iSixSigma

Calculating CSAT index using 2 parameters

Six Sigma – iSixSigma Forums Old Forums General Calculating CSAT index using 2 parameters

Viewing 9 posts - 1 through 9 (of 9 total)
  • Author
    Posts
  • #48481

    Shilps
    Member

    Hi,
     
    We are trying to calculate CSAT score using 2 parameters- first the importance ranking of each question( 1-5, 5 for Very Important/Relevant) and performace ranking( 1-5, 5 for Very good performance)
    I want to get the final CSAT score in the range of 1-5. However, it is coming in the range of 1-25.
    The way I am calculating is:
    For response 1, score is Performance Rating* Importance Rating. Eg, if reponse 1 is Very Important and Good Performance( V good being the max), then the score will be 5*4=20
    Now I add this score for all the responses and divide by the number of responses.
    What am I doing wrong/right?
     
    Thanks

    0
    #163554

    fake accrington alert
    Participant

    Csat index stands for  what?
    thanks

    0
    #163555

    Shilps
    Member

    Customer Satisfaction Index

    0
    #163560

    Yadav
    Participant

    Shilps
    Let me set the background. You have prepared the set os questions and your customer is rating the questions in terms of ‘importance’ and ‘how are you performing’.
    The purpose of Csat is to know in which parameters you are exceeding the customer satisfaction and does it matter to the customer. So I would advice you to follow the below approach:
    – Stratify the questions into 3 levels – High importance, Medium importance and low importance.
    – Now you can take the average of rating under each category.
    – It goes without saying that the rating under ‘high importance’ should be your key focus as that is ‘very dear’ to your customer.
    Hope this helps. If you are looking at arriving at a single score, just look at your score for the ‘high importance’ questions. That is what matters to your customers.
     

    0
    #163606

    Erik L
    Participant

    Shilps,
    Why not make the CSAT index additive and then use some weighting factor for the paremeters that you are going to use.  Are they a 50/50 split in imporatance?  80/20?  Taking each individual response, multiplying by the weighting factor and then summarizing should get you closer to what you are looking for.  This way you could potentially add factors to the index, just make sure your weighting ultimately sums to 1 across the various factors.
    Regards,
    Erik

    0
    #163696

    Shilps
    Member

    Hi Ajit,
     
    There are only 4 questions, but around 1000 responses. How do you propose we do it? Could you pls explain with an example.
     
    Thanks

    0
    #163697

    Shilps
    Member

    Erik:
     
    Just trying to interpret your reply. Supposing we have 60:40 ratio- 60 for performance(P) and 40 for importance(I).
    For Response A, where P is 4 and I is 5, the score will be (4*60+5*40)/100 ?
     
    Sounds good and logical , just wanted to know Stats guru have to say about it?
     
    Thanks

    0
    #163711

    Merry
    Participant

    Here is a suggestion: Why not divide the product by 5 to arrive at a figure. For example, someone would have given the importance as 4 and performance as 3. The product will be 12, and when divided by 5, it would be 2.4. Anything that scores between 4 and 5 are the highly important and best performing ones, and anything below 1 or 2 can be ignored. You can even round off or use floor or ceiling functions in Excel to remove the decimals and get an integer between 1 and 5.

    0
    #163732

    Erik L
    Participant

    I’d never claim to be a stats guru, just an eternal student, but your equation would work.  Not to splice hairs, but I think that your model would be more appropriate as an additive model vs. the multiplicative model that was alternatively suggested.  Other opinions out there?
     
    Thanks,
    Erik

    0
Viewing 9 posts - 1 through 9 (of 9 total)

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