Payment Behavior in the Default Phase - AFM and in3 Research — AFM — NL

OCCASIONAL PAPER

In brief
The AFM and BNPL provider in3 investigated the effect of adjustments to the customer journey on the payment behavior of 7,425 customers in the default phase. The research showed that extending the period before handover to the collection agency by 14 days has no effect on the proportion of customers with a payment default two months later. However, later handover proved beneficial: the proportion of customers dealing with a collection process and associated costs decreased by nearly one-third due to the adjustments. An additional reminder was found to be of great importance for this. The AFM encourages providers to conduct research into adjustments to the customer journey to reduce risks for consumers.

Authors: Job Krijnen, Annelot Wismans, Ilva van der Gragt Publication date: May 2025

Payment Behavior in the Default Phase
Research in collaboration with in3

OCCASIONAL PAPER

Table of Contents

Introduction 3
1.1 Promoting responsible use through adjustments to the BNPL customer journey 3
Research with in3 into payment behavior in the default phase 5
2.1 Later handover to the collection agency 6
2.2 ...including an additional reminder 6
2.3 ...including information on the consequences of paying late 7
Method 8
3.1 Procedure 8
3.2 Sample 10
3.3 Analyses 10
Results 11
4.1 Later handover does not have the desired effect on payment behavior but is still beneficial for customers 11
4.2 The additional reminder helps customers pay earlier 13
4.3 The additional reminder is more effective when it contains information about the consequences of paying late 14
Discussion 15
5.1 Adjustments in the default phase help in3 customers avoid collection costs 15
5.2 Additional attention needed for customers with structural financial problems 15
5.3 BNPL providers can research the effect of adjustments to the customer journey 16

OCCASIONAL PAPER
Payment Behavior in the Default Phase | 2025 3

1. Introduction

1 AFM, 2024; AFM, 2022
2 AFM, 2024
3 Volkskrant, 2024
4 BNPL Code of Conduct, 2024

Buy Now, Pay Later ('BNPL') has grown in recent years to become a widely used service among Dutch consumers.¹ BNPL makes it possible to postpone the payment for a purchase until 14 or 30 days after delivery or to pay in three installments.

The use of BNPL carries risks. For example, earlier research by the AFM showed that approximately 1 in 8 BNPL users stated that they had missed at least one payment term in the past 12 months.² Consumers who do not pay on time may face non-performance costs (such as reminder costs and collection costs). Debt counselors also signal an increase in the number of clients with problematic debts and link this to the increased use of BNPL.³ Such signals have led to media and political attention regarding the risks of BNPL. Plans to offer BNPL in physical stores have further stoked the social debate.

Due to new European regulations, BNPL providers will come under the supervision of the AFM in November 2026 and will, among other things, have to comply with rules regarding information provision, creditworthiness assessment, and default management. In the meantime, four major BNPL providers have drawn up a code of conduct to better protect their customers.⁴

The future regulations and the current code of conduct must contribute to increased consumer protection. In anticipation of the new regulations, providers can promote responsible BNPL use by making adjustments to the customer journey.

1.1 Promoting responsible use through adjustments to the BNPL customer journey

The BNPL customer journey can be roughly divided into four phases: the orientation phase, the purchase phase, the payment and return phase, and the default phase (see Figure 1). During the orientation phase, consumers explore products in an (online) shop, weighing the possibility of paying later in their choices. Around the time of purchase, consumers then choose to pay with BNPL. BNPL providers use acceptance criteria to test whether consumers are allowed to use their services. In the payment and return phase, customers who do not return their goods must pay within a term of 14 or 30 days, or within the two remaining 30-day terms. The default phase follows if customers do not pay in full within the set term(s). In this phase, there is a payment default, and if payment is not made, customers are handed over to a collection agency.

Figure 1. Phases in the BNPL customer journey Orientation Phase | Purchase Phase | Payment and Return Phase | Default Phase

OCCASIONAL PAPER
Payment Behavior in the Default Phase | 2025 4

Adjustments can be implemented in all phases of the customer journey to promote responsible use of BNPL by consumers. Earlier, the AFM, in collaboration with BNPL provider Riverty, investigated the effect of SMS reminders in the payment and return phase.⁵ That research showed that customers who received an SMS reminder paid earlier, resulting in fewer instances of reminder costs. In this report, we focus on the default phase. Based on new research, this time in collaboration with BNPL provider in3, we test the effect of adjustments to the customer journey on the payment behavior of customers who have missed one or more payment terms.

