Subscribe / Unsubscribe Enewsletters | Login | Register

Pencil Banner

How voice analytics helps iiNet hear customer problems on its network

Brad Howarth | July 27, 2015
Your customers are usually more than willing to tell you what is wrong with your business – and occasionally even what is right. Unfortunately, however, many businesses just don’t have the capability to listen effectively to all of that volunteered information and act on what they hear.

Your customers are usually more than willing to tell you what is wrong with your business - and occasionally even what is right. Unfortunately, however, many businesses just don't have the capability to listen effectively to all of that volunteered information and act on what they hear.

But for Perth-based telecommunications company iiNet, which prides itself on industry-leading customer satisfaction scores, responsive listening is a key requirement to providing good customer service.

Thanks to the introduction of voice analytics technology from Genesys, its customers have now become a principle source of information regarding service problems on its network, even before technical crews are aware of them.

iiNet's chief information officer, Matthew Toohey, said the company has long had a policy of recording calls for training and development purposes, as part of a continuous improvement program. But even with a team of five staff dedicated to sampling calls, they are only able to document and analyse between 100 to 200 calls over a fortnightly cycle.

"For us to be able to ascertain why a customer is calling and understand their perspective, we really need to listen to that customer's conversation. But when you look at the total call volume into iiNet, it is a thousand times that," Toohey said.

"We are potentially missing a significant number of data sets. Issues can arise due to weather, degraded infrastructure and all sorts of different things, so unfortunately with that model we are always at least two weeks out."

Toohey said the goal for iiNet was to find a way to mine information from its entire call volume at near-real time.

"What we are doing is trying to determine if there an underlying systematic problem or customer issue, and I want to know as quickly as possible if this is impacting 5000, 10,000, or 50,000 customers so I can redirect resources on to mitigating and correcting," he said.

In late September, the company began implementing voice analytics technology from Genesys. The system went live in December and is capable of monitoring all calls and comparing words used by customers to determine patterns that might indicate a service problem.

According to iiNet's voice operations manager, Nick Samios, in one instance the system detected an intermittent problem with the company's customer email service in WA.

"It was intermittent in nature, so it was hard to detect from traditional network and system monitoring," Samios said. "We learned about it quite rapidly using speech analytics, to see that west Australian customers had increasing contact about mail problems and were unable to access email.

"The time taken to actually analyse the particular situation was maybe half an hour, whereas in the past that would have been a multi-day exercise to achieve the same outcomes. And I doubt very much whether you would get to the same resolution in terms of the value of the data as well."

 

1  2  Next Page 

Sign up for CIO Asia eNewsletters.