Inventory Accuracy

Notebook-as-a-book illustration

By Joannes Vermorel, February 2013

The notion of inventory accuracy refers to all the discrepancies that exist between electronic records that represent the inventory and the physical state of the inventory. On of the most common form of inventory inaccuracy is phantom inventory. Such discrepancies can result in lower service levels, along with broader accounting issues and financial losses.

Measuring the stock on hand accuracy

The reality of inventory can diverge from its electronic representation in many ways:

  • records of nonexistent operations,
  • operations not accounted for,
  • mismatch of product codes,
  • erroneous quantities,
  • ...

While the root causes are diverse, those problems typically end-up generating stock on hand counts that diverge from their physical counterparts. Simply put: the recorded stock on hand does not match the amount of goods available at the storage location.

The accuracy of the stocks on hand can be measured with a variety of metrics classically associated to forecasting problems. Indeed, the electronic value hold for any stock on hand can be seen as an estimation of the reality.

Thus, one of the most simple metric to measure inventory accuracy from this viewpoint is the mean absolute error (MAE) with MAE = | Qe - Qr | where Qe is the electronic record and Qr the real quantity.

Then, when inventory records are fairly accurate, the hit rate metric, that is, the percentage of inventory records that happen to be correct, can also be used. In case of large inaccuracies however, this metric has the downside of not taking into account the size of each discrepancy.

Extent of the problem in retail

The problem of inaccurate inventory records is typically much stronger at the store level compared to the warehouse level. DeHoratius and Raman (2004) have carried a substantial investigation of the accuracy of inventory over 37 stores in the USA.

Although computerized tracking of inventory at the stock keeping unit (SKU) level is commonly assumed to be accurate, we found discrepancies in 65% of the nearly 370,000 inventory records we gathered from multiple stores of a leading retail chain. DeHoratius and Raman (2004)

15% percent of Gamma’s inventory records, nearly 55,000 of them, had an absolute error of eight units or greater, more than half the average target quantity of inventory maintained on the shelf for that SKU in a given store. In aggregate, the value of the inventory reflected by these inaccurate records amounted to 28% of the total value of the expected onhand inventory. DeHoratius and Raman (2004)

While significant progresses have been obtained at the warehouse level, inaccurate inventory records remains a widespread problem in retail at the store level.

Lokad's gotcha

Maintaining accurate inventory records is a requirement to achieve a satisfying level of inventory optimization. What is not measured cannot be optimized; and when a measurement is made, it has to be accurate. Yet, our experience indicates that most retailers suffer from poor inventory accuracy in stores. RFID represents a potential long term solution to this situation, however even RFID cannot tell if a product is damaged, misplaced or inaccessible.


DeHoratius, N. and A. Raman (2004). “Inventory Record Inaccuracy: An Empirical Analysis.” University of Chicago Graduate School of Business Working Paper.