5 AFM, 2024

OCCASIONAL PAPER
Payment Behavior in the Default Phase | 2025 5

2. Research with in3 into payment behavior in the default phase

6 AFM, 2024
7 Meier & Sprenger, 2010; Dawson & Henley, 2012; Grubb, 2015

In this report, we describe research conducted in collaboration with in3, a Dutch provider of paying in three equal installments. Customers using in3 pay the first installment directly when placing the order, the second installment after 30 days, and the third installment after 60 days. See Figure 2 on page 9 for a description of the regular process. in3 is particularly active in product categories such as home furnishings, white goods, and transport. in3 customers place an average order of approximately €500. This is relatively high compared to the average BNPL transaction of approximately €100.⁶

We focus in this research on customers who have missed one or two payment terms. in3 does not charge reminder costs to these customers, but after a certain period, customers in default are handed over to a collection agency that charges collection costs. We test three adjustments to the process that customers go through after the expiration of the third term when they have (partially) not met their payment obligations. The goal of these adjustments is for more customers to pay before they face collection costs. Missing payments can, after all, be the precursor to larger financial problems. Consumers who fall into default may also experience stress and uncertainty, which may negatively influence other financial choices and overall well-being.

It is important to understand why consumers in default do not pay when designing promising adjustments. We distinguish three possible causes, but note that this list should not be considered exhaustive. First, it is possible that consumers make purchases with BNPL that they cannot afford due to a temporary or structural shortage of financial resources. This situation can arise because people often have limited attention to long-term consequences or are optimistic about their current and future financial situation.⁷ Additionally, BNPL providers do not have complete insight into the financial position of consumers when accepting them for an order. For example, they have no visibility into existing credits and any defaults on them. Furthermore, a consumer's financial position can change unexpectedly after placing an order. Second, consumers who do have sufficient financial resources may still miss a payment term, for example, because they forgot payments, did not receive or notice payment reminders, or lack an overview of payment obligations. Third, it is possible that consumers are insufficiently aware of the consequences of paying late, such as additional costs. As a result, they may feel less urgency to take action and wait too long to make the payment.

Based on these possible causes, we investigate the effect of three adjustments to the process that in3 customers go through in the default phase on the payment behavior of these customers: later handover to a collection agency, later handover combined with an additional reminder, and later handover combined with an additional reminder containing information about the consequences of paying late.

With the proposed adjustments to the default process, we do not focus on customers whose payment default stems from structural financial problems. The chance that they remain in default despite the adjustments is high, because they simply do not have the means to meet their payment obligations. In the Discussion, we go deeper into the additional measures with which customers with structural financial problems can be helped.

2.1 Later handover to the collection agency

In in3's default process, as it existed before the start of the research, customers who were still in default three weeks after the last payment term were handed over to a collection agency. At that point, the collection agency takes over customer contact and charges the statutory collection costs of 15% of the outstanding balance, with a minimum of €40. The optimal timing for handover to a collection agency requires careful consideration. On the one hand, a later handover can provide customers with more financial breathing room, for example, if they receive their salary in the meantime. A later handover may also give customers more time to create an overview, making them less likely to forget payments. On the other hand, later handover may reduce the chance of payment because there are fewer incentives for customers to pay. For example, earlier research shows that charging costs for late payment of outstanding credit card accounts and owed taxes influences repayment.⁸ Other research shows that postponing the payment deadline for short-term loans can lead to deferral behavior.⁹

We test in this research the effect of a two-week later handover to the collection agency on the payment behavior of customers in default. In the existing process, customers were handed over to the collection agency three weeks after missing the last term; in the adjusted process, this period lasts five weeks. We expect that later handover to the collection agency will have a positive outcome in this case, because the benefits for customers with a temporary financial shortfall outweigh the potential negative influence on the payment behavior of customers who do have sufficient financial resources.

