As chatbots become table stakes in customer service, financial institutions are beginning to innovate beyond the basic bots that support customer inquiries and address FAQs. As the technology and use cases evolve, chatbots could be the key to accelerating omnichannel banking and personalized customer service across the industry, banking executives and analysts said during a webinar yesterday hosted by Informa Financial Intelligence.

“Bots have the potential to store together different areas and interactions that we normally have on a number of different devices and locations,” said Chris Ward, principal consultant and chatbot expert at Informa. “We’re actually creating a more holistic digital proposition that can sit across channels.”

NatWest’s artificial intelligence lead, Ankit Chhajer, looks to the Southeast for answers and inspiration about chatbot innovation for the U.K. bank.

“Innovation that’s happening in countries like India and China … they really upped the game in terms of creating a true ecosystem, where it’s not just about solving the conversation but it’s actually end-to-end needs for the customer,” Chhajer said during the webinar. “If you look at WeChat, it baffles my mind that 45% of all mobile payments in China are done on WeChat.”

NatWest, the London-based subsidiary of RBS, has been expanding its digital assistant, Cora, since its 2017 launch. According to Chhajer, Cora has answered 12 million customer inquiries and the bank has enlisted “an army of conversationalist analysts” to improve the chatbot’s ability to answer questions independently, a metric called “containment rate.”

Cora’s containment rate has doubled in the past year, Chhajer said, noting that NatWest has expanded the bot’s capabilities to answer account-specific inquiries and is using automation to make it easier for customers to complete day-to-day banking tasks through the bot. NatWest serves more than 14 million retail and commercial clients across England and Wales.

Also read: Tech focus: Do chatbots have a long-term future after COVID-19 bump?

Meanwhile, Boston-based Eastern Bank has taken a different approach with its chatbot strategy. Initially, the bank wanted to create one “mother of all bots” for every customer inquiry, but it swiftly switched gears after realizing this was not the most efficient way to solve customer service issues, according to Heather Allen, senior vice president and director of customer service at Eastern.

Instead, the $11 billion dollar bank built smaller, more agile bots for specific use cases, which afforded much deeper engagements that could be analyzed through frequent sprints. The bot also helped the bank add clarity to its website during the pandemic and would offer customers financial tips, such as “the top five things you need to know right now,” Allen said.

Despite the potential offered by chatbots, the financial services industry is still only at the tip of the proverbial iceberg when it comes to innovating beyond the basics and providing personalized and omnichannel customer service.

“Where I don’t see many [companies] playing yet is, obviously, that full virtual assistant, but this is still an important thing to keep your eyes on because this is where those nonbank big tech brands play,” Ward said. “They create solutions that can tie together many disparate solutions and products.”

Bank of America’s Erica is the financial institution closest to that ideal, Ward said, because “it never had the ambition to be merely a helping support bot or, indeed, a transactional bot.” Erica’s ability to proactively communicate insights to customers through the chat and complete tasks like scheduling branch appointments — in addition to being voice-enabled — are key features that make the bot stand out among its peers, he said.

As financial institutions look to provide a financial experience that’s embedded into customers’ lives, they’ll need to navigate the “white space of opportunity in terms of ‘how do we actually start to figure this out,’ and I think we just haven’t even really got to the point where banks are able to fill in data and then think about it and use it and deploy it,” Ward said. “There’s just a very long way to go.”