IVR Solutions

This section of our technical library presents information and documentation relating to IVR Solutions and custom IVR software and products. Business phone systems and toll free answering systems (generally 800 numbers and their equivalent) are very popular for service and sales organizations, allowing customers and prospects to call your organization anywhere in the country. The PACER and WIZARD IVR System is just one of many DSC call center phone system features..

Contact DSC today. to learn more about our IVR services and IVR application development software.

By Kurt Dahlstrand

Most contact center managers feel unrelenting pressure to reduce per-call costs without undermining customer satisfaction. To achieve these goals, contact centers are using the constantly-evolving sophistication of IVR and CTI technology to deliver ever-higher levels of customer self-service and agent efficiency. The economics of using these tools are compelling: a self-service transaction averages about 45 cents, while agent-handled transactions frequently run as high as five dollars per call in combined personnel and toll costs.

At the same time, however, customers expect—even demand—that contact centers deliver consistently high levels of quality, reliability and availability. If the customer’s Quality of Experience (QoE) falls short in any way (dropped calls, incorrect prompts, poor routing, prolonged silence or delay), there are only two outcomes: either the customer “zeroes out” to an agent queue, or the customer gives up and goes away—sometimes permanently. Either outcome kills the advantages of self-service applications, and erodes both a company’s image and its bottom line.

Fortunately, a new generation of end-to-end monitoring technologies can help identify and diagnose performance problems before they affect customer QoE, so that contact centers can gain the full benefit of self-service voice applications without incurring any of the hidden costs. By automatically dialing in and navigating through a voice transaction the same way a customer would, these technologies can track and measure the entire range of potential quality and performance problems -- from the public switched telephone network to the back-end databases and out to the agent.

The High Cost of Ignorance

Among the worst (and most avoidable) costs imposed by subtle, undetected performance errors are the added call lengths, toll charges and personnel costs that arise when call traffic that should have been handled by a self-service application are passed off to an agent.

Even a modest period of unnecessary congestion can generate significant added costs. At the contact center for a major U.S. bank, for example, managers discovered that a single 20-minute period of congestion in its IVR system (which normally handles about 75 percent of all customer queries) led to four hours of agent overload. Unable to complete their self-service transactions, customers zeroed out to agents – increasing the number of agent assisted calls by 300 percent. This burden on customer service agents not only raised the agent costs associated with these calls, but also boosted queue times and toll charges as frustrated customers waited to get through to an agent instead of using the self-service menus. The bank in question ran an analysis of these direct costs and determined that this recurring IVR performance problem imposed direct costs of over $2 million a year.

Indirect costs due to poor performance are harder to measure, but this data suggests that the dollar value of customers and business lost due to poor IVR/CTI application performance can be even higher than the direct operating costs.

To get a handle on these costs, and to keep customers satisfied with the increasing use of self-service applications, savvy contact center managers are beginning to embrace new monitoring tools and services that measure voice application performance. These automated solutions record carrier, host and database response times while verifying that the right prompt is playing at a sufficient quality level and that the call has been routed to the appropriate agent along with all the appropriate call data necessary to create a screen pop. If a problem is detected, these tools immediately issue an alert to the appropriate technical resource. The alert identifies both the type of problem and its location, and allows technical personnel to take corrective action before the problem begins to adversely affect customers, tie up technical resources and escalate into an expensive degradation of agent productivity and overall contact center efficiency.

“This Call May Be Monitored . . .”

Despite the many advantages of automated end-to-end monitoring from the customer’s perspective, these enhanced monitoring technologies are only now beginning to become an industry standard. Prior to 2001, the primary focus for call center monitoring had been on agent interactions rather than on the applications and infrastructure that play such a crucial role in improving agent productivity and maintaining the overall QoE for call center users.

With the rapid expansion of IVR/speech-driven self-service applications, along with the development of increasingly elaborate call routing and screen pop technologies to optimize agent interactions, this predominant focus on the human elements ignores the contact centers increasing reliance on the performance of self-service, CTI and screen pop technology to achieve its productivity and customer service goals. After all, who hasn’t dialed into a self-service application and been unable to complete the transaction or been asked by an agent to repeat the information already entered by voice or touch-pad?

