Banks often struggle to get a unified voice of the customer across channels, including feedback from call centers, chatbot messages, social media and branches. Much of the data they collect is siloed and static, coming from surveys taken at particular points in time.
The lack of a comprehensive, evolving view of customer feedback can impede improvements to products and services — a problem Truist set out to solve in early 2023.
The bank began work on a tool called Truist Client Pulse that uses artificial intelligence and machine learning to pull together insights across channels, executives said. The goal is to offer near real-time analysis of customer sentiment across business lines. To realize its goals, the team needed to iron out a few major considerations: in-house versus third party, internal approvals and what to do with the feedback — some of which they're still tackling.
"This was a big idea we had around really reimagining how we would measure client experience and taking all of those different data pieces and parts … [to] harness the power of the data and advanced AI to aggregate all of that feedback from different channels into a singular, comprehensive view," said Sherry Graziano, head of digital, client experience and marketing at Truist.
The effort culminated in two core offerings: a near real-time customer feedback insights tool based on an analysis of up to 30 million client conversations from multiple channels; and the Truist Client Pulse Score, a metric designed to measure customer sentiment and experience across client segments, channels and products.
The insights tool, currently being rolled out among select internal teams at Truist, will be launched companywide in the next quarter, Graziano said. The bank hasn't yet determined a launch date for the Truist Client Pulse Score.
A horizontal view of customer feedback
The bank's insights tool takes the form of a dashboard that brings up near real-time sentiment trends and highlights areas for improvements across segments, products and channels.
Client feedback analyzed by the dashboard could include information such as key themes that are raised, what's going well and what's not going well. It can break the data down by segment, geography and type of product, Graziano said.
The goal is to understand what a particular customer segment is experiencing so bankers can recognize patterns — including areas of concern or satisfaction — and suggest and implement product or service recommendations, said Tom Mazzaferro, chief data and analytics officer at Truist.
One hypothetical example is speeding up replacement card delivery.
"Maybe you're having a challenge with how quickly you're able to deliver a replacement card to a client. You will see that pop," said Graziano, referring to client sentiment analysis. "It's bringing that full view together to be able to go back to your payments leader and say, 'Hey, based on what we see happening, we might need to go and revisit a process.'"
A key advantage of this tool is the ability to gather near real-time customer feedback and make changes more quickly than traditional voice of the customer data-gathering methods, executives said. The bank so far doesn't have a release date for Truist Client Pulse Score, and the scoring methodology is still being developed, Graziano said.
'A small and mighty team'
The project was led by a core team of 10 and supported by more than 80 employees from across the bank, all reporting to Graziano and Mazzaferro, the executive co-sponsors. The team was able to move the project from ideation to execution in 12-18 months, a relatively quick rollout made possible by the determination of the team and the support they received across the bank, executives said. The initial proof of concept was rolled out in the spring of 2023, using three years of data.
"It was a small but mighty team," said Graziano, noting that the core group was supported by staff from business, technology and data teams. Team members "felt very passionately that we needed to really make sure we double down on taking all that data and bringing it together," she added.
The core team included Harveer Singh, chief data officer for consumer and small-business banking, and Ritesh Rao, senior vice president for digital strategy performance optimization.
Truist Client Pulse was built entirely in-house, a decision the bank made because of a desire to own customer experience wherever possible and significant cost avoidance, said Graziano.
Delivering the project in-house resulted in cost avoidance of up to $10 million, according to Graziano. The improvements from the project could deliver "upwards of $35 million in cost savings," tied to a 10% drop in customer complaints, she said. The long-term plan is to integrate the insights — and eventually the Truist Client Pulse Score — into CRM systems and other frontline tools.
Tiffani Montez, principal analyst at eMarketer, said the ability to aggregate and analyze data across channels is a major win for the bank, as customer feedback information is often siloed and challenging to put together in actionable ways.
"You could potentially comb through a transcript and identify key words that you're hearing in the conversation, quantify how many times that has happened and get the details on how that thing is occurring in real time, versus waiting a month or two," she said.
Creating a tool to unify various forms of customer feedback is an impressive undertaking, but the challenge is figuring out the best ways to use these insights to inform product and service improvements, said Kate Parker, senior director of customer intelligence at analytics and AI software provider SAS.
While Truist analyzes customer feedback data across channels, it still plans to listen to and understand direct feedback — including calls and other touchpoints — as aggregated insights only tell part of the customer sentiment story.
"We will always be listening to complaints, which is a compliance requirement," Graziano said. "Where the insights can become incredibly valuable is driving out some of the complaints."
