2 Replies Latest reply on Jan 8, 2015 1:53 PM by pvanderhyde@tableausoftware.com

    Support Case Deflection Algorithm




      I'm in the process of developing a support case deflection algorithm and was interested to hear if anyone has been able to successfully develop a defendable metric. At its lowest accurate form there is question and answer data that can be extrapolated with views to arrive at a percentage of potentially deflected cases. I'm targeting a higher accuracy form that combines the question, answer, view and user data, and is then combined with our Support case system data to monitor whether or not the user creates a support case within a set period of time (like 12 to 48 hours). Any thoughts, ideas, or experience on this matter would be appreciated.




        • Re: Support Case Deflection Algorithm

          We don't do anything like this, but I sure like the idea. I hope others share what they're doing so that I can borrow/steal from their good work and yours!



          • Re: Support Case Deflection Algorithm

            I utilize views of Answered posts - this is a combination of Google Analytics and Jive data to come up with numbers that I feel strongly about.  Specifically, I focus in on # of views of answered questions where the user avg's more than X minutes on page.  Most of our solutions take a few minutes to digest and I think its fair to assume that someone sticking around for X minutes of time is likely finding their answer.  


            Once Jive has a 'this answered my question' type option for users to select, I will modify our criteria to use this.