Hi ,

We are planning to implement a six sigma project on improving first call resolution. This is one of our primary metric for help desk performance.Now, my question is about data type. Currently , we are calculating FCR as no: of calls escalated to level 2 support for resolution which means my data wil be ( 3, 7, 8, 11,2 ,1,0).Should i convert this in to percentage( 78%, 38%, 24%, 36%) by considering my defects as no: of calls escalated to level 2 support and my opportunity is no: of calls i.e. FCR % = (No: of calls escalated to level 2 support / total no: of calls ) * 100

Which one is more appropriate, this data is non normal. Should i convert this data to normal. What appropriate tools can i use for analysis.

Waiting for your reply

 

Dear Anish,

Let me try and understand your objective. It seems you want to calculate the sigma level of the process.

To answer your question on normality, this data is not expected to follow normality as this is defectives data and not continuous data.

If you highlight the specific analysis that you wish to carry out, it shall be easier for me to answer your question.

Regards,

VK

Sir,

I have attached the sample data for reference.

I want to know if i should use No: of calls escalated( count of calls that did not meet the FCR - No: of defects / day) or FCR % for my data analysis.

Is there any possibility of using MSA here, the data is collected from the online tool.If yes, then how should i proceed.

Can i use 2 sample t test here to see if there is a significant difference in the FCR during week days and week ends( by taking count of FCR as continuous data) or should i use chi square by considering my defects as discrete.

Thanks in advance

Regards

Anish Mohandas

Dear Anish, 

Discrete data should be treated a discrete. Division of one discrete value with another discrete value can make it a ratio or a percentage, but it cannot be treated as continuous data. in your case, it is best to work with raw data as discrete data

 

Thank you Sir....

Hi Anish,

Hope you are able to resolve your query...

Is you process outbound or inbound?

Hi Anish,

Wanted to know how is calculated FCR% in your call center? is it on just per day basis or ratio of total calls resolved per day on total no. of calls received / day? depending on this FCR data can either be treated as continous or discrete.

Another point that i wanted to understand is why you need to relate FCR with Sr Management Escalantions? as all NFCR must not be getting converted into escalations!

May be FCR% can be related to RC% (Repeat Call) or response from a customer in a survey conducted on a sample basis.

What will be the count or number of Opportunity to calculate Sigma Level for project related to FCR% improvement?