When you create an inventory sampling, you must define in which area (Inventory Management or Warehouse Management) and according to which Inventory Sampling Profile the inventory is to be carried out. The profile determines on the one hand the currency in which the calculations are carried out; on the other hand, the default values for the various parameters of the inventory sampling.
Note
In inventory sampling, an attempt is made to minimize the extrapolation error. For this purpose, various parameters from the inventory sampling profile are required.
To create inventory sampling profiles, go to Customizing for Inventory Management and Physical Inventory under Inventory→Inventory Sampling→Create Inventory Sampling Profiles (OMCK).
After creating an inventory sampling by a separate document, the stock management units are divided step by step. After that, the samples are drawn and a batch input session is created in the process.
When the batch input session is run, the system creates the physical inventory documents. After you enter and post the count results, extrapolations can be created in the system and checked for recount.
The selection of the stock management units to be physically counted in an inventory sampling is carried out in several steps. (The individual steps are presented in more detail in the following sections).
- Selection of the stock levels
- Formation of the stock population and division of the stock management units of the stock population into value classes
- Formation of the strata
- Random selection of the stock management units to be inventoried from the strata
Form the Stock Population from Stock Management Levels
To define which stock management levels can participate in an inventory sampling, go to Customizing for Physical Inventory under Inventory Sampling→Define Stock Management Levels (OMCL).
The stock management levels, defined in Customizing, that actually participate in an inventory sampling are determined in the specific inventory sampling (document). All stock management units of the selected stock management levels participate in this sampling.
Caution
Stock management units from the selected stock management levels that are already part of an active inventory are excluded from the inventory sampling.
Stock Population
As sampling stock management units with high values unnecessarily increases the error, an upper value limit is set. Sampling is only performed on management units that are below the threshold value. The threshold value can be specified in either absolute value or relative to the stock population.
You can set five parameters that control whether or not a stock management unit is assigned to the complete-count area. Default values for these parameters are defined in the inventory sampling profile:
- A deletion flag has been set in the material master record for the stock management unit.
- The book inventory of the stock management unit is zero.
- The stock management unit shows a particular ABC indicator.
- The price of the material lies above a particular limit.
- The value of the stock management unit (material price x quantity) lies above a particular limit.
When forming the stock population, the system attempts to divide the stock management units of the sampling area into 1000 classes according to their value. Later in the process, this prevents a large bandwidth of stock values in the sampling. A large bandwidth causes a higher error rate than sampling from a set of stock management units with approximately the same value.
The system calculates the class interval from the set upper value limit. The class interval = upper value limit /1000, rounded up to the next whole number.
Classes are designated by using the class interval until the upper threshold limit is reached. The process of rounding up generally results in fewer than 1000 classes. The last class ends with the upper value limit.
Stratification
To calculate the stratification by the system, the following information is necessary:
- The number of elements included in a class
- The number of strata to be formed
Hint
For a detailed description of the Dalenius-Hodges calculation method, see the SAP Library under Inventory Sampling.
You can display all strata or just the optimal strata for the inventory sampling.
The number of samples to be pulled for counting is an important parameter for inventory sampling. For accurate prediction of the total value of the stratum using extrapolation, the sample quantities must be as large as possible. However, to decrease the inventory counting effort, a smaller number of samples is advantageous. You must determine the minimum sample quantity and the stratification parameters together with an external auditor.
Sample Selection
In random selection, the elements of the strata to be counted are chosen at random for each stratum. Random selection takes place with random numbers that are created in the system with an internal random number generator. This random number generator works in accordance with the linear congruency method.
When you conduct the random selection, the transactions carried out up to that point are binding. The stock levels, stock population, and stratification cannot be changed after that point. However, random selection does not automatically create inventory documents. Instead, a batch input session is generated with the stock management units to be counted. So, you still have the possibility to decide whether the inventory should be performed or not.
When the batch input session is processed, the physical inventory documents are created for all stock management units selected from the sampling area and for all stock management units within the complete-count area.
The physical inventory of these elements includes the usual three phases:
- Creation of physical inventory documents
- Inventory counting and entry of count results in the system
- Analyzing and posting differences
Update
During an update, the system performs the following actions:
- Reads extra posted count results
- Determines changes to book inventories and values
- Determines any necessary subsequent random selections
If you enter a count or post a difference for an inventory document of an inventory sampling, the entry and posting is retained in the physical inventory document. The system does not automatically update the information in the inventory sampling. This only happens during the update. Values for stock management units that have already been counted and for which differences have been posted are no longer updated.
A posting block for stock management units to be counted is put in place if planned in the inventory documents. It exists only between the creation of the inventory document and the entry of the physical inventory count. Stock postings are allowed before and after the posting block.
For stock management units that you do not have to count, postings can take place at any time during the entire inventory sampling without restriction. This may result in changes to the book inventories or book values during an inventory sampling. In this case, it is a permanent inventory sampling. The system determines the changes in the update and takes these into account in the extrapolation.
In permanent inventory sampling, there may be changes to the book inventory and book values in the course of inventory sampling. This can mean that a stock management unit belongs to a different stratum at the current time point than at the planned time point.
Switching between the sampling area and the complete-count area is also possible. If a stock management unit switches to the complete-count area, the unit must be counted. The system proposes a subsequent random selection for the complete-count area.
