I experimented with estimated wait times (EWT) but found the calculations unreliable. There are some formulas around for EWT in ICM but in experimenting, I found them to give wildly inconsistent results.
One of my customers insisted on using EWT even though the results don't appear to be anywhere near accurate. I created a custom function in ICM which is below.
I called the custom function: userEstimatedWaitTime. The formula is below:
(((SkillGroup.%1%.RouterCallsQNow+1)*(SkillGroup.%1%.AvgHandledCallsTimeTo5))/max(SkillGroup.%1%.Ready,(SkillGroup.%1%.TalkingIn+SkillGroup.%1%.TalkingOut+SkillGroup.%1%.TalkingOther)))
I've seen this formula and ones similar to it for calculating EWT in ICM. To use the formula in an ICM script, use a Set Variable node, assign either a PV or ECC variable to the value: userEstimatedWaitTime(Enter_The_Skill_Group_Name_Here); this assumes you created a custom function called userEstimatedWaitTime. Then, in IPIVR you can use that value to build a prompt or use it in some calculation to build a prompt, etc. Or, you can use the value in an ICM IF node to see if the EWT is greater than 20. If so, send the call to voice mail, else, send the call to the Queue to Skill Group node.
I've read in other forums that ICM doesn't support the EWT concept for a number of reasons I don't remember off the top of my head. You can search for ICM estimated wait time and find a number of discussions.
My experiences have been with UCCE 9.04 and IPIVR 9.02. I'm not familiar with PCCE so I don't know if what I told you applies to PCCE. Things may work quite differently in PCCE 10.0+ than in UCCE 9.04.
Bill
