Forecasts can be reviewed and adjustments can be made at the SKUL level, or for an entire grouping of SKULs designated in the GROUPING field in the Dynamic Header. To review and adjust several user-defined subsets simultaneously, please see the Forecast Summary screen. 

The Forecast Details Screen

 

Grouping SKULs

The Forecast Detail screen is capable of presenting six levels of demand planning information: Channels, SKULs,  Channel Groups, Groups, SKUs, and SKU SKULs. The tabs at the top left corner of the main grid can be used to toggle between these levels. The “SKUL” tab displays information for a single SKUL. The "SKU" tab displays aggregate information for a product across all locations. The "SKU SKULs" tab displays information for each unique location for the same item. Channels and Channel Groups are described in the Channel Forecasting section. The Group tab is dynamically labeled according to the grouping selected. It sums the historical/future demand data for all the SKULs that are included in the current displayed grouping. (If no grouping was specified during retrieval, the tab displays “Group,” and all the SKULs in the entire retrieval are summed.) Clicking on a tab switches the view to that level. The following paragraphs provide information on defining the desired grouping. 

To define a group of SKULs to be displayed in the Group tab, use the GROUPING fields in the dynamic header (shown below). Selecting a field in GROUPING sets that field as the primary grouping. For example, for a retrieval with 100 SKULs pulled from 3 Departments, selecting the Department field causes the Group tab to be labeled “Department” and to display the Department-level data. The Department-level data will be comprised of the SKUL totals for only the SKULs that match the displayed Department. Scrolling forward then displays the totals for the next Department. The SKUL tab will display all 100 individual results, and for 3 Departments, the Department tab will have 3 retrieval results. 

Groups of SKULs may be further divided up by selecting more fields. To select additional subgroups, select another field within GROUPING to define the first subgroup. For example, for a retrieval with 100 SKULs pulled from 3 Departments with 2 Planners each, selecting Department and Planner within GROUPING causes the Group tab to be labeled “Department/Planner” and to display Department/Planner-level data. The Department/Planner-level data will be comprised of the SKUL totals for only the SKULs that match the displayed Department and Planner. Scrolling forward first moves to the next Planner, then to the next Department. The SKUL tab will display all 100 individual results, and for 3 Departments with 2 Planners each, the Department/Planner tab will have 3 x 2 = 6 retrieval results.  

GROUPING Selection

Forecast Details Grid

Users may define which data columns to display in the main data table. To view the selection of available data columns, add or remove columns, or revise their order, right click on any of the columns in the main data table and select "Customize". Note the screen must be clear for the this feature to be available. Clicking this opens the Forecast Details Grid pop-up window.

Various column types can be set to appear with by checking the corresponding box in the "Set Visible" field. "Set Total", "Data Source Selectable", and "Actions" can be ignored as they have no impact on the Forecast Details grid. To change the display order, click and drag a field name to its desired location. 

Columns, column order, and graph selections are retained and remembered for each user name from session to session, even if the user logs out of GAINS 

Accessing Forecast Details Grid

 

Forecast Details Grid

Target Forecasts

Users may enter an unlimited number of additional forecast arrays for SKULs or groups of SKULs.  These additional forecasts types are user-defined.  Additional forecast arrays may be used to store budget or marketing forecast targets as well as customer-specific forecasts for a particular SKU or Group. Target Forecasts can be added to Forecast Details through the Forecast Details Grid pop-up window.

External recommended forecasts can be imported into a Target column array in GAINS by using Mass Upload function. See the section on Mass Uploads for more information.

Applying Adjustments

There are several ways to make a forecast / history adjustment. In either the Weekly View or Monthly View, the user may override the GAINS raw value, add to or subtract from the raw value, or multiply the raw value by a percentage. Also, forecast adjustments can be applied from 1-12 months at a time (forecast adjustments outside of lead time are not generally suggested unless the adjuster has specific information about future sales). Forecast adjustments should always be documented. The username and   time/date stamp are automatically logged when an adjustment is made; additional information should be included to note the reason for the adjustment.  

