Hi All,

I am doing an exercise for marketing. They feel that our Customer Complaints (CCs) have increased over the past two years and they are correct....but.... our dispatches have also increased (at a greater rate). Our CC's are running at approx 2% (2000 dispatches campared to 40 CCs).

What are your thoughts on this ?

I know you are going to ask what is the increase in orders vs CCs has been over thees years versus previous years, but I really want to know what an acceptable level is? (I know 0% would be good)

Greg B

Hi. I don't know if this counts, but are the nature of the complaints similar? What I mean is, are your dispatches correcting the same nature of complaints because if you do, then maybe you need to make corrective and preventive actions. I know this does not answer your question really, but I just got curious. I have this idea that CCs may be used as gauge in determining if we satisfied the customers or not; and if we didn't then we may need to correct something...

Best regards,

Naomi

encee98,

Yes we are correcting the issues and finding the root causes etc. The issues are not related and are across our entire customer range. The question really is, is 2% aceptable?

Greg B

The question really is, is 2% aceptable?
I'd say that you need more than one measurable here. Using only that one may lead you down the wrong alley.

From what you're saying I take it that your cc/dispatch ratio is going down. A positive trend can't be bad, and it would seem that your actions are effective.

On the other hand the number of CC's are going up, and every one represents a dissatisfied customer who couldn't care less if the above ratio is good. The number of complaints should be considered.

What about the financial impact? How much does the CC's affect your profit?

/Claes

Claes,

I agree with what you are saying and I do not usually crunch these type of numbers because they can be and usually are misleading (IMO).

Marketing think that there is a drastic increase in CCs. I say that there are also but other factors have to be taken into account such as the increased output. We have ramped up output by about 40% and the CCs have in actual fact gone down, as a percentage. Marketing just see the total number (morons) and I want them to see the whole picture which I wil give them on Friday but is 2% a good average? I have not seen a national or international study on the average complaints per company. I know this would be difficult as people see things differently but I thought I would plot it anyway.

I'll show the figures that show we are actually trending down but what I want from them in the future is a Risk analysis of how the CC affects the company both from an immediate dollar value and a risk to the customer and us (future purchases, next round cost increases etc).

I have been after these for years and I am still toget them.

Greg B

Just a word of caution here -
Are you 100% sure that you are capturing all customer complaints, or just those received formally in writing?
Our experience is that formal complaints represent only the tip of the iceberg. Customers usually make allowances for the odd supplier mistake, or don't even notice it. They only make a formal complaint when they suffer real inconvenience, or believe they can use it to lever down the price.
For example - we get our hauliers to report their actual delivery time at the customer. Using this data and counting how many complaints we get for late delivery, we find on average only 5% of lates result in a formal complaint. We therefore started logging all lates as nonconformances, and also capture concerns. The first result is a dramatic increase in 'complaints' - but from then on you can begin to make real improvements, with a real impact on customer satisfaction.
Incidently, our target for formal complaints is 1.5% of deliveries.

I heard that in FDA audits the first action is to check the number of complaints and compare this to a reference table, if you have to few then they start digging.
Maybe some one from this field can give some input.

This is similar to internal PPM levels, the numbers to compare just don't seem to be there.
I once asked about this in the field of hybrids and got no answer.

I think that people don't want to give numbers before they now an average and no one has an average cause no one gives numbers.

I hate to say this, Greg, but maybe there is no accepted number for Customer Complaints. We track complaints/ton for each of our four complaint types. Unfortunately, we can do nothing about the two types that are attributed to Sales...so we focus our energies on the remaining two types that are our responsiblity to address. We've noticed that here in the mill, we're getting better...wish we could say the same for Sales. The work we do is offset by what they don't do. :mad:

Obviously, the number of Customer Complaints you want is zero...but I remember on a Cost of Quality course I took a few years back, we discussed External Failure Costs. External Failures included Allowances, Warranty Charges, Returned Goods and Customer Complaints and the "accepted ratio" for External Failures for your Cost of (un)Quality is 20%-40%. (my brain is still having a hard time accepting that)

Just my 2 cents worth here.

