8 Success Tips for Debt Collection in the Pandemic

Across EMEA, most countries have moved past the peak of the current COVID-19 pandemic. Whilst the various stages of lockdown being reversed vary, what is almost certain is the harsh economic reality facing consumers and companies as government support lessens and even withdraws completely.

The International Food Policy Research Institute reports that the global economy has been and will continue to be negatively impacted by this pandemic. The implications of the pandemic in the rest of the world have resulted in a decrease on South Africa’s export demand. In terms of recovery, the ‘Quick’ recovery scenario results in a GDP decline of about 5 per cent by the end of 2020

In Britain alone the International Monetary Fund predicts Britons will have to borrow more than £400 billion in two years as the recession caused by the pandemic delivers a blow to the public finances.

Collections departments are now bracing themselves for a tsunami of people unable to meet their financial commitments.

Working with our clients across various sectors and geographies, we have seen first-hand what best practice in collections today looks like. I shared our top tips to help everyone working in debt collection meet this unprecedented challenge in a presentation last week at the Retail Banker International virtual conference.

Tip 1 – What data do we need to identify the vulnerable?

It is essential to proactively identify the vulnerable. It is worth repeating that collections departments are now tasked with dealing with thousands of customers who are not in arrears due to the usual reasons and without the usual market data to help identify and segment them. These clients have been furloughed with a rapid, surprising and significant reduction in income and these dramatic changes in circumstance have yet to be detected by credit bureaus.

As an industry we have always been proactive about predelinquency. But now we need to dig deep across our organisations to capture data and find the vital information needed to clearly separate the economic victims from the steady state collections customers.

The data is there – be it from call, collections or communications systems and we need access to this information as it will inform how we talk to our customers who need to be segregated carefully.

Information that tells us how reliant on credit the customer is, card utilization, renewal of UPLs
Information that tells us about their financial morality – transactor not revolver, direct debit payer, early settlements of UPLs, takes advantage of interest-free offers, low LTV, lower than average balloon payment in terms, number of credit lines
For issuers – the changes in card transaction spend type and velocity
Credit stress – number of new applications for credit internally and at bureau, frequency and degree of overdraft facility
Whilst rare, any information about their industry, occupation, earning potential, stability (number of address & telephone number changes)

Without much of the above, these customers will end payment holiday periods as low-risk customers. There is a real urgency to understand the true degree of risk on these customers — on which you know little, as they are not normally a collections customer.

Just as this is unlike any other crisis, COVID customers are different. We need to ensure we have or can capture at the earliest possible opportunity, in a format that enables empirical use:

The reason for delinquency
The duration of impact
The disposable income
Affordability over time
Industry sector and role within

additional to our standard data this information is necessary to help us understand not only their position now, but where they are likely going to be for the rest of this year and beyond. These are customers we are going to have communicate regularly with to understand the changing nature of their personal circumstances.

Three distinct collections phases

There are three distinct pandemic related collections phases – tactical, mid-term and the more strategic long-term. Each requires a different approach to customer segmentation and communication.

Tactical Phase – Day 1-90: Capturing the correct data to ensure COVID-19 is registered as reasons for the policyholder being delinquent will be critical in relation to both near-term relief and impairment reporting later in the timeframes outlined.
Mid-Term Day – 90-180: As payment holidays expire, customers will require a pre-contact to ensure the right options are put in place for their best outcome in the longer term. Interaction will be essential at this point and with better segmentation; the customer journey may change to either standard collections solution discussion or longer term forbearance options. In either case, validated affordability will be important.
Long-Term Strategic – Day 180+: Multiple touch points with the customer will be required to validate their financial/personal wellbeing and understanding the reasons for COVID-19 payment holiday/forbearance will be critical for regulatory reporting.

Volumes will still be high and customer journeys will get more complex as initial volumes flatline. There could be a bigger share of vulnerable customers needing forbearance. Four UK tier 1 lenders have announced an expected £20.2 billion increase on impaired loans. We expect an increased need for digital-led payment plans and a higher demand for loan modification so the longer-term financial impacts can be worked through. Ultimately, companies will have to manage higher impairments and losses, which will attract C-level attention to the overall collections and recovery performance.

Tip 2 – How can we better assess affordability?

We need to triage assessment levels — we need to know the degree of assessment necessary for someone who feels a short-term collections solution will suffice versus the amount of validation required for a customer seeking a long-term forbearance solution Dig more deeply into household balance sheets and cash flow. Where it is available, use Open Banking data to determine accurate affordability assessments. Credit bureau data and other bbehavioral indicators will signpost trouble with current debt, as will the use of forward-looking analytics to detect the effect of incremental debt on default risk

There isn’t a single generally accepted measure of over-indebtedness or affordability risk and a combination of criteria and approaches should be applied to evaluate whether a person is overindebted.

Again, COVID customers are different as we don’t usually know what caused this consumer to default. Credit scores don’t tell us, so we need to continue to collect more data to gain a deeper understanding of the household balance sheet and investigate past financial behavior data for signs of affordability challenges.

Tip 3 What data do we need to collect to manage performance and regulations?

The additional data to capture in context of COVID-19 includes:

Reasons for hardship: Unemployed, furlough/reduced income, medical, quarantined.
Type of relief applied: Payment holiday, reduced instalment – what relief the customer had and when did it stop.
Industry sector, so you can estimate need for prolongation.