Percentage adjustments made on the Group tab apply the selected percentage to each SKUL in the grouping. Addition, subtraction, and override adjustments distribute the adjustment among the SKULs in the subset according to the percentage of raw forecast that the SKUL comprises of the total. Adjustments made at the Group level immediately appear on the SKUL tab. Similarly, adjustments made at the SKUL level are immediately displayed in the Group tab. A SKUL-level adjustment always overrides a Group-level adjustment. Deleting any adjustment automatically removes the related adjustments on the other tab.   

Fast Override adjustments may be typed directly into a forecast or history cell by double-clicking on it. Other types of adjustments or those requiring documentation are made in the Adjustment Maintenance window. Both methods are detailed below:

 

Steps required to apply a FAST OVERRIDE adjustment to the GAINS raw forecast:

1. Double-click in the cell corresponding to the month in which an adjustment will be made.

Selecting a cell for a FAST OVERRIDE Adjustment

2. Enter in the quantity of the adjustment.

3. Exit the cell by hitting Enter or clicking elsewhere. The potential adjustment is now displayed in red font.

4. Click Save or (Ctrl +S) to save the adjustment. The adjustment is now displayed in blue font. 

5. To remove an adjustment once it has been applied, right click on the cell and select "Delete Adjustment."

Deleting an Adjustment

Steps to use the Adjustment Maintenance window:

1. Right click on the desired forecast or history cell and select Adjustment Maintenance. This opens the Adjustment Maintenance window. 

The Adjustment Maintenance window

 

2. Enter the quantity to be applied in the “Adj Qty” field.

3. Select the type of adjustment in the Indicator (“Ind”) field.  The quantity result of the Adj Qty and Ind selection are displayed in the “Adj” field. The indicator options are:  

• (blank): Override adjustment. The Adj Qty is applied as the new adjusted forecast.
   
• +: Addition adjustment. The Adj Qty is added to the existing raw forecast (shown in the Raw field). After saving, the revised quantity does not change even if the raw forecast is revised in the future.
   
• -: Subtraction adjustment. The Adj Qty is subtracted from the existing raw forecast (shown in the Raw field). After saving, the revised quantity does not change even if the raw forecast is revised in the future.
   
• %: Percentage adjustment. Percentage adjustments are multiplied times the raw forecast.  To increase a forecast by 25%, the user must type 125 and % into the “Adj Qty” and “Ind” fields, respectively.  Entering 25% as a forecast adjustment will result in a 75% decrease of the raw forecast.

4. Enter the number of periods to affect in the “Periods” field. The application period starts on the selected month (shown in the Month and Year fields) and affects future months for the number of months entered. 

5. In the Document field, enter notes related to the adjustment. These notes are saved in each month that the adjustment is applied to.

6. From the “Event Name” drop-down list, select the applicable demand event that this adjustment is related to. The Event Description and Event Notes fields are automatically populated based on information for the selected event.

7. Click Apply. The potential adjustment is now displayed in red font. Click Save (or Ctrl + S) to save the adjustment. The adjustment is now displayed in blue font. 

 

Description of fields in the Adjustment Maintenance Window

Field	Description
Raw	The raw forecast or history value
Adj Qty	The forecast or history quantity being applied to the raw value. The method of application depends on the following "Ind" field
Ind	Indicator of the adjustment application method: Blank (Override): Apply the Adj Qty as the new forecast / history value + (Addition): Add the Adj Qty to the Raw value to create the new forecast / history value - (Subtraction): Subtract the Adj Qty to the Raw value to create the new forecast / history value. Adj Qty's that would result in a value less than zero are not allowed. % (Percentage): Multiply the Raw value by the Adj Qty percentage. Adj Qty less than 100 results in a reduction, while Adj Qty greater than 100 results in an increase
Periods	Number of periods to impact. leaving this field blank indicates a 1-period adjustment only
Adj	This value will become the new adjusted forecast/history value after changes are saved
Unit Price	Unit price for the SKUL, if available
Document	The adjuster may enter any documentation regarding the adjustment here
Adjuster	The GAINS login name of the adjuster is automatically recorded and displayed here after saving
Adjustment Date	The date / time stamp is automatically recorded and displayed here after saving
Event Type	This drop-down list displays the event categories set up in GAINS. These are maintained in the Demand Management Screen, accessible through the DEM Screen button
Event Name	This drop-down list displays the possible events the adjustment relates to. These are maintained in the Demand Event Management screen, accessible through the DEM Screen button
Event Description	This read-only field displays the description associated with the selected event. These are maintained in the Demand Event Management screen, accessible through the DEM Screen button
Event Notes	This field imports notes related to the Event Name, if selected, and can also accept new notes specific to the adjustment

