Optimisation can help the credit and collection industry provide better outcomes for the customer
The credit and collections industry has always been quick to take advantage of the latest ways of thinking to improve the work that it does and to treat customers more fairly. So when a new methodology of ‘optimisation’ emerged in the marketing arena, it was quick to see the parallels with its own work and to take advantage of the best practice.
Optimisation is a technique which relies on complex algorithms and statistical analyses to model the potential outcomes of strategic decisions that a creditor or collector might take. Essentially, rather than going through a traditional iterative process of champion-challenger exercises, optimisation models the potential outcomes of a decision off-line and provides results upon which a final decision can be made.
Such a comprehensive process can be used at any point throughout the credit and collections cycle, from pre-qualifying, ID checking and fraud avoidance, through to collections and debt policy.
Optimisation empowers the user to take a ‘top-down’ approach to strategy design, focussing on the overall objective of the decision point and how the array of decisions, that can be made on the customer base, can deliver the greatest overall benefit.
Many creditors and collectors already make sophisticated decisions based on champion-challenger projects.
Optimisation can improve this process by modelling all possible outcomes in an off-line environment.
As such, many in the industry are keen to take this technology on, but are concerned about not having the resources internally to manage it. However, in reality, the requirements need not be complicated, and there is help at hand.
A successful optimisation project will require:
- A sufficient breadth of data on the customers.
- Adequate data quality with good historic information and data on the outcomes of various decisions in the past, for example how these have affected profitability.
- A good understanding of the decision point itself. Some decisions will have more possible outcomes and implications than others. Deciding whether to accept a new customer or not could be a very simple decision, but then looking at whether to accept a customer at a given price will open up the scope of possible outcomes, each with further implications.
- An understanding of any regulatory requirements, internal organisational and portfolio constraints, and the operational implementation of an optimised strategy.
From pre-delinquency to debt sale, senior collections professionals have been quick to look into this technology, because it is ideally suited to important policy decisions, which will have a range of possible outcomes.
Take, for example, a decision of what can be done to avoid clients tipping over from a pre-delinquent to a delinquent state. A whole range of lettering and contact options are available, but different considerations will impact upon thedecision such as:
- The culture of the industry sector.
- Each option will have potential benefits and costs; this is a classic situation for optimisation, where there are a number of possible outcomes, within a restricted budget.
- The long-term impact of the actions in terms of customer behaviour, churn or contract renewals.
- After delinquency, likewise, there are various decisions to be taken, including a choice of different DCAs and enforcement methods.
Methods of optimisation
In general, we offer two optimisation solutions. Neither is necessarily more sophisticated than the other, but each option is valuable in different ways:
- Strategy tree optimisation – many businesses will already use strategy trees to model potential outcomes of their decisions. Optimisation can be seamlessly integrated into this to provide better outcomes.
- Marketswitch optimisation – this relies more on making decisions at an individual level, allowing the creditor or collection activity to see the outcome of potential decisions on an individual customer.
Such concepts are no longer ‘out of reach’ or theoretical. Indeed we have worked on a recent example where a major retail bank wanted to achieve the right contact at the right time via the right channel for every customer. In doing so it realised significant performance improvements across its entire customer base of five million.
The strategy tree optimisation method was used to identify the best collections path to assign to each
customer, and a tree was produced for rapid deployment in their existing systems.
This brought a significant reduction in net impairment and path-specific customer level models driving better alignment of risk and return, and a return on investment in the project in less than three months.
Optimisation has also been used to assign debt cases to the best collection agency. One particular example of this resulted in a 15% increase in balance collected, as well as a 9% increase in revenue for each DCA.