These costly flaws were highlighted by a benchmark study conducted last year of contact center performance in the utilities industry. Created and published by Empirix, Inc. (a testing and monitoring firm that provides products and services for both Web and voice systems) the study revealed that self-service performance varied tremendously from week to week and from company to company. (See Figure 1.) Of the high number of transaction failure rates, slow or unavailable databases accounted for some 20 percent of all transaction failures. In addition, the study showed that more than 37 percent of all transaction failures were due to switch problems. Poor quality or wrong prompts accounted for roughly 26 percent of failures. The remaining 15 percent of failures were caused by routing errors, in which calls were accepted by the ACD, but were never connected to the voice application.

The results from this study were cross-indexed (of the financial services, utility and wireless industries) against relevant customer satisfaction data from Purdue. This combined dataset shows significant correlation between overall customer satisfaction and the reliability of contact center self-service applications—a strong indicator that high call failure rates damage a customer’s overall level of satisfaction. (See Figure 2.)

The Pitfalls of Manual Monitoring

Prior to the advent of automated monitoring solutions, call centers made do with manual methods that required agents to dial in to an application. Some centers also used network management tools to monitor individual system components. Unfortunately, manual monitoring is subjective, expensive and rarely performed with adequate frequency or intensity. As a result it yields little in the way of detailed, consistent data about system performance. Network management tools also have major shortcomings: they miss a majority of potential performance problems because they report only on individual components, not on whether these components are interacting seamlessly to deliver a high QoE to the end user.

Manual monitoring or stand-alone component monitoring can miss glaring performance problems are painfully apparent to customers. At a contact center for a large wireless carrier, for example, agents would sometimes be flooded with self-service calls even when the IVR was still reporting as active and below capacity. The company implemented a performance monitoring solution and quickly discovered that when customers called in they were greeted, oddly enough, with a sound clip of an opera singer rather than the expected opening prompt. A quick analysis also revealed that a load balancer was configured to think that each IVR only had a single T1 in each trunk group rather than the two T1s per group that were, in fact, available. The result was that, at only 50 percent capacity, the load balancer would think the IVR was full and route calls directly to agent queues, even though the components all appeared to be acting normally.

The Quick Fix

Discovering the causes, frequency and symptoms of call failures and other performance problems is only the first step in reducing costly slowdowns, downtime and congestion in contact center systems. Transmitting that information to the appropriate technical support person or system in real time is just as important. For this reason, automated monitoring solutions should include intelligent alerting capabilities that allow contact centers to quickly detect, diagnose and resolve issues. Speedier resolution not only improves customer QoE, but also saves on toll charges, agent costs and other contact center resources.

In addition to providing immediate access to performance data that helps contact center managers detect and respond to performance issues, well-designed automated monitoring solutions also offer archiving and analysis of historical performance data. Over time, this data can be used to set – and actively manage toward – specific service levels for each aspect of the contact center infrastructure. For example, individual service levels can be set to monitor carrier, IVR, and database response times.

With all these advantages—true end-to-end monitoring, accurate benchmark performance data, real-time alerting, and significant boosts in the availability, reliability and quality of IVR/CTI systems—automated monitoring gives contact center managers an important new weapon in the war to control costs while enhancing customer satisfaction.

The increasing complexity of contact center voice systems (network, switch, IVR, CTI, middleware and databases) has generated a steadily growing potential for performance bottlenecks, outages, failures and errors. In turn, these performance problems lead to frustrated customers, unnecessary toll charges, overinvestment in capital equipment and bandwidth, inefficient use of agent and technical resources and increased agent turnover.

In an increasingly competitive business environment, with customers standing ready to abandon companies that deliver inferior quality of experience—and as a matter of sound fiscal practice—automated monitoring solutions are rapidly becoming standard elements in contact center design and management.