Leveraging Complaint Data for Greater Consumer Protection
Complaints are stories shared by unhappy customers looking to release their anger, seek compensation, or improve a service. When Jose, for example, called his pre-paid card operator because no cash was dispensed during a withdrawal at an ATM, he was relieved to know that they would credit his money back to his card. While it is important to solve particular instances of a problem like Jose’s, if supervisors do not dive deep into how financial institutions identify recurring issues, their extent and risks, the same issues will continue to happen.
Photo Credit: Raju Ghosh, 2016 CGAP Photo Contest
CGAP research has highlighted the potential of complaint data to be used to identify patterns that may indicate certain consumer protection issues at the provider level, and collate data that are useful for monitoring trends across the market. However, for financial supervisors in many developing countries, appropriate data collection presents a challenge. To increase the effectiveness of complaint reports analysis and eventually enable supervisory action upon such analysis, key questions need to be addressed: namely, what information is needed and how should that information be inputted and reported.
What to collect?
Apart from identification of the financial service provider (FSP), each complaint should be composed of at least four core elements:
Demographics set the complaint scenario. Information about age, gender and location of the complainant can help supervisors identify variations of issues across different consumer segments. For example, the Financial Ombudsman in United Kingdom identified that, from April 2015 to March 2016, among young people (under 25), bank accounts were the most complained about service, while among older people (over 65), credit insurance was the major problem.
The proper classification of the product complained about is relevant for identifying not only the specific consumer agreements and information systems used by the FSP, but also the specific FSP responsible departments, facilitating the communication between the supervisor and the senior manager in charge for it. For example, Bradesco — one of the biggest private FSP in Brazil — has different departments to manage retail consumer credit, credit cards and accounts.
The same product can be distributed by various channels that may provide very different service levels. For mobile money providers, for example, knowing if transactions over the counter (through agents) or mobile wallets (usually remotely, without agent intervention) are causing complaints may change the focus of supervision. The Financial Mediator Bureau in Malaysia identified that ATM transaction complaints registered have risen 9.2 percent between 2014 and 2015.
The problem or issue itself must be captured.
Two other important elements are not captured from the person with the complaint, but should be generated automatically by the institution for each complaint filed in its complaint handling information system:
Processing time to measure timeliness of the complaint resolution process. For example, an FSP can provide the average time for each complaint type, allowing the supervisor to identify the most complex or resource intensive issues.
Escalation serves as an indicator of the types of complaints that are more often escalated to an ombudsman, regulator or the court, indicating that the solutions given by financial institutions did not satisfy the consumer.
Challenges in capturing data
Ensuring accuracy in complaint databases starts with the filing of each complaint at the FSP level. Complaints are input manually and subject to biases and failures. The complaint systems interfaces should be designed in a way that operators and consumers can easily classify their complaints. For example, “choice overload” — the reduction in cognitive capacity for a person to pick the best option if too many are presented at the same time — explains why a roll-down menu to choose between five product categories can work better than one with 90 choices.
Itau, one of the biggest private banks in Brazil, in its complaint filing webpage, asks the consumer to fill out a four-field form with standardized choices:
Customer status: if she is a customer, whether an account or other service holder, or a non-customer
The reason for contact: a request, a complaint, a suggestion or a compliment
The topic: 29 choices, ranging from branch to debt renegotiation to POS (for which they use the colloquial name “maquininha” or “little machine”)
A sub-topic, with choices changing accordingly to the topic chosen
When it comes to collecting data through complaint reports, granularity matters. Independently of the variables to be collected by supervisors, if data points are excessively aggregated, precious and needed information is lost. When designing the layout for complaints data reporting, supervisors should consider a format that allows cross-tabulation analysis so that supervisors can prioritize their actions.
Supervisors also should look for comparability of the data reported by different financial service providers to allow sector-wide and peer analysis, for due prioritization of supervisory actions. Standardizing the terms and definitions used for classifying products, channels, and issues in a way that is generic enough to fit all providers and yet meaningful for monitoring is key, though not always an easy task. The Financial Conduct Authority of the United Kingdom, for example, requires providers to report complaints in a standardized layout.
Regulators face other consistency challenges too. For example, a complaint for one FSP can be a simple enquiry, while, for another, it happens only when consumers dispute a decision or when it involves financial loss. To ensure that the whole market is on the same page, communicating reporting standards prior to collection of data is as important as verifying them upon data reception. All terms for reporting should be clearly defined, avoiding misinterpretations and generating comparable reports.
Standardization and intuitive interfaces for filing complaints are key to obtain high-quality data, whether at the FSP, ombudsman or regulator level. When this data is appropriately reported to supervisors, it can unlock valuable analysis. Instead of focusing on individual complaints and solutions, supervisors can identify riskier products, channels and processes at the FSP and market levels to effectively address them, requiring changes in product design and internal procedures. This way, Jose’s complaint will not be just a story of an unhappy customer, but an essential part of improving services to all consumers.