In 2016, Jacqueline Reses and her team introduced a product that makes customized loan offers to merchants who use Square each time they open its dashboard. The automated offers are based on cash-flow data that the payment-processing platform has captured through transactions. How does it work? Imagine you run a food truck. Lunchtime business is taking off and you need to place larger wholesale orders, but you don’t have enough cash. Square’s algorithm might predict the situation and generate a loan proposal, no application required. “We size the loan such that we know the seller can pay it back,” says Reses. “We won’t give you a loan that’s bigger than what you can afford.” Square’s tool is on track to lend more than $1 billion in 2017, with an average size of around $6,000 and term of nine months. Unlike competitors, Square structures its interest charges as a one-time fixed fee, ranging from 10% to 16% of the overall loan, and has also created an automated daily repayment system that calculates amounts based on the rise and fall of sales, reducing the chance of default and allowing merchants to pay loans off faster if business picks up.