Truist has a new digital assistant – could it be the best one yet?
- Truist’s new digital agent is built to cater to more than 100 queries, ranging from basic questions about account management to more complex inquiries about financial wellbeing.
- The digital assistant allows consumers to easily access human agents if need be, and works on cutting down the cognitive load for both customers and teammates.
Chatbots are becoming central to most customer engagement strategies, with the market expected to hit $102 billion in 2026. However, as discussed in our previous coverage, without the necessary research into consumer pain points, chat bots can add friction rather than remove it.
Researchers in retail and consumer services point out that although automation can save costs, it may make businesses pay in other ways:
“Despite the economic benefits for companies using chatbots in service encounters, they often fail to meet customers’ expectations, can undermine the customer service experience and lead to service failures.”
Although automation in conversational messaging may seem like a no-brainer from the outside, not all companies that implement chatbots are able to leverage their capabilities properly. Building on one of the biggest bank mergers in recent memory, Truist thinks it has cracked the problem with its new digital assistant, called Truist Assist.
Anatomy of a good chatbot
Available to Truist’s wealth and retail clients around the clock, Truist Assist can respond to more than 100 common queries. The digital assistant is built on top of Amazon Lex, which is an AWS product that helps companies build conversational interfaces. It is the result of a year-long effort from the Truist team.
“This is the first big proof point coming out of the merger, something new that takes the digital channels, tools and technology and couples it with the ability to have the human touch in our teammates’ channels, and our contact center team,” said Sherry Graziano, head of digital and contact center banking at Truist.
As a brainchild of Truist’s Innovation and Technology Center, the digital assistant is built through a collaboration of multiple teams that collectively prioritized client co-creation and teammate satisfaction. As a result, Truist Assist dives deeper into the chatbot-agent handoff than most implementations, and allows for the process to be more convenient for both teammates and customers.
One example of this is how the digital assistant provides teammates visibility into how a customer has been interacting with the system previously. This visibility ensures that customers don’t find themselves repeating information they’ve already fed into the system before. On the other end, teammates receive clients with much greater insight into their goals and history. Simply put, this saves time on both ends of the interaction.
Graziano also emphasized that the product is not built to resemble a decision tree, whereby consumers can’t access a human agent until they’ve already fulfilled certain requirements in the system. “At any point along the way, a client can say, ‘I want to talk to someone’, making an off-ramp that’s available at any time during the exchange.”
Associate Professor Paula Dootson, from the QUT Business School shows in her research, that this is a deceivingly simple feature to miss out on, and one that can have a key impact on customer frustration. “We found that in a chatbot service failure context, telling a customer late in the service interaction that a human employee is available to help can lead to a greater chance of customer aggression,” she wrote.
Perfect customer satisfaction is nigh impossible to achieve, but the successful pursuit of this ideal nearly always culminates in co-creation. According to Graziano, by focusing on co-creating, Truist Assist was able to narrow customer expectations from the digital assistant down to 5 main features:
1) General support: This includes queries that relate to finding information like the hours kept by an office, etc.
2) Transactions and top tasks: Customers can ask queries regarding their transactions and bills, as well as book appointments, etc.
3) Recommendations and insight: Here, customers can gain personalized recommendations about their financial wellbeing and transaction history.
4) General account management: Along with insights, the digital assistant also helps with enacting run-of-the-mill tasks such as keeping track of your cards and accounts, being able to block if needed, etc.
5)Promotions and rewards: According to Graziano, this category helps with “a client maximizing their benefits and the rewards or cashback with personalized promotions.”
Examples of the interface and its range of capabilities can also be observed in the pictures below:
The interface shown above does away with the excessiveness that has become common within UX reimaginations (think: Robinhood), and errs in favor of robustness and simplicity (think: Apple). In an age of confetti-hurling interfaces, it is a relief to come across interfaces that are meant to facilitate – not dominate – customer experiences.
By incorporating clients early in the process, companies can significantly cut down on deployment issues and avoid a mismatch between product features and client expectations.
A good measure of effective customer engagement and messaging channels is the attention the firm pays to the basics: customer-centricity and employee satisfaction. If unsure, breaking down the development at discrete stages and reassessing can be a good strategy.
“We brought in clients to understand what the friction points are. Not only through the creation of the tool itself, but we tested it and we learned before we went into production. When we did go into production, we started beta and a pilot, which were then brought to teammates who were the first users. So, we had the ability to learn what the tool was before we went through multiple geographies of our clients,” said Graziano.
Recently, its been raining chatbots with Wells Fargo coming out of its own through a partnership with Google. Very often, names of technological features or processes like Natural Language Processing and seamless connectivity are used to tout capabilities that have become quite commonplace in the market. While it may work at generating word-of-mouth and intrigue around new products, the strategy can backfire when tools don’t deliver on customer demands.
To avoid this, FIs must refocus on customer-centricity, or else doom their campaigns to the realm of empty hype. So far, Truist Assist has been able to check some of the most fundamental requirements. However, its continued success will depend on how well it can maintain its focus on client needs as the sands of economics and technology continue to shift in the future.