Her team completely revamped their customer care process to understand customers on a personal level
Knowledge on policies and good BDM relationships helped this mortgage director keep a single mother and her four children from losing their home
- 2018 Commercial Lenders Roundtable
- Top 10 Brokerages 2018
- 2018 Brokers on Aggregators
P2P marketplace lender Zagga has partnered with the University of Technology Sydney (UTS) to develop a customised machine-learning algorithm that can predict investor behaviour.
The algorithm matches the funding appetite of investors with loan opportunities from creditworthy borrowers, and tracks their behaviour on the platform over time and builds a unique understanding of this behaviour.
“Very few marketplace platforms use technology to match their investors with potential borrowers, relying more on a self-selection marketplace-style system,” Zagga CEO Alan Greenstein said in a statement. “Partnering with UTS, we were able to create an algorithm that goes one step further and helps us understand the underlying driving forces behind investor behaviour.”
The challenge that came with building a behavioural science algorithm for a fintech company delivered useful insights to the UTS team. Development began in February 2017, and the first phase of the project was integrated into Zagga’s platform in June this year.
UTS School of Electrical and Data Engineering associate professor Richard Xu, and senior lecturer with the School of Software Sam Ferguson found that investors’ decisions are influenced by a range of behavioural factors, which includes social biases, risk versus returns assessment, and group mentality. These factors “are possibly the difference between what an investor says they will invest in compared to what they are actually open to investing in.”
“The algorithm will start to help predict the probability of someone investing outside their selected preferences, meaning we can open up broader opportunities to investors, providing an investment menu more tailored to their personal preferences and behaviour,” Greenstein said. The technology will also help Zagga to discern whether a loan will be funded and how soon.
Once the next phase of the algorithm development is complete, Greenstein hopes they will also be “able to uncover new thinking on investor behaviour.”