We base our ratio on CCRs/shipments and are typically below 1%. In regard to tracking numbers of CCRs, there are a number of pit-falls you need to watch out for. The "new" zero defects requirement has triggered a large number of CCRs which would, in the past, have been overlooked by the customer. Everybody in the automotive industry is scared to death they will send a defect to the OEM and be placed on a penalty containment. We are getting complaints on perceived problems that aren't even a requirement. A good example of what I mean here is - we put a sealer under the head of a fastener and a customer finds a part with a spot that isn't coated. The requirement is to coat 90% and the spot is well within acceptable limits. The complaint is now on the non-requirement that the customer doesn't like the look of this part and we need to stop leaving spots. My only point here is that the perception of a lot of customers has changed due to stress driven down the chain.

For the other side of the equation I have to go back to my grandfather who always said "be very afraid if you don't get complaints from your customer". If you dig a bit into what he meant you find a bit of logic. His theory was that the majority of "complaints" were not really more than contacts from your customer letting you know of an issue that has come up. They aren't saying they will stop doing business with you, they are letting you know they found an issue and you should fix it. If a customer called and said he found a flywheel in his bin of parts it isn't because he wants an 8-D, it is because he is letting you know he found it and maybe you would like to fix it. He doesn't assume he will get future flywheels. If you can accept that every complaint is not necessarily a complaint but a communication of a working partner, you had better pray your partner doesn't give up on the communication. The day he stops letting you know the little issues is probably the day he decided to buy from somebody else.

I guess what I'm saying is that a numerical analysis of complaints doesn't mean much without a real in-depth look at all the other things surrounding them. Sometimes they are up, sometimes down. Anybody ever notice the spike in minor complaints when business is slow? There is a definate drop in them when business is booming and everybody is busy.

Dave

I heard that in FDA audits the first action is to check the number of complaints and compare this to a reference table, if you have to few then they start digging.
Maybe some one from this field can give some input.

This is similar to internal PPM levels, the numbers to compare just don't seem to be there.
I once asked about this in the field of hybrids and got no answer.

I think that people don't want to give numbers before they now an average and no one has an average cause no one gives numbers.Howard: Yes, and that is why I have so many problems with the basic concept of six sigma and errors expressed as parts per million. This thread, to me, is similar in that respect: a percentage of errors versus what? Per million what? If there is a list of 3 parameters each item must meet, then are there only 3 chances for an error on that particular part? How do we count being (or not being) late on delivery? Does that mean we really have 4 parameters? Error in packaging (but not spec'ed) counts, or not?

We can answer these questions internally, for our own particular situation, and use those metrics to see if we are in statistical control, or improving (or not). But, unless the parts (of the million) and the errors definitions are the same, trying to directly compare one situation (company) with another is problematic, to say the least.

Your company's "error" may be my company's "opportunity for improvement". Same problem, different accounting.

As long as it works for us, so what?

JMHO

Craig

encee98,

Yes we are correcting the issues and finding the root causes etc. The issues are not related and are across our entire customer range. The question really is, is 2% aceptable?

Greg B

I guess my $0.02 is would you tell your customers that a 2% rate for CC is acceptable? I might be over simplifying this - but what do your customers expect from you. (Reality is different than expectations - we all know that. But what would they tell you if you told them that you accept 2% CC?)

I understand and agree with what most of you are saying but IMO this percentage could be diffrent for each one of us. wouldn't the type of industry we are in increase or decrease this number. in other words we are in corrugated packaging and our percentage on C.C. is 1.44%. This is within an acceptable range ( Although of course we would like to get it lower ). Would this percentage be acceptable for someone who may have diffrent customer requirements?

Howard: Yes, and that is why I have so many problems with the basic concept of six sigma and errors expressed as parts per million. This thread, to me, is similar in that respect: a percentage of errors versus what? Per million what? If there is a list of 3 parameters each item must meet, then are there only 3 chances for an error on that particular part? How do we count being (or not being) late on delivery? Does that mean we really have 4 parameters? Error in packaging (but not spec'ed) counts, or not?

We can answer these questions internally, for our own particular situation, and use those metrics to see if we are in statistical control, or improving (or not). But, unless the parts (of the million) and the errors definitions are the same, trying to directly compare one situation (company) with another is problematic, to say the least.

