Chatbot: The intelligent banking assistant

The increasing sophistication of mobile technology has helped us exchange details, authenticate and conduct transactions seamlessly. Information is literally available at our fingertips, eliminating the need for human support. The only area where human interaction has had a lead over technology is the personal touch during a conversation, especially in the case of relationship-based interactions. However, with all major innovators, including technology giants,1 putting their weight behind technology that provides human-like conversation experiences, even that edge seems to be diminishing. A platform designed to understand, learn and converse like a human and answer ad-hoc queries in real time is commonly referred to as a chatbot. Chatbots have attracted the attention of firms across industries and are being viewed as a means to create differentiation in an increasingly crowded landscape.

The evolution of the chatbot: Raising the bar of intelligence over time

With the augmentation of computing power, the chatbot has managed to create a larger impact for stakeholders over time.

2010

A simple digital tool

 

Chatbots in banking
  • Reduce time spent on generic customer queries, FAQs to enhance productivity.
  • Resolve repetitive employee queries to IT helpdesk, legal and vendor teams.
  • Fed with a structured, relatively static database and able to recognise specific keywords in queries.
  • Direct all queries beyond data sets to humans for resolution.

2014

Conversational analytics platform

 

Chatbots in banking
  • Continuous learning through advanced machine learning and natural language processing capabilities.
  • Analyse structured and unstructured data (e.g. app store reviews, social media comments, emails) of high volume and velocity through conversational analytics and provide real-time insights for decision making.

2017

A digital assistant

 

Chatbots in banking
  • Initiate action on its own.
  • Perform multiple tasks on its own, such as scheduling of meetings, booking of cabs, and auto ordering of items based on voice and text commands.2

What are the characteristics of a well-designed chatbot?

AI chatbot, chatbot features, AIML in chatbots

A well-designed chatbot

Available round-the-clock for conversation with customers
Provides instant response to any query without any delay or making the customer wait. 
Capable of remembering customer preferences and uses order history to suggest products, learns from customer responses to the products advertised, and cross-sells effectively
Improves efficiency and reduces TAT through quick information delivery; performs mundane tasks at high speed
Provides specific user input at each point, learns from customer feedback and follow-up queries and improvises, thus enhancing the user experience
Restricts response format to easy-to-understand text, images and unified widgets for better interaction
Ability to seamlessly interpret languages commonly conversed in
Provides the same experience irrespective of the channel: mobile, web, etc.

With chatbots gaining more traction, many firms across the globe have started offering off-the-shelf products that help developers to build, test, host and deploy these programs using Artificial Intelligence Markup Language (AIML), an open source specification for creating chatbots3. A few platforms support integration with payment providers for seamless processing of customer payments based on a customer’s interaction with the bot.

Increasingly, chatbots are also attracting interest in the world of FinTech, and a number of companies have developed their own chatbots using proprietary technology and algorithms. Chatbots utilise application programming interfaces (APIs) to integrate with data management platforms. This allows them to analyse the extracted data as well as web- and mobile-based user interfaces and deliver the necessary insights to the end customer.

The typical architecture of a chatbot platform is shown below:

 

Typical architecture of the chatbot platform

chatbot platform, chatbot interface

Adoption of chatbots by financial institutions

Consumers of information in the financial services sector range from end customers of the bank to CXOs. Forward-looking financial institutions have taken a leap of faith by investing in chatbots to deliver ‘contextual insights’ to the right person at the right time and through the preferred channel. They have begun transitioning to ‘conversational banking’ and are viewing chatbots as new age contact centre executives, minimising TAT and costs along the way4. Chatbots are designed to answer queries such as:

  •  How much money did I spend on travel last month?
  • Can you share a list of ATMs nearest to my location?
  •  Can you transfer X USD to ABC vendor right now?

Financial institutions across the globe are assessing the viability of deploying chatbots for varied objectives

Proactive suggestions for superior customer satisfaction

Financial institutions with a digital-savvy customer base are testing out various approaches to proactively deliver insights to the customer based on her/his transactional history and digital profile:

  • Recommending investment options based on savings bank balance and risk profile
  • Providing market-related news and impact on portfolio
  • Suggesting ways to utilise reward points of credit cards

Speedy authentication

The days of entering multiple options on an interactive voice response (IVR) system for customer service seem numbered. Firms are experimenting with chatbots that allow customers to authenticate based on voice samples from natural conversation and help complete transactions quickly.

CXO dashboard

Typically, 30% of senior management time is spent on operations and other miscellaneous tasks. Out of this 30%, 60% is taken up in getting the desired metrics from MIS/IT team/multiple departments and follow-up on the same.