 

Group Level Adjustments

If a change is made at the Group level, GAINS will disaggregate the values proportionally to the SKUL's raw value. If a SKUL value is zero, GAINS will implement Zero-to-One during Group Level Adjustments and the disaggregation will result in a nonzero SKUL value for that cell. 

History Adjustments

Adjustments to the demand history can be performed in the Forecast Detail screen.  Adjustments are typically made to smooth out unusually large spikes and troughs in demand that are unlikely to occur in the future, or to correct for errors/discrepancies in the raw history.  Adjustments to history do not delete raw history that is interfaced to GAINS, only overlay it with the revised value. After history has been adjusted, GAINS may create a different raw forecast during the next forecasting process. 

The steps for applying, reviewing, and deleting history adjustments are the same as those used for forecast adjustments. Please refer to those sections for details

Pending Adjustments

Users can create pending forecast adjustments for SKULs or groups of SKULs in the Pending column.  Simply enter the forecast adjustment or use the Adjustment Maintenance window in the same method as a regular forecast adjustment.  

The username of the person who saved the pending adjustment is displayed in the column header. If more than one person has made pending adjustments, the column display can be updated to display a different user. Double-click on the column header label to open the “User Selector” pop-up window. All users who have saved pending adjustments are available.

Real-Time Processing

When Users make fast override adjustments in the forecast or history cells, they can hit the refresh symbol in the toolbar to trigger real-time processing. There are three options for the refresh: SKUL, SKU, and SKUL+. 

For SKUL, the update occurs at the SKUL level.

For SKU, the update occurs for not just the selected SKUL, but for all other locations where that SKU exists.

For SKUL+, the update occurs for all related SKULs throughout the Bill of Materials and any parent/child relationships with that SKUL.

 

Real-Time Processing: Refresh Options

Channel Forecasting

Channel forecasting allows the management of distinct forecasting streams within a SKUL. This is useful when, for example, several customers’ demands for a SKUL must be forecasted and managed separately. Forecast and history are tracked and managed for channels, and the replenishment plan is generated for the total of all channels, the stocked SKUL record. Details and examples of channel forecasting and forecast/history adjustments are available in the GAINS Reference Manual. 

If Channels are enabled in GAINS, users will have access to two additional Forecast Details tabs. The "Channels" tab disaggregates the SKUL view into the channel for each SKUL.  The "Channel Group" tab displays aggregate information for an item across all Channels.

Although a SKUL’s channels are forecasted separately, the replenishment plan is generated for only the stocked SKUL record. Therefore, when viewed in conjunction with a replenishment screen (Inventory Event Planning or Cross-SKUL Order Builder), the Forecast Detail screen automatically synchronizes all the appropriate channel records. The SKUL and Group tabs have different synchronizing behavior, as follows. 

SKUL Tab 

When synchronizing the Forecast Detail SKUL tab with a replenishment screen, GAINS retrieves all the current SKUL’s channels and displays them in the SKUL tab. When scrolling through to the next SKUL in the replenishment screen, all the channels are cleared and reretrieved for the next SKUL. 

Group Tab 

When synchronizing the Forecast Detail Group tab with a replenishment screen, GAINS automatically sets the grouping option to Item and Parent Location. This is the SKUL-level forecast (total of all channels), which drives the replenishment plan. 

Forecast Ownership

If the Forecast Ownership functionality is enabled at the database level, the “owner” of a forecast is the user with the GAINS login that corresponds to the SKUL's planner code. On the SKUL tab, the owner’s username appears in the “Planner” field in the SKUL Header. Only the owner can save adjustments to the final adjusted forecast from which inventory policies and replenishment schedules will be generated. All other users may make forecast adjustment suggestions, which will be stored in the “Pending” column. Pending adjustments are displayed in orange font.

