6. How Big Will Big Data Get?
What we saw in 2011: "'Big data' might seem like an obvious way to describe the wealth of information on policyholders that is available to insurers. But what it really represents is a change in how data is handled within the enterprise."
What happened in 2012? There are several "V" words associated with big data — volume and veracity among them. There's also variety and veracity, according to SMA's Breading. But more than anything, he says, the "V" word most associated with big data is velocity.
"Five years ago you could do analytics on large data sets, but it might end up taking you weeks to build a model and let it crank away for 10 days to come up with the answer. Now we're talking about building the model, running it it and getting the results in a couple hours — then running it again," he explains. "
SMA is in the midst of research that finds more insurers taking on big data projects this year than last. Breading says that it's those kinds of projects that are associated with speed that correlate closest to projects considered big data initiatives.
"In 2012 there was a lot of interest and education, strategies being studied, but not many projects underway," Breading says. "In 2013, it looks like there's going to be a lot more pilot activity on in big data."