Chatbots can reduce the time spent by CXOs on operations and provide readily available information quickly, thus allowing them to focus more on strategic business objectives. Queries can be as varied as:

CXO queries to which chatbots provide real-time insights

Chatbot metrics, KPIs for chatbots
Chatbot metrics, KPIs for chatbots

The next wave of chatbots (2017–2020)

In their current form, chatbots have reached a certain level of maturity. They are developed for specific tasks and are unable to suitably handle specialised queries requiring knowledge outside the functional domain. The time is now ripe for financial institutions to enhance the capabilities of chatbots to create a completely differentiated experience for their customers by combining knowledge across all relevant segments and providing better insights. This will give rise to a new level of conversational banking where results are delivered instantly through real-time conversations, thus facilitating better decision making.
 

Significantly drive customer loyalty

Chatbots can add a new dimension to the power of ‘personal touch’ and massively enhance customer delight and loyalty.

  • When a customer logs into his or her account, a chatbot can greet the customer and discuss customised offers with her/him instead of the conventional pop-up notifications or banners that occupy valuable real estate.
  • During the conversation with the customer, the chatbot can use advanced speech and natural language processing capabilities, and sentiment analytics to gauge the tone, emotions, and voice accent to offer solutions that are deeply customised to the overall context of the conversation.

Such an enhanced level of customer engagement and satisfaction will help drive customer loyalty without the need for any manual intervention.

Create a cognitive financial institution

Chatbots could be channelled to create an ‘insights-driven bank’. A typical day in the life of such a bank’s business user would involve decision making that is driven only by data-based analytics. The queries may be related to sales and marketing, impact of global trends, new product launches, and internal metrics such as employee attrition and sales targets.

A chatbot can be designed to respond to all kinds of requests and queries. Some examples are given below:

Developing cognitive capabilities and deeply customised offerings is key to moving to the next level of conversational banking.

  • Which region is likely to achieve maximum growth in terms of new customers in the next one year?
  • Who will be my biggest competitors in the microlending space?
  • Which employees from the technology team are likely to leave in the next three months?
  • Which global and domestic trends would have the maximum impact on my business and to what extent? Are there any measures successfully adopted by banks in other regions?
  • What are the best and worst scenarios for launching a contactless credit card in metro cities?
  • Which is the best customer segment to target for advertising the new mobile app? 
  • Can you automatically schedule meetings with the heads of low-performing regions at the beginning of every month and share the business numbers with me before the meeting?

Real-time responses to such queries through predictive analytics and recommendations based on prescriptive analytics can significantly enhance the quality of decision making in the firm.

Experiment with integration with other upcoming technologies

Having achieved a certain degree of maturity in chatbot deployment, financial institutions can seek newer ways to partner with technology firms and leverage innovative technologies.

These are a few potential areas of exploration:

  • Can facial recognition technology be used to carry out zero-click transactions through chatbots?
  • Will chatbots be able to showcase the impact of long-term savings in a virtual reality environment?
  • Can chatbots provide an immediate status update on a cross-border blockchain transaction?
  • Can chatbots be deployed using IoT to converse with customers?
  • How can banks leverage vernacular language solutions to reach the unbanked population of India?

It will become imperative to think beyond the obvious to assess use cases for chatbots in the future.
 

Immediate steps for a firm considering chatbot implementation

In order to fully realise the RoI from chatbot deployment, it is critical for a financial institution to define its business requirements and identify the use cases to be met through the chatbot. KPIs should be identified and measured to gauge the performance of the bot, and necessary alterations need to be made based on these in order to achieve the desired results. A data analytics driven Centre of Excellence (CoE) with a strong data operating model and with representation across business and product and technology teams can be established to adopt a centralised pan-entity approach towards the utilisation of chatbots for solving business needs and generating predictive insights.

Sources:

  1. Metz, C. (2016). A chatbot would like to help you with your bank account. Wired. Retrieved from https://www.wired.com/2016/06/new-banking-ai-now-chatbots/ (last accessed on 29 September 2017)
  2. Martin, J. A. (2017). Who’s smartest – Alexa, Siri, Cortana, or Google assistant. CIO. Retrieved from https://www.cio.com/article/3192139/consumer-electronics/whos-the-smartest-alexa-siri-cortana-or-google-assistant.html (last accessed on 29 September 2017)
  3. Jee, C. (2017). Eight of the best chatbot building platforms for developers. Techworld. Retrieved from https://www.techworld.com/picture-gallery/apps-wearables/seven-platforms-for-developers-build-chatbots-3639106/ (last accessed on 29 September 2017)
  4. Morgan, B. (2017). 5 ways chatbots can improve customer experience. Forbes. Retrieved from https://www.forbes.com/sites/blakemorgan/2017/08/06/5-ways-chatbots-can-improve-customer-experience-in-banking/2/#608ac2eb7545 (last accessed on 29 September 2017)
Follow us