HR Management, HR Technology

“A Fool with a Tool is Still a Fool” – 5 Guidelines to Make Big Data Work

Illustration by istockphoto.com

There’s a lot of talk about Big Data these days and many HR people are either excited or intimidated by it. Maybe both.

Regardless, there seems to be more talk about “it” than on how to use it in solving problems or how HR can benefit from it.

So the question is how to use Big Data — what can you do to make sure you get meaningful results? Here are five (5) guidelines that will help keep you on the “straight and narrow.”

First, let me introduce a case study we can use as an example in this five (5) step process.

Case study

There was a large, multinational, high tech company that was concerned about its leadership pipeline. Top management asked Recruiting to look at the profiles of past and current Director level managers and above. They hoped there were some common factors that would shed light on how to direct recruiting efforts in the future.

Recruiting requirements had for many years focused on Engineering graduates from major well-known universities. The criteria were Masters in Engineering with high GPA’s.

We’ll see how this case pans out as we go.

Rule No 1: Never choose what Big Data to use first

When you have Big Data, it’s easy to believe that all your problems can be solved — i.e. select some data and then search around for a problem it can solve.

Hmm, that’s not a good approach.

One of the greatest dangers of violating Rule No. 1 is that you define success as merely using “it” rather than whether or not you’re using it in a meaningful way.

Rule No. 2: Start by defining the problem you’re trying to solve

Clearly define the result you want first. What is the question you want an answer for? Only THEN choose the data that will help you get there.

The real value of defining the result you want is that it helps keep you on track and focused. Clearly defining the result you’re after is not always easy to do. But you can’t select relevant data until you do it.

Case study: The end result was to identify the common background factors of Directors and VPs that might be correlated to their rise through management ranks faster than their colleagues.

Rule No. 3: Identify what data you need

Once you have defined the problem/result, identify all the data elements you need.

Case study: The data pulled included: university graduated from, GPA, major, history of performance ratings.

Rule No. 4: Identify the cause of the problem

If you (and others) are satisfied with the data you selected, run with it and accept the result. Don’t keep trying again and again just because you didn’t get the results you want.

Case Study: When recruiting looked at where management leaders (Directors and above) had gone to school, what their majors were, and what their GPAs were, the results were surprising.

Most of the leaders had come from small, second and third-tier universities. A few majored in engineering. However most majors were in liberal arts and business. And most of the current management had GPA’s in the 2.5 range.

The majority did not have advanced degrees. As far as performance ratings, 63 percent had higher ratings than those who had Engineering degrees and graduated from first-tier schools.

Rule No. 5: Make your recommendations

Once the problem’s common background factors have been identified you’re ready for your recommendation. You can feel confident that the data you’ve captured is comprehensive unlike the “old days” when you had to search several databases or, worse yet, look through employee records manually.

Case study: Recruiting made their recommendation to management. They recommended that the company initiate a pilot program: Half of the college hires would have to meet the requirements already in practice.

The other half would need to meet new recommended criteria: no top tier university requirement, no requirement for either grade point average or type of degree.

Interview candidates more closely on what they have done in the past/achieved even though there may be no job-related work history. Look at achievements in college associations, volunteer organizations, etc.

Top management’s reaction? They thanked recruiting, but decided not to change hiring requirements even with a small pilot project. They asked recruiting to continue focusing on candidates with Masters in Electrical Engineering with high GPA’s from top tier universities.

A big disappointment for the recruiting team. We can only speculate on management’s reasoning — but it might stem from their discomfort in letting go of a familiar process.

Conclusion

Following these five guidelines won’t necessarily make your job easier or get acceptance of your recommendations. They will serve as guardrails to make sure you stay on the right road, avoid the “fool with a tool” syndrome, and deliver maximum value to your company.

Having Big Data ushers in a whole new era for HR. Don’t let it get away from you.

Learn to control it and make it work for you — not the other way around.

 

Jacque Vilet, President of Vilet International, has over 20 years’ experience in International Human Resources with major multinationals such as Intel, National Semiconductor and Seagate Technology. She has managed both local/ in-country national and expatriate programs and has been an expat twice during her career. Jacque has also been a speaker in the U.S., Asia and Europe, and is a regular contributor to various HR and talent management publications. Contact her at jvilet@viletinternational.com.
  • Dr. Wendell Williams

    The key to using big data in human resources mandates understanding what causes results. Degrees, schools, class rankings…none of these “cause” performance. They are all just correlations…Users need to narrow-in on things like specific cognitive ability, organization, interpersonal abilities, and attitudes/interests/motivations ….All the things that walk around on two feet…anything else is could be pure chance.

  • rajkumar

    wow great post. I am also looking such as blog . i like that post