8 Agarwal et al., 2013; Skov, 2023
9 Carter et al., 2022
10 DellaVigna & Linos, 2022
11 Campbell et al., 2022

The effect of later handover on the payment behavior of customers in default may depend on the communication customers receive during this period. Therefore, we also investigate the effect of later handover in combination with the effect of two additional adjustments to the communication.

2.2 ...including an additional reminder

Later handover to the collection agency can lead to customers unnecessarily delaying payment or forgetting their payment default during this period. An additional payment reminder during this period may offer a solution. Earlier research is known to show that reminders can have a strong influence on (financial) behavior.¹⁰ Sending a reminder to customers with 30 days of credit card payment default, for example, helped improve payment behavior.¹¹

Therefore, we test the effect of an additional payment reminder during the period of later handover (adjustment 1) on the payment behavior of customers. We expect that more customers will meet their payment obligations when they receive an additional reminder during the period of later handover than when they do not receive one.

OCCASIONAL PAPER
Payment Behavior in the Default Phase | 2025 7

2.3 ...including information on the consequences of paying late

The effectiveness of a payment reminder may depend on the content of the reminder. For example, it is possible that customers pay late because they are not aware of the financial consequences of paying late, or because they read this earlier but have since forgotten. Research on mentioning consequences in reminders shows mixed results. While some studies showed that emphasizing a fine for late payment in a reminder letter increased the percentage of on-time payments,¹² earlier research by the AFM in collaboration with BNPL provider Riverty showed that the effectiveness of an SMS reminder did not depend on the content of the message.¹³

Therefore, we test the effect of the content of the additional payment reminder: next to a regular reminder, we investigate a version that also states that the customer will be handed over to a collection agency if payment is not made, and that collection costs of at least €40 excluding VAT will be charged. Additionally, it is indicated that the customer will not have the possibility to make new payments with in3 for two years if payment is not made. With this adjustment, we try to motivate customers to make the payment by giving them clarity about the consequences of not paying during the period of later handover (adjustment 1). We expect that more customers will meet their payment obligations when they receive a reminder with information about the consequences during the period of later handover, compared to when they receive a reminder where these consequences are not mentioned (adjustment 2).

12 Dusek et al., 2022
13 AFM, 2024

OCCASIONAL PAPER
Payment Behavior in the Default Phase | 2025 8

3. Method

14 According to the intended (and predetermined) procedure, communication and handover were to take place at three moments: day 67, 82, and 97. However, in the execution of the experiment, some unplanned minor deviations in the procedure occurred. By adjusting the analysis days, we are still able to investigate the effect of the three adjustments on payment behavior based on the collected data. See the statistical appendix for more information. 15 By this, we mean the legally required final reminder that creditors must send by email or letter to customers before they are handed over to a collection agency based on the Act on Collection Costs ('WIK'). This reminder must indicate a payment term of at least 14 days after receipt and announce that collection costs will be calculated if the debt is not paid within this term.

3.1 Procedure

We focused in this research on customers in the default phase. in3 customers who had not paid one or two terms on day 64 – four days after the expiration of the last payment term – were randomly assigned to one of the four research groups. This experimental setup makes it possible to draw conclusions about the effect of the different adjustments to the customer journey. Figure 2 shows the process that in3 customers went through from day 64 after the purchase for each research group.¹⁴

In the control group (group 0 in the figure), customers went through the regular default process. One week after missing the last term, they received an email advance notice of the handover to the collection agency (the 'WIK notice').¹⁵ This email stated that collection costs would be charged if customers did not pay the outstanding amount within 14 days. Two weeks later, customers who were still in default were handed over to the collection agency. At that point, the collection agency takes over customer contact and charges collection costs. The collection agency approaches customers in default through various channels, such as telephone, email, and post.

In the other three groups, customers were handed over to the collection agency two weeks later. Customers in the group later handover without additional reminder (group 1 in the figure) received no additional communication during these two weeks. Customers in the group later handover with additional reminder (group 2 in the figure) received an additional payment reminder on day 67 – seven days after the expiration of the last payment term. Customers in the group later handover with additional reminder incl. information on consequences (group 3 in the figure) also received an additional payment reminder on day 67, containing additional information about the ultimate consequences of a payment default (namely collection costs and blocking of future in3 use). In all three of these groups, customers received the aforementioned WIK notice by email two weeks later – 14 days before handover to the collection agency.

