In the past decade, collections shops globally have faced different challenges depending on the markets they operate in. In many markets, historically low loss rates led lenders to reduce the size of their collections operations. Their shops are now running lean, focused on efficiency and minimizing costs. The low losses also led to relaxed acquisition criteria and offers of riskier products to customers with low credit scores or little credit history. As consumer spending—and revolving balances—rose, so too did delinquencies. With loss rates mounting, lenders suddenly find themselves without adequate tools or staff to address their at-risk accounts effectively. The new approaches discussed here can help all lenders achieve best-in-class collections operations, improving revenues and lowering costs.
Toward distinctive collections operations
The return of higher loss rates in North America over the past six quarters has not been accompanied by a commensurate rise in the sophistication of collections operations (Exhibit 1). Many shops are still organized by stage of delinquency, and many still rely mainly on the phone channel to contact customers. While risk segmentation may be used in the first few weeks of delinquency, as a means of identifying customers to whom a live agent will be assigned, thereafter additional segmentation is rarely applied. Contact channels and treatments, furthermore, are adopted without the necessary evaluation. Brute-force dialing is the norm, leading many customers actively to ignore phone calls. To a lack of sophistication, other challenges can be added—including consumer-protection regulations and the prevalence of debt-settlement companies determined to enlist long-term delinquent customers. Clearly, lenders need to invest in smarter and more effective collections operations.
Though major issuers in every market face the same macroeconomic pressures, they exhibit widely varying levels of collections performance. Many abandon (charge off) a high proportion of delinquent customers without ever directly contacting the account holder (making a right-party contact). Obviously, these lenders need to improve their strategies with, for example, better “skiptracing”—the process of locating good contact information for making right-party contacts.
The more advanced collections shops are deploying machine learning, advanced analytics, and value-at-risk segmentation. These powerful new tools can support lenders in prioritizing at-risk customers for closer and more productive attention. The value-at-risk approach, furthermore, is changing the way successful collections shops operate. Based on this approach, they are applying a variety of contact methods and more collector autonomy to offer tailored solutions—including more generous settlements. To be successful, the approach does not necessarily require major IT investment, but it does need a well-trained frontline staff that actively manages performance, utilizing daily data and reporting.
The seven pillars supporting best-in-class collections operations
In organizations where collector capacity is limited, effective segmentation is of vital importance. While all organizations keep track of the number of days accounts are delinquent, more effective collections operations prioritize customers by value at risk. This is a score calculated from the product of outstanding balances and some measure of collections risk (Exhibit 2). Lenders have long assessed collections risk by using models resembling those used for underwriting consumer credit. Increasingly, however, collections-specific models are being developed and employed. The models are designed to predict the probability of customers’ defaulting or remaining longer in delinquency (rolling through 30, 60, and 90 days of delinquency). Some models even venture to predict customer receptivity to contact. Developers are increasingly incorporating machine-learning approaches into the models to better assimilate customer behavior and improve predictive power. Better enterprise data warehousing has allowed developers to incorporate into models the nightly recalibration of collections-risk scores and the combined values of a customer’s accounts.
The purpose of all segmentation in collections is to distinguish delinquent customers requiring human contact from those who will respond to automated messaging or require no contact at all. Lenders naturally want to reserve valuable live-agent capacity for medium- and high-value accounts at risk. It is in their interest, therefore, to identify as many delinquent customers responsive to automated prompts as possible.
Behavioral segmentation, discussed in another article, can be the perfect complement to value-at-risk (VAR) segmentation, identifying the behavioral clues to customer receptivity (or aversion) toward particular contact approaches and treatments. Once the segmentation is established, collectors can determine the contact and treatment strategies. Well-executed VAR segmentation will align efforts with the needs of the customer to find the most suitable ways to collect from all types of at-risk accounts.
Compared with more traditional delinquency segmentation, the VAR approach has several advantages. It helps lenders identify larger numbers of customers who can be routed to automated channels, including interactive voice response (IVR), with no detrimental impact. With fewer customers routed to live agents, collections shops ensure that longer, more effective conversations can take place where needed. Another advantage is that the VAR approach permits early redirection of high-balance, high-risk customers away from low-skill early-stage teams to better-performing collectors. The approach also more quickly routes low-balance customers in the later stages of delinquency to less frequent contact by less skilled collectors.
