Debt management tends to be treated as purely an operational function even by lenders that operate sophisticated risk analysis processes in other areas of their organisations; that is to say that its core objective is seen to be managing its internal costs as efficiently as possible. The predominant philosophy seems to be that in order to deliver better results you can either try to get more effort out of each member of staff or, if you’re already at the point of maximum output, you can increase headcount. This philosophy is outdated. It focuses on working harder and fails to take into account the benefits of working smarter.
Debt management departments no longer add value only by being efficient but also by being effective. In fact in most cases improvements made to the strategies employed in a debt management function lead to significantly bigger rewards than improvements made in the efficiency of operations.
The best way to drive an increase in debt management effectiveness is to implement risk based collections. A risk based collections strategy is simply one that starts with the core belief that not all customers will respond equally to any one action and so a one-size-fits-all approach to debt management must lead to areas where too much is invested in collecting a bad debt and areas where too little is invested; areas where good customers are lost and where bad customers are retained.
A risk based collections strategy should start by seeking to measure the risk of each customer and to measure the cost and performance of each possible action. With this knowledge, a series of specific strategies can be created to optimise the investment made in collecting funds against the expected returns of those investments. Some customer groups will pass through a light strategy with few direct contacts involved, others may receive more contacts, some messages will be communicated in a light tone, others in a firm tone, etc. The total investment made in the debt management function may remain unchanged but the distribution of that investment will certainly change.
One analogy that might help to clarify the role of risk based collections comes from the world of card games. Let’s assume the simplest of card games: a deck of cards is shuffled and a player’s money is doubled whenever a red card is drawn. Because the card order is random, the player is forced to treat every card as equal and therefore to play the same strategy at each draw. After an extended period of play the player will end the night even having doubled their money (red card drawn) as often as the lost it (black card drawn). However, if the player can count the number of red and black cards dealt and can retain that data, they can begin to see patterns emerge and can begin to make bigger bets whenever the odds are more in their favour and smaller bets when the odds are against them. Or, in other words, they can apply differentiated strategies based on perceptions of risk. Following such a strategy the player could end the night as a winner.
Risk based collections seeks to do the same thing. If an organisation does not know the risk of each customer they must treat every customer as if they were the same and so apply a one-size-fits-all strategy, perhaps only varying their approach based on the size of the debt at stake. The results of such a strategy are muted. However, once the organisation begins to segment its customer base into groups of similar risk it can begin to customise strategies for each of these groups – leading to significantly improved results even with the same staff using the same operational systems.
Three major components are required to implement a risk based collection strategy: a scorecard; strategy management software; trained staff.
Scorecards sit at the heart of risk based collections. A scorecard is a statistical tool used for forecasting behaviour. It works by first examining a database of known data and identifying similar characteristics within groups of accounts and then matching the on-going behaviour of each of those homogeous groups with regard to a particular metric of interest – percentage in default after 12 months, percentage to make a payment within 1 month, etc. So then, whenever a new account enters debt management its characteristics can be analysed and compared to the characteristics of each of those groups, allowing the organisation to assign it to a particular segment and to infer its most likely future behaviour.
The output of a scorecard is a score that relates to a given probability of an event happening. Unless lots of detailed information is needed, the actual score is usually summarised into a score band, each of which represents a group of accounts whose performance falls in a similar position somewhere along a risk continuum. For example, it may produce a score between 0 and 100 which is summarised into three risk bands representing groups with a high, medium and low probability of default. This data is usually augmented with other data relating to the size of the debt outstanding or some other proxy for the impact on profit should an account be written-off. Combining the two measures provides a simple risk matrix which can be the starting point for a fairly complex risk based collections strategy.
Simply understanding the risk inherent in each customer segment does not add any value to an organisation in itself. That organisation also needs to change the physical manner in which it handles each of these segments in order to do that. Thus, individualised collections strategies need to be developed for each identified customer segment taking into account the differences in expected behaviour. This is the second component of risk based collections.
Building a individualised debt management strategy involves selecting the series actions to be taken, determining the timing of those actions and, in certain cases, the specific script or letter format to use. In order for these to lead to continuous improvement the use of test-and-learn experiments should also be ingrained in the processes and the results of those experiments, as well as the business as usual strategies, should be reported on and monitored by management. Small portfolios might get by using manual procedures to design and manage these strategies but larger business will need suitable systems. A good strategy management system must be able to assist in the design and the management of strategies.
During the design phase it is important to have access to accurate and flexible data analytics. The strategy management system should be able to access historical databases and to display that data in a way that makes trend spotting and advanced analysis easy. The most advanced systems will even allow simulation of possible strategies – were data can be fed through competing strategies in the same way they would be processed live and the results compared for any important metric so that, for example, the most profitable or the cheapest strategy of several possible alternatives can be identified even where complicated interactions exist. Such advanced features can save a lot of time in the early stages of the strategy design process and so can add significant value to debt management strategies. However, the lack of such advanced tools need not stop the implementation of risk based collections.
During the strategy management phase it is important to monitor the performance of each strategy: comparing the performance of the strategies in real life to expectations as well as the results of test-and-learn experiments to business as usual. Whenever a strategy involves people – as debt management strategies do – it becomes very dynamic and so procedures must be in place to facilitate fast and regular strategy changes.
These strategies become real in the operational environment where the lender’s staff deal with customers. The most common means of implementing these strategies in the operational environment is by making adjustments to the three Ts as discussed in another article in this section. I will not go into great detail again here but I have included a diagram below which summarises the sort of strategy differences one might consider for each segment.
Suffice to say, each of the three Ts will be adjusted according to the risk of the account and the size of the expected recovery/ loss. So, on the one extreme an outstanding debt that is high risk and has a large outstanding balance is one in whose recovery the lender should be prepared to invest by using reliable treatments (even if those are more expensive); contacting customers early and often; by using a sever tone of voice at a time that is convenient to themselves. On the other extreme, an outstanding debt that is of low risk and that has a small balance should be passed through a strategy with a different set of priorities. In this case the treatments should be, first and foremost, cheap (even if this means they are less reliable); contact can be initiated late in the process and need only be infrequent; while the tone should be kept conciliatory and messages should be delivered at a time that is convenient to the debtor.
In order for staff to be able to successfully deliver on these dynamic strategies a new skill set is often needed. Gone are the days of debt managers as bullies. It is no longer sufficient for staff to simply understand the technical intricacies of their front-end systems. In today’s environment debt managers must also be sales managers and so staff need to understand customer behaviour and how to adapt to it. Some good training courses exist that can impart such knowledge on teams and this should always be considered alongside any risk based collections implementation. The technical changes made above will come to nought if the staff expected to implement the changes do not know how to do so effectively or if they simply do not want to change from their old ways of doing things.
* * *
Customer education is always important and it should be a priority for all lenders to point higher risk customers towards websites like these and other services that can aid them with their personal finances and budgeting.