OCCASIONAL PAPER
Payment Behavior in the Default Phase | 2025 9

Figure 2. The regular in3 customer process up to day 64 (1) and the process during the research for customers with a payment default in each of the four groups from day 64 (2).

Advance notice of handover (WIK)
Handover to collection agency
Analysis day
Additional reminder
Regular reminder
Information on consequences of default

Legend
0 25 32 30 55 6260
€ € €
0
1
2
3
67 81
67 81 95
67 81 95
67 81 95

1 Customers follow the regular customer process up to day 64:
1 2 3
1
0
2
3

Later handover without additional reminder
Control group
Later handover with additional reminder
Later handover with additional reminder incl. information on consequences

Explanation: (1) With the regular reminders, customers received an email, SMS message, and app notification. With the additional reminders, customers received only an email. (2) In the second part of the customer process, the days on which communication and handover to the collection agency took place may deviate by up to 2 days from the days shown for some customers. See the statistical appendix for more details on deviations in the procedure.

2 After day 64, customers in default are randomly assigned to the control group (where the regular process continues), or to one of the experimental groups: expe

OCCASIONAL PAPER
Payment Behavior in the Default Phase | 2025 10

3.2 Sample

The sample consisted of 7,425 unique customers distributed across four groups (control group: 1,863, later handover without additional reminder: 1,825, later handover with additional reminder: 1,859, later handover with additional reminder incl. information on consequences: 1,878). These are customers who were in default on day 64 during the period from July 5, 2024, to October 10, 2024, and met the inclusion criteria.¹⁶ Customers in the sample were on average 36 years old, had an average order amount of approximately €500, and placed an average of 1.6 previous orders with in3.¹⁷

3.3 Analyses

To draw well-founded conclusions about the effect of a later handover to the collection agency and the sending of payment reminders on the payment behavior of customers, it is important to formulate hypotheses and perform statistical tests. Before the start of the experiment, the AFM and in3 jointly established and documented the hypotheses, an analysis plan, and the criteria for inclusion in the sample. These and all additional analyses not presented in the report are described in detail in the statistical appendix.

16 This includes, among other things, customers who had no more than one outstanding payment with in3 during the research and customers who were not granted a payment extension. The inclusion criteria for the sample are explained in the statistical appendix.
17 In the statistical appendix, we describe additional analyses showing that customers in the four research groups did not significantly differ from each other on these characteristics.

OCCASIONAL PAPER
Payment Behavior in the Default Phase | 2025 11

4. Results

In this chapter, we discuss the main outcomes of the research and some additional analyses. To test the hypotheses, we look at the proportion of customers in the research who are (still) in default at two specific moments (the analysis days in Figure 2). To test the effect of later handover to the collection agency on payment behavior, we look at the proportion of customers in default on day 119. This is slightly less than two months after customers were assigned to one of the four research groups and one day before the loan is marked as 'possibly uncollectible' by in3. To test the effect of the (content of the) additional reminder, we look at the proportion of customers in default in the three groups with later handover on day 78. This is two weeks after the start of the experiment and before customers in these groups receive the advance notice of handover to the collection agency.

In the additional analyses, we also look at the proportion of customers handed over to the collection agency, the timing of payment, and the proportion of customers in default on day 94. The results are described in more detail in the statistical appendix.

4.1 Later handover does not have the desired effect on payment behavior but is still beneficial for customers

Of the customers in default on the start date of the research (day 64), slightly less than a quarter (23.8%) in the control group were still in default at the end of the research (day 119). This proportion does not differ significantly between the control group and each of the other three groups where customers were handed over to the collection agency two weeks later (see Figures 3 and 4). Thus, later handover in this research does not lead to a reduction in the proportion of customers with a payment default on day 119.

Figure 4 shows the payment behavior of customers on each day during the research. Although later handover has no influence on the proportion of customers in default on day 119, this figure does show differences in the moments when customers pay. Additional analyses show that the average time to payment in the control group is significantly shorter than in the later handover group and the later handover with additional reminder group. Between the control group and the later handover with additional reminder group...

(Note: The provided source text ends abruptly at this point.)