Approach

While the large customer base represented a sizeable asset, it needed to be initially reduced into smaller sub groups before further refinement was applied. While not necessarily ‘customer first’, we did this by covering all of the levers the bank had available to grow home loan balances. This included acquiring customers that don’t have a home loan (usually First Home Buyers), acquiring customers that have a home loan elsewhere, retaining existing home loan customers, growing the balance of existing home loans through top ups or refinancing and cross selling an investment home loan. This resulted in the creation of four streams, each contributing to the core goal of increased home loan balances, but targeting different segments of the customer base with different offers.

1. Enhancing the Pre-Approval Experience: Acquiring a home loan is a protracted process that includes securing pre-approval, finding a suitable property, and aligning the purchase with bank valuations. This journey, spanning months or even years and often involving multiple banks, offers ample chances for customers to abandon the home loan application. However, even taking this into consideration, analysis identified 85% of pre-approved customers failed to convert to a fully funded home loan within the six-month approval period, despite going to the effort of providing their financial details and securing pre-approval. A crucial revelation as to why was the absence of any consistent customer communication throughout the six months, including nothing at the approval's expiration. After mapping out the home buying journey, a range of communications were triggered at different points of the pre approval window. These included guidance on what to expect throughout the process, home buying tips and tools and the option to easily extend approval. Importantly, branch staff were alerted at key points (particularly the expiry event) so that they could personally contact the customer.

2. Timely Life Event Triggers: CoreLogic property listing and sold data was tested to understand the impact of a property listing or sale event on home loan retention and acquisition. All existing home loans were monitored at address level while addresses of non home loan accounts (transactions, credit cards etc) were ring fenced separately to identify customers that may become open to switching their home loan from a competitor (renters were excluded by applying an implied renter vs occupier flag). Listing and sale events triggered proactive calls to home loan customers from the Home Loans save team and automated communications to both existing and potential home loan customers.

3. Utilizing First-Party Data in Propensity Models: Sub segments of customers without an existing home loan were identified as potential first home buyers or existing owners with a higher likelihood to refinance via the creation of custom propensity models using internal data. Existing home loan customers were separately scored against two additional models to identify their propensity to top up their home loan (for holidays, cars etc) or take out an investment home loan. All 4 models were proven to have up to 7 times lift over random targeting at the top deciles.

4. Property Investment Gamification: While the investment home loan propensity model identified customers with a higher likelihood to take out an investment home loan, research indicated that investing in property was both attractive and daunting in equal measure. While communications helped educate them, ideally we wanted them to somehow experience the process and get a feel for the types of decisions involved. This led to the creation of the world's first property investment simulator, which combined gamification mechanics and a technical mash-up using real data to deliver a real life experience.

Results

All four streams proved successful and were subsequently continued. Specifically:

• With an investment of $71,611 the pre-approval program yielded $2.524 million in incremental profit in its first rollout, representing an ROI of 3,526%. This success was recognized with two Effie Awards for marketing effectiveness

• While highly profitable for home loan retention, the property listings trigger campaign required further refinements to the refinancer component to better exclude renters and identify potential customers more open to switching. Further work was also required to understand whether earlier intervention was needed, even before the property was listed for sale. This later led to the creation of a ‘propensity to list’ trigger

• All propensity model campaigns delivered lift over control cells and positive ROI, especially the First Home Buyers and Home Loan Top-Up models.

• Combined with the investment home loan propensity model, Investorville paid for itself within 12 weeks, achieving an ROI of 221%. It was acknowledged as the Transformational Digital Business Platform of the year, attracting over 80,000 visitors, 20,000 registered users, and over 600 users taking out Investment Home Loans with the bank.