1
$\begingroup$

I have a data about customers and their activity on a website for a two-year period. Also, I have a customer support work evaluation data for a shorter period of that two-year period. The question is:

How to find out if inefficient customer support affected loyalty of customers and, in particular, how many customers were lost due to this factor?

So far, I just made a simple comparison on how many customers kept using the website after leaving the negative (unsatisfactory) or positive (satisfactory) evaluation of the customer support work. The comparison shows that the customer retention rate for customers who left negative feedback is lower than for those who provided positive (satisfactory) feedback.

I'm not an economist, but I'm sure there's a more scientific way to estimate the influence of one factor on another using statistical analysis.

$\endgroup$
1
  • 2
    $\begingroup$ With the information given it appears this could be modeled as a simple "Binary Response/Discrete Choice" economic/statistical model, a well-known tool in statistics and econometrics. $\endgroup$ Mar 4, 2018 at 11:53

2 Answers 2

1
$\begingroup$

Maybe you can do a Probit regression?

Let Y be a binary indicator. If the customer continues to use the website frequently (you can determine the threshold for that) after leaving the comment, then Y = 1, otherwise Y = 0. This Y is going to be your dependent variable.

Let X be the measure of satisfactory, the higher the better.

Your independent variables are X and some other factors that may be relevant. For example, the amount of time spent on this website prior to the survey, the amount of money spent on this website prior to the survey, number of friends and etc.

If the coefficient for X is positive, it means customers tend to stay if they are satisfied with the website's customer support (i.e. customer support inefficiency lead to loss of customer).

$\endgroup$
0
$\begingroup$

The best way would be to use a regression analysis, What i would do as an economic major would be use two independent variables such as in your case level of customer support, customer feedback to see if the dependent variable "customer retention rate". also you need to use parameters for customer support evaluation, customer feed back and customer retention rate for the years. Then run a regression analysis to find if the independent variable are co-related to your dependent variable, thus you can arrive at a statistical justification for your problem.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.