5 Ways Banking Chatbots Could Improve the Customer Experience

Dylan Azulay

Dylan is Senior Analyst of Financial Services at Emerj, conducting research on AI use-cases across banking, insurance, and wealth management.

5 Ways Banking Chatbots Improve Customer Experience

In banking and finance, chatbots have the potential to improve the customer experience by allowing customers to check their account balances, transfer money, learn about interest rates, change their billing addresses, and more.

Although banks and financial institutions will tout the chatbots they claim to offer customers, at Emerj, the AI Research Company, we have found that banking chatbots are far more hype than substance.

Our AI Opportunity Landscape reveals that banks are eager to make press releases for their chatbots, but the majority of these bots are no longer accessible, quietly removed after their initial pilots.

This points to the reality of the chatbot landscape: most of them are too basic to transform the customer experience in any exciting way.

But they do have potential, and executives at top banks are interested in automating customer service once they automate their banks’ risk, fraud, and compliance processes.

In this article, we break down five ways banking chatbots could improve the customer experience in the new decade:

  • Marketing
  • Answering Frequently Asked Questions
  • Onboarding Customers
  • Facilitating Core Banking Activities
  • Analyzing Customer Tone and Sentiment

This article is based on insights from our interview with Dheeren Vélu, Director of Artificial Intelligence at Capgemini Invent. Dheeren first gives a brief history of chatbots.

Leaders in finance interested in selecting a chatbot vendor may benefit from our comparison of 5 chatbot vendors in banking, which is based on our vendor selection services for global banks.

Chatbots: Definition and Brief History

Dheeren defines conversational user interfaces (CUI) as the new and latest form of a two-way conversation between a human and a computer through natural language processing (NLP). There are two forms of CUI: voice bots (voice interface) and chatbots (text interface).

The evolution of human-computer communication, says Dheeren, started with Microsoft Disk Operating System, where instructions were fed to the computer to carry out a task using command line interface.

Later in the 1990s, Microsoft introduced Windows operating system, which paved the way for institutionalizing the graphical user interface. This was a big improvement over CLI, and it was the birth of Windows, Icons, Menus, Pointer (WIMP).

5 Ways Banking Chatbots Could Improve Customer Experience

Customer expectations in the banking industry have continually evolved in the last two decades. Previously, people walked into a branch in order to carry out their banking needs. With the advent of the internet, banks created digital self-service channels via “internet banking.” With the smartphone boom, there became a need to bring banking to mobile.

All of this led to the birth of omni-channel banking services and, as Dheeren suggests, the potential loss of a personal connection in banking. Chatbots could in part bring this personal connection back to the industry.

Dheeren describes five broad use-cases for chatbots in banking:

Answering Frequently Asked Questions

Customers are looking for a fast and easy way to get answers to their banking-related questions, such as:

  • Frequently asked questions, such as how to freeze an account when they lose a credit card, how to add a cardholder to an account, and more
  • Product inquiries and information about loan interest rates, the best credit cards for them, and more
  • Quick and accessible links to FAQ pages so they can look up answers on their own

Banks could save on call center overhead by integrating chatbots that can handle these types of questions. Instead, call center agents could spend their time handling escalated calls and tickets that require more nuance and care.

According to Salesforce’s 2019 State of Service Survey, 64% of customer service representatives spend their time on complex customer service tickets when their workflows involve a chatbot; only 50% do so when their workflows do not involve a chatbot.


Banks could use chatbots in marketing efforts that also improve the customer experience and increase brand loyalty. Chatbots have the potential to:

  • Personalize the customer’s experience
  • Send the customer relevant notifications while they are on the bank’s website or logged into their accounts
  • Gather and analyze customer feedback, allowing customer service teams to address emerging issues
  • Moving customers through the sales funnel, increasing conversion rate and revenue

Onboarding Customers

According to Dheeren, chatbots could also play a pivotal role in banking by “conversing” with potential customers while they are visiting a landing page, filling out a loan application, and more.

Chatbots could pre-populate fields on landing pages and within application forms, which could increase the chance that a customer opts into that landing page and expedite the time it takes for them to apply for a loan respectively.

Facilitating Core Banking Activities

In some cases, chatbots can handle core banking activities, such as loans and payments, within a chat window. Customers could transfer money from one account to another, pay invoices, apply for a mortgage, and more by conversing with a chatbot.

Most chatbots can only handle basic questions and pull information from FAQs, but the few robust chatbots in banking that exist are integrated with customer accounts via API.

For example, unlike Wells Fargo and Ally Bank, which have discontinued their chatbot pilots, Bank of America’s Erica chatbot is still available as of 2020 and can handle banking services within its chat window:

Analyzing Customer Tone and Sentiment

Dheeren says that sentiment analysis features could allow CUIs to report on how a customer is feeling while it is engaged with a customer. 

According to Dheeren, algorithms are constantly being refined to better understand user sentiment. In the past, an algorithm may have analyzed a sentence with a double negative as one containing negative sentiment.

Sentiment analysis algorithms today can take more context into consideration and analyze the sentence as one containing positive sentiment. Some chatbots can properly analyze sarcasm, as well.

This could help the bank decide what to do next to best serve that customer. Sentiment analysis results could trigger a chatbot to route the customer to a human customer service agent who then knows to handle their concern with more care.

Emerj for Banking and Financial Services Leaders

According to a 2019 NPR and Edison report, 24% of adults in the US own a smart speaker device. This number is likely to grow in the coming decade, and banks have an opportunity to capitalize on this trend by integrating chatbots and voice bots into their customer service workflows. They also run the risk of losing millions on fruitless conversational interface projects.

Large banks and financial institutions use Emerj AI Opportunity Landscapes to assess the AI application landscape in banking and financial services. Emerj helps banks and financial institutions avoid huge losses that come when their executives fall for chatbot vendors that are lying about doing AI. With insights from our AI Opportunity Landscapes, banks can save themselves from building a costly chatbot that they remove from their website or mobile app after only a few months.


Header Image Credit: Enigma Swiss

Stay Ahead of the AI Curve

Discover the critical AI trends and applications that separate winners from losers in the future of business.

Sign up for the 'AI Advantage' newsletter: