Measuring Daily Cycle Counting Inventory Accuracy

[Chessie]

Measuring Inventory Accuracy

This may not be the correct place to raise this--sorry if it's not. A question was raised today about how to measure the accuracy from our daily cycle counting. One school of thought is that the measure should be based on total count less total of error count and divide this by the total count. For example:

Item 1, BOH shows 12, Count shows 10 (error = 2)
Item 2, BOH shows 10, Count shows 10 (error = 0)
Item 3, BOH shows 9, Count shows 10 (error = 1)

(30-3) / 30 = 90% Accuracy

To me this masks the fact that 67% of the items counted showed inaccurate quantities. I'm not knowledgeable about what is considered standard practice in this area but have been asked to find out if there is a standard practice and what it is. Does anyone know where I find this out? The financial folks want to know! I just want the measurement to help drive improvement and hope there is a standard procedure that will help me out.

 

[Wes Bucey]

This may not be the correct place to raise this--sorry if it's not. A question was raised today about how to measure the accuracy from our daily cycle counting. One school of thought is that the measure should be based on total count less total of error count and divide this by the total count. For example:

Item 1, BOH shows 12, Count shows 10 (error = 2)
Item 2, BOH shows 10, Count shows 10 (error = 0)
Item 3, BOH shows 9, Count shows 10 (error = 1)

(30-3) / 30 = 90% Accuracy

To me this masks the fact that 67% of the items counted showed inaccurate quantities. I'm not knowledgeable about what is considered standard practice in this area but have been asked to find out if there is a standard practice and what it is. Does anyone know where I find this out? The financial folks want to know! I just want the measurement to help drive improvement and hope there is a standard procedure that will help me out.

Interesting question on statistical usage and protocol. Many of our resident statistics mavens will address that issue.

I'm more interested in your plans for exploring Root Cause of the "why" of the discrepancies. Which count is accurate? How do you determine that? Is there a discrepancy in count or is there "inventory shrinkage?" Can units be lost in transit? Is there a difference in Who does the count?

Seems to me there is an opportunity for improvement regardless of whether the reported statistic is 67% or 90%. At the end of a day, week, or month, do the discrepancies even out or is there a net trend toward over or under count? Do you really have a handle on what variability you intend to examine?

 

[Chessie]

Interesting question on statistical usage and protocol. Many of our resident statistics mavens will address that issue.

Didn't get much of a response, but I did some research and came up with enough to make a recommendation.

I'm more interested in your plans for exploring Root Cause of the "why" of the discrepancies....
Seems to me there is an opportunity for improvement regardless of whether the reported statistic is 67% or 90%.

To give you a little more background, this is a mature cycle count process that does include a process for identifying and trending root causes. Since by our current accuracy measurement we in reality always exceed 99% inventory accuracy, taking action on the trends tends to be a low priority. I'd like to get it a bit more focus.