First Global Examples of Banking Chatbots – I –
In the mobile-oriented world, as both individuals and enterprises, we are all app fatigue. There are so many apps around and we can even see separate applications for different functions within a bank. The sustainability of these apps that create high costs for the banks and which are often insufficient in terms of the customer expectations seems to have decreased. On the other hand, the rise of messaging and even the fact that it is accepted as the most comfortable way of communication push almost every industry to take action in this sphere. As a result, conversational platforms are important not only for our personal relations but for our our interactions with the institutions as well.
Artificial intelligence technologies accompanying this trend, made chatbot a top trend and a bot economy has emerged globally in the last quarter of 2016 but especially in 2017. Globally, even as of 2016, it is possible to talk about 200 companies producing chatbot solutions, an investment of $ 22 billion and a valuation of $ 159 billion. Gartner estimates that by 2020, 85% of customer interaction will be without human interaction. Oracle predicts that 80% of the organizations will use chatbot in the same year.
Banks are aware of this trend and focus on artificial intelligence-based chatbot and virtual assistants to provide their customers predictive, contextual and personalized experiences. Now, the use of conversational banking and artificial intelligence in customer communication is an important component of many banks’ digital strategies.
In the last three years, after one another launches of chatbot by the global finance giants as well as the neo / challenger banks have been an evidence of the fact. In this blog, we will talk about the initial examples of the banking chatbot and virtual assistants offered by the leading financial institutions.
Bank of America – Erica
The virtual assistant Erica, introduced by Bank of America’s then-Retail Banking President Thong Nguyen in 2016 Money20 / 20, was a big hit. Erica, which can interact through voice and text within the Bank’s mobile application, has been positioned as a personal finance management tool aimed at providing customers with better financial habits and effective budget management. Bank of America, aims to deliver personal finance advice service to the masses, which is usually provided to the upper income group. Thanks to its artificial intelligence, predictive analytics and machine learning, capabilities, Erica enables customers to make payments, check account balances and make savings as well as receive notifications and warnings. It also provides tips on how to improve savings, budget management and credit score. It is possible to receive a proactive message from Erica -that was offered to a 65 million customer base, 22 million of which are mobile users- like “I found an opportunity to reduce your debt by putting $ 300 aside, would you like to hear?
By collaborating with a training platform, Erica is also the source of many videos and educational documents for improving credit score and better budget management.
Bank of America executives say that by utilizing the big data they have, they know how to train Erica about understanding the nuances of the language and acting aligned with the context and the intent. For this purpose, they use the questions that customers ask, for example, to learn their account balance, as a training data.
One of the creators of Erica, Michelle Moore, then-Head of Digital Banking, replies to the estimation that 6% of US work fields will disappear in the next 5 years, including customer service, along with the developments in artificial intelligence saying that “Erica was created, not to take away people’s jobs, but help them with their banking needs.”
At the beginning of 2019, Bank of America announced that the number of users exceeded 6 million, and so far Erica has met 35 million customer requests. This year, Erica’s proactivity muscle has been empowered by some new features. For example, Erica informs the customers if there is an unexpected increase on a regular monthly payment, or informs them when they are eligible to participate in the bank’s loyalty program, or notifies an upcoming payment in 5 days. Erica’s integration into voice assistant platforms like Siri, Alexa, Cortana, Watson, Viv is also planned for the upcoming phases.
Capital One – Eno
Eno, launched as a pilot by Capital One at the beginning of 2017 and rolled out in October 2017, is an AI-based chatbot. Eno is trained in terms of natural language processing capabilities by the data provided by the use of a specific group of customers in the pilot phase. She provides customers with account balance, account summary, available credit limit, billing deadline, recent transactions monitoring, and credit card payment. Eno, positioned as a shopping assistant, can monitor expenses and proactively report unexpected spending situations, thus enabling instant action on potential frauds. Another feature of Eno is the ability to generate a virtual card number during online shopping. It is enough to use 💰 emojis for account balance, and 👍 emojis to confirm a transaction for Eno who can recognize emojis and produce meaningful answers. Eno serves on the mobile application of the bank and can send notifications via sms or email if requested.
The bank says that within the creation process, a large team of diverse disciplines, including a filmmaker, an anthropologist, a journalist, and a user experience designer, have designed Eno as a complete character, taking into account the philosophy behind it. For the team, the manner Eno answers is as important as the content of the answer.
Ken Dodelin, VP of Conversational AI Products, says that thanks to Eno’s machine learning technology, she has imporeved itself in terms of NLP level, various communication methods, and the ability to create relationships through conversation.
Eno is not Capital One’s first project on conversational banking. Previously, the bank had provided the customers the opportunity to learn billing deadlines, account balances, pay credit card debts, track recent account movements and spending habits, via Amazon Alexa.
Citibank – Citibot
Han Kwee Juan, then-CEO of Citibank Singapore, said that they worked on the new methods to meet ever-changing customer expectations and played a pioneering role in the realm of innovation. He added that Citibot is the result of Citibank’s strategy of a simple, fast and enhanced digital experience, being present at the digital channels where customers are most active and want to be served.
Kahina Van Dyke, then-Director of Collaborations with Facebook Financial Services, says that banks, financial institutions and fintechs try to take part in the platforms that have become customers’ living space and Facebook Messenger is very convenient for this.
Citibot, which had a beta phase and has been used by a group of 600 people, was rolled out in March 2018. The chatbot is planned to be expanded in other countries after Singapore.
JPMorgan Chase – COIN
Unlike other banks that deployed chatbot for their customers, JPMorgan Chase launched a chatbot in 2016 for the use of its employees to improve operational processes. Called COIN (Contract Intelligence), the bot is able to review complex contracts in seconds with minimum errors. JPMorgan Chase announced that it saved 360,000 hours of labor thanks to COIN. COIN was not the first bot used by JPMorgan Chase, the bank previously used a chatbot to provide access to systems, examine emails and respond to simple technical requests such as password identification.
Absa Bank (former Barclays Africa) – ChatBanking
Absa Bank, a subsidiary of the Absa Group (former Barclays Africa), introduced Africa with a banking chatbot experience in 2016. Absa Bank’s bot is able to answer frequently asked questions based on artificial intelligence and perform daily transactions. The ChatBanking, that serves via Facebook Messenger, Twitter, and with an integration in 2018 via WhatsApp, provides account information, spending and transactions tracking, purchase of airtime and data, and making payments. The bank’s strategy is to leave the simple questions to the chatbot, to ensure that the customer service team focuses on more complex issues.
As this is a serious blog, we will continue to explore more examples. To reach the second blog on this topic: First Global Examples of Banking Chatbots -II-