Artificial Intelligence

Artificial Intelligence, Deep Learning, Machine Learning

Retail banks and GenAI: beware of “silent failure”!

Mar 12, 2024 | Ai | 0 comments

80% of retail bank executives believe that GenAI represents a significant step forward. However, only 6% have developed a roadmap for transformation at scale. Silent failure” looms.

A possible “silent failure” despite general enthusiasm for GenAI. Clearly, retail banks are hesitant even though, as the 20th edition of the Capgemini Research Institute’s World Retail Banking Report shows, executive committee members plan to increase investment in digital transformation by up to 10% in 2024.

The report is clear: most retail banks are ill-prepared to thrive in the future of intelligent banking. Globally, only 4% of retail banks scored highly for team engagement and technology capabilities. 41% had to make do with an average score.

Capgemini believes that focusing on intelligent solutions that incorporate AI-based capabilities will enable banks to overcome current structural challenges and ensure sustainable growth over time. However, success must be measurable. Of the banks surveyed, only 6% have established KPIs to measure the impact of AI on an ongoing basis. And here’s a surprise: over 60% of banks are still in the process of identifying and developing such criteria, while 26% that have already put them in place are not measuring them.

An “AI observatory” to monitor, control and report on AI

According to the report, banks are at risk of succumbing to the “silent failure of generative AI” due to the late achievement of disappointing results from their experiments with this technology. For example, only 2% of executives report that they regularly monitor the key performance criteria of the economic impact of their GenAI performance. Furthermore, 39% of executives say they are dissatisfied with the results of their AI use cases. This further reinforces the gap.

To avoid this, the study suggests that banks should set up an “AI observatory” to track, monitor and report on the real impact of AI and generative AI, when implemented on a large scale.

A year after GenAI became a major topic for boards, we see that banks risk falling behind technologically if they don’t quickly adopt these solutions to start leveraging its capabilities,” said Nilesh Vaidya, Global Head – Banking and Capital Markets, Capgemini. Generative AI can be a differentiator when used responsibly and wisely across operations. We also need to redouble our efforts to make generative AI explainable and transparent. We need to act now. To begin with, establish practices that reinforce trust and intimacy with customers. Success will depend on the development of a roadmap based both on enthusiasm and on a pragmatic, traceable and measurable approach.

Bank employees in favour of GenAI-based co-pilots

GenAI offers considerable potential for improving efficiency and customer experience across the retail banking value chain. More than two out of three bank employees (70%) focus on operational activities. Worse still, this proportion rises to 91% for employees responsible for customer integration. This leaves little time for interaction with customers. More than 80% of employees give an ‘average’ rating to the effectiveness of automation in their functions (reception, credit, marketing, contact centre), revealing a significant gap between the bank’s objectives and reality.

However, employees are enthusiastic about the potential of GenAI’s co-pilots to automate fraud detection, data visualisation and analysis, as well as writing and sending personalised content to customers.

The report shows that banks could optimise up to 66% of time spent on operations, documentation, compliance and other customer onboarding activities through AI-powered intelligent transformation and generative AI co-drivers.

Conversational AI could reduce customer call abandonment rates

The pandemic has shifted customer service offerings to digital channels, and self-service tools such as chatbots have become the norm. Despite this change, customers are expressing dissatisfaction. Nearly two out of three (61%) have contacted agents because they were dissatisfied with the solutions provided by chatbots. A further 17% said they were simply wary of chatbots, preferring human agents.

Traditional chatbots based on precise rules lack the flexibility and adaptability of advanced AI-driven systems. In fact, they are unable to handle complex or unexpected queries. Over 60% of customers consider their experience with chatbots to be only “average”. This means that the call abandonment rate is on the rise. Capgemini estimates it at 12% for large banks. And almost 18% for smaller banks, on a global scale.

According to the report, banks should create intelligent contact centres that leverage chatbots with conversational AI capabilities and intelligent co-pilots to help agents with their day-to-day tasks.