Building empathy in pension communications with AI

The transition to the new Dutch pension system is a massive one… and it is also an emotional one. Federica Foroni Lo Faro (BrightPensioen) and Pieter van Giels (IG&H) explain why emotions are key and how AI can help unlock empathy to empower pension participants.

With the transition to the new Dutch pension system, pension funds must ensure participants make well-informed financial decisions, as these will impact the rest of their lives. In the past, this social responsibility has been tackled with studies, facts and figures to help people make choices.

However, alongside fact-based information, emotions have always played – and will continue to play – a pivotal role in the decision-making process.

Understanding the emotional rollercoaster

Happier individuals make better, more rational decisions. Better decisions lead to increased happiness. Stressed individuals tend to have knee-jerk reactions, making riskier decisions with a more short-term focus. Alone in 2022, at least 20% of participants aged around 55 were unhappy about the choices they made.

A pension is an emotional product. In fact, most citizens seek out information about their pension during negatively charged emotional situations.

People will gather information about their pension when a big change occurs in their lives such as a partner passing away (43%) getting separated or divorced (36%) or if they become unable to work (33%). In contrast, a meagre 13% seek out information during a more positive experience such as starting at a new job. Especially during emotional moments like the loss of a loved one, empathetic customer interaction is essential.

Pension funds must communicate to their participants about the transition to the new pension system. This will likely lead to an increase in customer service requests. To provide the best possible customer journeys for participants, the pension industry can make smart use of AI capabilities.

Real emotions meet artificial intelligence

Chatbots powered by artificial intelligence can identify and respond to emotions. Technology can initiate empathetic dialogues 24/7 with hyper-personalization to help deal with customer service requests. It can understand participants' demographics, tech-savviness, knowledge level and cultural differences.

The recent developments in Large Language Models show that these models can explain information to different audiences. For instance, such a solution can offer guidance in a participant’s mother tongue using the appropriate vocabulary. However, it can also help to determine whether making a decision now is best or whether to postpone it, such as scheduling an appointment for the following week.

To make use of these technologies, funds can leverage the capabilities available in the market. For example, explaining a pension to someone does not require building a Large Language Model from scratch. Instead, you can use the likes of ChatGPT or BARD. These self-learning solutions don’t just learn from conversations with participants, they can also learn from an agent how to best approach difficult topics.

Furthermore, the technology can be used to approach a participant later via different channels, such as via an app or email, or provide the suggested responses to an agent.

AI can also play an important role in making the customer journey more dynamic. Rather than automation with pre-determined flows, emotions can be used as data points. These can trigger what is known as the ‘next best action’.

For example, if participants are happy, this presents an ideal moment for communication to be intensified. In contrast, if a participant is angry and distressed, this can be a dynamic trigger for them to get in touch with a human agent. AI can be used to gauge the emotions of participants and ultimately empower customer service agents by ensuring they are prepared for upcoming conversations and can get in touch with those who need it most.

By considering emotions, the pension industry will be able to provide better decision-making support and fulfill their duty of care to their participants.