Eight Classic Big Data Mistakes
-
Focusing on Technologies Instead of Business
CIOs and IT leaders too often lose sight of business requirements because they get caught up in the big data infrastructure planning. Work with business closely to ensure tech aligns with needed outcomes. -
Not Knowing What You're Looking For
Always precisely determine what you seek at all stages. Otherwise, your analytics team will get lost amidst the big data. -
Disregarding Context
All data should be assessed for context and relevancy or it will never translate to usable business value. -
Dismissing Bias
Preconceptions help guide teams toward effective analytics. But they can also skew conclusions. -
Shortchanging Data Quality
Deploy tools such as language correction libraries to process unstructured data in order to ensure its integrity. -
Not Securing Data Sponsorship
To avoid pushback somewhere high along the leadership chain, you need an executive influencer to serve as an advocate for the data. -
Not Executing a Cost-Benefit Analysis
Every project needs its own assessment. If you simply apply an initial use case model to all of those projects which follow, the cost analysis may veer off target. -
Dwelling on What Already Happened
While you need to understand why a customer behavior ortrend resulted in a certain metric, the point of analytics is translating that into what's going to happen next.
Demand for big data analytics will only continue to soar as the overall market is expected to grow at a compound annual growth rate of 27% through 2017, amounting to a $32.4 billion market, according to industry research. Yet, with all of the investments into analytics tools and talent, CIOs and other IT leaders can lose sight of what they're ultimately seeking to do. Collecting lots of big data, after all, doesn't amount to much if teams don't know how to effectively mine and translate it. Or if the quality of the data itself is highly suspect. Or if team members fail to understand the potential for missing context or ignore the existence of ill-advised personal biases in forming conclusions. These are among the following eight classic big data analytics mistakes we've compiled in order to highlight the common existing traps and lend guidance about how to avoid them. They were adapted from a number of online resources, including those posted by IBM and Oracle. Clearly, there is no surefire script which will guarantee success, given that many organizations still face a considerable learning curve here. But developing awareness of these trouble signs will certainly help minimize obstacles. For more about the mistakes from IBM, click here. For more about the mistakes from Oracle, click here.