Forrester estimates that, in general, firms use less than 5 percent of the data available to them. We also estimate that data is growing about 40-50 percent annually, but the average enterprise only captures around 25-30 percent of that. This means there is a bunch of data not being captured and used by your firm, and the divide is going to grow over time. So what?\n \n There are companies that have figured out ways to use the raw data your firm might be ignoring, which would allow you to leave your competition in the dust. Yes, there are vendors adding the term \u201cbig data\u201d to marketing materials just to drive hype. But we at Forrester think the future of your firm may depend on your paying attention.\n \n The data produced by your customer-facing systems is a treasure trove for targeted marketing and retention insight\u2014if your firm has the know-how to get that value. If you are not doing something to capture this value, you\u2019re losing ground to competitors.\n \n Forrester defines big data as techniques and technologies that make it economical to deal with data at an extreme scale. Big data is about three things: 1. Techniques and technology, which means your firm needs people who know what to do with the data to get the value. 2. Extreme scale\u2014data that outstrips your current technology due to its volume, velocity, variety and variability. 3. Economic value, or making solutions affordable and helping your business case for investment.\n \n CIOs need to take a pragmatic approach to big data. Establish a core set of data governance directives that account for the special requirements of big data sets. You might not have discovered and protected all the elements in the dozens, hundreds or thousands of terabytes of machine data collected as part of a big data project.\n \n There are no best practices yet, so start with policy and start small. Once you hit the right business value, your project could take off. Don\u2019t encumber that event with overly complicated rules, but plan to get to extremes quickly.\n \n Brian Hopkins is a principal analyst at Forrester Research, where he serves enterprise architecture professionals.