Grow Your Own Analytics Talent

Recruiters call me all the time, looking for analytics talent meeting long checklists of requirements. These lists read like ads for the Veg-O-Matic. They want people who can slice data, dice it, clean fish, and mend socks with it.

But wait, there’s more! It’s not unusual to hear ridiculous demands, like 10 years of experience in a technology that has only been around for five years.

The callers say all this with conviction, even though they clearly don’t fully understand what they’re asking. One guy explained that his client rejected a good candidate because it wanted “more of a pure data scientist.” I asked what that meant. “Gee, I really don’t know,” he said.

You want people who can analyze data and provide you with some useful information? You can have ‘em. But you’ve got to be reasonable. About those checklists: Look at every item on your list, and ask, “Why?” Are you asking the analyst to fill in tasks for others who aren’t doing the work very well (instead of addressing those issues directly)? Are you padding with nice-to-haves? Are you holding out for someone just like you? These things will only impede your progress.

Stop dreaming. Stop waiting for magic people who are rock-solid experts at database management, programming, statistics, accounting, presentation, and graphic arts, able to do it all at the speed of light and happy to work for what you want to pay.

You’re losing money while you wait for perfection. Begin the process by defining your business goals, identifying the capabilities necessary to meet them, and determining which of those you do and do not have in-house right now.

Now, let me tell you a little story. Once upon a time, an analyst quit her job. The boss did not say, “Woe is me, for verily I am up the creek without a data scientist.” Instead, the boss looked at her secretary and said, “You can do this.”

Then the boss did something very shrewd. She provided her former secretary with training in statistics and software. Although the secretary had neither previous experience in data analysis, nor a college degree, she had a brain, a good work ethic, and an understanding of her industry. The secretary became a very decent analyst and the boss lived happily ever after. This is not a parable; it is a true story about a student of mine.

The first resource for filling your analytics needs should be your own people. They already know your business, they want professional growth opportunities, and they may already have skills you haven’t tapped. When you need to look outside, seek new hires who meet key requirements and provide training to fill in gaps. Invest in training. Don’t skimp on quality or quantity! Look for the best and most relevant education through professional organizations, consultants, universities, and vendors.

Which do you think will provide better value for your organization -- top-flight training, or head-hunting fees?

Meta S. Brown, Business Analytics Consultant

Meta S. Brown is a consultant, speaker, and writer who promotes the use of business analytics. A hands-on analyst who has tackled projects with up to $900 million at stake, she is a recognized expert in cutting-edge business analytics. She has conducted more than 4,000 hours of presentations about business analytics, and written guides on neural networks, quality improvement, statistical process control, and many other statistical methods. Meta's seminars have attracted thousands of attendees from across the US and Canada, from novices to professors.

Tell Me a Story: Why Data Analysts Must Be Storytellers, Too

Data alone won't make an analyst's work memorable or actionable in the eyes of a business executive. A story puts it into perspective.

It's the Data, Stupid

When it comes to acquiring the data that will feed your analytics initiative, "free" isn't always the best approach.

Analytics Talent and Unreasonable Expectations
  • 11/29/2012 3:16:47 AM

Wonderful, spot on message Meta regarding fostering your own internal talent.  I so enjoyed reading your take on headhunters parroting the demands that even most employers do not truly understand.

I have seen these listings and while I don't mind studying towards something, what most companies are asking for is often unrealistic.  Almost to the point of being absurd.   

So while the position(s) go unfilled - these companies can continue to do apparently what they do best - hold unreasonable expectations.

And if they do find this person with it all - are they prepared to adequately compensate ?  

And I love your true life example of a secretary who made it big !  Good for her !

Re: What Wal-Mart does...
  • 11/28/2012 5:16:58 PM

Hi Beth,

 You are right about the risk of limiting creative thinking if we limit "what can be done with the data". But the source of the data and the features of its different instances should allow us to limit the scope of our predictive model. But it is true that ruling out only the obvious "can'ts" is the right thing to do.

Re: What Wal-Mart does...
  • 11/28/2012 12:48:56 PM

Hospice -- I can see some danger in asking "What can't we do with the data?" though -- in that it might limit creative thinking or innovative approaches to modeling (unless you're talking about ruling out the more obvious "can'ts" but not creating too many limitations?).

Re: What Wal-Mart does...
  • 11/28/2012 12:46:16 PM

I love it, Peter! I can see it scrawled across whiteboards in student classrooms, too. It's a great reminder for folks at any level. Thanks for sharing the pic!

Re: What Wal-Mart does...
  • 11/28/2012 10:36:05 AM

I love that graphic. It's an excellent illustration. The defense against torturing data is to truely understand the scientific method and the role of experimentation in numerical analysis. Further, understanding the inadequacies of the data you are working with is important in understanding what you can expect to get from it.

That quote went up on my white board (well, 1 of 3!) a few days ago. I guess there is something viral about it as I've seen it now several times in the last week. I placed it as a warning to myself and others. Tread carefully and learn how to respect the analytical process.

Re: What Wal-Mart does...
  • 11/27/2012 11:53:41 PM

Thanks for sharing. I think we should start by the asking the question: what can we really do with the data and what can't we do with it? We need to answer that question before trying to build any predictive model with the data we collect.

Re: What Wal-Mart does...
  • 11/27/2012 8:52:34 PM

Meta, I would agree there's danger in thinking that there's magic in the tools -- especially those delivering visualizations. It's too easy to think the data is good because it LOOKS good! You've got to understand what it you need and what the data is telling you.

Re: Home grown
  • 11/27/2012 8:48:19 PM

Great story Meta. Thanks for sharing!

Re: Home grown
  • 11/27/2012 8:45:59 PM

Peter, sounds like you've got a great work environment!

Re: What Wal-Mart does...
  • 11/27/2012 8:32:05 PM

@Peter Mancini, funny you should mention this quote. I heard it just myself a week or so ago, from Mike Swinson, an EVP at TrueCar, in a presentation on predictive modeling he delivered at IE's Predictive Analytics Innovation Summit in Chicago. I just wrote about his presentation today: Ask & Ask Again: Questions Matter in Modeling. I like the imagery he used in his slide presentation:



Page 1 / 2   >   >>