May 19, 2020

How the financial sector is using AI for customer service in times of crisis [Cases]

At the beginning of 2020, the vast majority of companies had to adapt and redefine their strategies in response to COVID-19.

The quarantine, along with other pandemic risks, led companies to migrate their in-person service to digital channels and shift to remote work. In the same way, the crisis shaped a new consumer model.

In this sense, goals and processes planned for the future had to be expedited in order to ensure continuity of service. And artificial intelligence, which in many cases wasn't a priority, became a critical ally to deal with the situation.

Banco Comafi and Rapicuotas, two Argentinian companies, pioneered the use of this technology and helped redefine strategies during the pandemic.

That’s why Martin Frascaroli, CEO of Aivo, talked live with Julio Fabian Mealla (Digital Banking Manager, Banco Comafi), Lucia Ferreiro (Operations Coordinator, Rapicuotas) and Jorge Reina (Managing Partner, Rapicuotas) on how they adapted their processes against this background.

Below are the highlights from their conversation.

Sofia, Banco Comafi’s virtual hero

Sofia, Banco Comafi’s AI-powered virtual assistant, was already a key part of their service strategy at the time of the outbreak.

But with the COVID-19 crisis, that strategy made a 180-degree turn. This was the case especially after the well-known “Black Friday” in Argentina, when at the height of the pandemic, thousands of seniors went to banks to collect their retirement pensions.

In the light of that situation, it was clear banks had to change their processes, particularly those designed for seniors.

New strategies for new needs

This is why Banco Comafi took quick action. With the goal of decreasing service volume at branches, they let Sofia take control of new tasks: This included providing a withdrawal code without a debit card at ATMs, PIN activation and managing appointments for in-person services at branches.

The process for requesting a withdrawal code without a card was carried out in five days. Mainly, this was the answer to anyone who didn't have a debit card and wasn't able to go to a branch to collect their pension.

“We developed an interactive process where the recipient enters their ID in the chat with Sofia and requests the code. After internal validation, an SMS is sent to the person’s cellphone with the amount to be collected,” explains Julio Fabian Mealla, Digital Banking Manager.

There's a similar process when a customer wants to schedule an appointment at a branch. “Sofia connects to the bank's system and lets the customer schedule a visit. The customer chooses the branch, day and time, enters their contact info, and that’s it,” he adds.

A problem-solving, empathetic experience

The above cases show how Banco Comafi quickly developed new features for a customer segment that usually isn’t too familiar with mobile and home banking technology.

By just entering their ID number, they can easily get through the process.

This is because Banco Comafi’s main goal is for Sofia to solve inquiries comprehensively. As Fabian explains, “the worst that can happen is that the bot transfers you to a traditional channel. That’s why we’re working on strengthening the structure, decision trees and above all, the APIs between Aivo and the bank so that every operation can be carried out through Sofia.”

As a result, inquiry volume increased by more than 300% since the beginning of the quarantine. On the other hand, Sofia's provided more than 1,000 withdrawal codes and 500 PIN activations so far. And in answer to the “Black Friday” in Argentina, Banco Comafi managed to decrease the transfer to branches by 50%.

“What we see is that now there's a change in implementation. People discovered a new channel and are already using it. We want to be able to use a more limited service model, but one that's useful for customers to complete 80% of transactions and inquiries,” concluded Fabian.

Super Rapi, the copilot for remote agents

Rapicuotas is a financial services company with 17 years of experience and 26 branches across the country. With a goal of promoting digital transformation, in November 2019 they added Super Rapi, their AI-powered conversational chatbot.

Before the quarantine, they had six customer service agents. They were also in the middle of setting up and training the bot, whose objective was to cover initial customer contact. But COVID-19 forced the team to reorganize their processes to address two main issues as explained below.

Related article: Coronavirus: What to do in Customer Service to face the crisis

A new content strategy

On one hand, they had to completely change the bot settings. Initially, content was targeted at branch service, answering questions like “What do I have to bring to apply for a loan?” or “What are the requirements?”

But when the quarantine started, customer needs and FAQs changed entirely. Prior FAQs were replaced by questions related to online issues.

The changes also affected service channels. During weeks of lockdown, 90% of traffic took place on WhatsApp, which became the most-used channel.

As Lucia Ferreiro (Operations Coordinator at Rapicuotas) explained, “we try to address everything through this channel, which is also the most customer-friendly one. They write a message and wait for us to reply whenever.”

Empowering human agents with AI

On the other hand, the second redefining dealt with increasing the agent team and their move to remote work. “We saw there would be an increase in the amount of inquiries and started recruiting people to customer service in order to address the volume,” Lucia explains.

This is because while the bot handles FAQs, agents issue payment links to carry out transactions. Thus, Rapicuotas had to increase its number of employees from six at the beginning of March to 40 in April.

The enormous challenge was training more than 30 people in just one week while working remotely from the cloud. “The situation forced us to bring the operation to home offices, something we had not anticipated. We had to do everything possible for Rapicuotas to keep functioning and be able to give answers to people. It was a major restructuring,” says Lucia.

Record speed and efficiency

After two weeks of reconfiguring the bot and expanding the service team, Rapicuotas launched its new strategy. As a result, they went from 14,000 inquiries in January to 118,000 in April, nine times the volume.

“We had a peak of 12,000 inquiries in one day, what we used to get in a month,” Lucia adds. “In reality, when it comes to the platform, we could carry out the entire process quickly and efficiently.”

For Martin Frascaroli, Rapicuotas succeeded in two different ways. First, they could “learn what customers want in this new context,” opening up channels that weren’t used before and answering new inquiries quickly.

Second, they could grow their service team in so little time. “They set an example of success by bringing on people remotely with record speed in just two weeks,” he concludes.

Are you looking for new ways to improve your CX?

Our customer service solutions powered by conversational AI can help you deliver an efficient, 24/7 experience  to your customers. Get in touch with one of our specialists to further discuss how they can help your business.