Reducing financial hardship by establishing contact

A TFI research project by Felix Uhl, Claartje ter Hoeven & Marieke Fransen
Posted on August 01, 2019

"Reducing financial hardship by establishing contact: A personalized and targeted communication intervention to prevent problematic debt" is one of the short-term research projects supported by the Think Forward Initiative.

Many households struggle with problematic debt or are at risk of running into serious financial problems. Some people facing such financial hardship ignore their situation and fail to repay their debt on time. As a consequence, people are charged additional collection costs and fines, elevating their debt burden even more. Felix Uhl (University of Amsterdam & de Gedragsstudio), Claartje ter Hoeven and Marieke Fransen (University of Amsterdam) believe it is time for action. The accumulation of debt is unnecessary and needs to be prevented. In this project, the researchers investigate how personalized and targeted communication could motivate debtors to contact their creditors and repay their debt.

Curious to read about the results? Download the full report below!


Problematic Debt in the Netherlands

Even though the Netherlands is a prosperous country, there are many people who can barely make ends meet. A few figures to indicate the magnitude of the problem:

  • 1 in 10 children grows up in poverty
  • 1.4 million Dutch households have problematic debt or a risk thereof
  • Average debt is approximately 43,300 euros divided among 14 creditors
  • 1/3 of the debt is the actual claim and 2/3 are additional collection costs

It can happen to everyone

Life events like a divorce, (health-related) incapacity for work, or job loss often contribute to the risk of running into financial problems (Schonewille & Crijnen, 2018). During such life events, people often have to deal with multiple challenges, which impedes cognitive function (Mani, Mullainathan, Shafir, & Zhao, 2013). This may result in neglecting problematic financial situations, leading to additional collection costs and fines. This unnecessary accumulation of debt can be prevented when clients proactively contact their creditors and ask for help managing their debt. Early detection and management of relatively small debt can therefore contribute to reducing serious debt problems in the future. More and more primary creditors are taking their responsibility in the growing debt problems and are looking for solutions. In many situations, companies and governments are willing to make arrangements for people who are (temporarily) unable to pay their bills (NRC, 2018).

Getting in contact: Personalization and Targeting

Establishing contact between customers facing financial problems and their creditors in an early stage, i.e., when no extra costs or fines are charged yet, is key to prevent further financial problems (NVVK, 2018). Therefore, this research focuses on how we can activate customers who are unable to pay their bills to contact their creditors to set a payment arrangement. In 2013, the Behavioural Insights Team in the UK tested the effectiveness of mobile phone text messaging as an alternative method of inducing people to pay their outstanding fines. Their results show that text messages are found to significantly increase average payment of fines and that text messages were especially effective when they use personalization (Haynes et. al., 2013). We expect that customers, who are unable to pay, will also be more inclined to contact their creditors to settle a payment arrangement when they receive a personalized message because these are often perceived as warmer and friendlier which may reduce perceived barriers of seeking contact.

Another strategy that may activate customers to take action is message targeting. Targeting or behavioral targeting can be defined as customizing a message on a specific group of customers (based on their previous behavior). In the current context we propose that customers who cannot pay their bill due to a possible life event will be more motivated to set a payment agreement when we explicitly ask them in a text message to do so than when they receive a standard message asking them to pay the outstanding bill or even threatening them with additional costs. We refer to this as a targeted message since the content of the message (i.e., set a payment arrangement) is targeted at the situation these people are in (i.e., unable to pay).

Research Question and Hypotheses

What are the effects of a) personalized and b) targeted mobile phone messages on closing payment arrangements with customers in debt?

Hypothesis1: A personalized text message will elicit more payment arrangements than a non-personalized text message

Hypothesis 2: A targeted text message will elicit more payment arrangements than a non-targeted text message

We conducted two separate studies to answer the research questions and test our hypotheses. Two large Dutch companies participated in this research. The first study on personalization was conducted among customers of a major energy provider with two different labels (A and B). We conducted a second experiment on targeting among customers of a telecom company.

The effect of Personalization

We found that sending a personalized text message reminder resulted in more payment arrangements compared to the non-personalized text message. However, this effect was observed only for label B.

We have two possible explanations for the difference between the labels. One is the type of customer. Label A and B are different brands. One is a real price fighter (Label B) while the other label is a premium brand (Label A). Perhaps more customers at Label B are in a financially difficult situation compared to the customers at label A and these customers might have a greater need for payment arrangements. Another explanation for the difference between the labels could be the fact that the content of the text messages differs on a few small points. The different text messages that were used in this study are as follows:

Text messages from Label B specifically mention the outstanding amount. By mentioning the specific amount, the messages from Label B are more tailored, i.e., adapted to the specific situation of the targeted individual. Additional research should be conducted to further disentangle the effects of personalization and tailoring.

The effect of targeting

At a big telecom company, we have been able to measure the effect of targeted text messages (i.e., pointing out to customers in need that they can make a payment arrangement) versus non-targeted messages. By sending these targeted messages to a group of customers who previously always paid well, the assumption was that within this group there are people who want to but may not be able to pay. These customers are expected to be more motivated to contact their creditors for a payment arrangement when explicitly asked in a text message compared to when they receive a standard payment reminder. We designed two different types of targeted messages and also simplified the original text message. We included this simplified text message for exploratory reasons. Below the four different text messages that we used.

The first targeted text message is a traditional style text message payment reminder including an opportunity to make a payment arrangement. In the second targeted text message reminder we deliberately reversed the order of the message, starting the message with the payment arrangement instead of asking the customer to pay their bill. Both versions of the targeted text messages led to an increase in the number of payment arrangements compared to the original text message reminder. For future research it could be interesting to measure the effect of the targeted messages between different groups of customers.


Our research shows that even small changes in text message reminders can cause more people to close a payment arrangement than the original text message reminders. Although the differences in percentages may seem rather small, in absolute numbers the effects might be quite large. The results of the current studies provide practical tools for companies. Not only helping the customer to prevent additional cost but also the company and the society at large can benefit from preventing people to accumulate debt. For now, it can be concluded that personalizing text message reminders and sending the right text message to the right group of customers might assist in helping people with their financial problems by setting payment arrangements. Our research provides good starting points for further research into the prevention of unnecessary problematic debt.

When people experience financial problems, they do not only have outstanding accounts at telecom and energy companies, but probably also with approximately 14 other creditors like tax authorities, insurance and banking companies (NVVK, 2018). It would be interesting to replicate our study at different companies with different products, in the Netherlands and abroad, to study the effect of personalization and targeting.