To accept a forecast adjustment suggestion, the owner can right-click on a value in the Pending column and select “Approve Pending Adjustment.” 

Another impact of the Forecast Ownership feature is that users who are not owners do not have access to the “Forecast Approve” icon. However, note that if forecast ownership is not enabled, any user may approve forecasts.

Demand Transaction Review

When demand transaction information is utilized in GAINS, the demand transactions for each item / location can be reviewed in the Detail table using the Demand Transaction Review window. To view the data, select any month of history that has demand transactions, right-click, and select Transaction History.

Selecting the Demand Transaction Review Window

 

On the left side are fields for including or excluding individual demand transactions from consideration in future forecasting. 

The Demand Transaction Review Window

Description of fields in the Demand Transaction Review window

Field	Description
Exclude	This checkbox can be marked to exclude records from being considered in creating forecasts. Click save after marking Exclude to execute
Include	This checkbox can be marked to include records that GAINS has marked as outliers in creating forecasts. Click save after marking Include to execute.
Ideal Location	The location where the product is ideally shipped from. Demand is captured in the system at this location regardless if it ends up shipping from here or not.
Actual Location	The actual location where the product is shipped from. This can differ from Ideal Location.
User Ship Loc Override	The location that the User can manually input as where the product is going to be shipped from.
Order Number	The number code identifier for the customer's order. Determined from the User's ERP system
Split	A number indicating how many times an order for a product was separated into various quantities (separate orders)
Request Quantity	Quantity requested for the transaction
Ship Quantity	How much of the product was shipped
Status	Code for the order status: 00 = Opened, 03 = Closed, 04  Canceled
Request Date	Date at which the request for shipment was received
Order Created	Date at which the customer placed the order to get the product shipped at a later date
Order Ship Date	Date at which the product was shipped to the customer
Customer Description	Description of the customer conducting the transaction
Customer Code	Code for the customer conducting the transaction
Sale Price	Price per unit
Line Number	The row number of the product within each order

 

History Begin Date Override

The History Begin Date Override function in GAINS allows Users to remove historical values from consideration for forecasts. This option can be accessed by right clicking on the monthly bucket you want to start the process from and selecting "Set History Begin Date Override". All monthly buckets prior to the one the User selected will now be ignored for Forecast calculations. Users can right click the cells highlighted in red and select "Clear History Begin Date Override". Forecast End Date Overrides set a date after which GAINS will apply a zero-forecast

Reasons to use History Begin Date Override:

• Prevent antiquated or non-repeatable sales patterns from impacting the forecast
• Reduce the noise contributing to upward or downward trending demand history
•  Quickly turn off a SKUL’s forecast in anticipation of the product stop selling

Set History Begin Date Override

History Begin Date Override

Clear History Begin Date Override

Set Forecast End Date Override

Forecast Approval

Each SKUL has a forecast approval flag, either “NAPP” for Not Approved, or “APPR” for Approved. The business purpose of these flags is to indicate which forecasts have been either automatically or manually determined to be satisfactory (and therefore which replenishment orders are valid). Forecast Approval is typically done at the beginning of each month after new forecasts are generated.  Owners should approve all reviewed forecasts to prevent unnecessary re-review.

Users can perform manual approval of forecasts for items that have been reviewed and possibly adjusted to planner satisfaction.  The user may approve forecasts on the SKUL tab only.  Clicking the “SKUL Approval/Unapproval" (white checkmark) button in the toolbar sets the forecast review status to Approved. Furthermore, clicking the "SKUL Approval + Next" (Right facing arrow icon) button in the toolbar (keyboard shortcut Ctrl+Enter) sets the forecast review status to Approved and brings the User to the next SKUL. 

A SKUL's approval status is shown in the “Approval Code” field in the Review Metrics section of Forecast Details. As noted previously, it shows “NAPP” for Not Approved, or “APPR” for Approved.

