Got Big Data? Where to Start?

Posted by HR Analytics on Thursday, October 30, 2014 Under: Data
"Big data" has been a big buzz word for the last couple of years now.  And the hype around it, for the most part, is appropriate because it is undeniable that there really is big data, lots of data, everywhere.  But for many organizations, especially the smaller size companies and smaller departments, big data might not be all that relevant.  And for organizations trying to get started with their analytics capabilities, this concept of big data might just be a distraction and possibly overwhelming.  Sure, we want to be able to handle big data eventually but for our immediate needs and purposes, it might be best to start, not by looking at what data you have but with first identifying what the problem is.  Start with a research agenda.  Rather than getting overwhelmed with the volume, velocity, and variety of big data, start with identifying what problems need to be solved and what questions need to be asked.

Another misleading avenue companies tend to take is the purchase the big data tool.  Big data vendors can have really fancy stuff and can really showcase their capabilities.  However, without having a research agenda, you might be putting the cart before the horse.  And in some cases, these fancy big data tools might just be super expensive over-the-top solutions.  Before looking at solutions, clearly get a good handle on what the problem is first. 

It might be best to start small and just continue to scale up to tackling big data.  In HR, for instance, it might be well worth the time to start with the research agenda.  Ask yourself what problems you are having with, say, your workforce?  Is it the upward trends in attrition? Is your organization facing a retirement boom?  Is it difficult recruiting millennials?  You must first know what problem you need to solve.  Don't jump ahead and buy a statistical modeling tool.  In Six Sigma's DMAIC, before you can Measure (or Analyze or Improve, or Control), you first must be able to Define the problem.  

The Define stage involves the following steps:
  • Articulate clearly the nature and the scope of the problem
  • Seek to clarify the facts
  • Set the goals and objectives
  • Project the targets
  • Identify the stakeholders/the customers
With a clear grasp of what the problem is, you should now be able to map out your process and your way through the data (big data if you have it).  In some cases, you might not need to deal with big data at all.  It is ok to start with small data.  You want to be able to get some quick win analyses and scale up from there.  By dealing with the low hanging fruits, you'll be able to identify the analytical tools you'll need before committing to large scale and super expensive tools.  Start with what you have--maybe your traditional spreadsheets can do the job.  Have your team of analysts learn some of the free open-source (and often just as powerful) softwares out there.  You might find that investing in your analytics talent might be more worthwhile than investing in some fancy software.  

Don't let big data cause analysis paralysis.  Big data doesn't have to be all that intimidating.  Some organizations might even find that big data is really irrelevant for them and that all they have is small data.  Use the DMAIC steps as your guide through your analytics process and you will know what data is relevant and irrelevant.  Start with defining the problem and you'll see what you'll need to measure and what data you'll need--whether big or small.



In : Data 


Tags: "big data" dmaic "six sigma" "small data" 
comments powered by Disqus

Copyright © 2012 Analytics-HR. All Rights Reserved   |   Privacy Policy  |  Contact    
Make a Free Website with Yola.