Customer service interactions can often be a source of frustration for customers, as well as a major cost center for banks. What if there were a way to resolve inquiries faster and more efficiently?
NDS worked with a major banking client to develop a chatbot that runs on IBM® Cloud and uses IBM Watson® technology to provide customers with fast, helpful answers to thousands of questions.
10 minutes savedper chatbot interaction when compared with call center
30% dropin inquiries to the bank’s call center one month after chatbot was launched
16,000+ conversationshandled by the chatbot at once—freeing up call center capacity
Business challenge story
Solving the customer service conundrum
Operating a call center can be a costly endeavor, requiring banks to spend significant time, money and effort recruiting, training and employing people. And even though banks may dedicate considerable resources to servicing customer support requests, many still struggle to resolve inquiries quickly and effectively, resulting in low customer satisfaction and high employee attrition.
Imagine a dramatically different customer service scenario. What if the bank could provide personalized, immediate answers to routine customer requests through an easy-to-use text interface enhanced with interactive media? Or detect the impatience in a customer’s tone of voice, and route them to a specialist right away, reducing frustration?
A major global bank in Mexico turned this vision for better customer service into reality by joining forces with IBM Business Partner Nearshore Delivery Solutions (NDS).
Gustavo Parés Arce, Co-founder and General Manager of NDS, sets the scene: “This bank serves millions of people in Mexico, with a call center team of 50 people dedicated to fielding banking web portal inquiries from customers. There simply weren’t enough agents to keep up with call volume; the result was that customers frequently experienced long wait times and many reported that they weren’t satisfied with the response to their inquiry.
“On the flip side, the experience is equally stressful for the agents who spend all day on the phone answering the same questions and dealing with frustrated customers—which is why the attrition rate in the call center industry is very high, at an average of 40 percent. On top of that, running a call center is expensive, with the average phone interaction costing between USD35 and USD50.”
Fit for the future
To help the bank transform its customer service operations, NDS proposed a forward-looking solution: an artificial intelligence (AI) chatbot, powered by IBM Watson Assistant and Watson Tone Analyzer, running on the IBM Cloud platform.
“We knew that AI had the power to transform the bank’s call center operations—and we also knew that IBM offered the ideal technology for enabling this,” says Gustavo Parés Arce. “We find Watson’s AI capabilities second to none and, crucially, the solution was able to work with Spanish—a must-have requirement for this client.”
Choosing IBM Cloud gave NDS access to both the infrastructure and platform services for building the bank’s innovative virtual assistant.
“IBM Cloud was the ideal platform for building and hosting the solution,” notes Gustavo Parés Arce. “It gives us access to an enterprise-grade infrastructure, as well as a one-stop shop for APIs, machine learning, cognitive technology and services. Plus, with IBM Cloud, we can start small and scale up effortlessly to handle rising chatbot interactions.”
NDS analyzed a year’s worth of emails, call transcripts and chat messages sent by customers to the bank’s support teams. The team trained Watson Assistant to identify common issues, questions and keywords, using this information to develop a base of knowledge for the chatbot, which it uses to answer customer questions.
Gustavo Parés Arce notes: “We found it very easy to build a chatbot with IBM Watson. It took us just 12 weeks to design and deploy a working solution, which really impressed our client!”
The chatbot acts as the first point of call for customer service requests, capable of handling many routine questions quickly and effectively. Where an inquiry is too complex for the virtual assistant to handle, it seamlessly routes the request to a suitable call center agent who can address the customer’s concern.
Similarly, the chatbot uses IBM Watson Tone Analyzer to discern the customer’s mood, and adjusts its response accordingly. For example, if a customer’s responses or tone seems angry or annoyed, the virtual assistant transfers the customer to a human operator, who can calm the situation. This helps reduce customer dissatisfaction and shapes a smooth service experience.
The bank has integrated the chatbot seamlessly with its website. For assistance, customers simply click on a “chat” button and a messaging interface pops up. Customers type in their questions or requests and the virtual assistant responds in seconds.
To make the chat experience more engaging and interactive, the bank has embedded multimedia into the chat interface. For example, if a customer wants to learn more about applying for a credit card, the chatbot delivers a video that demonstrates the different steps involved in the process. And if a customer asks about the location of nearby ATMs, the chatbot uses location services to provide a map showing ATMs located near the customer’s current location.
“The chatbot is currently able to answer around 1,000 different questions, and we are constantly training it to handle new inquiries,” states Gustavo Parés Arce. “Every month, we compile a list of words and phrases the chatbot did not understand, and work with the bank to determine which topics they would like to train the bot to answer. Then we program the virtual assistant with the necessary vocabulary and logic to reply to these inquiries.”
Better customer experiences at lower cost
By bringing AI to customer service, the bank has significantly increased the number of support inquiries it can handle. Whereas its live-agent call center can manage a maximum of 50 simultaneous customer conversations, the chatbot is capable of handling more than 16,000 interactions at the same time.
Gustavo Parés Arce adds: “The bank has been able to significantly increase customer service capacity without hiring any additional staff. The virtual assistant has been so cost-effective that the bank is expecting to achieve a full return on investment in the IBM solution in under a year.”
In the first month of the chatbot’s release, the bank reported a 30 percent drop in routine inquiries to its call center. Today, 75 percent of all customer inquiries are handled by chatbots. It takes the virtual assistant an average of two minutes to complete a customer interaction, which is 80 percent faster than the typical call center resolution time.
Not only does the chatbot reduce the total number of queries that agents address daily, it also fields most of the common, repetitive questions that burden call center teams. This frees up customer service representatives to focus on more complex issues, giving them time to answer customer questions thoughtfully.
“Feedback on the virtual assistant has been fantastic,” says Gustavo Parés Arce. “Customers love that they can get a fast, consistent response whenever they have a question, without having to waste their time on the phone. And customer service representatives spend less time answering routine, repetitive calls, making their working hours more enjoyable.”
The bank is eager to build on this success and will continue to enhance the virtual assistant. It also plans to harness new analytical tools to dig deeper into its data to unlock opportunities to cut costs and increase revenue.
For example, after examining its customer service records, the bank discovered that last year alone, it received around 50,000 inquiries related to credit card applications. However, because these requests were buried among masses of other data, they went unnoticed—depriving the bank of the opportunity to follow up with customers and encourage them to apply for a credit card.
Gustavo Parés Arce adds: “By harnessing analytics, the bank can more readily identify and act on cross- and up-sell opportunities, without spending any money on traditional marketing campaigns. This could potentially help the bank to increase share of wallet with existing customers and grow revenues.”
NDS continues to work closely with IBM to help businesses realize the full potential of AI and other groundbreaking technologies.
Gustavo Parés Arce concludes: “There is clearly enormous potential for cognitive, AI and cloud technologies to transform the way we do business—not just in banking, but across all industries. We work with IBM because they have the capability to deliver on the promise of innovative technologies, helping organizations the world over to solve problems and seize new opportunities.”
About Nearshore Delivery Solutions
Nearshore Delivery Solutions (NDS) brings talent and disruptive technology solutions to global companies, helping clients advance their digital, mobile and cognitive journeys. An IBM Business Partner, NDS is headquartered in Mexico City and has offices in California, United States.
To find out more about solutions and services from NDS, email: email@example.com or telephone: (52) 55 50 81 20 71.
Take the next step
To learn more about IBM Cloud services, please contact your IBM representative or IBM Business Partner, or visit the following website: ibm.com/cloud
To learn more about IBM Watson solutions, please contact your IBM representative or IBM Business Partner, or visit the following website: ibm.com/watson