In the context of managing debt collection, data makes it possible to address customers intelligently. As a result, recovery rates can be increased by an average of 15%, along with other more qualitative benefits, as explained by Roxana Racaru, Data and Digital Performance Director at Collection Services, the collection arm of Concentrix Payment Services.

The challenge is well known: to contact the right person, at the right time, through the right channel and with the right message. Easier said than done! To achieve this, it is very important to have reliable data. In this way brands will be able to manage their outstanding payments themselves or entrust them to a partner in an intelligent manner. Possible options include: implementing a self-care solution, which will give debtor customers a high degree of autonomy.

In practice, the first stage will take the form of analysing the contact file provided by the brand. A clear history of contacts, ideally over the past six months, will be needed to identify profiles or payment behaviors typical of a group of end customers.

We will then have two pieces of vital, strategic information:

• A reachability score, which will define the preferred contact channel and the time,
• Customer behavior with regard to debt (late payment, partial payment, etc.).

It is on the basis of these two calculated items of data that a recovery plan can be chosen or formulated. This will be based on a predictive model of behaviour, while minimising the effort required of the recovery teams and preserving brand image. Taking this approach will be more conducive to a smooth recovery of debt and increased customer loyalty to the brand.

Data, Automation…and Human Expertise!

To meet these requirements, data and automation are not enough: a human eye is needed. Specifically, in the context of a given recovery scenario, it will be necessary to analyse the failure that was encountered at a certain stage.

For example, a message placed on the answering machine has had no effect: is this a personal trait of the customer? Or do we need to create a new category: “does not respond to answering machine messages”? Only human analysis, in test & learn mode, can provide an intelligent answer to this question!

In this general context, the approach taken by the Collection Services department, which specialises in collection at Concentrix, is unique in two ways:

• The Concentrix teams have always had strong customer relationship expertise, allowing them to understand the customer in the round, not just from the perspective of their unpaid debt,
• The concept of scores and reminder procedures are not standardized but customized, depending on the criteria provided by the customer, as well as on the final customer/brand/business sector contexts. This does not of course prevent implementing automated processes, which are always necessary.

In addition to the feedback collected by the recovery negotiators over the telephone, this human expertise is explained by the fact that Concentrix makes use of data scientists experienced in the recovery business. They are part of a consulting unit of Concentrix, particularly in the area of data science and customer experience, which works daily with the Collection Services team.

As Audrey Gandoin, Data Science Manager at Concentrix, explains: “Concentrix has a particularly important feature: its specialist collection teams—Collection Services—and Concentrix work in close harmony at Concentrix’s headquarters. There are two advantages to this organization:

• Proximity and the complementary nature of their expertise—unique in France—which make it possible to effectively manage scoring activities and the creation of new functions,
• A choice of channel and contact points that are best suited to each customer.

To give just a few of the results, by way of illustration (in the telcos, utilities, bank insurance and BNPL sectors):

• A 20-40% increase in the reachability rate
• A 5% to 30% increase in the settlement rate

It can be concluded from this that the “old-fashioned” debt collection practices, which relied solely on repeated high-volume, standardized letters and calls, are no longer relevant! ”

15% better performance on average

To sum up, this methodology provides clients with two major advantages:

• Improved performance: this is clearly reflected in a better average recovery rate of 15% and NPS ratings above 50.
• A high number of new insights: predictive models often reveal behaviors or issues that were not previously identified. These insights make it possible to introduce improvements to the way in which the brand’s processes are organized, favoring a best-in-class customer financial experience.

Ready to improve recovery rates while protecting your customer experience?
Discover how our debt collection solutions combine data science, automation, and human expertise to deliver smarter, more effective debt recovery.