Skip to Content

Efficiency vs. Accuracy – Treatment of Store Research Counts for Replenishment


SAP is a close ended solution that supports transactions to be accurately, completely and fully reconcilable.  This is a very good thing. Retailers businesses across tens of sectors love SAP for this. This of course pre-supposes that the business processes are always robust and all information is gathered completely and accurately.

There are always forces at play that make this an almost no-win task. Some of these forces include:

  • Retailers big or small almost always process transactions that are order of magnitude more  than any of their size peers in other industries
  • Many of these transactions happen in a decentralized environment in stores and DCs many a time by associates of varying degree of skills and experience. One
  • Loss of units in stores and DCs euphemistically called shrinkage, when someone decides to walk with merchandise and not pay. Of course no transaction is created

There are many more examples of why retail business stretches the boundaries of accuracy more than other businesses.

Under this backdrop, I would like to present some of the challenges that retailers encounter and how best they can be addressed within the folds of SAP solution. In this blog, I will cover the replenishment situation for stores.

Treatment of Store Research Counts for Replenishment

Retailers have always found that an accurate count of what is in the store is critical for replenishment planning. Howsoever sophisticated the planning engine maybe, it is dependent on accurate information of units on hand.  Given some of the distorting forces that I mentioned above, the book count often does not represent the true availability in the store. Store associates make a “research count” to establish actual availability in the store.

Even though research counts are often good to trigger replenishment, they are frowned upon by inventory accountants, because they lack the rigor of a true physical inventory.  Research counts are known to be used by store managers instigate replenishment, when they know head-office would under-replenish based on the existing count. Further under-count of units at hand often happens because associates miss units wrongly stacked or laying in the backroom.

Retailers would put software filters to weed out errant research counts, but these tools go only so far. Many erroneous research counts make their way in.  Robust wall-to-wall annual physical count confirms the fragility of the research count process.

Two possible options on how to treat the difference arising due to the research count:

  • Accounted as inventory transactions and accounted for both in unit and dollar terms
  • Accounted for as unit only transaction and not accounted for change of cost and retail dollars

Retailers not using SAP often have separate unit system for replenishment and a separate system for stock accounting. The distortion between the two systems is known but largely ignored. Though these differences are known at an aggregate level, it is often not feasible to find out where significant differences arise.

Retailers using SAP and running replenishment based upon on-hand units in SAP have to find an answer to the question.

In my view, the first option, where the transaction is brought into SAP as an inventory loss or gain, thereby affecting both the units and dollars would often be a bad choice. Given the inadequate rigor on research counts, these counts will often not pass the test as a true accounting transaction. It is also known to cause a seesaw affect on the inventory valuation, especially if these research counts make the inventory go negative.

The second option, though in the face of it seems counter-intuitive, can be implemented in SAP by moving the difference to a separate “not-available” storage location that is not visible to the replenishment system.  Because the inventory valuation happens at the site level, the research counts do not affect the inventory valuation.

At the time of annual physical inventory, the units in the not-available second storage location are made zero and all units are moved to the primary storage location.

The second option also provides visibility to not-available stock both in units and dollar terms. High value or important items can be researched further with the help of stores and corrected. Shrink reserves can be directionally vetted against the dollar value in the not-available bucket.

1 Comment
You must be Logged on to comment or reply to a post.
  • Your points on the inventory accuracy and research counts are excellent.  With the economy as it is the value of tight inventory control and true inventory accuracy is very, very valuable.

    Thanks for the knowledge share.