Ten years of underinvestment in the collections and recovery core is coming back to haunt financial services.
The pandemic and consumer preference pushed companies even further into digital interaction and the digitization of operational business practices. New tech has transformed platform capabilities and service expectations. But now, without that investment, too many companies are trying to implement real-deal digital collections strategy without the architecture to support it, argues Michael Orefice, Business Practice & IT Leader at Bridgeforce.
Here's what those companies are missing. New, modular, cloud-based platforms are reinventing collections and recovery right now and many of the pain points that have been plaguing business leaders for decades will disappear, Orefice says. Instead, collections and recovery strategists will need to face a completely new set of challenges: artificial intelligence and advanced data modeling.
The Game Has Changed
The banking industry is facing significant transformative pressure from almost every direction. A global pandemic, political change, technology evolution, FinTech disruptors, and withering margins are fueling the need to redesign how banks do business. Business executives are inundated with internal and external sound bites about the promise of AI, the magic of Robotic Process Automation, the extinction of the telephony channel, and, of course, the need to embrace all things digital. The foundations of the business and traditional components of the buying decision are being reshaped and redefined.
The Future of Default
The collections and default business is a cyclical part of the credit lifecycle, experiencing considerable variation in investment and prioritization based on external economic conditions. Heading into pre-pandemic 2020, many organizations were beginning to reprioritize default capabilities with a general acceptance that the global economy almost certainly had to slow down from its record setting multi-year growth. Efforts to revitalize the collections and recovery core after more than 10 years of underinvestment brought to light some harsh realities. Principal among them was the fact that technology advancement had shifted the role of platform capabilities and service expectations. The pandemic pushed the envelope even further as the ability to perform non-human digital interactions, and the convenience it created, forced a change in consumer preference while creating a sustained focus on the digitization of operational business practices.
The end result is that most traditional platforms lack an architecture that can support the intelligence, dynamic workflows, and digital interfaces necessary for real digital collections.
The Next Migration Should Be Your Last
Since the early 1980s, the collections industry has been evolving in parallel with technology advancement. In the early 1990s, paper collection cards and rotary phones were mothballed in favor of mainframes and digital dialers. The millennium ushered in tiered and distributed systems enabling operational advantages and the concept of specialized segmentation for classifications like bankruptcy, deceased and repossession. In 2010, evolutions in workflows, business rules engines, and integration capabilities with external interfaces for Bankruptcy networks, Repossession networks, and other specialized activities helped streamline and optimize collections and recovery operations.
The Cloud is changing the rules…AGAIN. In traditional platforms, deceased, SCRA, repossession, foreclosure and others are attributes (often referred to as an account status) of an account in the recovery inventory. Associated with these account statuses are sets of processing rules, operational steps, and even integration flows. In other words, they are small subsets of functionality embedded in, and dependent on, the entire platform.
But what if they weren’t? The modularized construction of a cloud native solution enables critical functions like bankruptcy and repossession, to name two, to operate and exist as independent components. In doing so, it opens the door to a host of immediate benefits. Each module can be maintained and advanced independently at its own pace. Migration away from a current traditional platform can happen by function (i.e. bankruptcy, deceased, SCRA, etc.) rather than by portfolio to simplify the effort and reduce the risk and likelihood of time and budget overruns. The function itself, like SCRA or Deceased, can exist within recovery operations as its own entity instead of a bolt onto a traditional platform.
A true cloud platform isn’t a monolithic entity. It is a well designed set of technology relationships assembled together to facilitate outcomes. The underpinning of that type of architecture is the ability to update, replace, and optimize function and technology independently without having to change the whole monolithic stack. At the 1,000-foot level it is incredibly complicated, but at the 100,000-foot level, it means that your big-bang release events and platform migrations are a thing of the past.
If that isn’t enough to get everyone excited, the funding, planning and resource burn spent trying to prioritize, budget, and eventually execute these activities is gone, too!
