Tips to Balancing Contact Center Automation Costs and Quality

PAY ATTENTION TO PERFORMANCE

When it comes to automation, cost also needs to be balanced with system reliability. No matter what kind of technology is in place, it had better work properly.

Effectively serving customers is dependent on a highly functional self-service environment that includes not only the IVR and related applications but also the supporting speech recognition, text-to-speech, telephony infrastructure, and data connections. Each piece of the system must operate precisely as it was designed or the bottom line could suffer.

If the technology doesn’t work properly, it could send the caller down the wrong path. Once that happens, he’s often left to hang up the phone and call back. “Now your call abandonment rate goes up and your call volume goes up, which is the exact opposite of what you intended,” Herriman says.

The financial ramifications of such a scenario could be great. In fact, getting trapped in automated phone systems is the top customer service–related reason businesses lose customers, [24]7 reported in its “2016 Customer Engagement Index.” More than a third (37 percent) of the 3,500 consumers in the survey who said they’ve ended business relationships because of poor customer service cited frustrating experiences with IVRs.

Part of the reason for poor IVR design is that with the rise of digital channels, many companies have shifted their focus to improving the Web and other self-service channels, forgetting the phone channel.

The research found that 64 percent of consumers begin their customer service journeys on the Web and, of those, half (32 percent) pick up the phone when their issue isn’t resolved, only to be trapped in a never-ending IVR loop with little hope of reaching a real human.

Additional research from the Marchex Institute supported this idea and also found that as many as 11 percent of callers hang up during IVR interactions because the systems are too confusing or frustrating to use.

To draw that conclusion, Marchex used its Call DNA conversational analytics to map, classify, and score more than 2 million calls placed in 2015 by consumers to businesses. Busby says the reasons for the hang-ups usually come down to companies trying to do too much with automation. “A simple one- or two-question IVR usually passes most consumers’ acceptability test,” he says. “Once you get past that, customers start to get frustrated and hang up.”

Anadkat agrees. “If it requires more than one action on the part of the customer, those are the kinds of things you don’t want to automate,” he warns.

Other hang-ups can be attributed to people not understanding the IVR, not finding what they want, or needing to start over after winding up in the wrong spot, according to Busby.

Anadkat has an even simpler explanation: In their quests to move interactions along as quickly and efficiently as possible, companies can easily “start with any technology and make it completely unfriendly,” he warns.

And while it might seem counterintuitive to throw more automation at an unfriendly, already damaged, and poorly performing automated system, that’s exactly what experts recommend. Modern IVR testing solutions can provide key insights into IVR performance and the problems that consumers are having during their interactions with automated phone systems.

Such applications remove the guesswork from measuring caller satisfaction with IVR self-service, according to Michael Boukadakis, CEO of Enacomm, a call center technology provider that in March released the ViA 3400 IVR testing solution. “Organizations can determine what IVR functions callers use most, why callers start self-service but fail to finish, where and why callers transfer out of the IVR, which campaigns are most profitable, what calling or usage trends are developing, and much more,” he said in a statement.

Crichlow points out, though, that in many cases where customers abandoned a company because of bad experiences with automation, the IVR was likely “the final straw” in a series of other problems that customers had with the company.

“IVR is 30- or 40-year-old technology,” he adds. “It has successfully handled hundreds of billions of calls over the years.”

Companies used to be able to blame their IVR problems on faulty speech recognition, but that is no longer a valid excuse, according to Crichlow. IVRs, virtual agents, and routing systems have been enhanced with natural language processing, predictive analytics, and machine learning that enables them to work extremely well today, he says. “The fundamental issue is no longer with the technology itself. The speech [recognition] has gotten so much better, and, thanks to Siri and Cortana, consumers have gotten more comfortable using speech.”