NN

NN

Artificial Intelligence bringing NN’s contact centre to the next level

The Low Down

Salesforce Products

Region

Capabilities

Introduction

NN is an insurance company with a strong presence in more than 18 countries, especially in Europe. They offer their services close to 1,5 million customers each day, mostly thanks to a wide network of brokers and partnerships with several banks. Their customer service centre receives over 150,000 calls a year from employees of its banking partners as well as directly from end-customers.

The Challenge

NN aims for clarity, care and commitment towards their customers in everything they do. Their challenge is to handle a high volume of calls in a timely fashion, with the right solution at hand. This has a high cost implication for the business, so NN wanted to optimise their service centre by analysing and understanding the common topics for which they received calls.

Analysing text data from customer’s and banking partner’s calls is much more complex than just analysing numbers. Firstly, Dutch text data is particularly challenging because of the Dutch language complexity. Secondly, the outcome of the models are not simply numbers but human language with a lot of industry specific words. These require further interpretation by a business expert to deduce meaningful and actionable insights.

The Solutions

Set-up of an Artificial Intelligence model

NN has the ambition to be an innovative insurance company. They were intrigued by the possibilities of Artificial Intelligence to make their way of working more efficient by avoiding calls and increasing customer satisfaction. According to NN a happy customer is a customer who does not need to call their customer service centre at all.

Before starting with the implementation of the AI model TellMi, data was split over 2 different product branches: Life (for example outstanding balance insurances) and Non-life (for example home & family insurances).

For each product branch, a topic model was built. This is a machine learning model that analyses text data and summarises this information in different non-predefined themes. When call logs are fed to the topic model, it determines which topics appear most during the calls.

Topic detection thanks to TellMi

Thanks to a close collaboration with NN’s business users, 4C identified 14 high level topics for which customers and bank tellers call. 4C subsequently divided the 2 most important high level topics into 3 sub topic models to gain a detailed understanding of the call reason.

If all Dutch phone calls from the last 3 years would have been analysed by a person of flesh and blood it would have taken 2 years to complete the job. Thanks to the AI model TellMi of 4C, NN was able to achieve the same results in just 3 months.

Reduction of customer calls

It is important to NN that their Customer Service department can focus on those issues that occur most frequently.

With insights delivered by 4C, NN now has the advantage of tackling customer’s issues proactively for example by clarifying the terms and conditions of their products, establishing a specific customer platform for self-service or by sending out letters/emails that anticipate to customer questions.

4C estimated that NN will be able to potentially eliminate 35 to 40% of calls from customers and bank branches. This in turn corresponds to potential cost savings between €300K and €500K for NN’s contact centre.

“We found the 4C team to be passionate and flexible in their approach. With their clear, structured insights into automated data analysis, we will be able to reduce the number of unnecessary calls to our customer service centre. As a result of this, our customers will be the biggest winners of all!”
Gunter Van Meensel, Customer Experience Office at NN

The Results

18

Actionable (sub)topics

35% - 50%

Potential call avoidance ratio

€300K-€500K

Potential savings

Ready to become a customer company?

Let's Talk