By Rahul Kumar, GM & VP of Financial Services and Insurance at Talkdesk
The banking industry continues to face a number of growing challenges – and margin compression is high on leaders’ lists to tackle. Last year, pre-tax profits across the sector fell by more than £3 billion – the first major drop since the bounceback after the pandemic.
Commoditised product offerings have led to a race-to-the-bottom fueled by pricing, while in the background, rising operational costs, slow income growth and reduced productivity leave little room for manoeuvre. Combined with increased customer expectations and a volatile market, the urgency for change becomes ever more apparent.
But banks know change is necessary. In one survey identifying top priorities for banking organisations, using workflow improvements or automation to drive efficiencies came top, with 82% of organisations selecting this as a key area of focus.
Today’s financial leaders face a dilemma: reduce cost-to-serve while enhancing the customer experience. Historically, improving one has meant sacrificing the other.
AI, however, presents a brand-new opportunity to achieve them simultaneously, driving both efficiency and differentiation. In fact, Gartner predicts that by 2029, agentic AI will autonomously resolve 80% of common customer service issues, reducing costs by up to 30%.
The implications are huge: faster resolution times, a lower cost-to-serve, and the chance for brands to interact with customers at a level of personalisation that was previously impossible at scale.
What do consumers expect from their bank?
For banks looking to make their customer experience a key differentiator, there are several areas ripe for change.
Firstly, is the desire for more personal connection. In the UK, some 6,561 high street bank branches have closed in the last decade. While the rise of digital banking services like mobile apps mean that services now offer greater convenience, the consumer-bank relationship is at risk of becoming increasingly impersonal. Though self-service helps with basic tasks, it offers nothing to serve emotional or more complex needs.
Consumers want guidance, not just goods
How banks interact isn’t the only thing that matters to consumers, it’s also what they deliver. As the cost of living continues to be a major concern, the vast majority are looking to their bank to provide them with personalised financial advice.
But less than one in four (23%) consumers rate their main bank highly for its range of services and the competency of its tailored financial advice, highlighting a mismatch between what banks offer and what consumers desire from them.
How can banks deliver?
The real breakthrough comes from multi-agent orchestration, using multiple specialised AI agents working collaboratively together. This approach takes customer service automation away from single enquiries and instead towards automating entire workflows end-to-end.
Take fraud reporting as an example. If a fraud trigger occurs, one agent could contact the customer to confirm the activity is fraudulent, while another automatically initiates an account freeze. A third agent could open a case for review.
This coordinated approach allows for a smoother experience for the customer, with a much faster resolution time; the bank benefits from lower operational costs, fewer errors, and more satisfied customers.
And by integrating real-time data analytics and customer insights, these interactions can be personalised beyond generic personas and demographics, allowing banks to offer truly valuable customer experiences, efficiently and at scale.
Staying ahead of the competition
The onus is on banks to act quickly. Working now to build a secure AI foundation can help banks provide much-needed relief on margin pressures, while meeting the expectations of modern consumers.
Interactions that feel personal, even if they’re not in person, will be a key differentiator between the banks that simply offer a transactional service and those that are seen as trusted partners – and multi-agent orchestration is set to be a key enabler in the change.
