Posted On: November 16, 2006, 7:29 am CST

Posted By: stat guy

I don't doubt that predictive modeling is useful. What I doubt is that it makes any sense to invest in technology that only delivers new technology. Old fashioned responsible underwriting and claims analysis cannot be understated. Too often we get the sale of technology but don't have people with the skills to make any use it or who can't make any sense of the information they assemble. What good does it do for a claims associate to run a claimsearch match report if they can't interpret what is means? So a prior claim was found from 1993? who interprets it? Using tools to move faster, does not necessarily make one more efficient, that's my argument. Why spend millions on thin-client, when even ISO relies on a mainframe at the end of the data stream?