Tableau

Hiring Tableau Developers: The LinkedIn Recruiter’s Guide

Tired of struggling to find the perfect Tableau talent?  Look no more!  Here’s how to hire a Tableau designer easily, efficiently, and painlessly.

Step 1: Ditch the BI checklist

We all have them, the checklists that are supposed to make our lives easier.  For sectors and skills that change quickly, legacy lists will not do.  If you’re looking for someone with 10 years experience doing ETL, using a laundry list of other business intelligence 1.0 tools (hint: Crystal Reports, Cognos, OBIEE, and the like), and able to fix the kitchen sink, you’ll find that person, but it will take awhile.  (For an incomplete guide, check below.)

Other times, you get these lists from the client seeking talent.  Okay, MOST of the time.  And where did they get that list?  Google, or, perhaps even more dangerous, LinkedIn.  And so, what was dated the first time replicates and mutates into a monster where very few people fit that description.

What to do instead:

  • Push back. Do they really need someone with 20 years experience writing SQL for a 10 table database?  (No.)  Or a Tableau Zen Master? (Only about 30-40 people have ever had this recognition.)  Or a computer science degree?  (Besides excluding a vast number of very qualified candidates, including many females and minorities, this does not guarantee a good analyst.)
  • Clarify the complexity.  When you don’t do the work, it seems complex.  Technology in particular gets this label.  Figure out what the true skill need is of the job (people will tell you they need an expert when they don’t) and who would be willing to take it (skill need versus salary/opportunity).
  • Get comfortable.  Do you do a lot of technical recruiting?  Try to get familiar with the technology so you don’t get punked by Pokemon tricks or swindled by a good talker.

Step 2: Rewrite the job ad

You may not need to do this.  Maybe, just maybe, a simple edit will do.  But, as someone who has been on both sides of this, you probably want to start over.  Trust me, it ends up being easier.

Your clients do not understand the job, the applicant pool, and all the juggling you, as a recruiter, do.  Your applicants, particularly those at certain levels, get hit all the time by opportunities in the “big data/data” sector.  You have to somehow make the connection.  You can try to make both people happy (this rarely succeeds) or you can figure out the goal (we need someone to bang out reports, educate, and support users vs we need someone to manage the server, create data sources, and monitor and manage adoption).

Some things that may help:

  • Get the technology out of the equation.  Yes, it’s Tableau, but what is this person doing?  If the client can only get 3 things, what are the top 3?  What is peripheral?  What’s the long-term outlook of this role (if it’s just status quo ad infinitum, then you need a task rabbit, not an expert)?  What is the immediate need for this role?
  • Who is a good example?  Movie pitches are great for this.  It’s like Castaway but on Mars (the pitch for The Martian).  We need Zoe, but for Tableau reports.  Zoe does a lot of ETL work, requirements gathering, and server monitoring for this other product.  Pavel is usually in front users, educating and supporting them.  Or we need someone that’s a hybrid of the 2 of them.  You can also use LinkedIn for this – we like this person, but need someone in X-Y budget range.  Chances are, you can’t get the original person, but you may get a colleague or mentee.
  • What’s 80% of the job?  This is what’s most important.  Sure, the kitchen sink is tempting, but it will scare away some of the “dream people” you covet.  Those who specialize heavily in Tableau are far less likely to consider a job with loads of ETL work or lists of other seemingly related tools (to understand, see how this analyst reacted to using PowerBI).
  • Look at some good examples.  Joshua Tapley runs a stellar program at Comcast.  Paul Banoub has no shortage in finding talent.  PluralSight just nabbed 2 rockstars in one month.

Step 3: Focus on attainment and practice, not years

Too often we equate years experience with skills.  In the days of old, this was true.  There are a number of people who have done exceptionally well with Tableau in just a year’s time.  Sometimes, this is because of other experiences.  Other times, it’s because of practice and deliberate learning and failing.  Or, duos push themselves and others to new heights.

I see (too many) job descriptions for Tableau with a demand for over 5 years of experience.  Let’s give this a quick think.  Tableau was a software start up 13 years ago (that seems like ages!).  They had their first customer conference in 2008 with 187 attendees.  Tableau Public launched in 2010, which is right around the same time that I, in Ohio, started using Tableau.  There was no user group yet and the first one that we had, I was one of the more experienced users (thankfully, Michael Cristiani was there).  I am one of the more mature users, despite starting out so late.  It’s still reaching critical mass.  2-4 years is probably the median.

Additionally, years using can deceptive.  I started in 2010, but did not use it daily until 2014 when I started consulting.  Before that, it was just one of MANY things I did.  I didn’t get “comfortable” with what I knew until 2015, when I started participating in the community.  Others have done this too.

