To truly harness the potential of large amounts of unstructured information, companies must tweak management processes and workplace culture
One of the most powerful analytical approaches with big data is randomized, controlled experimentation, and it’s done on a huge scale in online and start-up firms’ websites. This type of analysis is powerful because it’s the only way to establish cause and effect. I’ve already mentioned A/B testing at LinkedIn—a technique involving running two different versions of a web page, randomly assigning customers to each, and then observing whether there are statistically significant differences in customer behaviour across the pages.
See on www.livemint.com