Can Empathetic Machines Create Unprecedented Customer Engagement At Scale?

By Supryo Sen, EVP and GM, Global Head of Financial Services, Optimal Strategix Group, Inc.

Supryo Sen, EVP and GM, Global Head of Financial Services, Optimal Strategix Group, Inc.

Much has been written about the emergence of AI as the next frontier in technology and the limitless applications of machine learning to unlock dark data hidden way in the depths of massively distributed global server farms or tucked away in the cloud. As we adopt more and more machine, and the machine mimics more and more of the human - it is important for us to remember why we are doing so. Simply put, to solve very deep human problems and to enhance human value. Robotic process automation is merely a start. For unless new domains of human value are created, true enterprise value cannot be created as we will continue to fight for a fixed or worse, shrinking share of the pie.

To create increased human value, enterprises must deeply learn real human behaviors in all its glorious complexities. And then let the machines learn to empathize. Imagine EAI, Empathetic Artificial Intelligence, software truly at our service, and not just as a service.

Companies that focus their AI initiatives and applications to engage with customers with a deeper understanding of drivers of behavior will achieve massive competitive differentiation”

Imagine the situation of a hypothetical health insurance company in an emerging market. Perhaps founded less than a decade ago. Aggressively pursued customer acquisition with less attention to portfolio risk. Result: a large customer base with adverse emerging health experience and substantial losses due to inadequate pricing and anti-selection.

A couple of CEOs later, the Board hires and charges a new CEO to prioritize the portfolio from hemorrhaging and to turn the ship towards profitability. How to right the wrong for shareholders and do right by the customers?

The new CEO starts with data to learn about the customers, but the data is highly fragmented, strewn across multiple systems and the quality is poor. Big data, little data, good data, bad data - after being inundated with all kinds of data, she challenges her team to find the right data to learn more about the customers and to improve portfolio profitability. After several tense meetings between functional heads, the team presents their plan to the new CEO. They proclaim:

• We will segment the customer base 
• Understand lifetime value and retention risk of each customer in the portfolio 
Launch differentiated interventions based on customer lifetime value (CLTV) and retention risk 
For instance, for high CLTV and high retention risk, we will aggressively pursue a retention strategy 
For low CLTV and low retention risk, we will launch health and wellness programs to mitigate claims risk 
And we will do so sensitively, with a deep understanding of customer expectations to promote loyalty

To keep costs low, the team then rapidly deploys data lakes with key variables predictive of the models that are built for customer segmentation, CLTV, claims and lapse propensity. The build of the expensive data warehouse is also put on hold. There is an intense effort to collect, cleanse and aggregate only predictive data to solve prioritized and focused problems and to demonstrate ROI rapidly. Lacking customer behavior data, it is directly collected from customers through primary means and bended with historical transactional data. This hybrid data set results in a view not only reflective of past behaviors but is also highly predictive of future customer behaviors.

Once the dashboards are deployed and the customer insights are obtained in real-time, the team focuses on ways to operationalize the data and analytics to empower the frontline into action. To do so, the team quickly reverts back to their framework of differentiated interventions measured by CLTV and retention risk. For high CLTV and high retention risk customers, dashboards are deployed at outbound call centers with embedded AI agents capable of tone and sentiment analysis. When customer service reps make an outbound call to an at-risk customer, the correct segment of the customer is profiled in real-time on CSR screens. Segment based differentiated call scripts and recommended talking points are displayed based on real-time customer tone and sentiment analysis.

The result of this highly contextual cognitive touchpoint and empathy is deep customer engagement. Human and machine working hand-in-hand to solve customer issues in a way neither could have independently, efficiently and in the shortest possible time. It is not always perfect. But the system is designed to evolve over time through deep machine learning.

On the other hand, digital health and wellness programs are launched for customers at high risk of claims. Evidence based cognitive robots replicate human coaching and judgment through digital health and behavior modification algorithms, lowering costs. By marrying behavioral insights, cultural sensitivities and responding to the differentiated needs and expectations of different customer segments, superior customer engagement and measurable improvements in health outcomes are achieved through a highly scalable deployment of machine empathy.

End result: A return to profitability with retention of high CTLV customers, and risk reduction of low CLTV customers. Most importantly, an internal cultural transformation begins to take hold at the firm based on obsessive customer-centricity.

AI is indeed the new giant frontier in technology. Companies that focus their AI initiatives and applications to engage with customers with a deeper understanding of drivers of behavior will achieve massive competitive differentiation. No doubt, there will be faults and failures. But with tenacity, organizations can overcome hurdles rapidly using traditional and cognitive learning processes.

Enterprises that learn to emphasize deeply with their customers, are aware of their behavioral preferences and priorities, continuously refine their assumptions through cognitive technologies, will be in the best position to engage their customers deeply. For lifetime value maximization occurs most efficiently across the spectrum of acquisition, up-sell/cross-sell, service, retention, loyalty and referrals when customers are deeply engaged. AI when done with empathy has the potential to create customer engagement at an unprecedented scale with very attractive ROI.

Read Also

Unleashing the Power of BI

Unleashing the Power of BI

Mark Pritchard, CFO, Data-Mail
New Generation of Travel & Expense at Gonzaga University

New Generation of Travel & Expense at Gonzaga University

Linda Wilson, Director, Finance System and Services, Gonzaga University
Tips to  Mitigate Technology Implementation Challenges

Tips to Mitigate Technology Implementation Challenges

Regis Quirin, Director of Finance, Gibney, Anthony & Flaherty, LLP

Business Expense Management Special