Your company's "error" may be my company's "opportunity for improvement". Same problem, different accounting.

As long as it works for us, so what?

JMHO

Craig

I guess ideally you would measure in errors per opportunity for error. The difficulty is then in determining how many opportunities for error you have. Is this the same as the number of measured items? Or is it more complex than that.

Just something to ponder.

I guess ideally you would measure in errors per opportunity for error. The difficulty is then in determining how many opportunities for error you have. Is this the same as the number of measured items? Or is it more complex than that.

Just something to ponder.

Welcome to the Cove, gerrybean! :bigwave:

Yes, and that is why I have so many problems with the basic concept of six sigma and errors expressed as parts per million..Welcome to the Cove!I guess ideally you would measure in errors per opportunity for error. The difficulty is then in determining how many opportunities for error you have. Is this the same as the number of measured items? Or is it more complex than that.

Just something to ponder.Oppurtunity as far as I understand it

Plastic part with 500 dimensions
Made 2000 with one dimension out
2000*500 = 1,000,000 opportunities
500 ppm

but to non statisticans the part is no good. 100% failure.

Who is right?

I guess ideally you would measure in errors per opportunity for error. The difficulty is then in determining how many opportunities for error you have. Is this the same as the number of measured items? Or is it more complex than that.

Just something to ponder.

Gerry

Well, each item has things we measure, and each measurement could be counted as an opportunity for an error. The problem I see is the chances for error that are not measured, but the customer would still have issues with. Say an order has 15 different items, in differing quantities. Do we count the accuracy of filling the order (part number and quantity, say) as one chance, or 30 chances? So you are exactly right, "the difficulty is then in determining how many opportunities for error you have."

But, to my thinking, there is no one "right" answer, except that within an organization decisions must be made concerning how questions like this are answered, and the resulting decision becomes the "rule" within the entire organization. Without this groundwork, it seems to me, the result would be confusion.

That said, I am no where close to being well versed in these concepts.

Welcome to the Cove, gerrybean, and I hope we have many more interesting conversations.


Craig

2% would be low in non critical consumer items - but for a safety critical use 0% complaints would be the aim. The cost of delivering a 0% complaint product or service would be higher - so, perhaps the question is what complaint rate makes financial sense for your market? Could you benchmark against your competitors?

Brian

Welcome to the cove
Oppurtunity as far as I understand it

Plastic part with 500 dimensions

Made 2000 with one dimension out

2000*500 = 1,000,000 opportunities

500 ppm

but to non statisticans the part is no good. 100% failure.

Who is right

I guess that depends on if the dimension is for a feature that hold the part together or one that makes it work. . . :2cents:

Agree, what type of dimension is out and how much. Is it a critical, major, minor, or incidental. As stated above if it involves fit or function, then it is indeed classified as defective. If this is a recurring problem on a minor or incidental dimension which has no significant bearing on the application, then suggest contact customer and try to obtain a waiver, deviation, or engineering change. Then it will reduce your internal and customer rejection rate and PPM. I had to deal with this years ago where edge trim dimensions had a tolerance of +/- .010, but did not mate with another part, and just hung in the wind under the car. Went to big 3 engineers and received engineering changes on all such dimensions to reflect the real world conditions, and save money on tool repair costs.

True Ralph, but what if you are a tier 19 supplier who's customer won't ask for any engineering changes? They just tell you it is out of spec so fix it. In the world of "Zero Defects" there seems to be no such thing as minor or incidental. The fact that the part is completely functional has no berring on the fact that it is OOS.

Dave

Greg,
Here's another question that you might want to ask yourself in regards to the 2%. Are the CC's that you're getting now, costing the company more money or less money than they did in the last 2 years? Point being that if you are losing more sales dollars (lost customers, etc), then IMHO the percentage is irrevelant. I work in the stamping industry, and from my experience with looking at things like scrap as a percentage, well, it doesn't show a true picture. I could have 2% scrap for May and only 0.5% for June, but actually loose more money in June because of the dollar value of the items lost. Hope this makes sense?
Jonell

Dave- In that case there are 2 options:
1)Go up the customer chain to find someone with "common sense"
2) Suck it up and initiate the root cause CA and contiue to monitor the gains.

