This article was contributed as part of Tearsheet’s new Thought Leaders contributor program. Amit Bhasker is omnichannel service line leader and Vaibhav Grover is assistant vice president, retail banking at Genpact, a global professional services firm focused on digital transformation.
It’s hard to escape artificial intelligence these days, particularly in perhaps its most common form, conversational AI. The days when conversational AI was limited to science fiction movies and television shows have long since passed. Now, this technology, which utilizes AI, machine learning, and natural language processing to enable people to interact with a machine by having a conversation either as text or as voice in natural language, is practically everywhere. The leading tech giants have already launched their own software for voice-based conversational AI. These include Apple’s Siri, Google Voice, Amazon’s Alexa, Microsoft’s Cortana and Facebook’s M. Conversational interfaces are here to stay and their use is growing with every passing day.
As always, the banking industry has been an early adopter of this technology. In June 2018, Bank of America announced that it had surpassed one million users on Erica, the first widely available AI-driven virtual assistant in financial services, within two months of completing its phased rollout. In a Genpact study of more than 6,000 consumers, 27 percent of respondents said they would be comfortable setting up a new account with a digital assistant, such as Siri or Alexa. This increasing level of acceptance provides financial institutions with an opportunity to use conversational AI as a key differentiator in the marketplace.
Why conversational AI makes sense for banks and their customers
Whatever the industry, consumers are looking for solutions that make their lives easier and have a smooth, trouble-free customer journey. Amazon Prime is a great example. As soon as Amazon removed delivery charges and long wait-times for Prime members, customers were happy to pay an annual fee and subscriptions skyrocketed. By June 2018, Amazon had an estimated 95 million paid Prime member subscribers in the United States.
Conversational AI has the potential to do exactly the same for banking customers – remove the friction and provide faster service. By simply talking to an interface in their natural language, customers can get an instant response. Imagine parents getting an advice on investment plans for children on a voice user interface while feeding their toddler; a business owner transferring funds to a supplier while traveling to the office; or a young couple applying for a personal loan while browsing for the location for their next vacation. Conversational AI can overcome the boundaries between structured language, computer literacy, and physical challenges.
How the technology is moving banking in a new direction
With customers ready to embrace conversational AI, there are a number of factors banks should consider as they look to implement the technology:
- Intelligent interactions, not transactions. Banks need to ensure that they position themselves for the future. As customers share more information, more data becomes available for voice- and text-based devices to process an increased number of interactions every day. When the data from these interactions is enriched with more internal and external data, it will create meaningful insights on the way we live, spend our money, and interact. This represents a shift in how we think about digitized banking – a transition from transactions to intelligent interactions. These insights will not only help banks design a personalized experience for customers, but when combined with an ‘always on’ and omnichannel service, financial institutions can create hyper-personalized customer interactions.
- An enlightened assistant to customers. With enhanced digital adoption, while customers have access to more channels, the demand for human service and interaction still remains high. Genpact’s research suggests that customers are most satisfied with the human touch of in-branch and phone interactions. For example, 57 percent of consumers surveyed say that speaking with someone in a branch makes it easier when opening a new account. The challenge for banks is to deliver their services through a digital channel while still retaining this personal touch. Conversational AI has the potential to support customers as an enlightened assistant throughout their journey. It can perform multiple tasks, including assisting customers with selecting the right product, helping them calculate their monthly loan payment, completing applications, conveying regulatory disclosures and taking their acceptance, responding to queries, managing account activation, and servicing their account. An AI-powered assistant can facilitate these functions very close to how a human would perform. By leveraging this type of assistance in day to day in-branch or phone activities, banks can retain the personal touch for customers while delivering services through a cost-efficient digital channel.
- A reliable associate to bankers. Some enterprises are now using conversational AI in their hiring process. Conversational AI-based automated interviews of potential candidates will likely become the norm for banks in future. Financial institutions are also using AI to provide faster, smarter information to their bankers and service representatives when interacting with customers. Cognitive interfaces are not only helping bankers resolve customers queries with a high degree of accuracy in real time, but also have the potential to go to the next level and provide recommendations on next best circumstances by using predictive analytics and sentiment analysis.
Why the journey is going to be complex
Although there are endless opportunities, the development and implementation of conversational AI is going to be a complex journey. There are multiple challenges that banks need to overcome. One of the primary challenges is knowledge engineering and authoring. The knowledge required to train AI bots resides in a variety of forms, such as repositories, standard operating procedure documents, and the minds of people handling these processes. It is therefore critical to create a mechanism that can not only capture this knowledge from these sources quickly, but also is able to continuously refresh the knowledge base on a regular basis. Hence, it is important that banks understand that developing and implementing bots is not just a job for techies, but should also include the correct amount of involvement from domain specialists. Process owners should be able to update the knowledge base directly and quickly to overcome this challenge. In addition, banking is a highly regulated industry. When bots perform financial transactions and provide financial advice to customers, it is critical to maintain an audit trail of each interaction. Banks must have a mechanism to trace and pinpoint the knowledge source used by the bot to facilitate any transaction or interaction.
What the industry needs to do
Even with advances in smart phone technology, today’s bank customers may soon find tapping on a screen as frustrating as dialing a phone number on an old rotary phone is for the current generation. The time is now for banks to accelerate the development and testing of voice interfaces. The technology is there and the industry has already seen some success. However, stronger use cases and value-added capabilities need to emerge in the industry. Banks don’t need to solve every challenge on day one, but rather should begin with simple and non-actionable queries and then start looking at applications beyond basic interactions.
To achieve this, banks and technology companies will need to collaborate and experiment together. Banks cannot afford to develop such solutions on their own, as they lack the necessary expertise and budgets. Technology companies do not have real-time customer data and access to end users. By partnering together, both industries can create multiple use cases, as well as more easily build proof of concepts and prototypes backed by customer journey design and experience analytics, which can speed the development of new products and services. JP Morgan recently entered into a partnership with Amazon to develop a new skill for Alexa that will allow the institutional clients of investment banks to access the firm’s analyst reports and stock information using the voice assistant. This type of repeatable model will spur more innovation and drive the use of new conversational AI solutions for the benefits of both the banking industry and its customers.