News Stay informed about the latest enterprise technology news and product updates.

Recipe for predictive analytics success includes one part storyteller

XL Insurance Inc. has a three-part formula for predictive analytics success: Be a programmer, learn statistical analysis and know how to spin a good yarn.

The secret to building a successful predictive analytics team is finding people with statistical analysis, programming and -- perhaps most important -- storytelling skills, according to one practitioner.

It's important to find multitalented people because, oftentimes, predictive analytics teams are rather small, said Jennifer Golec, the vice president of strategic analytics at XL Insurance Inc. Multifaceted individuals offer a higher level of flexibility, she said, and that comes in handy when resources are tight.

Ideally, predictive analytics professionals should be one part programmer, Golec said, because they'll be working with a great deal of information and conducting exploratory analysis. Commercial software can help in these areas, she explained, but some programming skills will still be helpful for tasks like manipulating or massaging data and creating new variables.

Predictive analytics professionals should also focus on developing statistical analysis skills because those are necessary for building multivariate models, statistical tools that use multiple variables to forecast outcomes.

"The third piece is that you have to be part storyteller. You have to be able to interpret those results," Golec said. "[That means] really being able to interpret the insight that you pull from the data. You have to be able to relay that because if you don't, you'll be sitting on this great model and you won't be able to implement it."

The popular 2011 film Moneyball -- which tells the story of Oakland A's general manager Billy Beane, who used analytics to find undervalued players and build a great baseball team -- might give the mistaken impression that analytics is all about crunching numbers. But it's much more than that, according to Golec. Organizations must also strive to understand how the results of predictive models translate to the real business world.

"Sometimes that is the danger with products like SAS," Golec said. "They make it so easy to push the data in and hit the button and have something come out. But if you're not trained to understand and interpret that output, you could end up with junk and you might not know it."

More on predictive analytics programs and predictive modeling

Read about the in-house skills needed to launch a predictive analytics and predictive modeling program

Overcome the fears associated with predictive analytics software

Learn to do predictive analytics on a limited budget

Turning to analytics to improve loss ratio
Golec, who has a doctorate in economics from the University of Missouri and ran a predictive analytics program for insurance provider The Hartford, first began working for XL Insurance and its global parent company, XL Group PLC, about six months ago.

One of her first tasks was to find a software vendor that could help the property and casualty company build out its fledgling predictive analytics program. XL Insurance launched the program to do a better job of avoiding unnecessary risk and, ultimately, improving its loss ratio.

"The loss ratio is losses over premiums," Golec said. "The lower it is, the more profitable you are."

Golec took a close look at SPSS, which was acquired by IBM in 2011, and Wolfram Research's Mathematica tools. But she worked with software from The SAS Institute Inc. in the past and decided to do so again.

"Half the battle is working with the data, manipulating the data and getting in into a form that allows you to actually do the modeling," she said. "SAS allows us to get the data into the shape and form that we want."

XL Insurance is using several SAS products, including SAS/STAT, a statistical analysis tool; SAS Graph, a visual tool that allows users to present information in charts and graphs; SAS Enterprise Guide (EG), which makes it easier to do exploratory analysis of data stores; and JMP, a data visualization tool.

"With EG, I can quickly start getting a feel for what I have now, how I want to clean up the data; where I want to collapse records, and what kinds of derived or predictive variables I can create," Golec said. "We also got JMP and that looks like a great tool in terms of making it easier to understand the data."

Implementing predictive analytics
The team at XL Insurance is in the process of building predictive models for risk assessment. The next step, according to Golec, is to implement those models and closely monitor and measure the results.

Golec said the toughest aspects of the implementation phase will likely revolve around change management and, specifically, getting the right people to adopt predictive analytics as part of their usual routines. Making sure that any workflow or architecture changes are properly documented is also a major challenge.

Another is "making sure that we've come up with how we're going to track it and make sure it's working," she said. "But I think the big thing in implementation is just achieving that buy-in and making sure that it's used."

Dig Deeper on Predictive analytics

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.