Using historical call result data for finding the best phone number

[TommyG]

Hello

Has anyone developed an algorithm to find the best phone number and the next best phone number using historical call result information of the customer?

Second question: when automating a solution using historical call result data, which approach is better - finding best phone number or finding the worst phone number?

Thanks!

[Adam G]

What you are asking is actually more CRM data than telephony data.  If you have (any) result sets which can be matched with your Inbound/Outbound callers, I expect you are talking about a Result Code entered by an Agent in ACW...?

Let's assume that you [i]are[/i] comparing your callers with your enterprise result data and go with a hypothesis;

Finding the most frequent number will not necessarily mean the "best" - similarly, the least-frequent number may not be the worse.  It's all about quantifying the result - the Frequency vs. Result.  For example:

IF a number gets a good result, [i]every [/i]time, it could be marked as "1"
IF a number gets a good result [i]half the time[/i], it could be marked as "2"
IF a number gets a good result [i]less than half the time[/i], it could be marked as "3"
IF a number [i]rarely [/i]gets a good result, it could be marked ars "4"
IF it's never had or good result or it's new - or if it doesn't fit any criteria above, it could be marked as "5"

Now you have a priority order to use for future business.  As long as you keep it up by adding a result code which can be used in your reporting "algorithm", then it should work indefinitely.

But this is just a hypothesis....