March 20, 2011

Analytics Sunday: Top Ten Analytics Staffer Challenges

If you are an analyst, you can relate to the issues listed here.

Top Ten Analytics Staffer Challenges Within Companies

(Not in any particular order):

Number 10 = Information Technology Misalignment.  If you work in an environment where you require support from your information technology team, oh boy, good luck.  When you have a great IT team, you hold on to them like grim death!  Otherwise, you have to fight around these folks.  I've had great IT teams that supported my group so well you wanted to adopt them into your team!  And, I've been in situations where I hired my own IT-clones because I wasn't getting the support I needed.

Number 9 = Being Alone.  Your company hires you, and gives you absolutely no support whatsoever, right?  You are alone.  This is where Twitter is really important --- you have to act like you have co-workers, and build relationships with psuedo-co-workers.  When you're alone, you can't effectively serve your Executive audience, especially if they have a pre-conceived notion of what your job is or how fast you should do your job.

Number 8 = Systems.  You don't have an analytics system if you use Google Analytics --- rather, Google has a system where they get to know your business, inside-out, if they choose to look at it.  Analysts really struggle to communicate needs to Executives.  How do you tell your CFO that your PC is too slow to run a 100GB dataset through SAS?  How could the CFO ever understand that?  I recently watched as a SAS analyst printed a document on that green and white tractor paper from the 1980s.  Yes, this happened.  This person was leveraging a mainframe computing environment from the 1990s.  It is terribly hard for an analytics expert to convince anybody that the computing environment stinks.  Honestly --- if you can at all afford to do so, buy your own high-powered desktop PC --- you can get a rocket for under $1,200, and connect a 1.5 terabyte external drive to it.  Create your own data extracts from your own crappy database system, and do the heavy lifting yourself!

Number 7 = Competition.  Somebody in your company believes s/he can do your job.  I ran into this at Nordstrom all the time --- there were folks who actively recruited their own analytics projects, or refuted the work my group did with their own database analysis.  Even if you're the only analyst at your company, somebody in IT strongly believes s/he can do your job, and at times, is asked to do your job without your knowledge.  You combat competition by doing the job faster, more accurately, and in a more unbiased fashion than anybody else at your company.

Number 6 = Accuracy.  One of my best analysts made a ton of mistakes, and was not capable of catching the mistakes before the client caught the mistakes.  This is one of those "three strikes and you're out" scenarios ... you make mistakes that other folks catch three times, and you don't have a lot of credibility.  Try hard to tie your numbers, where possible, to an existing report or financial document.

Number 5 = Career Path.  You may not realize this, but there are business leaders conspiring to hold you down ... not on purpose, mind you.  I once had a business leader tell me that I was not capable of doing one other job in his company because I did not have any relevant experience outside of analyzing numbers.  If I could do it all over again, I wouldn't pursue a VP - Database Marketing job with zeal, I'd try to gain merchandising skills, marketing skills, finance skills, store management skills, creative skills, just about anything possible to broaden my horizons.

Number 4 = Politics.  You are not given complete information when you are assigned a project.  In a meeting at 8:00am, two Executives got into a spat about Facebook, one individual said that Facebook customers are worth $124.00 each, one individual said that the social media team should be fired.  At 9:00am, you're asked to demonstrate what the value of a Facebook user is.  You're not told that somebody wants to fire the social media team if your analysis suggests that there isn't any additional value derived from Facebook use.  When you present your findings ($2.00 incremental value), one VP just keeps beating you up over the results.  Well, this person doesn't want to fire her staff, so you are the problem.  Of course, you're not the problem, but you got caught in the middle of a political battle.  Interview the person giving you an assignment, try to ferret out as much information as you can.

Number 3 = Boss Can Do Your Job.  Oh, this one is terrible.  I've been the boss (my poor staff!), and I've been micro-managed by a person who used to do my job.  When you are in this situation, you often have to do each project twice ... do it the way the previous boss wanted it done, then do it the way you want it done.  Always create an appendix --- if your boss demands that your work has to be done a certain way, then do it that way, but refer others to the appendix for a new way to look at the information.  Never give up, never give in.  Your job is to reflect the behavior of the customer, not the methodology of the boss.  Honestly, you can accomplish both tasks.

Number 2 = Misinformation.  What you read on Twitter is not representative of the actual challenges the average analyst faces in a company.  You have few places to get viable information ... on Twitter, a half-dozen or a dozen individuals control the discussion.  Be careful here!  Try hard to get a neutral, unbiased, reasonable view of what is happening.

Number 1 = Sanity.  Your organization is going to take you down an unpredictable path.  When I worked at Lands' End, everything was about measuring the incremental value of advertising (what is today known as attribution).  At Eddie Bauer, everything was about the 'small stuff', how many customers responded to a promotion.  At Nordstrom, it was generally about customer value in a store environment, or the incremental value of an online customer to our store division.  The company you choose to work for shapes how you develop.  If you want a well-rounded experience, if you want to maintain your sanity, you may need to work for 3-4 companies in the first decade of your career.


  1. Elizabeth Eckels8:11 PM

    Kevin - What a wonderful post! Being 2 years into the field, it's great to know what challenges I will face, and a bit relieving to know the ones I currently face may, in fact, be normal. Another thing I could add to this list is, no data is 100% accurate - and some people will dog you over it. (I dogged myself for it in the beginning.) It's important for us to share that directionally/proportionally accurate is what we're here for.

    I appreciate the advice and insight you've provided.

  2. I'm confident that your challenges are normal!

    Yup, data is inaccurate, everybody just needs to deal with it. We all make decisions based on inaccurate data, so it is ok.

  3. This is one of the most insightful blog posts I have ever read. Thank you, Kevin, for "telling it like it is." This should be Required Reading for all data analysts AND their managers!!!

  4. Anonymous4:45 AM

    An interesting read.

  5. Your post is pretty illuminating and practical,I completely agree with the above comment. Thanks for the sharing.Will you share more about it? I just want to know more.You are very talented.I guess, this site would be my farorite one. Thank you again!!!


Note: Only a member of this blog may post a comment.

Summer Schedule / Document Your Progress This Summer, Ok?

Most summers, I cut back on how often I publish. This summer will be no different ... and you may be subjected to a few videos instead of bl...