Customer Complaints Formula - Number of complaints/number of transactions × 100

[D.Salman]

Dear all,
Our company would like to measure customer complaints by using the following formula:
Number of complaints/number of transactions×100.
Now, if the number of complaints in the current month is 20 and number of transaction is 65, result will be 30%, and if the number of complaints in the next month is 25 and denominator are 100, result will be 25%, the rate went down but complaints went up.
I don’t think the above indicator will help us decreasing the number of complaints over time. I suggested monitoring the number of complaints instead of the above formula.
Any feedback to support my idea will be highly appreciated
Regards

[DsqrdDGD909]

I think a way of comparing the level of complaints to the level of business is a good idea.

I think keeping the raw data is a good idea as well.

One thing to think about - how are you defining complaints and how are you defining transactions?

[Phil Fields]

Customer complaints can be a tough one since you do not control the number of complaints, you can only react to them. Other items possibly to measure is:
• The time to reply to the customer to satisfy their complaint
• The number of similar complaints, is this a trend, does this trigger a CAPA
• The time to close a complaint

How does your QMS react to complaints, are complaints used as part of continuous activities.

Just a few thoughts,
Phil

[Jennifer Kirley]

Phil is correct.

The percent of complaints is an ineffective metric because many people will not complain; they will simply buy elsewhere. In addition, the information has no meaning; it does not tell you what went wrong.

But if your company insists on such a metric, it would be achieved with the ratio of #complaints divided by the #sales. 5/100=5%.

I recommend resolving complaints and then tracking customer satisfaction in subjects that can lead to action:

1. Our delivery was early or on time.
2. Our product always functioned properly.
3. Our assembly instructions were accurate.
4. Our user's instructions were easy to understand.
5. Our customer service representative satisfactorily answered your questions.

Learning how often you have repeat sales and referral sales is especially informative.

[sauhiew]

Without understanding the nature of your business, it's hard to judge if this metric is a good customer satisfaction measurement or not. For example, in the consumer market selling low price commodity, most customers will just turn away elsewhere. On the other hand, in the market serving OEM as a part supplier, most customers will file a complaint to you. Your company/management should best judge if this is a good measurement or not.

One thing for sure, measuring the percentage of customer complaint over transaction is better than the absolute value of the # of complaints. For example, you may have 30 complaints for 100 transactions this month, for a 30% rate of customer dissatisfaction. However, due to business growth, you may have 40 complaints for 200 transactions per month, 6 months later, for a rate of 20%. You may have more complaints, but the rate of complaints has decreased for a more positive trend.

[Bev D]

to the OP: what is a transaction?

a customer complaint metric is intended to determine if your improvements are making things better or if soemthing has changed to make things worse: basic process control and is a goo dthing. The trending of the metric is not intended to tell you what to do nor is it a substitute to appropriate response to individual complaints. However: if that simple ratio has an incorrect denominator the data it generates will be useless.

[Vassilis]

I think there is a simple way to monitor the percentage of complaints ( U chart)
and also the number of complaints ( C chart , number of complaints/month)

I hope that i gave an idea!!

Vassilis

[Murphy's Law]

I'd monitor both.

When I analyse Dppm on my components, I display both Dppm and volume shipped.

Other things you may want to look at is who is giving you the complaints and reasons. They may be same.

For a Quality tool on this, look at a paynter chart.