 

Forecast Approval

Forecast Simulation

The icon labeled “Forecast Simulation Popup” in the toolbar opens a window called "Forecast Simulation". The purposes of the Forecast Simulation feature are to (1) show users the potential results of all forecast models, not only the selected model, (2) enable application of a different model if necessary, and (3) show the impact of history adjustments on forecast immediately. 

By default the Forecast Simulation screen opens to the Sim Current tab, and the current GAINS-selected forecast model is displayed in the graph and highlighted in blue font in the list of models. All of the other forecast models (those simulated during the forecast process but not selected) are also displayed in the list. To display a different forecast model in the graph, click on the row to highlight it. To display multiple forecast models in the graph, hold down the Ctrl key while clicking the desired rows.

 

Forecast Simulation Popup

 

Forecast Simulation

Description of Fields in Forecast Simulation Window

Field	Description
Model Name	Forecast model name. The selected model is displayed first
Model	Forecast model number. The selected model is displayed first
Dmd Pattern Match	Y (yes) or N (no) based on which models are pre-selected for the SKUL's plausibility flag. More details on Model Plausibility can be found in the Reference Manual. Y models are preferred
Review Code	B, V, M, P, or blank based on the SKUL's review code. Blank models are preferred
MAD	Mean absolute deviation, a measure of the forecast model's simulation accuracy
RER	Relative error ratio, the ratio of MAD to total demand
Growth %	(12 Mo. Fcst – Past 12 Mo. Dmd)/Past 12 Mo. Dmd
Annual Fcst Dmd	Total 12 mo. Forecast
Months	12 months of raw forecast values

 

Apply a New Model

To apply an alternate model as a Fixed Model, highlight the desired row in the "Sim Pending" tab and select "Apply Fixed Model". A confirmation message appears that asks for an expiration date. Upon exiting the Forecast Simulation screen, note that the selected model is displayed as the Fixed Model. Additionally, the Fcst Adj column values are immediately updated to the relevant values for the new fixed model. However, replenishment records are not refreshed until the next GAINS replenishment planning process is completed.

Check Impact of History Adjustments

The Forecast Simulation window can be used to check the impact of history adjustments on the forecast before saving. The steps are as follows: 

1. Enter the required history adjustments in the Hist Adj column. Do not click Save. 

2. Open the Forecast Simulation window. 

3. The forecast model that would be selected under the revised history is displayed in the graph. 

4. For a reminder of the current history and forecast, use the “Sim Current” tab. 

5. If the revised forecast is acceptable, close the Forecast Simulation Detail window and click Save to commit the history adjustments. After the next GAINS forecast process, the Forecast column will update with the new values. 

6. If the revised forecast was not acceptable, close the Forecast Simulation window and revise the history adjustments as necessary. 

Saved history adjustments appear in the “Sim Current” tab – only applied but unsaved adjustments appear in the “Sim Pending” tab. 

Leading Indicators Analysis

Users can access the Leading Indicator Analysis (LIA) screen from the three-dots in the toolbar. This brings up the Leading Indicator Analysis window, which will be populated if multivariable regressions return statistical significance for the independent variables.

Accessing the Leading Indicator Analysis Screen

Accessing the LIA screen from the forecast summary screen enables users to view the Leading Indicator Analysis on a SKUL, SKUR and SKU level. SKUL level LIA is brought up automatically when selecting the window on a search that yields a single SKUL. When viewing an item with multiple locations (which will be shown in the details section of the window), SKU and SKUR level LIA can be done as well. To run LIA at a SKU level, the user should run the forecast summary for the item without other search criteria. To run LIA at a SKUR level, the user should run the forecast summary for the item with a specified region. If “region” is not in the header field, this can be added when the screen is clear using the “Define Header Fields” icon. The user should see the SKULs for this region in the detail table of the screen.

Leading Indicator Analysis Screen

After the desired SKULs are included, steps to complete the analysis are the same. The Variable Details and Statistics table will be prepopulated on the screen’s launch as described at a SKUL level. However, for SKU and SKUR level analysis, the user must check both check boxes in the “Analyze” column of the Leading Indicator Raw Data table of the Input tab to access the graphs and output table.