Cloudification of debt collection
13 MINS READ
Highlights
- Economic recovery post the pandemic-induced slowdown is bound to result in a credit surge.
- We believe the cloud could change the debt collections game for good, minimizing delinquencies and maximizing customer trust.
- A cloud-based AI platform will improve workforce management and customer interactions while enhancing operational effectiveness and risk management.
- Banks must take a phased approach to adopting cloud for debt collection and delinquency management.
In this article
Abstract
A wave of delinquencies is around the corner.
As economies reopen and the world recovers from the pandemic-induced shock to the financial system, banks must prepare for the upcoming wave of delinquencies. The emphasis should be on increasing collections at minimal costs and enhancing compliance—a tall task indeed. Financial institutions must reassess their collection capacity and capabilities and enhance them by adopting an end-to-end transformation approach encompassing strategy, processes, and applications. To accomplish this, banks must carefully evaluate legacy applications and systems, identify break points in the collections life cycle, and upgrade existing models by leveraging cloud technologies. We analyze the various factors that banks must consider and recommend a strategy to reimagine the collections function using cloud technologies.
The North American debt: An opportunity in adversity
Credit usage is only going to grow.
Consumer debt, especially the revolving debt from credit cards, forms a significant part of the overall debt in North America. During the pandemic, the credit card debt and subsequent delinquencies declined due to a multitude of factors like government stimulus, lower consumer spending, uncertain future, lockdowns, and more. However, as economies reopen, credit usage is bound to increase (see Figure 1).
While some banks have already reported that credit usage is exceeding pre-pandemic levels, many analysts have expressed concerns about slow economic recovery and predict an increase in delinquencies as the economic effect of the pandemic wanes off. For banks, this is a source of concern: a huge amount is at stake, and considering the sluggish economic recovery, a substantial portion has the potential to turn into unrecoverable bad debt. While this could be a potential roadblock to growth, it also presents banks an opportunity to reassess their capabilities and capacities in delinquency management. They must use technology to reevaluate risk and customer outreach strategies in order to improve operational effectiveness and resilience.
The need to transform collections
Given the digital era we live in, consumers use a variety of channels to fulfill their banking needs.
The increased usage of smart phones has shifted customer preferences across communication, banking, and payments to digital channels. While many banks have digitally transformed their customer-facing applications, the core back-end and peripheral applications still use legacy systems and processes. This presents banks a huge opportunity to transform legacy back-end applications in the collections function using cloud technologies, advanced analytics, artificial intelligence (AI), machine learning (ML) techniques, and ready-to-consume application programming interfaces (APIs). To significantly strengthen their collections capabilities, banks must reevaluate three broad areas and seamlessly integrate them with front-end customer-facing applications.
Risk management
Traditionally, banks have relied on their internal risk scoring models, mostly using customer data residing on their systems. Rising delinquencies and finite resources require better segmentation and a more holistic customer view, leveraging data from external sources as well.
Multi-channel strategy
Many banks have invested heavily in technology solutions for multichannel capabilities like SMS, chatbots, emails, and more, to enhance the collection function. However, these channels still operate disparately in the absence of a solid, sequential multichannel strategy. Cloud, AI, and analytics enable tailored segmentation for tangible outcomes in terms of higher promise-to-pay and increased collections at optimal costs.
Collections operations
Using AI capabilities on the cloud to uncover patterns hidden within the bank’s data enhances collections operations. For example, ML can significantly improve the accuracy of volume forecasts, saving millions in costs as workforce schedules are based on better demand data. Also, compliance can be improved by using highly trained speech-to-text ML models to monitor calls and automatically identify cases that violate government regulations applicable to collections. This saves millions of dollars in manual call monitoring and avoids lawsuits.
Reimagining collections: Cloud shows the way
A phased transition to the cloud is the answer.
Cloud offers a range of benefits; however, success will depend on the right cloud platform. The target cloud platform must seamlessly integrate with the existing applications and cause minimal disruption to business as usual. Banks must adopt a step-by-step approach for their cloud migration journey and consider factors like flexibility, ability to manage complexity, investment, readiness to deploy new capabilities, and innovation appetite while transitioning to the target collection model (see Figure 2).
Figure 2: Target collection model
Embracing the cloud in the collections space can help with:
Timely identification of potential delinquencies
Existing risk scoring models use customers’ financial and behavioral information available internally, in combination with micro- and macro-economic indicators. API platforms on the cloud offer a variety of customer information across transactions, payments, behavior, and more. Leveraging this can significantly augment enterprise data lakes and bolster legacy risk scoring models. This results in a single comprehensive risk score that appreciably enhances the predictability and veracity of existing models.