Try this instead:

  • Get practical examples.  Public is free and there’s numerous projects out there to use data, like Viz for Social Good and Makeover Monday where people can practice and show of their skills.  Start here by making a Twitter account and finding their hashtags (or click the links).  Or, find a data set to use as a benchmark.  School sites are often a good place to start.
  • Seek an expert referral or opinion.  Keep in mind, the blind request or the one sided relationship does not build for a long term favorable outcome.  Consider buying some consulting time to review a job description.  If you’re a recruiter, add this to your fee.
  • Go where the people are.  Tableau people play in lots of spaces, the most common being Twitter, the Tableau Forums, and even Slack.  There, you will find people enamored with the product.  Again, mind your manners and don’t blast these channels.  Engage in conversations and learn.

Step 4: Break the mold

We also get locked in on specific criteria, such as degrees.  I’ll let you in on a little secret: many people change careers a number of times and SPECIFICALLY women and minorities rarely start in IT.  By limiting yourself to only certain degrees, you miss some amazing people.  Consider this – you’re reading this blog, likely because you found me on LinkedIn and I can almost guarantee I don’t fit your checklist.  But, yet, you’re here.

Escaping the hard mold may cause your client to panic.  Likely, they’re stuck in the same staid waters, which is why the hired you.  Sell the oddballs, and also make sure you’re finding them in the first place as well.

  • Check your data.  Do your candidates all look the  same, either on paper or in person?  There’s likely something amiss as variations on a theme still comes with some variety.  Your client wants a cookie, which means chocolate chip, peanut butter, shortbread, and orange blossom all apply.
  • Explore other avenues.  Find some of the voices you want and check their network.  Again, you can ask, but remember to keep the balance.
  • Try A/B testing.  Swap up the emphasis on certain skills and play with minor rewrites in different places.
  • Seek feedback.  If you’re feeling really brave, post a redacted/simplified description to a Tableau specific channel and ask why people would NOT apply to that job.  Besides some of the standard answers, you may get valuable insight into where you’re turning people off.

Step 5: Track success

The more you do this, the more you learn.  And, you can keep track with data!  Which means you – yes, you recruiter – could use Tableau as well.  Or, you can hire a consultant to help you make dashboards or understand the industry.


Terms you should know

(in no particular order)

  • Data – information stored in a way computers, and sometimes humans, can read.  Usually permanent and liable to be used against you.
  • BI – a term from the ’90’s when data came in 1 size and usually flopped around in cubes.  Some will say it stands for business intelligence, but people smarter than me call it an oxymoron.
  • Cubes – a fast, easy way to scare off many consultants.  These things are like a car boot for the best parts of Tableau.
  • ETL – this is the clean up work around data.  “Extract (just like teeth), transform (think those awful makeover shows), and load (you know that overnight FedEx job – exactly like this with smaller muscles and bigger boxes).
  • Tableau – this comes in 2 primary flavors: desktop and server.  Desktop is on the desktop and the primary place people build dashboards.  Server is where people host, disperse, and (sometimes) edit dashboards.
  • Dashboard – both a specific term in Tableau and a more generic term for the end product that software like Tableau creates.  Sort of like the dashboard in a car, it tells you the important stuff.
  • Report – a generic term sometimes used in place of dashboards, but more often used to denote an old-school mostly table-centric build.
  • Table – a way of showing in data in (often) the most inefficient way possible, or as part of a database.  Great for CFOs and storing data, bad for most humans.  Confused?  Go here.
  • Charts – a way of presenting data so most humans can (hopefully) understand it.  Can be used for great evil as well, most often on the news.
  • SQL – usually pronounced like those secondary movies – sequel.  It requires writing things like SELECT, GO, CREATE TO, JOIN, and UNION.  Some of us call this code.  It looks a lot like some of the images in this post.  Go figure.  Or perhaps SELECT FIGURE GO.
  • Big Data – when regular data just won’t do.  Some people try to use this for scale, but it’s wholly arbitrary.  Some people call their 20-table databases big data.  Others call their Hadoop implementation big data, regardless of size.  Just like the difference between turquoise and teal or maroon and burgundy, the line from one to the other is subjective.
  • Hadoop – this is one of many databases Tableau can connect to.  It comes in many flavors.  My favorite is mint.
  • Viz – Informal speak for visualization/visualisation.  Plural = vizzes according to Tableau Public.
  • BI 1.0 – First or legacy BI tools.  They’ve been around awhile.
  • Crystal Reports – A tool we all probably used back in the ’00’s or the ‘aughts’ if you want to be fancy.  It changed names to fool us.  But we know better.  BI 1.0 tool, AKA old school.
  • Business Objects – AKA Crystal Reports. BI 1.0 tool, AKA old school.
  • Cognos – Also a BI 1.0 tool.
  • OBIEE – No.  Just no.
  • MicroStrategy – BI 1.0 with ability to make mobile apps.
  • PowerBI – Microsoft’s (latest) foray into analytics.  See above link.
  • Qlik (Sense or View) – BI 2.0 tool.  For a non-biased view of these tools and more, see Jen Underwood.
  • Jen Underwood – BI expert extraordinaire.  If you want to understand analytics and all the software players, VISIT HER SITE.

Leave a Reply

Your email address will not be published. Required fields are marked *