Some of the biggest decisions we make as consumers throughout our lives rely on the analysis of our financial data by banks and credit reference agencies to establish how good a customer we might be - from applying for a mortgage to buying a car.
The value exchange of financial consumer data has traditionally taken place between banks and credit reference agencies, leaving consumers as mere spectators, and sometimes victims, of a highly profitable industry. The debacle of the payday loan industry is a good example of such behaviour.
However, Open Banking changes the game and for the first time, consumers are in control to decide how their data should be used to improve their lives.
Categorising banking data lies at the heart of making Open Banking work for society. Developing a deep understanding of consumer behaviour is fundamental to creating relevant product and services that address people’s real needs. The approach taken by incumbents and challengers alike is over-simplistic, futile and delivers no real value to consumers or financial institutions.
The examples below from two British retail banks illustrates this challenge - classifying transactions into 10-12 high-level categories such as “personal” or “shopping” adds no value to consumers seeking to gain insights into their financial circumstances or to financial institutions looking to use this data to make better product and services decisions for consumers. This problem is widespread across the industry and it affects incumbents and challengers alike, as well as fintechs and other lenders.
We, at Ducit.ai, have focused our initial energy to resolve this problem once and for all. We have worked with Nationwide’s Open Banking for Good programme to create a solution to make Open Banking data work for both consumers and the financial services industry.
Our first use case has been the development of an income and expenditure capability for debt advice charities looking to help consumers to become debt-free based on our transaction categorisation engine. This is available free-of-charge to the debt advice sector.
Our approach includes 5 key components:
- A taxonomy of 89 categories for income and expenditure enclosed below that truly creates an accurate picture of people’s financial lives. Its sole purpose is to generate intelligence from Open Banking data to aid decision making. The range of categories goes from unauthorised overdraft charges to water, mortgage payments, payday loans and gambling. Whilst this approach significantly increases the technical complexity in the development of the solution, it does provide the depth of insights required to make intelligent decisions
- An artificial intelligence engine that mimics human behaviour and categorises transactional data in real-time across the entire Open Banking ecosystem to eliminate the current inaccurate, inefficient and manual processes
- A transaction categorisation API available to banks, fintechs and lenders that offer a real-time on-demand service to categorise their own transactional data for their own purposes - see our APIs documentation here
- An affordability score API based on the categorised data which is available to lenders looking to make credit decisions for consumers
- A consumer mobile app to empower consumers with access to their banking data to decide how it should be used to improve their lives and with whom it should be shared
Our approach will benefit two main audiences:
- Consumers looking to consolidate their banking data to gain a deeper understanding of their financial lives and better access to financial products and services
- Banks, fintechs and lenders seeking to improve the quality of their decision making based on deep transactional intelligence to drive better outcomes for consumers
As we complete our work in Nationwide’s Open Banking for Good programme with our debt advice charities partners by the end of the year, we are now getting ready to actively push for the adoption of our solution across the wider industry throughout 2020.
Looking forward to the challenge!
Note: This is a repost of the article originally posted on LinkedIn on December 6, 2019.