Los Chatbots de la banca en general son útiles para dirigir a los clientes en tareas transaccionales, ayudándolos a llegar al recurso online o humano adecuado para resolver la solicitud, pero esto debe evolucionar hacia la solución de problemas complejos, donde la IA se convierta en un asistente virtual que realice los procesos por sí mismo, como lo realiza NOMI de RBC y Erica de Bank of America.
A continuación, toda la información en inglés:
The use of conversational AI in banking has come a long way, but it has an even longer way to go before achieving widespread utility.
Chatbots — for those financial institutions that even deploy them — in general are helpful in directing customers in transactional tasks, such as how to use mobile remote deposit capture or finding an ATM. But most lack the ability to handle more complicated conversations.
There are exceptions, of course, and institutions with more sophisticated applications, like RBC’s NOMI and Bank of America’s Erica, tend to refer to their applications as a virtual assistant rather than a chatbot. But there are relatively few such examples so far in banking, and “chatbot” does not necessarily convey a lesser level of AI capability.
All these conversational “bots” use varying degrees of data, machine learning and natural language processing, and they all fall under the category of conversational AI. Most interactions are by text or chat, although some also use voice response.
From Chaperone to Sherpa
So how do banks and credit unions move beyond simple chatbots that help with just completing transactions to a more nuanced conversational AI?
One way is by thinking of AI as more of a “sidekick” to already existing customer service teams, observes Can Kekevi, a managing director in the financial services practice for Accenture.
“Virtual assistants can and should go well beyond a ‘point in the right direction’ or FAQ to handle complete transactions — from changing an address to cancelling a payment and updating a standing order to identifying a specific need and routing the customer to correct specialist,” says Kekevi.
Furthermore, the pandemic and related lockdowns have made the average consumer more willing to interact with businesses in general digitally. Pre-pandemic conversational AI “guided customers through a service interaction,” Kekevi notes. “It functioned like a chaperone, helping a customer get to the right online or human resource to resolve the request… Customers are now far more willing to engage with virtual assistants that can actually get things done.”
Banking Embedded in an Agnostic Assistant
The future of conversational AI in banking may not simply be only about banking, but rather banking would merely be one aspect of how the average consumer uses conversational AI in their daily life, says Peter Wannemacher, principal analyst, digital banking at Forrester.
The Next Step:
In the future, virtual assistants could be integrated into our lives in such a way that they help get everyday tasks done, including banking.
“The most impactful future advances in conversational AI in banking will most likely be built upon AI capabilities not exclusively related to banking,” Wannemacher told The Financial Brand.
For example, this could be a hypothetical future scenario where the average household has a kind of all-in-one virtual assistant that can help answer questions about banking, along with grocery shopping, home maintenance and many other aspects of everyday life.
If you’re thinking Alexa, Amazon’s popular voice-based assistant, Wannemacher is envisioning something with more utility, something that could go beyond just simple commands like “Alexa turn the lights off.” A virtual personal assistant could help you do your taxes or rebalance your investment portfolio along with other life tasks.
Right now the firms and vendors that are gaining traction in conversational AI in banking are using very specific and bespoke conversational tools that recognize banking as a unique need, Wannemacher says. “The next big horizon in AI is likely to be a generalist shift if it happens. It won’t be specifically about banking but some fundamental breakthrough in AI that happens.”
Making Chatbots More Popular and Personal
It’s perhaps because of this limited utility that chatbots and similar conversational AI tools have not yet seen widespread use by most consumers.
According to a Forrester report, Design Better Chatbots, which polled consumers on their digital interactions with all businesses, even though a chatbot is often a channel within a channel like a website or app, chatbots are falling behind native mobile apps in usage.
Three years ago, the two were neck and neck, with customers used a chatbot about as often as an app as a customer service channel, according to the report. Since then, mobile apps have grown twice as fast as chatbots. Both channels are on the rise compared with website usage, which has declined for customer service use, yet websites are still used by almost 50% of customers versus about 23% for chatbots.
Talk to Me:
Chatbot use, while growing, still lags far behind mobile apps and websites as channels customers use for customer service.
One reason for this disparity is that when building chatbots or any conversational AI tool, most firms focus on the tech aspect and not as much on design. Simply put, many aren’t appealing to use.
“This is basic, but if designers are not yet an integral part of your chatbot efforts, it’s urgent to make it happen,” Forrester notes. “If you don’t have the organizational power to make it happen unilaterally…make the case to those who do. If you’re a product or dev leader, recruit your design colleagues immediately to join the organization’s chatbot efforts.”
Personalization Expectations Keep Rising
Conversational AI can (and should) also be used to deliver more contextual and personalized information. As it currently stands, these tools offer information mostly presented in a static view, notes Vishal Pisari, vice president, digital solutions at Fiserv in a BAI post.
“Users will soon expect to be able to ask any question about their financial health and receive an accurate response,” says Pisari. He points out that when users use a conversational AI-capable app, they regard it as a blank slate. “They can ask whatever they want and have relevant information and support — such as help unlocking a credit card or displaying spending habits — available on demand.”
Personalization is also something customers are asking for. Only about 30% of bank customers that use mobile banking report that their financial institution offers personalized content and recommendations, according to Javelin. Conversational AI should reflect this need.
Consumers’ are craving customized advice based on their specific individual situations, notes financial technology provider NCR in a blog, and that offers banks the opportunity to stand out among the crowd.
“Customers are demanding more,” NCR says. “And FIs want to deliver. To do so means being able to provide here-and-now, individualized support in the digital channel. And to better anticipate the needs of their customers. That’s where conversational AI … can shine.”
The elements conversational AI needs to accomplish this are the ability to offer contextualized conversations as well as identify or even hypothesize intent, says Forrester’s Wannemacher.
“It’s gotten better in the last couple of years, but for the most part conversational AI in banking is still very clunky — or even nonexistent for many institutions,” he says. “But they have gotten better at understanding the most common and straightforward needs of the customer.”
Wannemacher adds that Forrester’s research shows that banks that do the best job at conversational AI have customers that report higher satisfaction of in-branch experiences as well. This is due to these institutions being able to use data to anticipate customer needs and offer contextualized advice.