“Big Data” is simply when you have more data than a typical relational database can handle. It requires new ways of storing and analyzing information (which creates new technological hurdles in terms of systems design and employee skillsets), but for companies who get it right the result is actionable analysis in near real-time. Team a Big Data file architecture (like Hadoop) with a solid Analytics platform and you’ve got a tool that can analyze social network information or make predictions of customer demand across a multi-national corporation.
As you might suspect from an IT project, there is plenty that can go wrong with a Big Data implementation, from incorrectly specified hardware to lack of talent in key areas. Interestingly, the biggest problem seems to be the one that is most easily prevented–not knowing what to do with the data. Do you know what problem you’re trying to solve?