To qualify, applicants needed a business account with the lender, meaning new customers faced a longer journey, especially smaller businesses which regularly operate from a personal account. The majority of lenders to this sector had also prevented further personal borrowing, meaning that SMEs were forced to apply for BBL via approved banks in order to stay afloat.
The result was a huge surge in applications in a short period of time. But applications also had to pass strict Know Your Customer (KYC) checks before funding was approved, including proof of trading prior to 1st March 2020.
With bank staff shortages due to Covid on top, operational resources were under huge strain and turnaround times for applicants to receive funds lengthened to between three and four months. To address this, the data and analytics (DNA) team in Barclays Business banking devised a new model based on transactional data which could automatically profile an applicant’s trading history. Millions of transactions on personal accounts needed to be analysed, segmented and scored, requiring multiple data engineering iterations which ultimately generated a model accuracy of 90+%.
Approval to deploy the model was given by the British Banking Association which saw it embedded into the new business account opening process by July 2020. Automating KYC validation of applications removed a huge volume of demand on both bank colleagues and systems and ultimately led to Barclays dispersing £300 million of funds to new sole traders opening business accounts between July and December 2020.