menu SIGN IN

Inventory Accuracy

Accuracy of the book inventory versus the counted inventory.



(|book inventory – counted inventory|)/book inventory


Most Advanced Planning Systems calculate net inventory requirements. If the book inventory used as the basis for these calculations has a high error, the net inventory requirements generated will not reflect the true inventory needs. The inventory error should be factored into the safety stock calculation to protect service levels from variance in inventory due to inventory count accuracy. Assertive continuous improvement programs should be in place to support a decrease in inventory count errors.

Unit: Percentage

Direction: Minimize

Benchmark Results

This KPI is most used for:
Operational Excellence


  1. Usik P.

    Dear: rking842000,
    How much TOD (tolerance of Discrepancy) in % ?
    Is it any standard ?

    Usik Pitono

  2. David Wendt

    I have worked in a number of manufacturing industries where 98% accuracy was standard to work from.


  3. chander jyoti

    can anybody help me out with better understanding by taking hypothetical example

  4. chander jyoti

    can anybody help me out with better understanding by taking hypothetical example?

  5. Neil Waterhouse

    Amongst the readers of this website, when counting stock errors, do you count an error as 1 or 2?

    Example if the system states stock is in location A01A but physically it is actually located in A01B, is this one error or 2 errors. (Dependant on the qty of course.

    For me to arrive at the true stock accuracy, I count as 2 errors.

  6. Ahmed Gharaibeh

    Dear Neil,
    Yes you had to consider it 2 errors.the formula it can be used for Qty or value, and you need to consider the absolute value of the difference whether the counted inventory more or less than the book.

  7. Cemre Nalbantoglu

    I believe that the best figure should be included in the “Benchmark Results”. This will encourage people to target higher points.

  8. Neil Waterhouse

    Thanks Ahmed and Cemre for your comments.

    Like you I totally agree with counting errors as 2 to arrive at a true benchmark on Inventory Accuracy. It’s amazing how many of Contract or Logistics Managers want to count errors as 1 to make their Stock Management skills look better than they actually are.

  9. Tony Johnson
    about 10 years ago by Anonymous

    How was it determined that the inventory items were recorded on the wrong book (location)? Whatever the means, the book should than be adjusted to show the correct location, or the item is sent to the location that it should have originally been shipped to.

    The importance/weight of the errors should depend on the cost, and or customer demand to correct the problem. For example, if the locations are close in proximity and shipping the item to the correct location is at marginal cost, then the error could be consider as 1 error. However, if the customer demand is high, and the available of this product affected customer demand, then the weight associated with the error should be high, and “2 errors” can be away with associating the appropriate cost impact of this error.

    There are many examples to use such as, if it is an overseas shipment and the transportations and equipment cost is high, as well as the quantity being substantial, then the weight associated with the equipment being at the wrong location may cost someone their job.

    Therefore, the KPI associate with inventory items being recorded on the incorrect book, and or location should be approached with a thorough understanding of the consequences the impact (error) will have on the company.

  10. Neil Waterhouse

    Thanks for your detailed explanation Tony. I totally agree with your comments and like to ensure that the benchmark is recorded and to measure the rate of improvement for Inventory Management

  11. yair S.

    I think that one should consider all adjustments to inventory before and after stock-take, otherwise, warehouse people can “arrange” the goods to get minimum gap in physical count, and A day after, they mismatch the inventory again.
    To avoid that and to get real accuracy of stock my suggestion for the KPI is: |Adjustments|/Count. Adjustments should include all of the adjustments done from the previous counts + the adjustments done in the current count.
    For those who use this KPI in manufacturing companies, calculate this KPI on Qty as well, compare Value accuracy Vs. Quantity accuracy and you will find out in what place you should focus.

  12. Neil Waterhouse

    Thanks for your comment Yair, I think consistent monthly Cycle Counts are the answer to stop the warehouse people trying to manipulate their stock results by “arranging” stock. With Inventory Management Best Practices, incorrect counting, incorrect arranging of stock etc etc are always found out either at the picking or putaway stages in the operation.

  13. john pena

    Cycle counting is a on-going process that validates your material management process. I conduct cycle counts on a daily basis in addition to reviewing pick or pulls from warehouse locations….Our goal is 99.5%. Our actual is 99.9%

  14. irfan ahmed

    Perhaps it depends on volume of transaction that enter daily. If higher need to measure it on daily basis for timely knowing accuracy. If volumes are lower then you can count them on any desired frequency that can be controlled. Tolerance should not be over 1%, i guess.

  15. Frank Hill

    Like John above we do daily cycle counts with the goal to count the entire warehouse once a month for two reasons 1) maintain bin accuracy as much as possible to avoid mis-picks and 2) to maintain a 99.8% to meet auditor requirements to for go a yearly PI.

  16. Mateusz Labecki

    To Frank Hill. I am interested in achiving 99.8% in that KPI but I have a few quastions to determine process differences.
    1. Do you manage goods flow using serial numbers ?
    2. How many SKU do you have ?
    3. How many bins/locations do you have ?
    4. How many pieces on average ?
    5. Are transactions booked in WMS online ?

  17. Frank Hill


    1.We don’t use serial numbers but Lot or Batch number which requires FIFO so oldest lots have to be picked first if it is not expired (if product does not have six months shelf life left it is expired).
    2.We have about 1200 which only about 825 are active.
    3.In rack storage a bin holds 2 48 inch high pallets stacked and we have 37,000 pallet positions, in the forward case pick area we have 825 bins one pallet per bin.
    4.Are smallest pick is a case we do not do eaches currently. In a single shift we pick 38,000 cases which is 70% of our volume the rest being full pallet picks.
    5.Yes all WMS transactions happen in real time thru RF scanners.

    Hope this helps, sorry for the late reply but it has been very busy lately.


  18. Ian Bruggemann

    In my distribution centre operation inventory accuracy (database vs Locator inventory units) was measured daily and all discrepancies had to be resolved within 24 hours to ensure that the replenishment team were working on accurate figures when allocating stock to retail stores.
    This daily routine/discipline ensured an accurate inventory at year end.
    In addition cycle counts were also done to further audit accuracy with fast moving and promotional SKU’s a priority.

  19. Ryan V.

    I am new to this site and was rather disturbed by the benchmark results of this KPI. I manage a small DC with about 6000 stock locations and just under 14000 SKUs. Over the past 6 years that I have managed it, our inventory accuracy averages 99.97%, our lowest being 99.301%. We do not cycle count, only a one-time year end invnetory, without shutting down mind you. I couldn’t imagine operating with accuracies in the 50’s and 70’s.
    I do have a question for the regular cycle counter. Do you account for all the adjustments you make throughout the year or is your accuracy based on your actual cycle counts?

  20. Dominique

    Can someone assist. We are on the gross margin method with one cycle count a year over 800 locations, and over 80,000 sku’s. What KPI’s you recommend we use. There are no systems therefore we do not have a perpetual count… I only know that we have at year’s end :(

  21. Dominique

    Can someone assist. We are on the gross margin method with one cycle count a year over 800 locations, and over 80,000 sku’s. What KPI’s you recommend we use. There are no systems therefore we do not have a perpetual count… I only know that we have at year’s end :(

Sign in to KPI Library

Please sign in or join for free to view the complete KPI Library.

(forgot password?)
Remember me on this computer

Not yet a member? Join Now!