Show TOC

Sales Data Enrichment Locate this document in the navigation structure

 

The purpose of sales data enrichment is to provide you with insights into current out-of-shelf situations, lost sales due to out-of-shelf situations, or durations of out-of-shelf situations. In addition, it enables you to get an overview of out-of-shelf numbers and out-of-shelf rates from the past.

Prerequisites

You have defined periods with sales deviations and assigned them to locations. For more information, see the Customizing for Data Enrichment under   Cross-Application Components   Demand Data Foundation  .

You have defined a location calendar and assigned it to the locations. This enable the generation of zero sales records. For more information, see the Customizing for Data Model under   Cross-Application Components   Demand Data Foundation  .

Features

You can define the following sequence of functions for sales data enrichment:

  • Initialize Worklist

    Selects the data you want to enrich from the Aggregated Sales Propagation DataStore object (DSO) and generates a worklist.

  • Pre-Cleansing (optional)

    Marks a data record in the worklist as a non-outlier for the given combination of location and date if the date is within a period with sales deviations and the period of sales deviations is assigned to the location. Records that are marked as non-outliers are ignored in the subsequent function (Outlier Detection).

  • Outlier Detection (optional)

    Marks a data record in the worklist as an outlier. This function is based on the interquartile range test that detects significant deviations depending on the threshold value for significance defined in Customizing.

    Note Note

    Outlier detection should be used only if a sufficient sales history is available.

    End of the note.

  • Generate Zero Sales

    Generates zero sales records in the worklist, if an expected sales record is missing and the location was open for business. If a certain number of days with zero sales is reached (defined in Customizing) the product is most likely not sold anymore and no new zero sale records are created.

  • Calculate Long-Term Average

    Determines the long-term average for every product-location combination by calculating the arithmetic mean. The calculation is based on the number of days defined as a threshold in Customizing.

  • Calculate Out-of-Shelf

    Determines whether zero sales records should be marked as out-of-shelf situations. Two different functions can be used — simple ratios and normal distribution. Both functions are based on the long-term average that was calculated in the previous step. The result of the function is compared to the corresponding threshold value defined in Customizing to decide if an out-of shelf situation is given. The threshold value controls how many out-of-shelf signals are generated.

  • Calculate Average Lost Sales

    Calculates the averages (quantity, gross value, and net value) for the identified sales records as average lost sales for records that are marked as out-of-shelf situations.

  • Get result

    Posts the enriched data records back so that they can be transferred to the DSO Aggregated Sales Propagation.