2degrees Improves Efficiency and Insights with CXone Mpower
Full-services telecom company 2degrees offers a range of mobile, broadband, and electricity services to more than 2.3 million customers in New Zealand. Its customer base had numbered about 1.6 million, but it grew after merging with the Vocus Group in 2022 and migrating its mobile customers onto the 2degrees network this summer.
Having that many customers can pose significant challenges for any customer service department, and 2degrees was having problems with call summaries and post-call work by agents, even with technologies from NiCE in place.
Then NiCE upgraded its CXone Mpower solution with generative artificial intelligence a little more than a year ago, and the results have been amazing, according to Wayne Anderson, 2degrees general manager of digital delivery.
“It’s been an interesting journey—a little bit of a roller coaster, Anderson explains. “We were using the technology before it was actually production-ready. We were using it in advance of the rollout, and we had middling success with it.”
Among the issues, “it didn’t seem like our agents really trusted what was being generated, and we could spot lots of errors. We thought that the technology had a lot of potential upsides, but there were a lot of problems as well,” he recalls.
That all changed in mid-2024, when the product received a generative AI update, Anderson says. “Overnight, that product went from sort of interesting to [something that] could actually replace what our humans are doing.”
After piloting the product with great success, 2degrees rolled it out across all its contact centers. That was in late 2024, with positive return on investment realized as early as February of this year.
However, getting to positive ROI required more than simply adding the technology. “When we were running this software as a pilot, the team members who were involved were very engaged. They wanted to use the tool and get feedback; it was all very positive. And it showed great results.”
But when the technology was rolled out to the entire team, most behaviors didn’t change, Anderson explains. “The system would generate the summary, and the agents would spend a fair amount of time correcting minor details, adding information that was superfluous to our requirements.”
So 2degrees spent three months focusing on change management, explaining to the team what was expected from the technology and how it was to assist agents in their jobs. Once the agents embraced that, the technology became a true time saver, according to Anderson.
Without the automation of summaries and decreased call handling time, 2degrees would have needed to add 50 full-timers to reach the same efficiency levels it has today.
NiCE’s CXone Mpower has also unlocked business insight that 2degrees didn’t have before, Anderson says. “We had used a crisis interaction analytics model. Having access to the auto-
summarizations has enabled us to get better insight.”
For example, team leaders can ask about potential issues, such as problems with prepaid customer billing, appearing in the last week or any other time frame.
The automated summarization also captures more critical points about the interaction. Initially, the automated summarization and human summarization each captured six or seven critical points. CXone Mpower with genAI captures about nine critical points about each interaction, providing 2degrees with much more customer insight, according to Anderson. “That is a significant step up with our data. It’s not a specific KPI measure, but that’s probably what I value most from the whole project, even more than the productivity savings.”
Anderson expects CXone Mpower to provide more analytics in the future.
“My ambition is to use order summary and interaction analytics to reliably determine what is driving the call. Then we can see through expert and copilot integrations what our standard operating procedure is for that call; if the agent is following the procedure; if the call was resolved; and if the sentiment was positive.”
The Payoff
Since incorporating the generative AI update of NiCE’s CXOne Mpower platform, 2degrees has seen the following results:
- the average handle time has decreased by 90 seconds per call;
- the number of calls escalated or transferred has gone down; and
- the increased contact center performance has equaled the work of 50 full-time employees.