One-size fits all
When customers log into their Chase account, they can see their card and account details, auto loans, credit score, etc and a tile for home loans.
The tile is a generic home icon with content which saw very less CTR compared to other tiles.
40%
Only 40% of customers who have a mortgage with Chase ever visit and engage with ChaseMyHome
2.06%
The current click through rate (CTR) is much lower for Chase MyHome compared to other tiles
5%
Out of 2M customers who have transacted (got a loan) with us, only 5% engaged with Chase MyHome 6 months prior to transcation
How does personalization make the customer feel?
In other research, we saw that when we don't target the right customer it makes them feel alien to a bank they trust. While campaigns that are tailored to customers make them feel special.
“I got an email yesterday about lowering my payment to 5% but I got my rate during the pandemic of 3%. So I don't know what CMS you guys use, but you might want to be more targeted"
“I got a video from Chase that addressed me by my name which made me inclined to watch it. It makes you feel special, candid… My mortgage company took the time to make me a little video. In a world where there's no customer service left, it felt nice to see that”
1-Day Discovery Workshop
The design team led a workshop bringing together product, tech and marketing stakeholders. The business goal was to increase engagement into Chase My Home (CTR) and increase applications (Lead initiate rate)
Business Goal
Problem Discovery
We did assumption mapping exercise for desirability (do people want it/need it), feasibility (can we do it) and viability (will it result in long-term profit)
Use Cases
We narrowed down a few use cases to start with so the backend team can plan parameters for recommendation model
Assumption Mapping
Are we solving a real customer problem, and will people actually want what we’re offering? Can we pull this off with the data, tech, and resources we have right now? Will this move the needle for our business, creating real value for customers and for us?
Currently, generic
But what if John doesn't want to refinance?
What if John wants to get cash and HELOC is better? Or pay off mortgage sooner so he can retire? Or buy a second home because he wants an investment property?
If we only market a rate & term refinance to this customer based on data 30 days ago, they might be high propensity for refinance but still not click on it due to different goals. To be timely and goal-oriented, we proposed a dynamic approach to personalization.
Proposal
What if we tracked recent data and behavioral indicators that indicate intent and show dynamic tiles to understand customers' needs rather than only selling one product like refinance?
Need based personalization
This way, we categorize personalization by needs not products. Our messaging and design is based on specific situations and not just what data we have about the customer.
Creating Use Cases
The proposal became our north star, and for now based on customer segments, existing and new data features in the SONAR model we created 6 use cases for MVP.
For the top decile customers (ready to apply), we had most data on as they had a mortgage and were buying their second home (common use case) or refinancing. The last two use cases were for nurturing our low decile customers (explorers) to improve return rate and engagement.
