Don't let a focus on containment (full automation) blind you to the potential benefits of partial automation
A common (but flawed) metric of the success of call-center automation is containment -- the extent to which callers complete tasks in the IVR without transferring to the call center. One reason this metric is flawed is that it fails to account for the significant benefits of partial automation (Larson, 2005; Suhm, 2008). Of particular importance is the agent time saved by every element of a task completed in the IVR before transfer, as long as that information gets to the agent (e.g., via the technologies known as computer-telephony integration, or CTI -- see http://en.wikipedia.org/wiki/Computer_telephony_integration).

Lewis (2011) reported that for a reservations application the containment rate was about 15%. On average, however, the application provided five of the eighteen elements required to book the reservation, resulting in significant saving of agent time (a reduction of about 25%). Kaiser et al. (2008) reported similar savings (15%–20%) for partial automation of member calls to a health insurance call center after revising the business case based on the value of partial automation.

References

Kaiser, L., Krogh, P., Leathem, C., McTernan, F., Nelson, C., Parks, M. C., & Turney, S. (2008). Thinking outside the box: Designing for the overall user experience. From the 2008 Workshop on the Maturation of VUI.

Larson, J. A. (2005). Ten guidelines for designing a successful voice user interface. Speech Technology, 10(1), 51-53.

Lewis, J. R. (2011). Practical speech user interface design. Boca Raton, FL: CRC Press, Taylor & Francis Group.

Suhm, B. (2008). IVR usability engineering using guidelines and analyses of end-to-end calls. In D. Gardner-Bonneau & H. E. Blanchard (Eds.), Human factors and voice interactive systems, 2nd edition (pp. 1-41). New York, NY: Springer.