What is an IVR system and how does it work?
An Interactive Voice Response System (IVR system) is likely to be your company’s first point of contact for inbound callers. The IVR system allows inbound callers to access help through a set of self-service options or to be routed to a live agent for real-time assistance either through either touch tone (“Press 1 for service”) or automatic speech recognition.
IVRs are designed using one or more dialogs which control the flow of the conversation with the caller. An IVR system can communicate with callers by either playing pre-recording voice prompts or using text- to-speech which reads text using an automated voice. IVRs can also collect information from callers through speech recognition or DTFM touch tone.
Dialogs can be designed to engage a caller in a conversation to provide information or self-service. For example, a caller could ask for her account balance or request to transfer money from one account to another. Normally, callers can also ask to be connected to a live agent for more personalized service. Typically, dialog designers maintain standard options for callers to request live help, either by pressing zero or by saying “operator”.
IVR systems can also answer calls placed to multiple phone numbers. Dialed Number Identification Service (DNIS) can be used to determine which number was dialed and then launch the appropriate IVR application. IVR systems are used primarily for handling inbound calls but can also be used in outbound call centers.
History of IVR
As call centers grew in popularity starting with the advent of the ACD in the 1970s, IVRs also became increasingly valuable to call centers. Initially, they helped organizations collect information that could be used to route calls to agents through the ACD. As they matured, businesses began to use them to provide self-service options to callers thereby replacing human operators. Operator replacement become an important driver of adoption as call centers started handling a growing volume of inbound calls and the number of people required to answer those calls increased. However, early IVRs were complicated to program and configure, requiring IT staff to implement and maintain.
Life became much easier for call center managers as a new generation of IVR development tools were released that allowed non-technical users to create dialogs using visual, drag and drop interfaces. Initially, vendors developed proprietary tools that allowed customers to build IVR scripts that were only support by the vendor’s IVR platform.
Voice User Interfaces
As speaker independent speech recognition was released commercially in the late 1990s, IVR platforms began to augment DTMF touch tone interfaces with rudimentary voice interfaces (“Press or say one to be connected to sales”). The IVR application would place an API call to a speech recognition server whenever they needed to interpret a spoken utterance. The speech recognition server would then return a result in a digital format that the IVR system could process.
The VoiceXML forum was founded in 1999 by a group of companies including AT&T, Lucent, IBM, Motorola and Nuance Communications. The group developed the VoiceXML standard which was later contributed to the W3C. The goal of VoiceXML was to build a standard language based on familiar XML markup which allowed voice dialogs to be built that could run on any platform that supported the VXML standard. The standard, based on a widely understood XML web development paradigm increased the number of voice developers and applications.
Companies that seek to get the most out of their IVR systems track the performance of their IVR and continually tune their applications for maximum performance. Given the volume of calls handled by modern IVRs, even small improvements can offer significant benefits. Some common KPIs include:
- Customer displacement rate – The percentage of calls fully processed by the IVR without agent intervention.
- Customer drop-out rate – The percentage of calls that leave the IVR and require agent assistance.
- ASR recognition rate – Measures the accuracy of the speech recognition engine.
- Average time in IVR – Time callers spend in the IVR.
Benefits of IVR Software
Cost reduction and call deflection
IVRs reduce the number of agents needed to answer calls and responde to service requests. Companies often look at the average cost of a service agent and how many calls they handle. The cost of the IVR system is then balanced against the savings in terms of a reduction in the required number of agents.
Increase customer satisfaction and experience
Increasingly customers want the ability to get information and service without speaking to someone directly. A well designed IVR allows customers to get the information they need without having to wait to speak to an agent.
Increase first call resolution
The IVR is able to identify callers and collect important information about their problems. That information can then be used to pull information about the caller or their past interaction history. This allows the agent to be more informed when they accept the call, enabling them to quickly and easily resolve the callers issue.
Reduce customer effort
Custom effort score is becoming an important call center KPI. Essentially, it measures not whether the customer’s problem was resolved but how easy it was for the customer to have there problem solved. If properly designed, the IVR can be used to reduce the number of steps required to get an issue resolved.
Manage high call volumes
For call centers that experience periods of extremely high call volume, the IVR can offload calls that would normally go to an agent.
Provide after-hours customer service
As an automated system, the IVR is available to answer calls 24/7 enabling you to provide service even durning off hours.