Identifying Customer Needs, Sans Small Talk

Written by

For such a seemingly trivial aspect of social gathering, small talk has provided significant economic value to banks over the years.

Transactions allowed bank staff to interact with customers and to learn about their lives, anticipate their needs, provide information or a listening ear, or to offer a well-timed referral to a personal banker or loan officer. Even when those conversations didn’t result in new business, they still cultivated a relationship and trust.

Consumers Stop Conversing
The pandemic hastened what was already a longtime trend: Consumers want a bank with a branch nearby, but most prefer not to visit that nearby branch unless they must.

In 2020, the cohort of customers who still preferred the branch received a new incentive to begin using their bank’s mobile app — safety. Branch sign-up lists and capacity caps only made using a branch that much more inconvenient. Although some customers have returned to visiting the branches, the pendulum shifted for many who are now acclimatized to digital banking.

Customers now also clearly prefer to do digital research on banking products, according to 600 banking customers polled by Total Expert.
They say they’re nearly twice as likely to search for a lender online versus contacting a lender directly. They are four times more likely to search online rather than ask a real estate agent for a referral for a mortgage lender. And they go to their financial institution’s website first when they have a new financial need.

Web activity, however, is not a two-way conversation. Unlike a teller who can ask follow-up questions, interpret customer responses and make referrals to a personal banker or mortgage loan officer, knowing what customers need depends on their activity: applying, initiating a chat, filling out a form or contacting a banker. Customers are increasingly “going dark” on small talk; where they do show interest, the bank must wait for them. Bank leaders should be wondering how to revive two-way, active conversations.

But where to start? Consumers can sense sales quotas in a branch. And they can’t be forced to fill out a form on a website any more than they can be forced to volunteer their financial needs. Banks must look to another way of conversing: data.

Data as Conversation Starter
Customers volunteer opportunities to serve them every day through their data. As account holders and borrowers, they provide significant information to their bank in exchange for financial services.

Understanding and using this data, though, has long seemed too intricate for local, community-focused banks. Advances in technology have changed that; using data to inform and to initiate customer engagement is far more attainable than ever before. Banks are moving back into active engagement because data allows them to intuit needs not vocalized by customers.

For example, every bank has an address for their retail depositors’ home. But when does that matter? It’s central to selling a home; when a customer’s home goes up for sale, the address is listed on a Multiple Listing Service (MLS), and it sends a signal to their bank. Customers selling a home often buy a new one, or they need to safely invest the proceeds of the sale. The MLS listing is the customer vocalizing a set of possible needs. Once a bank catches that signal, technology can allow staff to advise, interpret, engage or refer, depending on the bank’s strategy.

Even outside of mortgages, knowing a customer is selling a home can be both a revenue and relationship opportunity. The National Association of Home Builders found that customers are more than 2.5 times more likely to make large purchases within a year of buying a new home — items like appliances, furniture and home improvements — compared to consumers who did not. Would these customers appreciate savings through credit card rewards? Do they want to use their equity to buy appliances? Were they waiting until their new mortgage closed to purchase a commuter car? Even simple, widely available data points can become the basis for highly engaging and productive interactions between a bank and its customers.

Eighty-four percent of Americans report stress about their finances, according to a recent ValuePenguin survey; bank customers want help reaching their financial goals. Banks may not be able to stop the decline in small talk, but they can revive and even surpass it with new tools made for banking. There are so many more opportunities for banks to use their data to anticipate needs and to engage customers about their desired outcomes. The upside is lifelong loyalty within each customer relationship.

What Bankers Should Know About Conversational AI in 2019

Written by


omnichannel-12-21-18.pngWe’ve come a long way since filmgoers watched nervously as the computer “Hal” struck out on his own with the bland yet threatening response, “I’m sorry Dave, I’m afraid I can’t do that,” in Stanley Kubrick’s “2001: A Space Odyssey.”

Today, humans are comfortable interacting with machines. Twenty-five percent of customer service and support operations will integrate virtual customer assistant (VCA) or chatbot technology by 2020, up from less than 2 percent in 2017, according to Gartner, Inc. And in some cases, consumers seem to prefer machines to humans. Therapy bots like Woebot are successful in part because users don’t experience the fear of judgment that may exist speaking with another human.

The technology that enables machine-to-human interactions is known as conversational AI. It powers virtual assistants across apps, websites, messaging and smart speakers. In 2018, we saw virtual assistants take off in banking – finding their way into the apps and websites of the world’s largest banks. Pilots turned into production, and virtual assistants started engaging with real consumers at scale.

This technology is a growth engine for banks by servicing customers more efficiently, engaging customers to boost brand loyalty and acquiring customers to increase their lifetime value. But all conversational AI solutions are not the same.

Here are three key trends for banks implementing conversational AI in 2019.

Think omnichannel, not multichannel
Consumers’ expectations for banking are evolving from siloed multichannel experiences to deeply personalized omnichannel experiences. They expect the experience with their bank to be consistent and informed, no matter which channel they interact on, and they expect to move smoothly between channels. Banks implementing conversational AI should support “channel traveling” and never lose sight of who the customer is – not just their unique ID, but their preferences, history and more.

