Case Study: Banco do Brasil

The Central Bank of Brazil (BACEN) supervises the country’s financial institutions, authorizes the issuance of currency, and publishes a definitive ranking of financial institutions based on their volume of consumer complaints and the judgement of these claims, whether they are valid or not. These rankings represent a good or bad reputation for a financial institution and for many years, Banco do Brasil rode reputational waves as their complaint volumes rose and fell.

In 2021 and 2022, the Banco do Brasil, also known as BB, concentrated its efforts on reducing the number of complaints upheld by BACEN and improving its brand reputation. Using NiCE Interaction Analytics, Banco do Brasil automated the workflow of high-risk interactions, and dramatically improved its internal conflict-resolution rates and its external BACEN complaints ranking.

01 Before

The high cost of first place

Banco do Brasil is one of the largest banks in Brazil, with a reputation for customer service and reliability that extends back more than two centuries. Beginning in 2015, a downturn in the Brazilian economy and increased competition from other banks challenged the bank’s reputation for excellence.

During this time its position on BACEN’s customer complaints report fluctuated, and Banco do Brasil ultimately hit the number one spot in 2017. This dubious distinction as the financial institution with the highest volume of customer complaints motivated leadership to get control over the complaint-resolution process. “[In this way] We don’t want to be number one. In fact, we want to be off this list all together,” said Denilson de Assis Melo, IT team manager at the Banco do Brasil.

02 Desire to change

Motivated but manual

Motivated to no longer appear among the most claimed institutions, Banco do Brasil did its best to implement procedures in which high-risk interactions were flagged for supervisors and properly escalated through a system that included investigation, follow-up, and resolution. However, these procedures were manual, and largely reliant on the expertise and interest of the initial customer service agent. The potential room for error, in terms of issues slipping through the cracks and not reaching resolution, was massive. The bank managed an average of 2,500 complaints a month, and soon realized it would need automated support to resolve more complaints consistently and sustainably, avoiding the overflow of complaints to the Central Bank.

Not just aspiring to resolve more complaints in general: The Banco do Brasil set out to earn fewer BACEN complaints than any of the country’s top five banks.

“In Brazil, these rankings are published in all media, and to be on the top of this list is not good. If you are on the top of the list it means you are not treating clients like they deserve,” de Assis Melo said.

“This refinement and curation process with additional queries is a work in progress and always evolving, however based on our initial success we have high hopes for generating continued value with Interaction Analytics.”

Edgardo Ladeira

General Manager of Customer Service, Banco do Brasil

03 NiCE solution

An automated workflow for smarter conflict resolution

Based on its ongoing relationship with NiCE, and their success with solutions including NiCE Workforce Management and Quality Management, the Banco do Brasil selected Interaction Analytics to help them amplify, streamline, and error-proof their contact center-based conflict mitigation efforts.

The potential scenarios were complicated. As a first line of defense, BACEN suggests customers to attempt to resolve their conflict directly with their bank before the regulatory authority accepts the case themselves. Therefore, some customers came through Banco do Brasil’s contact center directly, while others called the BACEN directly. If the issue couldn’t be resolved internally at Banco do Brasil customer service center, the customer was redirected back to BACEN in a game of conflict ping-pong neither the bank nor the customer wanted to play.

Interaction Analytics had the potential to serve as both referee and coach, helping to keep customer complaints from bouncing to the regulator or other external agencies by putting into play an automated workflow triggered by customer mentions of “Banco Central” or “BACEN.” Whether the terms appeared in voice or text, these queries triggered the immediate engagement of a new back-office agents, which investigated the validity and risk level of complaints while the frontline agent continued to provide resolution. Based on the agent’s findings, the customer was routed through an efficient, internal resolution track based on risk level and conflict type.

Despite the seemingly broad queries, Interaction Analytics helped make this process incredibly efficient. Approximately 93% of the cases flagged by Interaction Analytics had some mention of the Central Bank. Of these mentions, 37% of the situations contained real threats.

For many of these threats, the attendant was able to provide first call resolution. However, for the threats found in the online solution that were not addressed, a case had been opened. A team of Back Office agents prioritized the treatment and monitoring of these occurrences, seeking an efficient path to resolution based on the level of risk and type of conflict. By precisely narrowing all customer interactions down to those that were high risk, then further refining to those that required high touch follow up, Interaction Analytics helped Banco do Brasil efficiently target its efforts where they would generate the greatest value.

04 Results

From first to last in a race to the bottom

From the dreaded number one spot in 2017 and 2018, Banco do Brasil ended 2021 in the 11th position. Since then, and through expansion of Interaction Analytics queries, the Banco do Brasil continued to work its way down the list, where it hovered between 13 and 14 (at the time of publishing. And most importantly, amongst the five largest banks in Brazil, Banco do Brasil reports the fewest BACEN complaints, an honor it has maintained with Interaction Analytics since the third quarter of 2022.

This last place spot was hard earned: From 1,000 BACEN complaints in fourth quarter 2021 judged to be valid, Interaction Analytics helped Banco do Brasil reduce the volume of legitimate complaints by 40%, dropping from 1,000 complaints in the fourth quarter of 2021 to approximately 600 the following year.

“Our primary objective was to make sure the minimum amount of cases left the bank, and we have accomplished significant results in that regard,” said Rely Evangelista de Oliveira, Solutions Manager of Ombudsman Office, Banco do Brasil.

Interaction Analytics provided other opportunities for the bank. For example, because of the back-office analysis and quality monitoring, analysts and supervisors identified opportunities for incremental performance and operational improvements, everything from malfunctioning headsets to reskilling needs. “It had nothing to do with the automation, but it created a new vantage point from which our supervisors could performance manage,” de Assis Melo said.

05 Future

From resolution to prevention

Banco do Brasil is further refining the “BACEN” query to continuously improve its complaint resolution rates with more precise workflows. Moreover, the bank is expanding Interaction Analytics to not just mitigate conflicts, but to prevent them all together, with eight new queries that align with critical KPIs along a successful customer journey. By mapping, monitoring, and strategically escalating flagged interactions internally, Banco do Brasil believes it can improve the overall customer experience.

Additionally, they are extending Interaction Analytics impact from the contact center into branch locations, where the majority of complaints originate. “By expanding Interaction Analytics to other points of contact we’re enabling a lot more prevention,” said Ladeira.

Amplified by NiCE Quality Management, Banco do Brasil has the confidence to continue growing its Interaction Analytics -automated workflows. Previously manual processes were difficult to quality control as there was no way to accurately measure and compare inconsistent processes. “Our workflow now controls everything, and adds a lot of value, not the least of which is the data and insights from which we can continue to improve and expand workflows,” said de Assis Melo.