Health insurance provider NIB engages in tens of thousands of online chats every year, personally handling customer queries through a team of online agents.
As it pored over a year’s worth of chat data, looking for way to improve customer service, the company realised that about half of the conversations it processed related to very specific enquiries from customers or potential customers that required a faster route to resolution.
NIB engaged DiUS, which saw a means for customers to engage with their health insurer in another way. DiUS took home the coveted Customer Experience award at last year's CRN Impact Awards and is now vying for another win in the category.
With NIB, DiUS saw the potential to quickly direct customers' enquiries to the correct service area without the need to push them through a live chat agent, and the NIB chatbot program was born.
The Nibby chatbot enabled NIB to proactively direct customer enquiries and streamlined the customer experience by reducing the amount of rerouting required, it also freed up live agents to deal with more complex issues and high-value customer service delivery.
The chatbot project was NIB first foray into natural language understand and automatic speech recognition technologies, powered by deep learning, in Australia.
DiUS developed Nibby using Amazon Lex, the Amazon Web Services, as-a-service interface for building conversational interfaces using voice and text. DiUS worked with AWS and NIB to implement the solution using a lean and agile approach, integrating Lex updates and new user experience updates as the project moved from the proof-of-concept stage all the way through to delivery.
DiUS was able to create that first proof-of-concept within two weeks, giving NIB confidence that Amazon Lex would be a suitable platform that provided enough tools to ensure that the case management solutions NIB were after could be realised. DiUS worked with NIB to refine and create new user flows for Nibbyu to work with.
New proofs of concepts, user testing, new interfaces and ongoing refinement continued and the solution went live three weeks after the project was kicked off, with DiUS and NIB still working together to refine the chatbot as it went into practice.
Two weeks after it was launched Nibby was completing 24 percent of the tasks it received; after 10 weeks it was completing 70 percent of its conversations, with 26 percent of those handled entirely by the chatbot. DiUS said since launch and from continuous improvement, Nibby was handling customer interactions at a peek of 77 percent, and deflecting calls by resolving customer enquiries through directing to online services.
The companies estimate Nibby has to date deflected about 80 hours worth of customer interactions from a consultant and routed the customer to the appropriate specialised channel for improved customer service. With continued development, Nibby is expected to deflect more than 1500 customer interactions per year from contact centre consultants.
NIB Health Insurance chief information officer Brendan Mills said “Thanks to our partnership we have been able to build Nibby in a faster, more agile way, which has reduced the operational complexity and the cost of running and managing the chatbot.”
NIB is now looking at how AI might be able to improve its backend systems as well as other customer-facing activities.
DiUS is a finalist in the 'Customer Experience' category in the 2018 CRN Impact Awards. For a list of all finalists and further details on the awards, please head to the CRN Impact Awards hub. The awards take place during the CRN Pipeline conference. You can get more information and purchase tickets here.