All types and sizes of companies still spend a great deal of time and effort carrying out their inventory manually and traditionally at the end of the year in the form of a key date inventory: Printed lists are filled manually from shelf to shelf with check marks. The use of so-called handhelds, such as RFID scanners, does not prevent people from still carrying out the counts. This recording costs companies between two and ten euros per item. A warehouse with 10,000 items therefore quickly leads to inventory costs of 20,000 euros per year. Compensating for counting errors also takes up additional time. This is because warehouse staff are often unable to carry out a full inventory count on their own. Student assistants or colleagues from other departments are recruited for the inventory period. However, the fact that so many people are involved means that errors are multiplied.

An initial attempt to fulfill the inventory obligation more efficiently was the 'permanent inventory'. With this method, the count of all items is spread over the entire year. Although this no longer results in additional costs for personnel, a closer look reveals that the effort is merely distributed and is ultimately just as high as with a full inventory. There is therefore no real reduction in costs.

However, the legally compliant inventory sampling makes it possible to carry out the inventory almost completely IT-supported. Purely manual counting by people is no longer necessary. Only a few items still need to be recorded by hand. Algorithms thus make work easier, far from any worrying scenario. This is because this modern process shifts the tasks of those responsible for stocktaking in the company from counting to stocktaking and inventory management.

Procedure and benefits

The inventory manager now only supplies the stock data to the software for an inventory sampling. This includes the material number, the current stock quantity and, as a rule, the unit price. To clearly identify an item to be assigned in the warehouse, it can also be helpful to enter warehouse information, such as coordinates or warehouse numbers, and other identification features, such as container or batch numbers. The algorithm uses this data to calculate how many samples from the warehouse still need to be counted in order to extrapolate an overall result. This can often reduce the inventory time by up to 90 %, and the count quantity by up to 95 %, depending on the warehouse.

The procedure is similar to forecasting election results in politics: it is possible to present a provisional election result shortly after the polls close. Only a small number of representative votes are analyzed and extrapolated very reliably using mathematical and statistical methods. If someone wanted to count a jar full of peas according to the principle of inventory sampling, the result would be as follows: Instead of taking out one pea at a time, he could determine the entire contents of the jar using a few representative peas. First he would weigh this group of peas, then the full jar and finally the empty jar. With this information, an algorithm can calculate the exact number of peas.

The Pareto principle

Inventory sampling can significantly reduce the tedious counting process in warehouses.

© Shutterstock, Marcin Balcerzak

The so-called Pareto principle is also important for inventory sampling: in a normal warehouse, the majority of the value of goods is accounted for by a manageable proportion of all items. The complete count of these highest-value items provides a very reliable and stable basis for determining the less valuable items. A system for inventory sampling automatically divides the rest of the items into value strata, from which it then draws the samples. Using the so-called difference estimation, the software then determines the extrapolated accounting deviation and the maximum relative error. The procedure is not only suitable for the inventory on the reporting date, but also for a permanent inventory.

Inventory sampling drastically reduces the time and effort required for counting. This means that warehouse closures and production stops can be avoided. This saves costs and minimizes counting errors. It also gives users a quick yet reliable overview of the quality of their inventory management. The result is at least equivalent and often even better than that of a full inventory. In the end, the auditor is also satisfied.

Inventory sampling: suitable for which companies?

As software for inventory sampling, 'Invent Xpert' divides all items into meaningful value strata. The counting of representative samples from each layer minimizes the counting effort.

© Inform

The inventory sampling procedure is suitable for companies of any size and in any industry. The specific requirements that must be met in order to use the appropriate software are as follows:

• An IT-supported, reliable inventory management system must be implemented in the company (ERP or warehouse management system).
• The warehouse must contain at least 1000 items.
• The distribution of items in the warehouse must correspond to the 'Pareto principle' (20% of the items make up around 80% of the warehouse value).

With these requirements, it is possible to carry out item counts much more quickly. Companies thus become more flexible and can use freed-up capacity for more meaningful tasks. Inventory sampling systems such as 'Invent Xpert' from Inform are based on the existing ERP system and obtain all the necessary data via secure, standardized interfaces that function either in text form with a fixed record format or as an Excel file with a variable column layout. This allows a wide variety of external systems to be connected to the inventory sampling software. At the same time, a fixed text format offers maximum revision security, as the data can be defined at a point in time and its integrity can then be easily checked.

Once the interface file has been created and transferred to 'Invent Xpert', the totalized inventory values, inventory quantities and the number of transferred records can be compared via a flow log. The interface data is handled via the local file system in conjunction with the web interface. Thanks to the browser-based use, it is not necessary to install the software on site. This saves additional work and costs in the IT department. The software also prepares the inventory result after the count has been completed so that those responsible for the inventory can forward the report to the auditor simply and easily at the touch of a button.

Software is not a danger

A healthy skepticism regarding the automation of human work is not unjustified in some cases. However, in many cases, IT-supported solutions merely relieve people of routine tasks. This was already the case in 1977 when inventory sampling was established as a legally recognized procedure. Today, inventory sampling is undoubtedly the most efficient method for annual stocktaking. More than 40 years after it was enshrined in the German Commercial Code, it is web-based and therefore quick and cost-effective to use. The possibility of drastically reducing the manual counting effort and the associated costs through intelligent algorithms proves the relevance of this topic in rapidly changing times.