While media coverage focus on a few spectacular Big Data successes, giving the impression that Big Data analytics is “as easy as collecting some data, deploying an application or two and waiting for the ‘game-changing’ insights to start rolling in,” the reality is that the average project returns just 55 cents for every dollar invested, according to Wikibon’s just-publishedBig Data Analytics Survey 2014 by analysts Jeff Kelly, David Floyer and Ralph Finos,
This does not mean that Big Data lacks value or that companies should ignore it. The spectacular success stories of companies transforming by applying Big Data are true, and survey respondents expect to realize $3.50 on average for every dollar invested over the next three to five years. Getting there, however, isn’t easy. Understanding the barriers to success, the analysts write, requires knowledge of the context in which they occur, including the state of Big Data analytics across vertical industries.
For instance, a high percentage of IT practitioners reported that their initial Big Data trials have been successful, while most business users judge those initial trials to be failures. The reason, Kelly says, is that the two constituencies have different criteria for success. For IT, success means getting the Big Data stack, including Hadoop and the analysis tools, working reliably. For business users, it means deriving business value from the data, an entirely different problem that starts with defining questions that will yield business value and the right data to analyze. Big Data also requires new skill sets that are in high demand.
It’s also important to understand Big Data challenges based on role, the analysts write. The challenges facing IT professionals tasked with running the technology are different from those facing applications developers, for example.