Yet another advantage of VAR segmentation is that it favors contact intensity over accounts per collector. Traditionally, collections departments call accounts in early-stage delinquency most frequently, steadily deprioritizing as the accounts move closer to the charge-off date. This usual approach is based on the assumption that all customers are less likely to pay the longer they remain in delinquency. Staffing is consequently geared toward managing the early-stage calling load, with accounts per collector an important measure of efficiency.
VAR segmentation reverses this emphasis. Customers representing high value at risk are assigned to a dedicated high-VAR team, which seeks to maximize the intensity of contact. Each agent spends the time needed to work out a tailored repayment plan. Medium-VAR customers are assigned to teams with targets of less contact intensity, while remaining customers are assigned to low VAR agents or automated outreach. Dialing penetration and saturation targets are achieved and maintained by regularly reviewing the volume of accounts assigned to the high-, medium-, and low-VAR teams and adjusting thresholds as needed. Even in collections shops with severely constrained capacity, customers with the highest value at risk remain the top priority, while low-VAR customers can be deprioritized or left alone to self-cure.
An added tool to the VAR model is an account-ownership team for the highest-VAR customers. Analogous to the classic recoveries model, this team comprises collectors who are each assigned a portfolio of accounts. It is an expensive tool, designed to address the accounts with the greatest value at risk to the bank. Team members treat their own lists as would true account managers. They are responsible for finding customers, building rapport, and providing permanent solutions. Given the low account-to-collector ratios, ownership teams can emphasize cures over taking payment. The primary goal is to eliminate accounts from delinquency, either by bringing them current or assigning them for an early exit. Collectors are encouraged to show initiative, sorting their portfolios not just by risk and balance but time since last contact, frequency of right-party contact, or even most recent payment activity.
Banks need to design skiptracing practices in compliance with all applicable privacy restrictions, which can vary significantly from one geographical jurisdiction to another. Where permitted in the United States, for example, three levels of activity are usually pursued, depending on account importance and the interval since a successful contact. For accounts without any contact details or without a successful contact over the first 30 days or so, the use of semiautomated batch skiptracing is common. Files containing customer identifiers are submitted to credit bureaus overnight, and the contents of the returned file (with any additional contact information) are loaded onto the dialer in time for the next morning’s calling campaigns. Automation creates a routine process that is also cost effective, with reported successful location rates in excess of 10 percent.
Collections shops can look to the wider enterprise as a source of customer-contact details. Siloed operations at many financial institutions prevent the routine sharing of details obtained through customer-service channels. A collections team can often locate at-risk customers simply by assembling these details from across the enterprise. Specialized skiptracing using recovery agencies can also contribute to the effort, as these agencies often find customers quickly that the bank cannot. Given the cost, however, this solution is usually reserved for customers representing the highest value at risk.
Frequently, banks fail to make successful contact with half of the accounts that are charged off. A portion of these may be linked to fraudulent activity or belong to customers declaring bankruptcy, but many are charged off simply because collectors were unable to make contact. Some customers in this group were not actively avoiding contact; rather they were not contacted due to poor onboarding or faulty skiptracing. It is worth the effort to contact these customers properly, since once located, they may be as willing to pay (or not) as the group with good contact details.
3. Contact strategy
The means to make contact include email, text messaging, posted letters, and telephone calls. In recent years, successful collectors have been using a range of newer digital channels as well; younger customers may be especially responsive to mobile apps (including chat) and social-media sites.
When it comes to calling, mobile phones rather than land lines are today’s norm. Collectors recognize that many customers now screen live voice calls, making brute-force dialing at all hours less effective for making contact. Research indicates that varying call intensity can help, along with the use of self-serve and voiceless options that create a more engaging experience. Collectors have found that customers gain a sense of empowerment when they can choose the contact channel, their own delinquency solutions, and the timing and pace of repayment.
4. Treatment choices
Rising delinquency rates strongly correlate to the rising proportion of customers who are unable to make full past-due payments and who may ultimately default. For lenders, it is essential to identify these customers early and offer them a range of treatment solutions. Many issuers reduce interest or extend the term of a delinquent debt to make monthly payments more affordable. More advanced collections operations will offer customers the opportunity to “name their own price.” This approach allows customers to set truly affordable terms and helps to extricate them early from delinquency. Some issuers have been able to develop advanced behavioral models, enabling the early identification of customers with the highest likelihood of rolling straight to charge off. These can be presented with settlement options, often at levels well below traditional 70 to 80 percent targets.
