Newcastle Building Society

Using data to improve customer retention

Newcastle Building Society has been helping customers in the North East of England to buy homes, save for the future and protect what’s important to them for over 150 years.

In today’s highly competitive market it is seeking to cement long-term relationships with its customers as well as attract new ones. One of the ways it aims to do this is by gaining deeper insights into its customers’ current and future needs using the data it holds.

Whilst the building society has some colleagues with data skills, for example in its credit risk modelling function, we identified an opportunity to work closely with them to build on these skills in some other key parts of the organisation.

We worked with the product development team, who previously had little coding experience, helping them develop an innovative data model to identify, and categorise, mortgage customers approaching the end of their fixed-rate period.

The insights which the model provides will be used to help the society retain those customers.

The company

Newcastle Building Society is a mutual, which means that it is owned by its customers.

It is the eighth-largest building society in the UK and the largest in the North East, with 31 branches stretching from North Yorkshire to Northumberland and across to Cumbria.

As part of its commitment to being a great place to work, the society strives to develop its people and was recently given a platinum Investors in People award.

The goal

Developing long lasting relationships is core to the building society’s strategy and in a very competitive mortgage market they wanted to understand the customer characteristics that deliver a long-lasting relationship and the commercial return from those relationships.

Having expanded in recent years, the building society now has around 36,000 mortgage customers, many of whom are on two-year or five-year fixed rates.

As these people come to the end of their fixed-rate term, the society wants to be able to offer them attractive mortgage products that meet their needs and therefore encourage them to stay rather than consider moving to an alternative provider.

Whilst retention levels were already high, the society felt there were opportunities to improve customer retention even further.

The society came to us with the aim of achieving two goals:

• To highlight the customer characteristics that encourage a long-lasting relationship and apply this information in support of new mortgage customer acquisition.
• To use both our own data and market data to optimise pricing offered at the end of a customer’s product term.

The results

Our team worked with the product development team to develop a data model that gives an overarching view of the society’s mortgage customers. During the project they:

• learnt to use the statistical programming language R
• cleaned and analysed six years of mortgage data
• developed a model to improve understanding of mortgage customers
• integrated both internal and external data into the model
• learnt how the same approach could be applied to other projects

 The model has already proved a useful tool in board-level decision-making.

Developing data specialists in-house

Newcastle Building Society already had a clear view of the data that it held but was not able to make full use of it because of a lack of effective tools and because its people initially lacked some of the data skills required.

As a purpose-led organisation, it wanted to develop those skills in-house, not just to give a better outcome for customers but also to give its staff the opportunity to meet their potential and to add to the regional skills base.

CEO Andrew Haigh explains: “To sit down and talk to the National Innovation Centre for Data about how they could help nurture and grow talent from within was a really compelling argument for us and a really strong alignment with our purpose.

"We knew there was potential and we knew we could be a more efficient and more effective business if we understood our data better. But the fantastic thing is that, rather than have to find new resources or outsource that process, we could get properly into skills transfer and develop the people that we already had within our existing team."

Working alongside data experts

To make that vision a reality, Product and Planning Specialist Jack Pattullo and Data Analyst Graduate Will Fisher spent a day each week with the building society, developing the model using the sophisticated statistical programming language R.

While Jack already had an in-depth knowledge of the society’s data, analysis had previously been carried out using products such as Excel and Microsoft Access, which provided limited insights. And while Will had recently graduated in economics, he lacked coding skills.

The model they created aims to help the building society to understand how to identify the right mortgage products at the right price so that as customers approach the end of their term, they receive an offer suited to their needs and stay with the society.

Jack explains: “This is the first time we’ve used the R statistical software. It’s open source so we can do literally anything we want with it, which is why NICD recommended that as an approach to take. With the software being free it also has a very large global community of users to offer support and innovative ideas on how to look at things.

“We worked alongside NICD’s data scientists at their site, with Will and I developing the model but with the NICD team explaining what each bit of code meant, what each bit of code was doing and when to apply certain techniques.

"More generally we also learnt just how to approach a data project. We’ve never really done anything of this size or complexity before so we learnt how we can take that way of thinking back to the business and apply it to other projects.”

Creating a belief within the team

The model has already had a positive effect on customer retention and the success has led to Jack becoming a Fellow of the National Innovation Centre for Data. While Will was planning to study data science part time, we helped him find a more challenging course that will meet his needs better and he has enrolled on a course at Newcastle University. 

Steve Caughey, Director of NICD, says that organisations often have a huge amount of data but have no idea how to get value from it. He explains: “People needing help with their data will frequently turn to a consultancy. That may solve an immediate problem but it’s not a long-term solution. Data is at the heart of digital innovation so it’s important to be able to explore your own data and identify operational improvements and new opportunities. We make sure that our clients get the skills they need to get ongoing insight from their data.”

Andrew Haigh adds: “The effect of this project is in part about better use of data and new skills being developed within the society, but the exciting thing is to see the enthusiasm it has created and the realisation within the team about what’s possible and how it has motivated them. There’s a real spark of passion about what can be done with data and a belief that within the team they do have the skills and capabilities we need – they just needed the support.

“It still feels like the beginning of something rather than the end of something. There’s a lot of potential in the relationship with NICD and the journey that the team is now on. And it’s so much more rewarding and so much more valuable to the business to be able to do that from within rather than find some other company, somewhere in the world to come and complete a task for us and then go away again. It is the beginning of a process and the beginning of a journey that will continue to develop and add further value in the years ahead.”