Frank Fiorille, senior director for risk management at Paychex, a $2.2 billion payroll and human resources services company, says that the trend in risk management is “moving toward offense.” So when increased competition started hurting the company’s top line, he set up an ERM Predictive Analytics group.
Several models were developed to combat loss rates and enhance cross-selling. The first of these cross-selling models, the Predictive Algorithm Targeting Retirement Investment Clients Knowledge (PATRICK), proved to be a hit with the sales team because it helped them prioritize the sales of 401(k) plans to the core client base at Paychex, improving the dial-to-appointment ratio for the company’s telemarketers by more than 50%.
Fiorelle says the concept works a bit like Amazon. “With Amazon, you buy one book, and then the next time you go to their site, they’ll say, ‘You might like these books based on your recent purchases.’ What our algorithm does is churn through our existing customer base and identify up-sell opportunities.”
“In the past, our sales people might have called Joe’s Dry Cleaner on a cold call,” he says. “Now we can say, ‘Hey, based on this guy’s score—the velocity by which he’s added employees in a month, or the number of transaction bank accounts he’s added—he’s probably ready to buy a 401(k) product.”
The big challenge, Fiorille says, was getting the idea of using risk management tools opportunistically in front of senior management “so that they could see the value-added function.”
The trick was showing sales-focused managers what a bunch of mathematicians in risk management could do to boost sales. “Once we showed them we could increase sales, it was over,” he says, though he adds, “Our operation is still a small skunk-works, out-of-the-garage operation, and I’m happy with that. It keeps us nimble.”