Jonell-Ypu are right about how to quantify the percents, convert to dollars and look at where you get the most bang for the buck in prioritizing corrective action efforts.

Getting back to Greg B's original question of what is CC rate, or Return Goods rate acceptable.
We used to use a "rule of thumb" target of 0.05% max., but this was I was not in the automotive market.

I guess that depends on if the dimension is for a feature that hold the part together or one that makes it work. . . :2cents:

I agree but wanted to show the level of absurdity that some can go to show PPM levels

I agree but wanted to show the level of absurdity that some can go to show PPM levels

Howard,

I missed your sense of humor in that post. . . I'll be more alert next time. . . .and yes. . .some people got to great lengths. . .

If we are a six sigma society I would "assume" that anything outside those limits would be unacceptable.

Howard,

I missed your sense of humor in that post. . . I'll be more alert next time. . . .and yes. . .some people got to great lengths. . .

This is my dry British humor :topic: :o

i would just like to share our measurements for Customer complaints.
what do measure is the 'response rate' of the department/unit that the complaint is endorsed.

Since we are in the financial services industry, we welcome an increase in the number of complaints received. This will also show that our Customer feedback system is recognized by our customers (policyholders).

We also take note on the improvements that were ACTED upon by the process owner based on customer feedbacks i.e., turnaround time improvements and/or cost savings.

:)

Welcome to the Cove, Phillip! :bigwave:

i would just like to share our measurements for Customer complaints. what do measure is the 'response rate' of the department/unit that the complaint is endorsed.

What do you mean by 'response rate'? Is this the time to reply/resolve a Complaint or the % of complaints replied to/resolved?

Since we are in the financial services industry, we welcome an increase in the number of complaints received. This will also show that our Customer feedback system is recognized by our customers (policyholders).

Hmmm....I don't think I've heard anyone say they actually welcome an increase in complaints! :) But I do understand your logic. However, I would prefer simply asking Customers if they had a problem with our product, what would they do as a means of demonstrating the existence of our complaint system.

I agree that customer complaints is a hard number to allow at all, but reality is quite different. We're a software company, and we look at a couple of measures -- the # of bugs we "fix" after the newest version is released, and the customer returns as a % of sales. I'm happy to report that the % of sales number is 0.3%. We're working on the bug fixing number.

One Cover posted a cost of quality % that seemed high to me. too. To my knowledge, the total costs of quality at a company -- external failure, internal failure, prevention and appraisal typically runs 20-35% for a manufacturing company in the US.

--Jodi

I agree that customer complaints is a hard number to allow at all, but reality is quite different. We're a software company, and we look at a couple of measures -- the # of bugs we "fix" after the newest version is released, and the customer returns as a % of sales. I'm happy to report that the % of sales number is 0.3%. We're working on the bug fixing number.

One Cover posted a cost of quality % that seemed high to me. too. To my knowledge, the total costs of quality at a company -- external failure, internal failure, prevention and appraisal typically runs 20-35% for a manufacturing company in the US.

--Jodi Jodi, that no. is a little higher than what is stated in this article.
Taking control of quality costs (http://www.manufacturingcenter.com/qm/archives/1198/1198qual.asp)

Welcome to the Cove, Phillip! :bigwave:

What do you mean by 'response rate'? Is this the time to reply/resolve a Complaint or the % of complaints replied to/resolved?

opps. sorry for the confusion.
yes, response rate = time to resolve the case.
We have actually identified the classification of feedbacks received from the customer and defined a standard time to response on that.
for example,
Complaint letters - three days.
Telephone calls: premium defaults within 30mins.
policy inquiries withing 45mins.
these are the quality standards.


Hmmm....I don't think I've heard anyone say they actually welcome an increase in complaints! :) But I do understand your logic. However, I would prefer simply asking Customers if they had a problem with our product, what would they do as a means of demonstrating the existence of our complaint system.

I have to clarify that i refer to customer feedbacks. but not actually a complaint-defect for the actual product. since ours is in the insurance business most of the time, the complaints is just in inquiries in the policy services... though of course we still do received calls/feedbacks from unsatisfied customers. :)

thanks for d warm greetings. :)