This is not a static process but a dynamic one. Communication is key to achieve an update on the customer’s circumstances and to segregate those who are falling into delinquency.

Tip 4 – What is the best engagement strategy?

The best engagement strategy should be led by the customer – so how open can you make your systems to customers?

The ideal is an omnichannel approach – contacting the right customer through the right channel at the right time. Can you change your contact strategies quickly to enable customers to auto-resolve on omnichannel platforms including auto voice, iSMS/WhatsApp, digital direct API, email, mobile app, online or through edocs?

Tip 5 – How should we manage payment holidays ending?

The key here is to keep reviewing, as the customer debt profile will change significantly. COVID-19 customers will have a higher financial morality and be uncomfortable about being in debt, with a higher intent to maintain their standard of living and service their debts.

Review your restructure toolkit. Can you optimise the selection using the right tool for the right customer segment? For example, can you provide a temporary solution which can be automatically reviewed frequently and perhaps free up creditor teams to focus phone capacity on those with most need?

It is essential to maintain contact to understand situational change and generate monthly updates via Open Banking, or chatbot to identify any changed circumstances.

This is a balance as you don’t want to over structure your procedures. Ask yourself these questions:

Is your process fit for purpose? Can you save consumers and creditors multiple calls whilst also ensuring appropriate forbearance that enables continued debt servicing, which in turn will mitigate the long-term impacts on individuals, businesses and the economy?
Can you protect consumers’ credit ratings during the crisis to enable access to credit? When the crisis ends, can you prove validity of credit scores and plan for a controlled way to enable people to return to “normal”?
As we move past the crisis, can you offer an updated budget to enable seamless transition from COVID relief (e.g., payment holiday, other) into mainstream creditor repayments?
Finally, for those with longer-term financial stress, can you enable a seamless digital transition into debt solutions through a virtual agent? In other words, can you take all the data insights, information validation and decisioning into your customer auto-resolve tools? Can you do so to the satisfaction of the regulator, the customer and your portfolio performance?

Tip 6 – How can we better segment customers for treatment?

Segmentation is a vital dynamic process. We need to look at pre-delinquency segmentation, lockdown and payment protection segmentation and post-payment protection segmentation and how the segments change across these categories. Our decision science teams need to provide regular insight as to what they are seeing and how these insights should influence customer treatment decisions. In the segmentation below, we can see how, as the phases progress, we are segmenting not necessarily by risk but more specifically by who we can let run with no need for help, who we need to assist or provide advice and stay close to, and who we need to assist and periodically monitor. This is about really strong prescriptive analytics.

Tip 7 – Advanced: How do we use data-driven strategies?

The current debt tsunami provides the perfect opportunity to use mathematical optimization, which has long been available to debt collectors but too rarely used. This is the most sophisticated analytics available to determine decision strategies and specific actions that will best meet objectives given all the data available, and given the constraints.

Optimization in collections can be applied across the different stages of the C&R lifecycle to solve different problems at each stage by defining the decisions that need to be made. Optimization helps evaluate and discover successful decision strategies which meet your business objective and cater to the conflicting constraints.

It can be used to explore different analytic techniques varying in sophistication, but it is vital to capture the right data for each technique and overlay decision modelling to optimize collections actions.

In early-stage collections, the key objective is to prevent accounts or consumers from rolling forward by being laser-focused on high-risk but collectable accounts and identifying low-risk self-cures for relaxed or no-action.

So in early stages, collections treatment strategy is about optimizing which channel, channel timing and channel messaging should be most suitable to which account to achieve these objectives whilst also knowing who will accept and pay using the auto resolve channel, and who will go outside the channel. Also, who will not accept and roll forward?

In late-stage collections, the focus is more on collections and less on retaining customers. Management of non-performing loans and the identification of consumers that require related restructure / settlement is a key focus in this stage.

You need to design the process around the decisions. Optimization of restructure and settlement offers will include reviewing take up, expected level of default in order to aggregate the impact on the portfolio quality in order to optimize what restructuring or settlement offers should be given to individuals.

Post charge-off, the objective is to maximize recoveries by determining which accounts to work, which to place and which to sell.

When it comes to third-party placement, optimization can be used to determine which accounts to best place to which agency in order to maximize recoveries, based on the performance of each agency on different quality and types of debt placed with them and the commission rates charged.

Optimization can also assist in strategies to determine which accounts are best to be worked internally vs accounts that should be placed or sold keeping in mind the expected value of the account , the probability to pay and the cost to work or place the account.

Tip 8 – What should we look for in a review of our operations?

A comprehensive analysis of policy, insight, people, processes, tools, technology, data, management information and data and analytics and strategy will deliver the path from current state to desired state. One of the best examples of operational negation is when analytics insight is ignored in order to gain operational efficiency — for example, ignoring the risk segmentation of accounts to build a larger dialing campaign! There are many common areas of poor practice because few organizations really have sight of policy, process and execution from data to insights, through workflow into operational execution and performance measurement.

Given the cost of bad debt on the balance sheet, it is still surprising how many organisations do not do a full annual assessment of their C&R capabilities and actual vs. potential performance to plan & forecast.