Friday, January 24, 2014

The Problem with SOME BI Pundits

The Problem with SOME BI Pundits
Late night ramblings by Chuck Hooper

Setting the stage:  my dictionary defines “pundit” as: 
        “a critic or authority on a subject, especially in the media

There are too many well known “experts” when the topic of Business Intelligence or Big Data or Data Warehousing etc. come up.
They have written books on the subjects.  They deliver key note addresses.  They tweet (sometimes way too often).  They do webinars.  They offer seminars.  They do white papers.  They profess their keen insights into all areas of data driven analytics.

I’ve not written a book - been busy, working.  I am planning one, now.
I’ve done too few speaking engagements - too busy, working.  I plan to do more, starting now.
I don't tweet, often – too busy working.  I will be doing more.
I’ve hosted few webinars / seminars – again, too damned busy, working.  Might do more.
I published few white papers – still too busy, working.  Might do one, someday.
BUT, for anyone interested, I am now willing to share many of my insights, extracted from 49 years of real work.

I’ll start with a few of the simple ones:
1.    Simplicity succeeds.  Complexity fails.  When the pundit puts up the diagram you can’t understand after 15 seconds, believe it is too complex, and, whatever it’s preaching won’t work - even though some vendor convinced him/her it would.
2.    Innovation is NOT building a better mouse trap.  We catch mice, today.  True innovation is solving something not solved, before.  When the pundits offer innovative solutions, BEWARE!  It’s, most likely, just a more complex way of getting to old results.  Probably costs more, too.
3.    Many people want answers, not capability.  Beware of pundits offering new ways based on little more than flawed research and poorly designed surveys.
4.    If we were not, all, in the same situation, Dilbert will not be such a popular cartoon.  It’s likely your problem has been solved, before. Avoid reinventing the wheel!
5.    One of our biggest problems is there are too many solutions.  Consider a full study of all current BI tools, running on all possible data bases, running on all possible operating systems, running on all possible hardware platforms.  Tens of thousands of possibilities.  Do you study them all?  Do you do your due diligence?  (I define “due diligence” as a delaying practice initiated by those that don’t know what to do, and, won’t trust those that do.).  If we start, and find out we are heading in the wrong direction, we can turn around.  If we never start . . . . .
6.    Talk is cheap.  Action can be expensive.  NO action is, typically, VERY expensive!
7.    Common sense is not that common! 

As for specifics on some of the better known pundits.  There are ones I think are good and ones I think are wasting my time.  I won’t name any of the bad, nor will I try to name all the good ones.  I will name two that I follow, on Twitter, and appreciate their many contributions and insights:
n  Neil Raden (Twitter: @NeilRaden). Lots of insights, and, a warped sense of humor –I like that.
n  Howard Dresner ( Twitter: @howarddresner ).  Howard “pulls it all together” - partially through crowd sourcing & partially through a wealth of experience.    

That’s enough, for now – I have to get back to work…


Chuck Hooper is a business intelligence and data visualization consultant. Chuck brings 49 years of business and IT experience with him. In addition to the current consulting offerings, Chuck does speaking engagements, and, conducts training sessions on the use of Tableau Software products, visual analytics, data warehouse design, and other business intelligence topics, at both the technical and the executive levels. 
For further information: email: chooper@bialytics.com   Twitter:@chuck_hooper   web:www.bialytics.com 
Material presented Copyright © 2012,2013,2014 by Chuck Hooper. You may quote / use any of the material by giving the author credit.


The Problem With BI Tool Implementations

The Problem With BI Tool Implementations
Late night ramblings from Chuck Hooper
 

First problem - there are no BI tools! 
A better title for this might have been “The Problem With “so called BI Tool” Implementations”.
All those we call "BI Tools" are really nothing more than aids to analyzing, formatting and disseminating information.  The REAL intelligence comes from people, and nowhere else!  So, the goal needs to be to get bright people, the right tool(s), the right data, and, to support them!

What are the problems with many BI tool implementations?
1.     Unrealistic expectations!  No one tool does it all - in some situations, you may need to consider more than one of these tools. That’s OK!  NONE are “magical”.  ALL require some level of work to implement.  The goal is to pick the tool that minimizes the implementation process.
2.     Long learning curve / too complex.  Both lead to delayed productivity, and, low adoption!  The low adoption can be a killer.  Potentially good analysts might be left behind.   Pick the tool that has the easiest start up learning curve.  It is, likely, you’ll never “master” any tool.  So, work to get productive, sooner!   
3.     High implementation cost.  Talk is cheap. Action can be expensive. NO action can be REALLY expensive!  There’s a question used by application designers: “Do you want that quick, cheap, or good?  Pick any 2 of the 3”.     If you want it cheap, it is either not going to be good, or, won’t be done quickly.  Part of the cost is the learning curve in the use of the tool; part is based on the adoption rate.  Do NOT confuse price and cost. 
Faster implementations tend to be cheaper and have greater ROI!    Hmmmmmm, MAYBE we can pick all three!
4.     Too many options!  There are a lot of so called “BI tools” available.  Be careful of doing your “due diligence”.   That’s what I call a practice initiated by people that don’t know what to do, and, won’t trust those that do!  Your problem is not unique – most likely it has been solved, before.  Do not reinvent the wheel.  If we were not all in the same boat, Dilbert would not be so popular!  
5.     Keeping it simple!  It’s easy to get hung up on complex processes that can lead to extended and expensive implementations.  Work hard to keep your efforts simple!  Even then, it is likely your end results will be overly complex!  Beware of tools requiring a bunch of third party add-ons, or too many individual pieces be installed.  Simplicity succeeds.  Complexity fails.  Believe it.  Live it. 
6.     Lack of actionability.  Many BI implementations have pretty dashboards and scorecards – lots of flash & pizazz.     If your outputs are not actionable, you’ve wasted your efforts.  This is not an issue with the tool(s) – lack of actionable insights is bad planning, design and implementation.  Too often, it’s the tool that is blamed (Probably because it can’t fight back!).  So, be prepared to take ownership of your successes, AND, of your failures.  Fix the failures.  Rejoice in the successes.  Move on.
7.     “Get it done and we can rest” mentality.  You are NEVER done.  A successful BI environment is one where all aspects of it are fluid / iterative.  Think of your competition as a 10 foot long crocodile.  You stop fighting AFTER he’s stopped fighting!     

