Show TOC

Process documentationDetecting Price Change Events

 

Price change events model the short-term effect of changes in price on the demand. For example, demand might rise directly before a price increase, and drop directly after it. In the Automatic Outlier Detection and Event Generation report, you can have the system create price change events in the past, based on historical price and demand data. The system uses the effect of the price change events in the past to estimate the short-term effect of planned price changes in the future, and uses the results in forecasting.

Prerequisites

You can use this function if the following applies:

  • Forecast settings for planning areas

    On the SAP Easy Access screen, choose Start of the navigation path Advanced Planning and Optimization Next navigation step Demand Planning Next navigation step Environment Next navigation step Administration of Demand Planning and Supply Network Planning End of the navigation path, open your planning area, and in the system menu, choose Start of the navigation path Extras Next navigation step Forecast Settings End of the navigation path.

    • In Price Settings screen area, in the Price Key Figure field, select the key figure that contains the price data.

    • In the Price Change Event Type field, select a non-reoccurring even type that you have created in the Demand Planning Worklist to model price changes.

      The bucket you use to define the event type duration (for example, week) must match the bucket you use to detect price change events in the Automatic Outlier Detection and Event Generation report. (You specify the bucket in the Periods field in Automatic Outlier Detection and Event Generation.)

      You can specify periods of impact before or after the actual event for the price change event type. You specify this in the Periods Before and Periods After fields in the Demand Planning Worklist, in the Event Types query.

    • In the Price Change Threshold field, you specify a threshold value in percentage. The system detects price change events in periods where there is a (positive or negative) change in the price key figure compared to the previous period that is greater than the threshold value in percentage you specify.

  • Forecast profile

    You must assign a univariate forecast profile to the CVCs for which you want to detect price change events, as follows:

    • The bucket you specified in the Period Indicator field of the master forecast profile must match the bucket you use to detect price change events in the Automatic Outlier Detection and Event Generation report. (You specify the bucket in the Period Indicator field in Automatic Outlier Detection and Event Generation.)

    • You must specify an MLR profile in theMLR for Univ. Fcst (MLR Profile for Univariate Forecasting) field of the univariate forecast profile. The MLR profile you specify must contain the price key figure in the MLR independent variable list.

    • You can specify a price change threshold in the field Price Change Threshold in the univariate forecast profile. If a value is entered here, it overrides the value set in the planning area forecast settings described above, for CVCs that have this univariate forecast profile assigned.

    • If you have selected the Allow T-Test to Disregard checkbox for the price key figure in the MLR profile, the system creates price change events only in case the t-test does not disregard the price key figure as an MLR independent variable.

    • During the calculation of price change events, the system only applies seasonal correction in case the Periods field of the univariate forecast profile is filled, and the detection time interval you specified for the Automatic Outlier Detection and Event Generation report contains at least twice as many periods as you specified in the Periods field in the univariate forecast profile.

To use this function, you must activate the business function SCM-APO-FCS, Events and Outliers, Causals, ABC/XYZ Classification (SCM_APO_FORECASTING_1). For more information, see SCM-APO-FCS, Events and Outliers, Causals, ABC/XYZ Classification.

Process

The system performs price change event detection as follows.

  1. The system starts the price change detection process within normal outlier detection, after it has cleaned the demand history from past non-price change events.

  2. The system detects changes in the price key figure.

  3. The system cleans the history from the effects of existing price change events. In case a price change event needs to be reestimated (for example, in a previous detection run, only part of the event was in the past, so that not all of its effect could be estimated), its effect is not cleaned from the history.

  4. The system runs MLR forecasting to calculate the effect of MLR independent variables on the demand history. Such effects are then subtracted from the history. The system uses the MLR forecast profile you specified in the MLR for Univ. Fcst field of the assigned univariate forecast profile. In case the price key figure is not specified as an independent variable in the MLR profile, or it is disregarded based on the t-test, price change events are not created for the processed CVC.

  5. The system smooths the history in the periods where price changes are detected, and price change events are to be created. Smoothing is done using moving average calculation. The system takes the window for moving average calculation from the values you have specified for the planning area - periodicity combination in the Outlier Detection Period Before and After Settings view of the Customizing activity Configure Outlier Detection and Event Catalog Assignments in Customizing for Advanced Planning and Optimization, under Start of the navigation path Supply Chain Planning Next navigation step Demand Planning (DP) Next navigation step Basic Settings Next navigation step Events and Outliers Next navigation step Configure Outlier Detection and Event Catalog Assignments End of the navigation path. A window the size of a full season year (taken from the Periods field of the univariate forecast profile), is used if the Customizing settings above are not maintained. In case the Periods field is also empty, the system calculates with one calendar year.

  6. The system smooths the corrected history calculated in the previous step once with moving average calculation for all periods within the detection interval. The window used for the moving average calculation is the same as in the previous step.

  7. The system creates price change events. It calculates the effect of price change events as the difference between the history the process started with in step 1, and the corrected history calculated in step 6. The effect cannot be greater than the original history of the corresponding period. The system calculates the effect only for the periods within the detection interval specified for the Automatic Outlier Detection and Event Generation report. If some periods of the events are outside this interval the system proceeds as follows:

    • If the price change event effect for the particular price change has never been estimated, the system sets the effect outside the detection interval to 0.

    • If the price change event effect for the particular price change has already been estimated by a previous detection run, or forecasting, the system does not change the previously estimated value. In case the effect of a price change event cannot be fully estimated, because some of its periods lie outside the detection interval specified for the Automatic Outlier Detection and Event Generation report, the system marks the event for future reestimation. The system processes and updates such events in later runs of the Automatic Outlier Detection and Event Generation report, until they become fully estimated. Reestimation means the deletion of the existing event, and the creation of a new updated event.

      The system creates price change events at the points in time where it detects a price change as described in step 2, except for the following cases:

      • If a price change event exists already at the exact same time.

      • If multiple price changes are detected so close to each other in time that the price change events would overlap each other if they were created; one event would start before the other ends. In such a case, the system creates only the last price change event, and only if the overall price change is above the threshold. It adds up the consecutive price changes, for example if the price increases by 5 and then decreases by 8, the net price change for which the event is created is -3. This only applies for price change events that are created in the current price change detection run. The system does not change existing price change events in this way.

  8. The system cleans the demand history of the updated price change events and executes detection of non-price change events.