By Donna Fluss
When interactive voice response (IVR) applications are well designed, implemented, and maintained, they automate anywhere from 20 percent to more than 90 percent of incoming calls. Many enterprises would face a major financial hit if they had to employ agents to handle the calls that are automated by their IVRs.
Despite these facts, IVRs are often neglected and under-resourced. Too many companies have IVRs that were implemented years ago and are enhanced only when something breaks or there is a major change in business requirements. Additionally, IVRs are often relegated to the nonessential category by CIOs, who were happy to “outsource” them long before hosting became acceptable for other mission-critical contact center solutions.
IVR Dependability Results in Its Downfall: While this is a seeming contradiction, the success and dependability of many IVR solutions has caused them to be treated as “second-class citizens.” IT groups take them for granted because they generally do not require a great deal of support to keep them in production. However, while call processing continues without interruption, the vast majority of IVRs are not performing at optimal levels, according to recent DMG Consulting research. It estimates that more than 80 percent of IVR users around the world would be able to improve their automation rates and dramatically increase customer satisfaction if they invested in routine optimization of their IVR solutions. Another way to look at it is this: If an IVR was installed more than three years ago and has not had an overhaul of its script or voice user interface (VUI) since then, it’s time for a full “health checkup.”
Scenario – IVR at Work: The following example proves the point. A financial service organization receives a million calls per week. This organization has an IVR that automates 60 percent of their calls, or about 600,000 weekly. Their non-fully loaded (variable) cost per agent-handled call is $5.50. If they automated as little as 2 percent of the remaining 400,000 calls per week, they would displace an additional 8,000 calls from agents. This would save them $44,000 per week, or $2,288,000 per year. At the same time, the quality of their service would increase and complaints would decrease.
Of course, there are additional factors to consider. By automating some of the easier calls, agent’s average handle time is likely to increase for the remaining calls by an average of two to three seconds. Assuming a 200-second average handle time, this will increase the cost per call by approximately 1 percent, although this gain could easily be offset by a corresponding reduction in average handle time due to a decrease in customer complaints about the IVR.
Additionally, there is also the cost of the optimization project to consider, which runs from approximately $50K to $150K, depending upon the resources required. The payback period from an IVR optimization, with a cost of $150K for the project and a 1 percent increase in the cost per call, was less than one month. It saved the organization almost $1.9 million during the first year.
Why Are Organizations Waiting? Given these returns, one would expect many organizations to jump at the opportunity to enhance their IVRs. There are three primary reasons why end-user organizations are not making the investment, all of which are based on fear:
- Vendors scare away prospects by pushing them to make major investments in expensive speech recognition-based platforms.
- Enterprises do not have the appropriate resources available in-house and are trying to avoid using “expensive” professional services for an “uncertain” return.
- End users do not see a compelling enough reason to take a chance on disrupting what they already consider to be a highly effective solution.
Separating Fact from Fiction
Myth 1: End users with touch-tone or non-VXML-based speech recognition applications must upgrade to a new platform in order to realize benefits from an IVR optimization.
Fact: No, they do not need to upgrade to a new platform. In some situations, upgrading to a new IVR platform could be a good idea, but if the application works, it is not a necessity. As long as there is a way to evaluate the performance of the IVR application by putting in trackers that identify where and why customers drop out and/or request an agent, the current environment can be optimized without replacing the underlying system. Premise-based IVR vendors make most of their money from selling new and upgraded solutions and charging maintenance fees, so this is what they push. Moving from a touch-tone IVR application to a VXML-based speech platform may have substantial long-term benefits, but it often requires a hefty up-front investment that could cost more than $250k, particularly if speech-recognition application-development fees are included.
While many organizations see the benefits of a self-service platform refresh that may have a payback in one to two years, it is often very hard to justify when the current solution still works, the CIO has a tight budget, and there are other critical investment priorities. In other words, the vendors often price themselves out of a job; instead of helping organizations undertake an optimization initiative, they scare off prospects by pushing an expensive “rip and replace” solution. This is one of the major reasons why sales of premise-based IVR solutions have fallen during each of the last few years and why this downward trend is expected to continue for the near future. This opportunity has been seized upon by more flexible hosted/managed service IVR providers.
Myth 2: End users have little to gain and a great deal to lose from trying to improve an IVR application that delivers an acceptable automation rate.
Fact: As long as experts are used to implement the improvements, the benefits should be significant, relatively rapid, and mostly risk-free. An IVR optimization exercise includes two major components: figuring out what does not work well in the current application (i.e., where people get confused or stuck and drop out); and identifying agent-related tasks that can be automated. The first step is to identify and document the problematic activities and new automation opportunities. The second phase, which has some risk, involves rewriting the application to take advantage of these opportunities.
The Bottom Line: IVR is not a new or sexy application. It’s perceived as a dependable core contact center application that runs without requiring a lot of attention. Most contact center leaders would prefer to invest in new solutions, like speech analytics, rather than in their old and dependable IVRs. The “catch” is that a small investment in IVR might result in major cost savings. Any organization that has not optimized their IVR application in the past three years should undertake an IVR assessment to identify ways to improve their existing system and estimate the potential long-term benefits of an optimization initiative. IVR optimization should not be a one-time exercise. Leading IVR users continuously strive to enhance the performance of their solutions; it’s part of their annual budget and their corporate culture. If the necessary resources are not available in-house, find a hosted/managed service IVR provider that can help enhance the operating environment with little up-front investment and risk. The world of IVR has changed dramatically in the last five years, as have many business requirements.
Donna Fluss is the founder and president of DMG Consulting LLC, a provider of contact center and analytics research, market analysis, and consulting. She is the author of industry reports on contact center hosting, IVR, speech analytics, performance management, workforce management, surveying and analytics, and quality management/liability recording. Contact Donna at firstname.lastname@example.org.
[From the December 2010/January 2011 issue of AnswerStat magazine]