Which metric is better for Cost of Poor Quality (COPQ)?

M

MGMTREP

As a consistent means to measure one aspect of the Cost of Poor Quality (COPQ), we count how many Nonconformance Reports are written on incoming product and multiply this number by a calculated dollar figure. The belief is that, for this aspect only, COPQ rises and falls with the number of NCR's written.

Another different method now being suggested is to not just count the number of NCR's, but also count how many purchase order line items (i.e separate product numbers) were issued. The thought is that it's more meaningful to know the percentage of line items that were written up. This means that even if were writing up more lines items, if it's a lower percentage of the total number of items ordered, it's good news.

The counterpoint is that this moves the focus to the wrong factor. While it's good that the percentage is going down, the costs incurred by having to write up more items is going up and that means the COPQ for this aspect is going up. To reduce this COPQ, the focus should be on reducing the number of NCR's that have to be initiated by eliminating the root cause of the NC's. Doing this will take care of the percentage.

Thoughts?
 
Last edited by a moderator:

Bev D

Heretical Statistician
Leader
Super Moderator
Re: Which is metric is better?

BOTH metrics are important as descriptors: total cost and cost as a rate. why not track both? I track both - of course it's automated so we can focus our efforts on improvements.

On the other hand I always get my radar pinging when questions arise about the metric rather than the work done to improve the metric...just sayin.
 

Jim Wynne

Leader
Admin
Re: Which is metric is better?

As a consistent means to measure one aspect of the Cost of Poor Quality (COPQ), we count how many Nonconformance Reports are written in incoming product and multiply this number a calculated dollar figure. The beleif is that, for this aspect only, COPQ rises and falls with the number of NCR's written.

Another different method now being suggested is to not just count the number of NCR's, but also count how many purchase order line items (i.e separate product numbers) were issued. The thought is that it's more meaningful to know the percentage of line items that were written up. This means that even if were writing up more stuff, if it's a lower percentage of the total number of items ordered, it's good news.

The counterpoint is that this moves the focus to the wrong factor. While it's good that the percentage is going down, the costs incurred by having to write up more items is going up and that means the COPQ for this aspect is going up. To redcue this COPQ, the focus should be on reducing the number of NCR's that have to initiated by defining the root cause of the NC's. Doing this will take care of the percentage.

Thoughts?

Welcome to the Cove. :)

One thing you need to resign yourself to, if you haven't already, is that in the end you're going to be dealing with relative numbers, not absolute ones at least as far as dollar values are concerned. Spending too much time trying to squeeze precision out of the numbers usually isnt' helpful.

Another thing you should look at, if you haven't already, is whether or not all of those NCRs actually need to be written. Often the problem is with the specifications and not the product.

In the end, the effort has to be put towards identifying as well as possible how much things cost and prioritizing the results. I agree that counting PO line items doesn't seem like a particularly useful approach.
 
M

MGMTREP

Re: Which is metric is better?

Thanks Bev D - you're the first Heretical Statistician I've encountered! An automated process would be nice to have, but for now, we're stuck with the two approaches I cited. The question wasn't intended to convey that the metric is the only focus - it just happens to be the one under discussion. I believe both sides are only concerned with the method only so far as the data gathered will help lead to a signiifcant reduction in NC product, and not as end itself.
 
Last edited by a moderator:
M

MGMTREP

Re: Which is metric is better?

Thanks Jim. I completely agree with your points. If it wasn't already obvious in the way I characterized the sides, I tend to think that tying the metric to line items clouds the issue by drawing attention away from the fact that the COPQ is increasing. Mind you, I like clouds - just not in my data.
 

qusys

Trusted Information Resource
As a consistent means to measure one aspect of the Cost of Poor Quality (COPQ), we count how many Nonconformance Reports are written on incoming product and multiply this number by a calculated dollar figure. The belief is that, for this aspect only, COPQ rises and falls with the number of NCR's written.

Another different method now being suggested is to not just count the number of NCR's, but also count how many purchase order line items (i.e separate product numbers) were issued. The thought is that it's more meaningful to know the percentage of line items that were written up. This means that even if were writing up more lines items, if it's a lower percentage of the total number of items ordered, it's good news.

