</span>"]

      materialCharacteristics --> materialCharacteristicValues --> materials
    end

    subgraph locationRegistration["Location Registration"]
      direction LR
      locationCharacteristics["Location<br/>Characteristics<br/><span style='font-size:10px;color:#64748b'>E.g.: Channel, Region</span>"]
      locationCharacteristicValues["Characteristic<br/>Value by<br/>Location"]
      locations["Locations<br/><span style='font-size:10px;color:#64748b'>Customers, Plants,<br>DCs, Suppliers</span>"]

      locationCharacteristics --> locationCharacteristicValues --> locations
    end

    subgraph unitRegistration["Units of Measure Registration"]
      direction LR
      globalConversions["Global<br/>Conversions<br/><span style='font-size:10px;color:#64748b'>E.g.: TON->KG</span>"]
      unitsOfMeasure["Units of<br/>Measure"]
      materialConversions["Material<br/>Conversions"]

      globalConversions --> unitsOfMeasure
      materialConversions --> unitsOfMeasure
    end
  end

  sellout["Sales<br/>(Sell-Out)"]

  materials --> sellout
  locations --> sellout
  unitsOfMeasure --> sellout

%% Keep the outer registrations wrapper visually transparent.
style registrations fill:transparent,stroke:transparent,stroke-width:0px

Accessing Data Registration and Extraction

In the Community Edition, open the side menu → Data → Data Operations. The catalog is organized from left to right as Theme, Group, Section, and Topic. Selecting a topic opens its API endpoint and the Download, Import, and, when applicable, Delete operations below the catalog.

Load order

Numbering is global: it indicates the order in which files are presented and loaded in this tutorial.

01. Units of measure

Units of measure are references used by materials, sales, inventory, routings, and bills of material. This tutorial registers UN, KG, and TON.

Loading the units-of-measure dataset

X01_unit_of_measure.xlsx

Import the three rows through Import.

02. Material characteristics

Planning characteristics classify materials and will be used in the next training to build forecast clusters. This dataset registers each characteristic's name and type; its values are later provided in the material file columns.

Loading the material-characteristics dataset

X02_material_characteristics.xlsx

03. Materials dataset

Materials represent raw materials, intermediates, and finished goods. The Estágio do material, Família do produto, and Formato do produto columns are included in the same file because those characteristics were registered in the previous step.

Loading the materials dataset

X03_materials.xlsx

Import the 18 materials through Import.

04. Global unit conversions

Global conversions record equivalences that apply to every material. They are used when the relationship between two units does not depend on the SKU, unlike the material conversions presented in the next step.

In this scenario, Sales history is in KG, while the forecast is generated and analyzed in TON. The global conversion between kilograms and tonnes is therefore required for the Demand Planning execution.

Loading the global unit-conversion dataset

X04_unit_conversions.xlsx

The two rows explicitly establish 1 TON = 1,000 KG and 1,000 KG = 1 TON. Both directions are required when the KG sales history is planned in TON.

05. Material conversions

This dataset records equivalences that depend on the SKU. For the eight finished goods, one UN corresponds to 0.25 KG or 0.40 KG, depending on the material.

The current history is already in KG, and the scenario's productivities are in TON. The global KG ↔ TON conversion in file 04 is therefore sufficient for the current forecast. File 05 completes the UN → KG → TON network so that a future input or analysis expressed in units can also be converted.

Loading the material-conversion dataset

X05_material_unit_conversions.xlsx

06. Location characteristics

These characteristics classify demand locations by region and channel. As with materials, this dataset registers the characteristics, while their values are provided later in the location file itself.

Loading the location-characteristics dataset

X06_customer_characteristics.xlsx

07. Locations dataset

Locations represent suppliers, plants, distribution centers, transshipment points, customers, or demand regions. Região do cliente and Canal de venda values are loaded in the final columns of this same file; there is no separate load for those values.

Loading the locations dataset

X07_locations.xlsx

The parameters needed by Supply Planning are already explicit in this file: Production Planning Book availability is 1 for the two plants and 0 for other locations; Supply Planning Book availability is 1; the default UOM is TON; and safety stock considers indirect demand (1) at every location. Preserve these values when reimporting: a blank field can clear an existing configuration.

The last two columns follow characteristic-ID order: Customer Region (REGIAO_CLIENTE) and Sales Channel (CANAL_VENDA).

08. Sales

In OpsFactor, Sales history may represent sell-in or sell-out. The Community Edition provides sell-out, which records a past sale—typically associated with billing—from a location. This tutorial uses the term Sales because it is more direct for the person running the process.

Sales history is the input to Demand Planning statistical models. For monthly or weekly projections, load at least two years of history so that models such as Holt-Winters and ARIMA have at least two complete seasonal cycles.

The file contains 38 observed months, from November 2023 through December 2026, for ten finished goods and six customer locations. There are no observed-sales rows in 2027; January 2027 is the forecast reference month. Every quantity is in KG; the global conversion in file 04 makes it possible to generate and analyze the forecast in TON.

Loading the Sales dataset

X08_sellout_history.xlsx

Import the 2,280 rows last, after materials, locations, and units of measure have been registered.

Return to the course map or continue to the 3. Demand Forecast Overview.