5. Frontline capabilities
A skilled, well-informed, and motivated workforce is always at the heart of successful collections operations. Frontline collections is a difficult work environment, much more challenging than ordinary call centers. Collections staff need a financial education as well as skill in negotiating with distressed customers in a highly regulated calling environment. These parameters define the complexity of employee selection, training, and motivation and indicate that assembling an exceptional staff is a challenging process.
Collections shops are under constant pressure to lower operating costs, and best practices are hard to come by. Collections service centers are often moved to low-cost locations where the pool of skilled candidates is usually more limited and competition for their services can be stiff. In such locations, successful collectors can be hard to find. The desired social skill set, furthermore, is elusive. An ability to connect on a human level is important, but some outgoing and gregarious candidates will not be the most successful. The ideal collector is calm, methodical, and self-starting, with a propensity to solve problems and an ability to engage and empathize with different kinds of people.
Successful onboarding involves training in the technical side of operations but also in the art and science of negotiation. Collections centers with low attrition and solid early-tenure performance will usually provide new collectors with an extended “nesting” period for integration. Trainee graduating classes are supported by their own dedicated coaches providing side-by-side instruction and monitoring.
Once on the job, many collectors come to think that they have little control over the frequency with which they are able to speak with the true account holder (make right-party contact). Experience suggests, however, that the most skillful collectors find a way: collectors achieving the highest right-party-contact rates consistently perform in the top quartile among peers month after month. Top performers are the most persistent and best able rapidly to locate the right party in order to begin fruitful discussions. They also tend to approach customers in a familiar, congenial way, building rapport. Their techniques can be taught and can lead to rapid and dramatic increases in RPC rates.
Empowering collectors to have intelligent, self-guided discussions with customers will yield performance improvements, more promises to pay, and better customer experiences. In a world focused on compliance and efficiency, many collections shops have resorted to tight scripting for calls. A boilerplate approach is followed for virtually all customer interactions, and collectors are trained to abandon conversations the moment customers do not agree to resolve their debt with such an approach. Yet a one-size-fits-all tactic can leave a lot of value on the table.
Collections shops improve effectiveness and customer and employee satisfaction by training collectors to rapidly assess a customer’s situation and empowering them to use that information to guide customers to better resolutions. Ability and willingness to pay are the most important parameters, and within that context, collectors can be trained to recognize four groups of at-risk customers, based on their willingness and ability to pay: able and willing (“Angelic”), able but not very willing (“Burdensome”), willing but barely able (“Caring”), and little able, whether willing or not (“Dire”). This segmentation focuses collections strategies and helps prepare collectors to overcome typical objections (Exhibit 3).
Finally, collectors should protect customers’ promises to pay by taking immediate payment when possible. Maximum value to the bank is secured when collectors set up automated payments and recurring payment plans. At some institutions, however, immediate payments are not accepted for some accounts. In these cases, collectors can reinforce promises to pay, in a brief dialogue about the payments process—when, how, and where customers will pay. This tends to instill a strong sense of obligation.
6. Performance management
The work of a collector can be monotonous and mentally draining. Repeatedly speaking with customers in dire financial circumstances can wear out nerves, while shifting aims to align with frequent policy and strategy changes can tax the skills of even the most talented new collectors. Collections shops can best ensure the success of their frontline staff by providing effective, ongoing coaching. This needs to be backed up by performance metrics to reward successful collectors and identify those who are struggling, so that they may receive focused and timely support.
With all good intentions, companies often promote their best collectors to supervisory roles. They should be aware, however, that those whom they promote might not be well suited for the burdens of administration. To get the most out of these promotions, companies should refit the role of collections supervisor, relieving it of many administrative tasks. Activities that are not essential to daily shop operations can be assigned instead to administrative professionals. Collections supervisors can then be assigned to spend the time they need to share their collections skills, wisdom, and enthusiasm with their teams. At least half a supervisor’s time should in practice be spent in one-on-one coaching sessions with team members. Teams should be small enough, furthermore, for supervisors to have meaningful personal interactions with each individual on a weekly basis.
