Vertical Markets Spotlight: Speech in Financial Services

Speech technology’s use in financial services continues to grow as banks, insurance companies, and brokerages seek to improve fraud prevention; cater to customers’ increasing use of Amazon’s Alexa, Google Assistant, Apple’s Siri, and other personal virtual assistants; and offer faster, better service via the contact center, all while maintaining compliance within an increasingly complicated regulatory environment.

According to Nuance Communications, multinational financial services firms are investing billions in their own conversational artificial intelligence infrastructures, while smaller firms are using tools from customer engagement technology vendors to offer voice-enabled and other AI services.

“They’re actually doing more than you might think,” says Eric Miller, founder and CEO of LumenVox. “Voice biometrics leads the way for uses of the technology because it is superior to other types of authentication, and maintaining the security of account information is critical to maintaining a customer’s trust—and business—in this industry.”

Other uses of speech technology in financial services include conversational commerce and contact center training.

Voice Biometrics

Financial services is the leading industry in the use of voice biometrics, Miller says. “Financial services, more than anything else, is about protecting the money.”

In the first quarter of 2021 alone, the number of suspected digital fraud attempts in financial services, notably identity theft and phishing attempts, grew by 109 percent in the United States and by 149 percent globally, according to a report by credit reporting firm TransUnion.

Other authentication methods, like asking a mother’s maiden name or other personal information, are too easy to beat and not all contact center agents might abide by all of the authentication procedures all of the time, Miller explains.

Voice biometrics provides another line of defense.

“It’s very accurate,” Miller says. “We’re working with a Tier 1 bank in North America to provide the entire voice stack. We are providing the speech recognition engine, which is grammar-based; transcription; the active and the passive voice biometrics; and progress analysis. All of these things are combined, and sometimes the bank is using them all at the same time.”

According to Nuance, more than 160,000 Virginia Credit Union members have signed up for its biometric authentication program, which uses Nuance Gatekeeper to recognize characteristics unique to their voices.

The member just starts talking, rather than remembering PINs, passwords, or answers to secret questions. If the voice isn’t recognized or matches a watchlist of fraudsters’ voices, this is flagged for the agent to take appropriate next steps.

Beyond more accurate authorization, voice biometrics is faster, according to Nuance. Verifying a caller used to take more than two minutes, but with voice biometrics, that’s now down to as little as three seconds, saving 90 seconds, on average, per call.

“Financial services was one of the early ones innovating in this area,” says Rebecca Wettemann, principal and CEO of Valoir. “They are taking the lead in determining how to use this data ethically. How do I make sure that my models for fraud detection don’t have bias?”

In a soon-to-be-published report, Valoir found that 23 percent of respondents said they trusted financial services more than any other industry.

So while voice biometrics in financial services might not have advanced—though the penetration certainly has—financial services firms are advancing their culture and strategy around the technology to ensure the ethical principles are in place and that their use of biometrics is transparent to all customers.

Conversational Commerce

Some financial services firms have added what Capgemini refers to as “conversational commerce,” the ability to conduct simple transactions and retrieve certain information via speech technology.

The capability has been around longer than many might suspect. ING Netherlands launched voice banking in 2014. Two years later, Capital One became the first U.S. bank to offer such a capability. Many others have since made the capability available to their customers.

According to a Capgemini report, 44 percent of consumers have expressed an interest in using voice assistants for banking transactions, while one in five said they would use voice assistants instead of visiting a bank branch or location. The latter figure is expected to grow to 31 percent within three years.

“We’re still at the very beginning of conversational commerce,” Capgemini says in the report. “The extraordinary rapid early adoption will drive investment and innovation.”

Yet many financial institutions are still hesitant to use the technology due to privacy concerns, according to data from Forrester Research. Consumers are only half as likely to use voice assistants for transactions as they are to use them to access media, Forrester analyst Julie Ask says in the report. “Consumers worry about their privacy and seldom prefer voice for any task unless they need to be hands-free.”

Capgemini acknowledges that there can be some hesitancy in adopting voice banking in some situations. Its research found that two-thirds of consumers prefer using voice assistants in their living rooms, while only 32 percent are comfortable using voice assistants around unfamiliar people, suggesting that there is a shyness among users to engage in dialogue in public.

“Brands need to take proactive steps to ensure privacy and ensure security of consumer data,” the consulting firm says in a recent report. “Over half (56 percent) of all respondents in our survey did not trust voice assistants with the safety and security of their personal data. Nearly half (48 percent) of the respondents felt that voice assistants are intrusive and seek too much personal information. Organizations need to alleviate these concerns by designing robust and transparent privacy policies.”

But even with some hesitancy on the part of consumers, numerous financial services firms are advancing their use of the technology.

“Capital One has been a pioneer in adopting voice in the financial services sector,” according to Capgemini. “It has launched an Amazon Alexa Skill to allow consumers to access their account information and even pay their credit card bills simply by using voice commands.”

The use of conversational commerce goes beyond banks and credit unions. Capgemini points to an insurance client that implemented an IVA to help customers 24/7. Using natural language, the technology has been able to provide customers with fast, personalized answers to a variety of queries.

“This has not only enhanced the customer experience but also improved operational efficiency for customer service representatives by allowing them to focus on more value-added requests and pre-qualify some customer demands through an exchange made by one of the voice assistants in advance.”

IVAs/Chatbots

Swedbank, one of Europe’s largest financial institutions, deployed an automated virtual assistant from Nuance on its website to answer customers’ questions while simulating a human conversation. The IVA answers 80 percent of questions successfully.

Nuance expects financial services use of voice technology in the contact center and elsewhere to evolve for expanded use in the future.

Financial firms “will have to evolve beyond traditional IVR systems,” Nuance says in a recent white paper. “They’ll need to provide a consistent, seamless, and personalized conversational experience for each customer across every channel—in digital channels such as messaging and chat, as well as in voice channels.”

Contact Center Uses

Though there are certainly the mundane, easily answered calls (i.e., what is my balance), many callers to financial services companies need to speak to live agents, according to Wettemann.

Voice technology enables contact centers to more quickly route customers to people with the right skills. A specialist in credit cards has far different skills and knowledge than a specialist in personal loans or mortgage loans.

Florius, a mortgage lender based in the Netherlands, saw the need to add speech technology to its contact center to better serve customers. Using Verint Speech Analytics, which is tied into the lender’s knowledge base, and Verint Real-Time Agent Assist, the lender now has real-time information on callers during the conversation and can better answer queries, says Seif Alhamrany, a team manager at Florius.

During calls, agents receive up-to-date information on what is being discussed, enabling them to better analyze and respond to customer needs, benefiting the customer and the lender.

Since adding the technology, Florius has improved its Net Promoter Score by 5 points and increased first-call resolution by 5 percent.

Though not unique to financial services, other contact centers are using the technology to capture both sides of customer-agent interactions for improved agent training and to ensure regulatory compliance. The latter is where many other speech technologies, like interaction recording, transcription, captioning, search, analytics, archiving, translation, and discovery, are being used heavily by financial services firms.

Experts expect financial services firms to continue to expand their use of speech technology as they look for ways to better protect customers and themselves against fraud and provide improved services. 

Phillip Britt is a freelance writer based in the Chicago area. He can be reached at spenterprises1@comcast.net.