Don’t Hire Data Scientists Until You Are Ready for Data Science

by   |   August 13, 2015 5:30 am   |   1 Comments

Greta Roberts, CEO and co-founder, Talent Analytics, Corp.

Greta Roberts, CEO and co-founder, Talent Analytics, Corp.

I had a call recently with a data scientist working in the human resources department of a major business. This HR data scientist has both a strong analytics and predictive analytics background. She has a bachelor’s degree in Statistics and a master’s in predictive analytics. She excels in R, math, predictive modeling, machine learning, and all things quantitative. She is also excited about applying data science from other domains, to solve interesting workforce optimization challenges.

She applied for a quantitative HR role that promised to let her use her skills and interest in solving difficult, employee-based challenges. She was hired for this role.

But HR won’t let her do data science.

Over and over again she has suggested a data-science approach to help solve employee-focused challenges that have plagued the organization for years and cost it many millions of dollars. And over and over again, she is denied the ability to move forward.

She told me that HR seems to be scared or hesitant to move forward with a new way of solving problems. Her main concern is that the reason for not moving forward was never fully discussed.

Instead of applying a data-science approach to the company’s problems, she is asked to generate the weekly and monthly reports to which the organization has grown addicted. When she is allowed to solve an interesting problem using analytics, and does so brilliantly, the executive HR leadership won’t give the solution executive visibility or implement it in production. Results are considered “interesting,” but not deployed. Then, she’s back to generating reports.

She isn’t alone. Not by a long shot. This sort of thing happens all the time. I can see it on LinkedIn updates as brilliant data scientists move from one company to another. And I hear it in the conversations I have with them about why they leave their jobs and their angst before they leave.

Related Stories

How to Put HR Analytics to Work.
Read the story »

From Information to Insight: Human Capital Analysis for HR.
Read the story »

Analytics Leaders Discuss Care and Feeding of a Successful Data Team.
Read the story »

Is People Analytics The New Humanizing Force?
Read the story »

My plea to companies that are considering hiring a data scientist: Stop hiring real data scientists until you are ready to do real data science.

I think I understand some of the problem: Perhaps a department has been under pressure to begin using an analytical approach, and so hires data scientists. But when it comes to actually using this approach, suddenly it’s too foreign, scary, or just “not what we have done before.” But companies need to learn from these people they are bringing into their domain or stop hiring them to begin with.

Anyone can hire a data scientist. But not every organization is ready for data science. Generating reports is not analytics, even if they are prettier or faster reports. Dashboards are not analytics, even if they are really pretty dashboards. More than any other department, HR should understand the devastating impact of changing the job description on someone who has been hired.

Hiring the data scientist mentioned above could end up being one of the most intelligent and strategic hires that HR department has ever made. But only if they let her do what she was hired to do. Data scientists can help move HR departments from being tactical to strategic by using an analytics approach to highlight previously overlooked patterns, make hiring decisions based on data, and improve retention.

The following seven tips can help you help that HR data scientist you hired become one of your most successful hires:

  • Assign reporting to someone else. It’s a very important task, but it doesn’t require a data scientist. Reporting will quickly bore a data scientist, and she will resign.

 

  • Don’t block data scientists from talking directly to your business areas. (I often hear they have to go through the HR business partner, who protects the business leader and blocks them from access).  Working with the HR business partner of course makes sense. Being blocked by the HR business partner doesn’t.

 

  • Task HR business partners with finding either high-turnover roles or low-performance roles that your data scientist can help with. These are great projects for a data scientist.

 

  • Have the data scientist focus first on solving business challenges (like financial advisor turnover), not HR challenges like compliance issues. This will give visibility to the great work they do and introduce HR’s new expertise to solving business challenges that affect the bottom line.

 

  • When the data scientist completes an analytics project, give her a chance to talk and present the results, regardless of the outcomes. Did it help or not help? Don’t keep the results inside of HR.

 

  • Admit that you are a little nervous about what they do. They are nervous about what you do, too.

 

  • Trust your data scientists. Stop being scared. You hired them because they have an area of expertise that traditional HR doesn’t. Embrace their area of expertise. You need to trust their advice and approach, or they will leave.

 

And mostly, don’t hire a data scientist if you are not ready for data science. If you thought you were ready and later realize that you really aren’t, let them know and let them go. Be honest. Don’t put them in a different role and block them as they keep trying to be successful.

Greta Roberts is the CEO and co-founder of Talent Analytics, Corp. Follow her on Twitter @gretaroberts.


Subscribe to Data Informed
for the latest information and news on big data and analytics for the enterprise.






Tags: , , , , ,

One Comment

  1. Lucas
    Posted August 14, 2015 at 10:02 am | Permalink

    When I worked in HR, I found many employees to be resistant to processes automation projects because they thought it would cost them their jobs. I think it is important to sell process improvement as enhancing current operations, not replacing them. Any process improvements should be driven strategically from the top down, with lofty goals that cannot be achieved the old fashioned way, this way automation and new techniques are seen as the saviors, not something to be shelved.

Post a Comment

Your email is never published nor shared. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>