IVR and Service Levels

[Carrigman]

We use a number of automated, self service options in our inbound contact center and they work fine. Callers who need to speak to agents however are experiencing long wait times because of understafffing. In our weekly and monthly reports we show an average Service Level (TSF): the percentage of calls answered in 20 secs. This figure includes 100% of the automated IVR options as being anwsered in zero secs and so the overall TSF average looks better than it really is. My argument is that TSF should really be used only to measure callers who need to speak to agents as it gives a good mathematical indication of what the wait experience is really like. Is it general practice to include IVR transactions in the measure or am I correct?

[mikael]

Your instincts are right on this one -- the important factor is indeed the actual waits experienced by those customers who need to speak directly to a rep.

Anybody going to the IVR gets put right through (assuming proper capacity on your switch), so you learn nothing useful by including them into the overall measurement of your service level. Every time I have seen centers combining both IVR users and direct callers, the reason was indeed to make the stats look better. Unfortunately, while you might succeed in pulling the wool over Management's eyes by that route for a time, you do your company no favors by the deception and the truth sooner or later comes out.

You may be sitting on a revenue opportunity -- in some scenarios, customers are willing to pay extra for priority handling.