The Adaptive Audio – VUI Cloud Blog

January 10th, 2012 – Reported From:
The Department of Psychology
Human-Computer Interaction Lab
Houston, Texas

Rice University has been studying the feasibility of using IVR Systems for voting in political elections.

One concern with allowing voters to use this means of casting a ballot is knowing whether or not the voter clearly understands what and for whom they are voting.

Older voters, or the hearing impaired, may not be able to understand audio messages spoken at playback speeds better suited to typical callers. Alternatively, certain callers may be more comfortable with a dialogue pace that moves a little faster – especially if there are many candidates, elections and voting options.

The University chose Interactive Digital’s Adaptive Audio software to enhance their existing IVR System with the capability to match audio playback speed to suit each individual caller.

Danae Holmes of the University’s Human-Computer Interaction Lab says  ”I have been working with Interactive Digital’s Adaptive Audio product for over three months now.

Our voice application is a prototype voting application used to study accessible voting in political elections.

We wanted a product that would allow callers with varying telephone script navigation skills the ability to speed up or slow down the caller dialogue as needed. As a very important requirement of the project, we needed a way to do this without making expensive custom changes to the voice application itself. Interactive Digital’s Adaptive Audio product accomplished this for us with only a single line of code added to our application.

The software performs as advertised, speeding up the dialogue flow for more skilled callers while allowing elderly, hearing impaired and less skilled callers to have a more productive call experience. I am impressed with both the product and support we have received from Interactive Digital.”

To learn more about how dynamically adjusting the call experience to suit individual callers helps your bottom line, contact Interactive Digital today.

Apple, Ally Bank, Amazon and Zappos. What do all of these companies have in common?

Besides great products, innovative marketing strategies and extensive product eco-systems, they all deliver on one more essential ingredient: Exceptional Customer Service.

Take Apple for instance. I brought my MacBook into the Genius Bar at my local Apple Store a few months ago. Two of the four little feet that keep the MacBook off the desk fell off. Despite the fact that the machine was more than 3 years old and that I did not have an Apple Care service plan, they still took the machine in the back and glued on four new ones for me. And while I had to make an appointment for this, but how about that there’s even such a thing as a Genius Bar in a retail store to accommodate me in the first place?

I’ve had many similar first rate customer experiences at Amazon, Best Buy and even some small retailers I have done business with through eBay. Responsive email and chat support, free shipping both ways, customer reviews right on the retailers own web site, extended product return and replacement policies and so on. Todays recession weary, highly connected consumer is demanding and changing business models and the customer service landscape as we speak.

Read more…

By TMCnet Special Guest
Luis Gonzalez

Busy contact center managers often overlook their customers’ feelings. They tend to focus on easily quantifiable metrics such as abandonment rates, service levels, staffing levels, etc. While these metrics are important to the bottom line, there is also tremendous value in focusing on the consumer’s experience. This is where behavioral science comes into the picture. Businesses can leverage psychology to improve customer service in contact centers.

In 2001, Richard B. Chase and Sriram Dasu published the Harvard Business Review article “Want to Perfect Your Company’s Service? Use Behavioral Science”.They detailed a study that examined encounters between customers and service providers to understand how these experiences make customers feel. It was the first look into how behavioral science principles can advise customer service operations. In 2010, John DeVine and Keith Gilson built upon this research with the McKinsey Quarterly article, “Using behavioral science to improve the customer experience”.These articles provide a roadmap for contact center professionals to improve customer satisfaction at low costs.

Perception is Reality.

Read more here

Are you losing business with an IVR that does not “listen?”

My 18 year old son was making a call to buy concert tickets with his new smartphone not so long ago. He looked up the concert’s web address on the phone, got the toll free number and rather than give the credit card information online and rack up fees on his data plan, he decided to call. Fairly common these days since many voice apps receive more than fifty percent of their calls from a mobile device.

At the start of the call, James had a strong enough cell signal and he navigated the IVR call script fairly easily. After a minute though, the signal weakened and he started to get frustrated with the fact that the IVR was not allowing him enough time to respond. It was bad enough that he has to deal with the degraded audio quality and latency issues, but now the IVR is constantly asking him to repeat himself. He’s a paying customer with valid credit card, ready to make a purchase and about to get even more frustrated as he tries to secure his concert tickets. As the IVR started it custom designed, tiered re-prompting dialogue, he hears only that he needs to keep repeating himself. He starts to think about how Google might find him a better price on a concert ticket from another source.

The call experience James is having is not at all uncommon in today’s retail environment.

Read more…

Two independent trials were conducted to determine the effects adaptive technology would have on the performance of the IVR.

The voice application used was designed to handle caller inquiries for medical insurance claims, benefits, member coverage and general information. The voice application served primarily members (generally novice users) and providers (generally expert users).

During Trial 1, Audio Playback Speeds of 100, 110, 114, 117 and 119 were used. During Trial 2, these were changed to 85, 88, 92, 100, 110, 114, 117 and 119. For both trials, the Adaptive Audio speed changing algorithm was used. The data in Tables 1 – 6 that follow illustrate production results gathered during these trial periods. These data indicate improvements in IVR Utilization and Average Handle Time (AHT), along with a reduction in the number of errors experienced by callers.

As Table 1 shows, with adaptation, the mean number of IVR turns was 22.95% (19826/86405), with a 95% adjusted-Wald binomial confidence interval [6] ranging from 22.67% to 23.23%. For the standard implementation, the mean number of IVR turns was 31.41% (428820/1365172), with a 95% binomial confidence interval ranging from 31.33% to 31.49%). Because the binomial confidence intervals did not overlap, the difference was statistically significant (p < .05).

Table 2 shows the mean number of IVR turns for adaptive and standard were, respectively, 4.4 and 3.2. This represents a 36.90 percent increase in utilization of the automated system when adaptation is used. Read more…

Every caller to a speech application has his or her own individual set of aural, speech, hand-eye coordination (as used in DTMF keypad entry) and material comprehension skills. Add to this environmental variables such as background noise, poor mobile phone signals and caller distraction, and it becomes clear that each call to the IVR system is truly a unique interaction.

This is one of the main reasons human operators are so good at handling any type of call – they can handle the dynamics of human conversation intuitively and with ease. Callers know this all too well and will often opt for an agent the first time the speech system fails to be productive for them.

To the extent that the speech application can monitor and adjust to the behavior of a particular caller during the call (an adaptive approach), a proportionate number of automated calls can be more efficient and productive.

Based on production metrics gathered at various installations, Figure 1 illustrates the relationships between average call length, the number of script levels and the effectiveness of Adaptive Technology at optimizing the call process. These data indicate improvements in the Average Handle Time (AHT), Average Handle Rate (AHR), IVR Utilization (IVR turns per call) and Caller Input Error Rates.

With call centers, enterprise IT departments and ASR-based hosting centers recognizing the economic benefits of automated calls versus using an agent, the trend is towards longer, more complex and information-rich speech applications. Read more…