IVR Redesign Case Study

Six different usability tests have been performed on PGE's Interactive Phone System (IVR) since 1996. With each test, design problems and opportunities for improvement were identified. For example, in redesigning PGE's phone menu in 1996 the following improvements were made:

Reduced Wording (26% less verbiage)
Counting the number of words that customers hear is an excellent means of evaluating the effectiveness of an IVR. Empty phrases, conversational pleasantries and repetition serve no purpose other than to increase listening time. Each additional word also increases the possibility of frustration, confusion or error. In redesigning the IVR and eliminating superfluous wording PGE was able to reduce the verbiage by 206 words:

  • Greeting 5 words (was 46 words)
  • Customer service main menu 40 words (was 118 words)
  • Field service menu 27 words (was 114 words)

Increased Task Times (60% faster)
Measuring task time for customers attempting to get through to the right person or department is also an effective way to evaluate an IVR design. For key tasks such as start, stop or move, payment arrangements and high bill inquiry the average speed of throughput was 51 seconds in the redesign (37 seconds faster than the previous design). This is a 60% increase in throughput for customers calling in.

¨ Start, Stop or Move 0:48 (27 seconds faster)
¨ Payment Arrangement 0:50 (39 seconds faster)
¨ High Bill Inquiry 0:55 (44 seconds faster)

Reduction in throughput is directly associated to the number of words presented, sequence of choices and terminology used. As a side note, with the introduction of automated outage reporting-speaking with a rep (not including hold time) 3:19 compared to the automated system which takes 1:01.

© 2002 Usability Concepts, LLC