A case study about how Heron Data helps reduce underwriting risk from Matteo, a Product Manager at Juni. Juni is an all-in-one financial management platform for e-commerce businesses, and one of Europe’s fastest-growing fintech startups
How do you and your team use Heron to reduce underwriting risk?
We use Heron Data to analyse all bank transaction data from customers applying for credit. We use the outputs from Heron Data to inform our proprietary scorecard. Heron Data allows us to correctly identify revenue and company risk factors and gain a sense of the applicant’s business model and size. We use the Heron Dashboard on a daily basis.
What outcomes have you achieved with Heron?
- Quickly and reliably decline highly-indebted businesses: We decline around 15% of loan applications from businesses with good revenue and trading history, but where Heron’s analysis showed high levels of debt. This means that just in the last few months, we have avoided issuing millions of dollars in loans to companies that looked good on paper. Heron’s classification of debt is so accurate that we can identify these debts even if they are not otherwise disclosed, or hidden across many smaller transactions.
- Decline high-risk businesses and catch fraud: We use Heron’s output to identify companies that have specific risk factors, like betting, crypto investments or a lack of tax payments. We reject 5% of all applicants due to these risky behaviours that we identify using Heron.
- Underwrite faster: Heron Data allows us to save the insane amount of manual work of manually going through and individually tagging bank transactions in the process of getting an application to an underwriting decision.
- Innovative underwriting: Auto-categorization fits perfectly in our general approach to underwriting with innovative data points allowing us to build and deploy predictive models and systems at speed and with confidence.
What’s your favorite way to use Heron?
We use Heron Data’s identification of taxes as a way to evaluate a company’s operational stability and general risk. Heron Data classifies tax payments with high accuracy, therefore if Heron Data does not find any tax payments, we need to look more deeply.