Intelligence and Automation
Collections and recovery efforts are a series of personal interactions evoking a variety of customer emotions from anger to hopelessness and everything in between. In most cases, people don’t choose to be delinquent but have succumbed to either external factors or a little short-sightedness. Financial institutions have well-defined treatments that can be used to help customers remediate their financial duress.
To date, most collection strategies place delinquent accounts into buckets based on models and scores and try to facilitate an engagement event to gather updated information and setup payment arrangements. In most instances, accounts run through a series of processes overnight to recalibrate the scores and, as appropriate, alter the placement of the account based on the anticipated next best action for the following day.
The only thing more tedious than this batch process is having to write about this batch process for what is now a third consecutive decade.
Modern digital solutions weren’t borne from the batch process mentality. Digital processes are the by-product of human-less web and app interactions in which the bi-directional flow of information has to occur in real-time or the expected business outcome doesn’t materialize. In essence, it took the serialized process flow of the batch process and flattened it so that instead of a single linear journey with a defined start and end point, it is a series of micro-decisions driven by events within the journey.
Digital interactions are designed to exchange information as part of an engagement journey in which a general business outcome (bring the account up to date) gets decomposed into a specific outcome (offer a payment plan) that is mutually agreeable based on the updated information provided during the interaction.
A real-time digital micro-flow example might include the following discrete steps:
- Get new or additional customer information
- Feed that information to existing intelligence entities
- Refine the desired outcome precision
- Previous desired outcome: Full-cycle payment
- New desired outcome: Re-age
- Respond with digital re-age content
- Wait for customer response
Micro-flows are repeated until there are no more alternatives furnished by the intelligence entity (Customer is referred to a specialist) or one of the provided remediation options has been accepted.
The ability to begin, facilitate, and complete recovery outcomes via the micro-decision process opens the door to leveraging non-telephony channels in more productive ways. To date, incremental use of specific channels like SMS and email have been constrained by the inability to interact with the customer in a dynamic way. The lack of real-time decision execution and the ability to extricate the human actor from the flow has made true digital engagement too risky.
A micro-service architecture eliminates those problems. Virtual agents can replace the human role and leverage natural language understanding (NLU) and natural language response (NLR) to facilitate digital conversations. What’s more, virtual agents will create less operational risk because they will deliver consistent treatment across every channel. The agent intelligence is focused on the quality of the NLP and NLR while the customer treatment intelligence remains a function of model responses based on real-time information exchange.
A fleet of virtual agents operating under the architecture and decision intelligence highlighted above will help drive down cost to collect in multiple ways. First and foremost, human interaction costs may be reduced by as much as 70% making sure only the most difficult and complex customer situations are handled by recovery agents.
Another significant expense reduction will come in the form of rationalization. Dialer platforms, IVRs, chat and SMS providers have all developed and licensed some form of autonomous capability. These overlapping solutions can be dropped from the respective agreements in addition to any of the associated costs like the maintenance of scripted narratives and virtual agent training functions.
In the next few years, collections and recovery will be completely reinvented and many of the pain points plaguing business leaders for decades will be replaced by the challenges of artificial intelligence and advanced data modeling. There will be plenty to work on and the development of digital talk-offs, decision intelligence, and fully functional collection journeys will not be completed overnight. But, in total, the collections environment will start shedding the weight of aging systems and a lethargic rate of change adoption.
Customers will be able to execute complex recovery journeys across any channel. Rather than suffering the embarrassment of sharing financial problems with a stranger, people will be able to make meaningful payment arrangements free of social pressures and the perceived judgment naturally accompanying human interactions. Instead of ignoring financial problems, customers will readily and proactively modify repayment conditions as their financial situation fluctuates.
All the while, business leaders will enjoy the benefits of constant optimization across business functions and modules. They will grow to miss the enjoyment of trying to scope and prioritize major system upgrades and replacements. Business teams will long for the day that they created workarounds to solve the real problems that can’t be accommodated on their decades old platform. A new breed of challenges will come into focus and the manner in which the business evolves will reinvent itself…again.
Thanks for reading! If you liked this story, make sure to subscribe to Collections & Recovery. You can unsubscribe at any time.