Make sure your solution supports sophisticated customer journeys and hand-offs between channels. Your customer should be able to start a conversation with your virtual assistant on Amazon Alexa, and the virtual assistant should be smart enough to follow up with more related details in the mobile app. The virtual assistant knows the optimal interaction model for each channel and generates the right response for the channel of choice.

Conversations that explain “why”
By now, consumers are accustomed to automated assistants that respond to them. A virtual assistant that answers questions has become table stakes. In 2019 and beyond, we’ll see consumers gravitate toward services that can give them answers to questions and explain their finances. People will come to expect answers to “why” in addition to “what.”

For example, customers will want to know their balance, but also why it is lower than expected. Or, they may ask if they can afford a vacation now, and if they could still afford it in six months. They’ll want to know their FICO score, and why it is lower than last year.

Banking customers already know chatbots can give their balance and move their money. In 2019, their expectation will be that conversational AI will do more to help manage their money with context and insights.

The era of available data is here
After years of waiting for banking data to be available, the future is finally here. Inspired by regulations such as PSD2, or the Payment Services Directive, in the European Union, large banks around the world are adopting open banking standards and launching modern developer portals that enable a new world of banking services. This is good for conversational AI, because its real value comes from personalized, actionable experiences—experiences that require data and services. With financial institutions such as Wells Fargo, Citi, Mastercard and Standard Chartered streamlining access to APIs, building meaningful conversational experiences and integrating them with the banks’ other services will be much easier and faster.

In 2018, we’ve seen conversational AI is here to stay, and in 2019, we need to make virtual assistants do more than respond to FAQs and complete simple tasks. Banks implementing conversational AI should remember consumer expectations are growing every year. To meet those expectations, leverage the abundance of available data via APIs to create omnichannel customer journeys that can understand your customers and explain the context to them.

More Than Your Average AI

Written by


artificial-intelliegence-6-6-18.pngUSAA was looking for a financial technology firm to tell them they were dead wrong.

They found that candid firm in the summer of 2017, and the resulting partnership has generated one of the first technologies the large financial services provider has rolled out that allows its members, active military personnel, veterans and their families, to interact with USAA on Amazon’s home devices that feature the digital assistant Alexa. USAA wanted to solve a problem: “How do we create a scalable conversation engine that can talk about something as sensitive as personal finances?” says Darrius Jones, assistant vice president for enterprise innovation at USAA, in describing what led them to their partner, Clinc.

Working together, Clinc, an Ann Arbor, Michigan based fintech that has grabbed the attention of national outlets like CNNMoney, and USAA developed a “scalable conversation engine,” as Jones describes it, that goes far beyond a binary question-answer interaction between a human and a “talking silo.” The two companies formally announced their partnership to create a conversational artificial intelligence solution in August 2017. USAA was the first major national bank to partner with Clinc, which had raised nearly $8 million in multiple funding rounds before the announcement.

“From the beginning, our teams worked together to create a very different experience for delivering content that is complex … and trending,” Jones says.

Those interactions propelled the work USAA and Clinc have done to be named in March as a finalist for Bank Director’s FinXTech Innovative Solution of the Year, an award presented at the FinXTech Annual Summit, held this year in Phoenix.

The truth is, several banks have worked with fintechs or internally developed some version of conversation capabilities with in-home devices like the Amazon Echo, Apple HomePod or Google Home. But most of these interactions are basic, limited to rudimentary questions about account balances and other simple, mostly binary, inquiries. But $155 billion asset USAA uses Clinc’s technology to offer broader conversations and analysis than just binary sort of answers. Jones calls it “three-dimensional” because of its ability to infer intent from interrogatory statements based on contextual evidence proposed in the interaction with its human counterpart.

“Our Alexa skill really has the ability to disseminate what you’re saying and, in some cases, answer a question most humans wouldn’t answer without proper context,” Jones says. So instead of just getting simple responses, the engine can analyze spending trends at specific places, for example, and aggregate data across several accounts, making the responses more holistic in nature. The technology can also be predictive at times when the user asks questions in a vague way, according to Jones, and can respond with a suggested prompt with a perceived answer, a capability that is so far rare in other similar AI interfaces.

USAA had wanted to wade into the AI and conversation engine area before signing on with Clinc, Jones says, and had developed a strategy they thought would have been effective, efficient and competitive, but then Clinc’s CEO, University of Michigan Professor Jason Mars, chimed in when they met at a conference. As Jones recalls, he told the team at USAA, “I think you guys have a great idea, but I think you’re doing it wrong.” It was exactly the assessment USAA was looking for. “We love partners who are willing to challenge us and make us better,” Jones says.

The conversational technology is still a ways off from administering payments or other products that might add to the bank’s bottom line, according to Jones. But USAA has already identified opportunities to leverage the technology to increase member loyalty, and potentially work in soft pitches for other products the bank offers and advise members of possible risks.

Jones says USAA has “really struggled with the success” of the pilot programs, so much so that they had to check and recheck the data and reporting to ensure it was accurate. Eventually, he says they hope to continue the scaling of the technology, which he expects to involve additional updates later this year.

“I have a belief that the days of typing or touching as your primary method of interaction are numbered,” Jones says.