Financial insurance firm, Santam Insurance has saved more than $2.4 million using predictive analytics to curb fraud.
Tracy Mckechnie, solution sales and special projects manager at Olrac SPS, demonstrated how predictive analytics can identify risks, accelerate claims settlement and help to curtail fraud, at this week's ITWeb BI Summit.
Santam, one of SA's largest short-term insurance companies, embarked on a transformation programme to modernise its legacy systems. The company deployed predictive analytics as well as IBM document manager software to build fraud-detection models.
Olrac SPS implemented an IBM SPSS real-time predictive claims solution at Santam.
Mckechnie pointed out that a fraudulent syndicate was identified within the first month after implementation; the syndicate has since been stopped. The solution has also helped to identify high-risk claims for proper merit validation to limit fraud.
Neels Maritz, Predictive Analytics analyst at Santam, said the system also enabled clients' claims to be paid faster. “Around 70% of claims can be settled within 48 hours versus weeks. With the savings we've been able to achieve and the detection of fraud, we can decrease our rates and insurance premiums.”
Says Mckechnie: “The results of predictive analytics can support and automate decisions that are made every day on the front line of organisations.”
She says there are six ingredients necessary for a successful predicative analytics approach. These include: business goals, enterprise buy-in, ROI, data, models and the ability to take BI to the front line.
“Predictive analytics is not new. It exposes trends and patterns that are non-intuitive. It uses all data and computing power to predict what's likely to happen. Because of the volumes of data we're dealing with, and the variety of data, traditional BI systems are more difficult to use.”
Share