The counterpoint is that this moves the focus to the wrong factor. While it's good that the percentage is going down, the costs incurred by having to write up more items is going up and that means the COPQ for this aspect is going up. To reduce this COPQ, the focus should be on reducing the number of NCR's that have to be initiated by eliminating the root cause of the NC's. Doing this will take care of the percentage.

Thoughts?

I know that there are lots of models for COPQ. One of this these is P.A. F. model where P stands for Prevention, A for Appraisal, F for failure ( internal and external).
In the technical literature there ar emany books on this topic as well on internet are available interesting suggestions.
At the beginning you could focus on external and internal failura such as return from your customer, internal scrap rework and misprocesses and excursion in your production line line, trying to give a value in money for loses.
The suggestion is to try to identify an order of greatness and not to be precise up to the cent , It is important to track the metrics over the time setting goals for the organization.:bigwave:
 
K

kgott

As a consistent means to measure one aspect of the Cost of Poor Quality (COPQ), we count how many Nonconformance Reports are written on incoming product and multiply this number by a calculated dollar figure. The belief is that, for this aspect only, COPQ rises and falls with the number of NCR's written.

Another different method now being suggested is to not just count the number of NCR's, but also count how many purchase order line items (i.e separate product numbers) were issued. The thought is that it's more meaningful to know the percentage of line items that were written up. This means that even if were writing up more lines items, if it's a lower percentage of the total number of items ordered, it's good news.

The counterpoint is that this moves the focus to the wrong factor. While it's good that the percentage is going down, the costs incurred by having to write up more items is going up and that means the COPQ for this aspect is going up. To reduce this COPQ, the focus should be on reducing the number of NCR's that have to be initiated by eliminating the root cause of the NC's. Doing this will take care of the percentage.

Thoughts?

Do a search in the attachments list for Patricia Ravanello's work. She has provided some excellent materials on this sort of thing.
 
M

MGMTREP

Thank you, I'll check it out.

Perhaps a clarification is in order. While the responses confirm many aspects of COPQ sources, technical details and the like, my question is really very specific.

Considering only the aspect of the COPQ attributable to writing nonconforming product reports (NCR's) on product from suppliers and using a known, consistent dollar cost associated with writing and processing the NCR's, which is more meaningful in terms of a metric of this aspect of COPQ for, say, a 3 month period:

1. How many NCR's are written?
2. How may NCR's are written as a percentage of the total number of line items ordered?

For example, if 50 NCR's were written in the first 3 month period and it cost $10 each time, the COPQ, again for this aspect only, would be $500.

If there were 100 line items, that would be 50% of the total lines items. The $500 cost would be the same.

The next 3 month period 75 NCR's are written on 300 line items. The COPQ would increase by $250, but the percentage would "improve" to 25%.

In my mind, since COPQ got worse using the percentage is a waste of time.

Agree? Disagree?

Thanks everyone!
 
Last edited by a moderator:

Jim Wynne

Leader
Admin
Thank you, I'll check it out.

Perhaps a clarification is in order. While the responses confirm many aspects of COPQ sources, technical details and the like, my question is really very specific.

Considering only the aspect of the COPQ attributable to writing nonconforming product reports (NCR's) on product from suppliers and using a known, consistent dollar cost associated with writing and processing the NCR's, which is more meaningful in terms of a metric of this aspect of COPQ for, say, a 3 month period:

1. How many NCR's are written?
2. How may NCR's are written as a precentge of the total number of line items ordered?

For example, if 50 NCR's were written in the first 3 month period and it cost $10 each time, the COPQ, again for this aspect only, would be $500.

If there were 100 line items, that would be 50% of the total lines items. The $500 cost would be the same.

The next 3 month period 75 NCR's are written on 300 line items. The COPQ would increase by $250, but the percentage would "improve" to 25%.

In my mind, since COPQ got worse using the percentage is a waste of time.

Agree? Disagree?

Thanks everyone!

There might be a good reason to count line items and number of NCRs, but because you've established dollar values for NCRs, that would be the way to go as far as COPQ is concerned. A purchasing manager might be interested in PO line items, but that's a different metric.
 
Top Bottom