CSAT project

Six Sigma – iSixSigma Forums Old Forums General CSAT project

Viewing 8 posts - 1 through 8 (of 8 total)
  • Author
  • #48283


    Hi all,I recently finished my 6sigma training and i am starting with my project.We have a web based process called as e-gain wherein agents reply or assist the customers with their queries either via email/chat. And we have CSAT survey for the same. we asked 5 questions to the customers as below:Question #1 Your overall experience with the chat service was good.
    Question #2 The site was easy to use and navigate.
    Question #3 The agent was able to answer your question or resolve your problem.
    Question #4 The agent was helpful and friendly throughout the chat session.
    Question #5 You would prefer to chat live with an agent rather than call. Based on which customers give their feedback in 5 ways as mentioned below:strongly agree
    strongly disagree
    somewhat agree
    somewhat disagree
    no opinioni am working on question # 3. can you help me as how can i start the project. pls give your expertise–
    Mangal Das



    ok…let’s see.
    So your metric is either x% of my customers need to rate the question 4 or 5 OR you would want to convert the customer’s feedback into numbers like Strongly Agree would be 5, strongly disagree would be 1 – in that case disregard all the “no-opinions” and take an average of the scores – say then you may want to increase your scores to 4.5 on that scale.
    Just my thoughts….Wait for the advice of the more learned on this forum before you act.
    This is an interesting project – I will appreciate if you could share your learnings on this forum when you close the project.
    All the Best!



    Hi neal,
    Thanx for ur info.
    Basically, we asks 5 questions to the customers and customer either reply their feedback as in ‘strongly agree’, strongly disagree’, ‘somewhat agree’, ‘somewhat disagree’ and ‘no opinion’.The questions that we asked to the customers are:”Question # 1 Your overall experience with the chat service was good.” “Question #2 The site was easy to use and navigate.” “Question #3 The agent was able to answer your question or resolve your problem.”
    “Question #4 The agent was helpful and friendly throughout the chat session.” “Question #5 You would prefer to chat live with an agent rather than call.” currently we are focusing on Question # 3.
    I want to know how can we analyse the data and improve the process or the agents on question number 3.Pls share your viewsMangal Das



    Dear Mangal,
    this is Transaction Feedback wherein u take a feedback from customer just after a transaction happens with the organization in the form of call centre or sales person. 
    CSAT deals more holistically with guaging the customer satisfaction levels . 



    Convert the responses to integers from 1 to 5. Then use Ordinal Logistic Regresssion in Minitab.



    If I understand your question regarding question 3, might I suggest exploring any relevant segments of the surveys to indicate troubled areas (e.g. topic of the chat).  However, sample size needs to be considered.
    It sounds as if you are approaching a point where you need to identify the drivers (input) of what the customer is experiencing.  Maybe eventually mapping out the processes once segments are identified.



    I agree, the number value system will allow you to measure progress but you still need to identify the causal effect (inputs).
    If the problem is solved that is good but can it be avoided in the future? If the problem is not solved then you need to ask “WHY” questions. Again so you can get down to the cause of the problem and correct it.
    There are many questions to ask – What type of defects are you measuring? What scope of problems should your employees be capable of solving? Are they? what happens with out of scope issues? Is there an escallation? should there be? Are we providing the right competencies in our people for the problems we anticipate? Are we anticipating the right problems?
    If you don’t set up to measure the correct data the scale won’t matter. You need to define the parameters of the project.



    Like a lot of  surveys your first one will give you insight into what questions you need to ask in the next survey in order to drill down  and attempt to identify the problems and what is driving them.  From your questions you essentially have 2 areas that you are addressing. One is based around the site and its navigation, the remaining questions are focussed on the agent or “soft ” side which is dealing with personal preferences and interactions. Based on the results of your first survey you may be able to narrow your project scope to either a “hard” element which is the site design or the “soft” side which is the people component.  By doing this you can the develop a more targeted and specific survey, i would strongly recomend the use of focus groups if you are going to do this

Viewing 8 posts - 1 through 8 (of 8 total)

The forum ‘General’ is closed to new topics and replies.