The insurance industry, as we all know, is a data-drivenmachine—being able to effectively analyze numbers is amake-or-break, live-or-die issue.

Trouble is—and it's one of theindustry's most pressing challenges—a great deal of that data istied up in legacy storage systems—some 30 years old, notes IanCampos, vice president of insurance with Capgemini FinancialServices, a consulting service.

To modernize their systems, insurers face essentially two paths.One is to make the massive investment in a wholly new,"best-in-breed" data platform, Campos says.

With this strategy, the costs are found not only with thehardware and software, but also with the gargantuan task of datamigration.

"To translate all that information is quite a challenge, and ittakes years to do it properly," Campos says.

Others are using "wrappers" or stop-gap measures, says Campos,where they build a newer system on top of the old main-frameapplications.

With either approach, time is of the essence—not only because ofthe competitive pressure from carriers that have already made thetransition, but also because the people who know how to operatethese older systems are "walking out the door" as they retire.

As tough as both choices are, the modernization effort simplyhas to be undertaken in order for underwriters to realize the fullpredictive potential of their data.