The work of an effective supervisor is greatly aided by daily reporting of team-member performance metrics. Depending on the prevailing culture, supervisors can make performance details available for all to see (and compare), or share them with individual collectors each day. In this age of routine digital communications, performance reporting is sometimes relegated to a passive email push. This method misses a vital opportunity. Effective performance management really begins with pull from the collectors, eager to know how they have performed (relative to peers) and how to improve to meet challenging monthly goals. The performance data should be the basis for regular performance dialogues, which are most effective when performed in stand-up huddles, with colleagues encouraged to suggest performance-enhancing improvements.
For some collections shops, VAR segmentation is a brave new world, requiring wholesale organizational restructuring. Others will need only to route their customers to dedicated low-, medium-, high-, and ultrahigh-VAR-collections teams. One cautionary note: experience suggests that in operations with existing high dialing intensities, routing early-stage accounts to medium- or high-VAR teams (with lower daily penetration rates) can lead to lower performance of such accounts, even if the resultant excessive account rolls are cured in later delinquency. Instead, a more traditional early-stage collections strategy will be more effective. This consists of large teams of collectors engaging in high-intensity dialing, followed by VAR segmentation applied to those accounts that enter mid-stage delinquency. In this approach, medium- and high-value collectors are free to focus on accounts most likely to roll to charge-off, capturing incremental value from lower default rates.
Within collections approaches, emphasis tends to oscillate between aiming to collect fees and interest from delinquent customers (as their risk of default rises) and seeking payments to bring accounts current, which reduces the delinquent population and shrinks collections operations. In times of historically low losses, ever-richer card rewards programs demanded a larger revenue stream from fees and interest. Now with delinquencies rising, greater liquidation rates enable collections shops to limit risk as well as operations costs.
In a changing economic environment, shifts in collections strategy will become necessary. Success depends on a well-informed frontline staff, well trained in new approaches and behaviors. Change is always difficult, but the chances of new strategies succeeding are far greater when collectors understand their purpose and share a sense of ownership in the results.
Where is collections going?
Consumer-lending delinquency began to rise in mid-2015, with collections operations expanding in its wake. Economic conditions globally were mostly positive during this time, and prevailing wisdom suggests that much of the increase was the result of shifts in bank-lending policies. Especially consequential was the relaxation of lending criteria, as financial institutions competed for more customers and more revenue. To avoid a sting from the tail of such expansive policies, collections shops need to adapt operations. This will involve more than simply scaling up capacity to align with new demand. Collections needs to get ahead of future losses by employing next-generation techniques, which include VAR segmentation, strategies informed by behavioral economics, and advanced digital collections tools.
Collections-management teams commonly find it difficult to reduce losses. Many initiatives lose lift before completion; others fail even to get off the ground. The decisive obstacles are many and particular, but three general issues can be called out. First, experienced managers sometimes persist with solutions that worked through the last cycle, despite getting poorer results in the present one. At the same time, they seem not to recognize the value of new techniques like machine learning. Second, an emphasis on trivial measures, such as cosmetic changes to the desktop, can sometimes be prioritized ahead of major initiatives like hiring staff. Finally, at many institutions, senior management continues to view collections as a cost center rather than a value generator, and minimize new investment.
A practical approach to creating effective collections solutions
Experience indicates that the development of newly effective collections solutions is best approached in the following practically oriented three-stage process:
- Assess capabilities. Start by identifying strengths and opportunities. Compare current performance with historical performance and competitor benchmarks; normalize performance for monthly delinquent dollars. Paydown rates can be indicative of overall collections effectiveness, and gap analysis against competitors will reveal the biggest opportunities for improvement.
- Create project teams. Assign team members to work on specific design elements for the new solutions; assign project managers to keep projects on track. Remove bottlenecks quickly with agile decision making.
- Implement solutions in test mode, revise as needed, and go live. Quickly identify successful initiatives, improve as needed, and scale up. Speed is important in implementation, but so is sustainability over the intended life of the initiative.
This approach will help whether the goal is to hire and train new talent, shift to a value-based strategy, or incorporate behavioral insights and machine learning into collections operations.
To respond to the changing demands of today’s credit environment, collections shops need to pursue a small number of substantive initiatives that will ultimately transform operations in digital and value-based directions. There is no passive formula for success. Within collections, leaders must be committed advocates of the new solutions if employees are to take ownership and succeed. The role of business leaders is to ensure success by creating the compelling business case for collections to operate as a center of value. With the needed investment in the seven pillars, collections can generate positive returns ahead of market competitors.