Hitting the Bullseye: Revolutionizing Acquisition

Author
.
2023

Overview

Step into the world of customer acquisition as the marketing team seeks to scale originations for unsecured personal loans through cost-effective direct marketing. With limited direct response history and the need for guidance on prospect data sources, a solution was paramount. Join Blend as they embark on a data-driven journey to revolutionize acquisition targeting. Through a meticulous DataLab study, a suite of machine-learning models was developed to predict response and approval probability among available prospects in the market. Rigorous model validation ensured accuracy, leading to the successful implementation of a direct mail campaign. Witness the impact as high-quality applicants and booked loans soar, with response rates in decile 1 doubling compared to decile 5. Approval rates in decile 1 surpass industry averages, providing invaluable customer profiling insights for strategic decision-making. Prepare to witness the transformation of acquisition targeting as consumer leading reaches new heights of success.

Challenge

The Marketing team faced a challenging task of scaling originations for unsecured  personal loans through cost-effective direct marketing. Their historical  sales had predominantly relied on cross-selling and non-targeted media,  leaving them with limited direct response history to draw insights from. To  effectively plan future campaigns, the team required guidance on prospect  data sources that would serve as a foundation for targeting. A key  consideration in their targeting strategy was the responder's ability to be  approved for credit, as it typically correlated inversely with demand in the  personal loan market. The bank aimed to minimize brand risk by avoiding a  high rate of declines resulting from the campaign. This complex set of  requirements presented the Marketing team with the challenge of finding an  optimal solution that would drive scale while maintaining brand integrity.

Solution

Blend conducted a thorough study of non-credit sources in the market and developed  machine-learning models to predict response and approval probability. The  models were rigorously validated and implemented in a direct mail campaign,  allowing for analysis and comparison of loan applications. This solution from  Blend maximized the potential for scaling unsecured personal loan  originations.

Impact

The launch of the program has yielded remarkable results, generating a  significant influx of high-quality applicants and booked loans. The program's  success is evident in the response rates, with the decile 1 group  outperforming the decile 5 group by a factor of two in terms of actual  response. Additionally, the approval rate in the decile 1 group surpassed the  decile 5 group by an impressive 50%, exceeding the category average. These  outstanding metrics indicate the effectiveness of the targeting strategy and  the program's ability to attract creditworthy individuals. Moreover, the  creation of customer profiling has provided valuable insights that inform  important marketing and strategic decisions, enabling the team to make  data-driven choices and further optimize their approach. This impactful  outcome showcases the significant value and success achieved through the  implementation of the program.

Key Data Points

2x
higher acual response
0.5
higher approval rate