Utilizing generative AI to revolutionize customer service in banking
Since the debut of OpenAI’s ChatGPT, organizations and industries have been actively exploring the potential of generative AI to enhance efficiency and gain a competitive edge. With its ability to create new text, images, code, and other content, generative AI can enhance marketing and sales efforts, software engineering, customer support, and more.
In a June 2023 report, McKinsey highlighted banking as “among the industries that could see the biggest impact as a percentage of their revenues from generative AI.” The report highlighted that if fully integrated, this rapidly advancing technology could contribute an additional $200 billion to $340 billion annually across the banking industry.
While there are many use cases for generative AI in banking, we can expect most banks to take a cautious approach to implementing this emerging technology as they continue to assess its potential and risks. One of those risks is accuracy: ChatGPT and other generative AI algorithms currently are prone to inventing facts, or “hallucinating.”
Prioritizing the customer experience
Given this existing flaw, it would be irresponsible of banks today to allow customers to interact directly with generative AI with no guardrails in place. However, banks can still vastly improve the customer experience (CX) almost immediately by using generative AI to make customer service agents more effective and efficient. As banks learn more about the technology and with the right training mechanisms in place, Gen AI can then be incorporated into customer journeys and processes.
The generative AI opportunity comes along just in time because bank customers have become increasingly dissatisfied in recent years. Data from Forrester Research shows CX quality remaining flat for direct banks, while CX scores for multichannel banks fell for the second consecutive year.
For banks, the ability to deliver a superior CX is inextricably tied to customer loyalty and deposit retention. Today’s consumers simply have too many options to tolerate a substandard CX. They want convenience, choice, and personalized service.
With its ability to respond to queries in an increasingly human-like fashion, generative AI enables banks to meet and keep up with the growing expectations of their customers. Below are some of the ways banks can use generative AI to dramatically improve customer service.
Enhancing workforce productivity. With its ability to create diverse content, including text and audio, generative AI is able to automate routine tasks that consume significant human resources. Generative AI, for example, can quickly produce concise summaries of customer interactions, including tone, resolution, or recommended next steps.
This type of “after-call work” traditionally has been done by an agent immediately following a customer call. Such post-call summaries are time-consuming yet essential to coordinated customer service. By automating data entry, document processing, and more, banks can shave off 30 to 60 seconds off each interaction and redirect human efforts toward strategic activities, ultimately enhancing productivity and operational efficiency.
Delivering exceptional customer experiences. In a competitive banking landscape, customer experience is a key differentiator. Generative AI facilitates personalized experiences on a larger scale by analyzing historical data and real-time interactions. This enables banks to provide tailor-made solutions and services, strengthening customer relationships and loyalty.
Gen AI analyzes every interaction in real-time and can proactively identify areas of friction in the customer experience and where automation could better assist customers. More specifically, banks can use Gen AI to detect fraud by analyzing interaction topics to uncover patterns, anomalies, and signs of compromise.
Fostering innovation. Innovation is a cornerstone of sustainable growth in banking. Generative AI expedites product development by rapidly generating and testing a multitude of ideas and prototypes. The result is faster innovation. As an example, banks historically would train their virtual agents to recognize every phrase a customer could use to check their account balance. With the learning capabilities of Gen AI, the virtual agent now automatically understands the intent of the phrases and no longer needs to be trained on every variation.
Additionally, generative AI aids risk management by identifying potential risks and opportunities through complex data analysis. Embracing generative AI cultivates an innovation culture that keeps banks ahead of market trends.
Strategic implementation of generative AI for customer service
Once banks understand the capabilities, limitations, and risks of generative AI, they can begin building a strategy and taking the steps necessary to ensure implementations that enable support agents to better serve customers. These include:
Harmonizing generative AI with human efforts. It is important to understand that generative AI’s role isn’t to replace customer service agents. Rather, this technology is designed to augment human capabilities. By empowering employees with AI tools, banks enable them to make better decisions, solve complex problems, and focus on tasks requiring creativity and emotional intelligence.
Generative AI can help bank customer support workers do their jobs by:
● Providing automated responses to routine customer questions, freeing up time for support agents to work directly with customers
● Creating personalized scripts for support agents to follow when talking to specific customers
● Analyzing customer feedback to spot trends, detect potential problems that must be addressed, and identify opportunities for new products or services
● Analyzing customer interactions to provide agents with actionable feedback to improve performance
Defining clear objectives. Successful integration of generative AI begins with clearly defined objectives. Banks must identify specific business goals that AI can facilitate, such as reducing average handle-time for contact center agents or enhancing fraud detection. These objectives provide a roadmap for effective implementation and evaluation.
Preparing and integrating data. Data quality and relevance are paramount for generative AI’s effectiveness. Robust data governance practices, data cleaning, and integration with existing systems ensure AI models generate accurate insights.
Choosing the right partner. Implementing generative AI necessitates a multidisciplinary team. But rather than spending money hiring data scientists, AI specialists, and domain experts, banks should consider working with a third party that has the tools, expertise, and experience to maximize generative AI’s potential. This allows banks to focus on the business of banking, not on technology management.
Third-party vendors not only can ease integration and the scaling of Gen AI initiatives, they can help banks meet regulatory compliance and ethical requirements as AI becomes integral to banking operations. Ensuring data security, privacy compliance, and ethical decision-making builds trust with customers and regulators.
Continuous learning and adaptation. To function optimally, AI models must constantly be fed data from which to learn and adapt. Establishing “human in the loop” feedback tools that allow AI systems to learn from outcomes and adapt over time improves accuracy and relevancy.
Laying the groundwork for long-term success
Generative AI initiatives will falter without adequate technological infrastructure, collaboration, and optimization along the way. Here are areas banks must address to support:
Risk assessment. While generative AI has the ability to transform banks’ customer service, a rushed implementation without proper risk assessment can lead to costly problems. Therefore, it is essential that banks identify potential risks and develop mitigation strategies. Banks will need guardrails to prevent generative AI algorithms from producing inaccurate information for customer service agents or customers. Those guardrails should include “human in the loop” trainers who can identify and remove information hallucinated by generative AI.
Scalability and infrastructure. Scalability is pivotal as AI initiatives expand. To support long-term implementations, banks must make sure their systems and infrastructure can handle growing demands. Bank contact centers that run on legacy systems lack the scalability necessary for Gen AI. Fully leveraging Gen AI at scale will require banks to migrate their contact centers to the cloud.
Pilot programs and testing. Small-scale pilot programs and testing validate generative AI solutions before full-scale integration. This approach identifies challenges, gathers feedback, and fine-tunes AI models for optimal performance.
Conclusion
Generative AI has the power to reshape banking by enhancing customer service, boosting productivity, and fostering innovation. By embracing this technology strategically, banking leaders can pave the way for genuine transformation. A balanced approach in which AI harmonizes with human efforts can enable a seamless and successful integration. The synergy between human expertise and generative AI capabilities will bring the banking sector into a new era of unprecedented success.
About Author:
Rahul Kumar is the vice president and general manager for financial services at Talkdesk.