I’ve only been here two years, but data is familiar territory for me. I used to work for a company that offers an insight into consumer and market behaviour using data analyses. After that, I worked for a business services provider that assigned me to financial institutions, where I worked on some great projects. One project I worked on was a fraud model for a health insurance company. Essentially, what you do is strip out the old structure of rules for analyses and risk estimates and rebuild it. Once you’ve finished the first version, you start testing. And you use the results to hone the model. But as an external member of staff, I’d usually left by that point, whereas what’s really great is if a project’s your baby, you get to see it grow.

There were great opportunities at Nationale-Nederlanden. Here, I can build, programme, test and evaluate to my heart’s content. One of the things I’m currently working on is the concept of the Recommendation Engine. Thanks to a machine like this, if you buy a book from the online store Bol.com, it can show you what other books people who bought the same book as you are interested in. Using data analysis, we essentially translate current consumer behaviour, such as buying products and reading articles, into future consumer behaviour. This approach is also of interest to insurers. For example, anybody with a home contents policy may well need an additional policy in the future.

As a team, we’re very results-driven. If you can increase the percentage of clicks on a banner from one to two per cent by isolating customer groups interested in the insurance policy offered, that doesn’t seem that significant. But translate that growth into absolute numbers on a daily, monthly and annual basis, and then measure how many customers take out that insurance, the results are considerable. Generating significantly better results is a challenge for me. And there’s another thing to bear in mind. We approach customers via all kinds of channels, but we don’t want to send our customers unnecessary information. The more targeted the approach, the more personal and relevant the messages for the customer. This is how Nationale-Nederlanden can get itself on the map as a reliable financial partner that knows its customers well, instead of as a seller that bombards the customer with offers in the hope that they buy something.

Nationale-Nederlanden has taken great strides in the field of data in the past two years. When I started here, we received a monthly, and in some cases, even an annual update of our customer data. We now get a weekly update. On our new platform, we’re even heading towards real-time for some sources. This means that we have much more and more detailed information. Which customers have joined and why? Which customers have left and why? Are they pleased with us? Or are they annoyed with us? We extrapolate all this data to the future.

The realisation that we can do great things with data is well and truly established here. Of course, it’s difficult every now and then to make sure that you’ve come up with solutions that actually work and to encourage people to use data insights and not to stick to the old methods. That’s logical. But we’re an open and flat organisation. It’s easy to talk to colleagues and to kindle their enthusiasm for data and analysis. At Nationale-Nederlanden, we’ve got the scope and the mandate for our solutions. All facilities, such as software and hardware, are of a good quality. In short, I enjoy working here.

Vincent van de Ven
Data Scientist Nationale-Nederlanden Customer & Commerce