This article first published in Feb 2018 issue of Asia insurance review and reproduced with permission from Asia insurance review.
It’s never been more important for life insurers to offer their policyholders high-quality service. Digitisation and intensified competition with other financial service verticals have blunted prospects for growth and made customer acquisition and retention ever more challenging. Meanwhile, consumers’ overall service expectations continue to rise as retailers in other industries set a higher bar. Insurers have struggled to keep up because of the inherent complexity of both their products and their systems’ environments.
Bringing cognitive computing to customer service will allow insurers to move past these challenges and deliver a level of digital service comparable to that of any other industry.
Insurers have had a mixed record in their service quality journey, making progress through the adoption of better systems and training, but also making mistakes in self-service and over-investing in interactive voice response systems. Customers grew to fear that an attempt to resolve a problem via the company’s call centre was likely to result in a runaround, and those getting shunted from one operator to another were likely to press zero to talk to a person anyway.
Improving life insurance service during the age of e-commerce has largely been a matter of getting right what humans and machines do well and understanding the important role that the customer service representative (CSR) plays. Insurers have made significant gains in that regard, for example, with call-scheduling applications that reduce holding time and the use of dashboards that inform CSRs about call history and can surface potentially relevant account information.
Cognitive computing has the potential to improve some CSR functions – but more importantly, it can transform the CSR role as we further identify what humans and machines do best and augment human processes through the mediation of machine intelligence.
The way it is
To understand the potential for cognitive computing to drive customer service evolution, let’s consider the way insurance CSRs operate today. CSRs aim to resolve a customer’s request in the least amount of time. The prevailing state of technology defines most customer interactions as a discrete linear process – beginning with validation and authentication, proceeding to accessing customer data, and concluding with a transaction request and closing out the call.
The technology places a variety of constraints on CSRs, which in turn affects the customer experience. CSRs typically work with multiple ‘green screen’ interfaces capable of retrieving very limited information. They need to memorise a variety of codes and they must often supplement their computer interface with paper sources about products and systems that can be difficult to find quickly. These challenges are exacerbated in an insurance systems environment where CSRs need to navigate manually among multiple systems.
These challenges have shaped the particular skills required of today’s life insurance CSR and they have saddled the role with a training burden that’s both onerous to the individual and costly to the employer.
CSRs generally spend eight hours a day on inbound and outbound calls, aiming to meet service targets for individual calls and production goals for their cumulative work. They spend a considerable amount of time learning to access a variety of source systems to address customers’ requests. Typically, they are assigned to specific products or ranges of products and there are limits to the complexity of questions they can be trained to handle.
In short, the way insurance companies are organised leads to system inconsistencies that are reflected by multiple, insular service environments. Within these service stovepipes, individual CSRs have a high ratio of training and meeting time to work hours and a low ratio of CSRs to the number of products serviced – that is, CSRs must train and work hard to respond to customer requests for a limited number of products.
The insurers that employ CSRs face the burden of high initial and ongoing training costs for them, as well as attracting and retaining talent. The job is relatively demanding, proportionate to the way associates are compensated; insurers struggle to plot a career path for those who succeed; and young talent entering the workforce find the technology outdated. According to the 2016 US Contact Center Decision Makers’ Guide from ContactBabel.com, the average annual turnover rate for a CSR was 29%, with an ‘average life span for a call centre worker in the United States of about 3.3 years.
In contrast, cognitive computing has become an attractive new area of technology that covers a range of capabilities that can improve CSR effectiveness. It can introduce significant efficiencies, facilitate systems integration, ease the management of service departments and make important advances in the quality of service.
Augmenting human intelligence
Cognitive computing, as the name suggests, refers to computational processes that mimic human cognition, but it’s the kind of machine intelligence that provides decision support and other assistance to a human. Computer scientists think of this as ‘augmented intelligence’ rather than artificial intelligence.
Cognitive computing reserves the highest kinds of human intelligence for the tasks that most require it, leaving computers to perform lower-order cognitive tasks of a more administrative nature. It serves as a force-multiplier, enabling a person to perform at a much higher level, potentially executing many more tasks and at far greater speed, consistency and accuracy than an unaided human could accomplish in the same amount of time.
