Predictive model for mortgages
It’s important to never stop learning because technology is moving so rapidly
When is it the right time to approach a potential new mortgage customer with an offer from ING? My latest assignment, which I’m doing in the Lead Generation squad within Mortgages, revolves around answering precisely that question. I’m investigating how we can predict when a customer’s mortgage with another lender is nearing the end of the fixed-rate period – because that’s the perfect moment for us to help the customer by making them a new offer.
My first step was to research the Dutch mortgage market. Although the starters market has declined, a growing number of people are switching. The extremely low interest rate means it can make financial sense for customers to shop around. I’ve been working on building the predictive model for the past few weeks, and it will soon be time to decide whether it adds sufficient value – in other words, whether it can generate enough high-quality leads. If so, the next step is to test whether our business case will actually work. Will the email that we send to customers present them with the right offer?
This is a full-time assignment for me, and one of my senior colleagues is working on it for 50% of his time. It’s an excellent set-up, because I have lots of autonomy but my colleague is also involved if I have any questions or just want to check that I’m still on the right track. Besides that, he’s helping me to think about whether this model could be beneficial for other colleagues too… such as in the Pricing team, or to predict future behaviour of ING customers, and identifying the best time to make them a new offer. I’ll be presenting a demo of our project to the various teams within Mortgages in the near future.
This assignment ticks all the boxes for me right now: I’m part of a squad, collaborating with a senior colleague and also learning a lot about mortgages, which had been a bit of a grey area for me up until now. As it happens, my boyfriend and I are actually looking for a mortgage as we speak, so it comes in handy to know more about the topic!
My data science course has now got underway. For the coming year, I’ll be spending one full day a fortnight attending lessons along with eight others, including a couple of people from the ING IT Class. We’ve already tackled Advanced Python 1 and Docker, which is virtualisation software that’s used to deliver software in packages called ‘containers’. It’s not really relevant for my assignment, but the Python programming language definitely is! In fact, I’ve run Python training sessions myself in the past, although they were focused on teaching the Python basics to non-tech colleagues. Needless to say, my data science course goes a lot deeper. It’s really great to be able to spend a day expanding my knowledge every two weeks. I also see it as part of my work, because it’s important to never stop learning. After all, technology is moving so rapidly. That’s why ING encourages us to go on courses and keep ourselves up to date on new technologies and methods.
Within our own department, we develop new ideas during a monthly hackathon. For December, I’m going to suggest that we work together on a project for a good cause – in the spirit of Christmas, and because there must be plenty of charities that could benefit from our help. I’m currently exploring which charity it could be, and then I’ll check with our team leads whether they think it’s a good idea – although I can’t imagine that they’ll say ‘no’…!
On sunny days, Nikki heads out on her road bike whenever she can and she has recently taken up yoga again. She also spends hours on her computer finding out more about machine learning, including via Kaggle.com, a data science platform where companies post challenges for data scientists from around the world to tackle. For a moment of relaxation, she enjoys reading. Most recently she read Factfulness and Pachinko. “I can highly recommend both of them!”