Back in the mid-20th century, Aristotle Onassis said “The secret of business is in knowing what no one else knows.”.  That was OK, 60-70 years ago.  We call this the “information age” for a reason.  It’s likely you don’t know more than your competition.  There’s a new secret, today.  “Take what you know and do more with it, sooner, faster, cheaper.”. 
n  More:  Pure people smarts, nothing else.
n  Sooner:  A decision – are you going to start that project now? Or, wait until competition eats your lunch?
n  Faster:  This is NOT “fastest response” times.  It is “faster to insight”.
n  Cheaper:  This is not the price of the tool or the infrastructure.  It is the total cost of ownership, including training, support, training, support, training, support etc.  And, do not forget training and support.  

To be successful…  
I believe it was Christopher McDougall  that said:  “Every day, in Africa, a gazelle wakes up, knowing he must run faster than the fastest lion, or, he will be eaten.  Every day, in Africa, a lion wakes up, knowing he must run faster than the slowest gazelle, or, he will starve.  It matters not, whether you are the lion or the gazelle.  When the sun comes up, you’d better be running.”.     Is YOUR competition already up and running?


Chuck Hooper is a business intelligence and data visualization consultant. Chuck brings 49 years of business and IT experience with him. In addition to the current consulting offerings, Chuck does speaking engagements, and, conducts training sessions on the use of Tableau Software products, visual analytics, data warehouse design, and other business intelligence topics, at both the technical and the executive levels.  
For further information: email: chooper@bialytics.com   Twitter:@chuck_hooper   web:www.bialytics.com 

Material presented Copyright © 2012,2013,2014 by Chuck Hooper. You may quote / use any of the material by giving the author credit.

Analysis Lesson From a 5 year Old

Analysis Lesson From a 5 year Old 
Late night ramblings from Chuck Hooper

I was taking care of my 5 year old granddaughter for the day.   She wanted to paint some pictures.  I pulled out her water colors and set them up on the kitchen table.  She took a brush, dipped it in water, then into a few different color paints.  She proceeded to put ten blobs of color on ten different places on the paper.  
She held up the final product and asked “Papa, do you like this?”.  
I smiled and nodded a big YES!  She then asked “Do you know what they are?”.  I said they looked like pretty birdies.  She frowned and tossed the paper aside: “NO!  They’re kitty cats!

We repeat the process – ten more blobs of new colors, in new shapes, all over the paper.  
Same questions asked.  
THIS time, I am going to be smarter!  The last ones were kitty cats.  She’s 5.  These MUST be puppy dogs!  
She says “No!  They’re horsies!“.  Another work of art tossed to the side.

One more time…. 
I am an analyst.  I am a bright guy.  No 5 year old will stump me a third time!  
This time, about a dozen colorful blotches on the paper.  Same questions asked.  I ask if I am allowed multiple guesses.  She says “Yes.”, so, I start: ”Are they aardvarks?  Abacuses?  Accountants?  Adams apples?  Affidavits?  Aggravated grandpas?…”  
I go through the alphabet.  I go back through the alphabet.  And again.  
30 minutes later, I am sweating.  
She is bored with saying “No!” so many times.  
Finally, I am done.  I say: “Papa gives up.  What are they?”
  
She shakes her in disgust, rolls her eyes (more disgust), and says:  
            “Papa!  Sometimes they’re just dots!“.

How often, as analysts, do we look for insights, and not find them, only to keep trying until we pass out?  
How many times is the data we are looking at just dots in a scatter plot?  
How do we know when to look elsewhere?  And, where is that elsewhere?  
What other data do we need to add to the mix, to make the data we already have, useful?  
The so-called BI "tools" can’t answer these questions.  These are the questions that need human brain power to answer.  The tool you use to analyze the data should let you “play with your data”.  My friends at Tableau Software like to say “Party with your data!”.  It’s the same idea.  It’s the easiest way to find those gold nuggets hiding in the numbers.  It’s also the fastest way to find out if you’re just playing with dots!



Chuck Hooper is a business intelligence and data visualization consultant,  He brings 49 years of business and IT experience with him. In addition to the current consulting offerings, he does speaking engagements, and, conducts training sessions on the use of Tableau Software products, visual analytics, data warehouse design, and other business intelligence topics, at both the technical and the executive levels. 
For further information: email: chooper@bialytics.com   Twitter:@chuck_hooper   web:www.bialytics.com 

Material presented Copyright © 2012,2013,2014 by Chuck Hooper.  You may quote / use any of the material by giving the author credit.