A first step for companies just starting out with big data is to identify opportunities to use their existing technology and data in new ways, evaluate public cloud and open-source options, and start experimenting with proof-of-concept exercises and prototypes, according to the analysts.
Over the following one to two years, those companies should look to use early successes with big data projects to justify funding for larger efforts. In the same period, it would be wise to also look for sponsors within business departments who will champion big data projects, they said.
Some 80 percent of what can be considered big data is unstructured or semi-structured information, said IDC analyst David Schubmehl during another presentation Wednesday. These sources could include anything from clickstream data to patent records, research archives and even video, he said.
This diversity will give rise to what IDC is calling unified information access technology, evidenced by such products as Oracle's Endeca and IBM's Vivisimo, as well as specialized vendors like Attivio.
Big data's challenges will also continue influencing the database industry, with growing prominence for technologies like graph and in-memory database platforms, said IDC analyst Carl Olofson during the presentation.
Traditional relational databases will also change, with capabilities expanding "to the point Ted Codd wouldn't recognize them," Olofson said, referring to the "father" of the relational model.
Chris Kanaracus covers enterprise software and general technology breaking news for The IDG News Service. Chris' email address is Chris_Kanaracus@idg.com