Applied to customer service, this kind of augmented intelligence can transform the CSR’s role, while increasing productivity and service levels.
Traditionally, a CSR works within a closely circumscribed process aided by systems of strictly limited output. The CSR needs extensive training to use any given system, learn a range of source systems, create manual workarounds and paper sources of supplementary information and work within the department’s protocols and toward its productivity goals. In the old paradigm, a high proportion of the CSR’s intelligence tends to be consumed in managing each system that’s used and in navigating between them. Within a cognitive computing paradigm, most processes are automated and training is significantly attenuated.
In the new paradigm, the CSR focuses on the customer’s problem through a single interface, while the computer does the low-level, administrative navigation to sources of needed information and presents them at the speed of thought to the CSR in an easily consumable fashion. Instead of being a retriever of information from multiple systems, the CSR becomes a consumer of information at the centre of the act of communication with the client. Using cognitive computing, the CSR receives new insights from all the data and interactions, gains easier ways to ask for the information, and uses the power of cognitive computing to find the most probable answers to satisfy the request. This results in increased customer satisfaction, increased CSR efficiency and lower overall costs.
DXC’s cognitive journey
Computer scientists think of cognitive computing as open-ended because of the nature of the technology; like human intelligence, it adapts and its capabilities can be combined in synergistic ways. The capabilities of understanding, reasoning and learning are the distinguishing features of cognitive computing.
The initial phase of the journey centres on a CSR’s unified dashboard that consolidates access to capabilities and sources in a single interface. With its focus on automation of the basic elements of the CSR’s job, it cuts time to execute tasks and reduces training. Behind the scenes, the dashboard consumes data from multiple sources for advanced ‘smart search’ and topic indexing. It automates the caller-validation process; initiates work orders during calls; and records, stores and analyses call summaries. Among the benefits demonstrated by the dashboard are lower costs, improved service levels and greater first-call resolution, as well as reduced average call-handling time.
There are also advanced analytics on the dashboard that focus on CSRs’ activities by individual or group, to see where a CSR spends his or her time, illuminating customer concerns and exposing any needs so as to refocus training. Analytics also reveal patterns in calls to help CSRs anticipate customer needs. For example, analysis may reveal that 82% of customers requesting an address change will also want to know the current value of their policy. While useful in the short term, this analysis ultimately points to the long-term gathering of intelligence in the company context – constituting a cognitive journey of its own.
The second phase of DXC’s cognitive journey is CSR Assist, which supplies more sophisticated cognitive capabilities to help a CSR talking with clients about life insurance. The cognitive capabilities speed and smooth the CSR’s interactions with clients through Conversational Assist, a CSR-facing chatbot equipped with natural language capabilities and insurance knowledge that can take direction from simple written queries. It can generate relevant responses to customers’ inquiries, find and deliver to the CSR the evidence to support those responses and provide a ranking of the responses. CSR Assist collects data from every interaction and uses it to power continual learning (vs training).
The last stage planned for the cognitive journey extends the kinds of natural language capabilities and insurance knowledge used by CSR Assist into a customer-facing chatbot. Although forcing customers into self-service was one of the great mistakes of recent history, consumers are now demanding increased opportunities for self-service – and not just the youngest consumers. Our capabilities will deliver on that demand, giving customers the opportunity to figure out problems on their own by suggesting solutions through abbreviated responses around-the-clock and providing the choice of interacting with voice or text.
For the foreseeable future, customers will still expect to speak to humans, at least for some purposes, and through our cognitive journey, we look forward to assisting our clients on theirs. Our capabilities will help them to refine their customer service and our scale and accumulating expertise will help us accelerate implementation of advanced capabilities and share insights from our installed base. CSRs will enjoy a less daunting, more rewarding role, with fewer handoffs and more first-call resolutions. By focusing their human qualities on customers’ needs, CSRs will provide more confident and effective service – achieving the ultimate goal: greater customer satisfaction in a competitive business environment.