Merging Open Source Banking & Credit Bureau Data
Objective
Create scaleable infrastructure that could handle open-banking data.
Obstacle
Budgets were tight for this start-up and needed to have a clear impact.
Outcome
Lower default rates, and long-term infrastructure set up.
Background
According to research by PWC, around 20 million adults in the UK are held back by inaccurate or incomplete credit data, leaving many individuals excluded from accessing financial products or receiving unfairly high interest rates. To address this issue, Plend sought to revolutionise the lending process by moving beyond traditional credit scoring systems. They aimed to combine open banking data with credit bureau data to provide a more accurate and fairer assessment of an individual’s creditworthiness, leading to loans with lower default rates and better terms for consumers. Understanding the complexity of integrating these diverse and large datasets, Plend turned to 173tech for their expertise in setting up the necessary infrastructure.
Challenges
Data Volumes: Processing open baking data involves handling significant volumes of detailed information, such as transaction histories, spending patterns, and account balances. The sheer size of these datasets translates into increased costs for both extraction and storage. For Plend, ensuring scalability and efficiency in managing these data volumes was critical to keeping operational costs under control.
Which Data Is Valuable: The abundance of data from both open banking sources and credit bureaus presented a unique challenge: identifying which specific factors carried the most weight when assessing loan applications. With a myriad of variables, from income patterns and spending habits to repayment histories, it was difficult to determine which metrics truly correlated with creditworthiness and loan performance.
Solution
Scaleable Infrastructure: We designed and implemented a scalable data stack from scratch, ensuring it could handle large volumes of data without driving up costs. This infrastructure was built with flexibility in mind, allowing Plend to adapt to future needs without overhauling their systems.
Data Modelling: Once the data stack was in place, we modelled the open banking and credit bureau datasets, creating a unified framework that integrated seamlessly into Plend’s existing analytical infrastructure. This setup enabled Plend to extract actionable insights, empowering them to make fairer, data-driven decisions on loan approvals.
The Proof Is In The Numbers...
£8m
10
£640
Implementation
Long-Lasting: Our work directly contributed to Plend funding over £8 million in fairer loans during their first year of operations, marking a significant step forward in their mission to make lending more inclusive and equitable. By building a robust data infrastructure and integrating open banking data with traditional credit bureau information, we enabled Plend to make more accurate lending decisions.
Default Rate: This approach not only helped reduce their default rate but also allowed them to proactively identify pre-arrear customers (individuals showing early signs of financial difficulty). With these insights, Plend could take preventive measures to support borrowers before they fell behind on repayments, fostering trust and financial stability for their customers.
Sustainability: The success of this project also positioned Plend for long-term growth. Our initial implementation provided the technical foundation and analytical tools they needed to scale efficiently. With these systems in place, Plend was able to begin hiring their own in-house data team, focusing on strategic work rather than heavy technical lifting.
