The majority of collections teams now have a range of creative options in place to help customers in debt manage their situation and find the best possible outcome, quickly.

But all this is irrelevant if customers can’t or won’t engage.

Getting customer communication right is a fine art but can science offer a fresh perspective on human behaviour that could inform the way we reach out to customers? 

The Behavioural Insights Unit believe it can and their work with organisations such as the Money and Pensions Service, Metropolitan Thames Valley Housing, and the Housing Repossession Taskforce (on behalf of two mortgage lenders in Northern Ireland) has demonstrated that small, cost-effective ‘nudges’ can prove highly effective.

Put simply, nudges are ‘designed to align actions with intentions and to help overcome the human urge to act against our own self-interest, altering people’s behaviour in a predictable way without forbidding any options or significantly changing their economic incentives.’

When it comes to a debt episode, the potential is clear and these interventions have a strong correlation with regulatory guidelines as the overriding motivation in their design is to encourage positive outcomes for the customer. 

Once a compelling message has been created, it needs to be delivered in a way that optimises the chance of a successful engagement, getting the most appropriate message in front of each customer based on everything that is known about them at the time. Most organisations hold a wealth of data but may lack the sophisticated technology to extract value and apply it to the collections process.

While traditional economic cohorts are useful, the ability to make use of real-time consumer data has changed the nature of segmentation. This means rather than using occasionally updated ‘clusters’, consumer data can be continually evaluated and new information absorbed, computed and optimised on a rolling basis. 

In addition, machine learning algorithms are being used to predict a number of possible customer actions including the likelihood of self-resolution, of making a full or partial payment, and whether that customer is likely to return to the collections process, for example. These predictions can then be used to inform the nature of the digital or agent-driven journey presented to customers, not just once, but every time they fall into arrears.

In the interests of fairness, this doesn’t preclude any options but initially presents the options most likely to meet that customer’s need on that occasion. In the contact centre, agents would receive smart insights on the next best step, while keeping all available options open. When business requirements change, the algorithms can be tweaked to operate within new parameters or to answer different questions.

But can machines be trusted to do our thinking for us?

In order to be ethical and compliant, artificial intelligence must only ever augment existing human decision-making. At no stage should AI or machine learning make definitive or irreversible decisons that might adversely affect any person’s future. Technology can be be harnessed to empower its users to make more informed and effective decisions but not to replace existing knowledge or experience.

What is the end benefit?  Creating a real-time alliance between data and technology is vastly reducing unnecessary agent activity, reducing wasted effort and repetition, reducing friction, significantly reducing cost and helping organisations to deliver the kind of seamless experience customers expect. 

As technological advances filter through to the collections space, creditors stand to see the biggest changes to debt management since the introduction of automation.

As part of our quest to liven the debate and to bring new and interesting ideas to the collections community, Flexys is inviting Pantelis Solomon of the Behavioural Insights Team to share some of the unit’s experience in an interactive webinar, Engaging customers in debt: the science of successful messaging, on 28th July at 10 am.  As a complementary addition to Pantelis’s talk, the Flexys machine learning and data science team will be discussing how to maximise engagement and reduce wasted effort through simple but effective algorithms designed for the debt management sector. 

Collections professionals can register for the free webinar here:
https://www.workcast.com/register?cpak=7202560772473995

Jon Hickman, Chief Executive Officer, Flexys