However, getting all this information as a perennial real-time feed is beyond the capabilities of most legacy systems. In addition, cloud providers offer resources like higher compute power, ML models, and a set of easy-to-deploy tools, which are expensive to obtain and deploy in-house. Moving to the cloud, thus, makes it easier for risk management teams to quickly refine and adjust their models to enable higher accuracy and precision, which facilitates data-driven decisions.
Smart workforce management
An efficient collections model must strike the right balance between risk analytics and workforce management. Risk teams continuously deliver segmentation, forecasting, and modelling to the operations team so that the right resource with the right skills is provisioned at the right time.
Cloud enables rapid analysis of resource modelling, historical data on bank systems, industry trends, and other external indicators across multiple dimensions. This, in turn, defines a solid collections strategy based on roles, responsibilities, and skills of the workforce, as well as the risks associated with the debt. This enables operations teams to allocate accounts based on a combination of strategy and other aspects like the capacity to optimize the quality of operations, speed-of-change in risk strategy, governance, and controls as per prevailing regulations.
Effective omnichannel orchestration for multichannel play
Cloud platforms infuse flexibility into legacy collection systems, which enables organizations to try, test, and adapt risk strategies faster. Companies must leverage advanced analytics for micro-segmentation and customization of channel strategies based on customers’ preferred digital channels and time of day. Experience shows that more customers pay on their own using digital channels, especially those in the early stages of delinquency or predelinquency.
Personalization can start with using customers’ preferred channel to send a simple text message before moving to a chatbot. In the absence of a response, the communication can transition to a human agent. This significantly shifts the load from voice channels to alternative channels, leaving the voice channel free to focus on late-stage buckets and charged-off accounts.
A case in point
A large North American bank was operating with legacy platforms and multiple channels resulting in channel overlap and lower returns. In addition, the load on voice channels was high. All of this led to poor customer engagement and high cost of collections. We ran a pilot to transform the front office collection processes using digital and cloud technologies, and delivered the following benefits:
- 78% promise-to-pays met through digital channels
- About 31% reduction in voice channel usage
- 40-75% improvement in digital channel penetration
- Reduction in intrusive calls in early delinquency cases
- Reduction in manual calls
Natural language powered conversations
Cloud-based platforms help redefine the possibilities of AI-powered conversations. For example, when a customer is unreachable, it triggers an SMS to initiate a conversation with the customer through an AI-powered chatbot, which can offer personalized payment options in a non-intrusive manner. The chatbot sends links with personalized payment options based on the customer’s personal and financial situation. This helps the customer to evaluate the options and make payments and/or set up direct debit instructions without logging into the bank’s portal.
Operational effectiveness
Moving to the cloud empowers banks to adopt intelligent technologies with the potential to ramp up operational efficiency across customer experience, process, compliance, and training.
- AI assisted agents support human agents during customer calls and chats by identifying intent and providing real-time, step-by-step assistance. Natural language processing (NLP) techniques identify call drivers and sentiment to glean insights from customer interactions and improve call outcomes.
- Speech-to-text and text mining technologies analyze millions of offline call records and chat transcripts to identify customer sentiment, behavior, customer preferences, and regulatory fails. Such quality analysis delivers insights to improve compliance management, design and deliver targeted solutions that suit customer preferences, and train operations staff.
- Computer vision technologies transform manual, paper-based workflows. Computer vision and image-capturing ML models automate the process of reading and extracting data from scanned documents, cleanse and validate data from structured and unstructured documents, and process that data across various legacy systems.
A case in point
A leading US bank’s legacy back-office processes led to increased correspondence, manual operations with high cycle times, and higher costs. We ran a proof-of-concept (PoC) to modernize their back-office collections operations using computer vision technologies to automate document review, categorization, and routing. The bank realized several benefits from this PoC:
- About 80% of the process steps automated
- 60% reduction in cycle time
- 50% reduction in manual effort
- Increased accuracy
In summary
A cloud makeover for the debt collections space will bode well for all stakeholders.
The pandemic has proven to be an inflection point for collections, underscoring the need for a business-driven transformation approach. Success in the collections space will depend on banks’ ability to quickly adopt a robust cloud framework with the flexibility to seamlessly connect with legacy applications, a well-executed cloud roadmap, right skills and talent aligned with the risk strategy, risk analytics, and more. Banks that act with alacrity in moving to a cloud-based collections model stand to gain, especially given the possibility of rising defaults in the coming months.
Contact
Contact
